ALLAN HANCOCK FOUNDATION PUBLICATIONS c^. ^ OF ^n^ THE UNIVERSITY OF SOUTHERN CALIFORNIA First Series ALLAN HANCOCK PACIFIC EXPEDITIONS Volume 13 1946-1950 THE UNIVERSITY OF SOUTHERN CALIFORNIA PRESS LOS ANGELES 7, CALIFORNIA 1950 ALLAN HANCOCK FOUNDATION PUBLICATIONS ALLAN HANCOCK PACIFIC EXPEDITIONS Volume 13 1946-1950 THE UNIVERSITY OF SOUTHERN CALIFORNIA PRESS LOS ANGELES 7, CALIFORNIA 1950 CONTENTS 1. The Bryophyta of the Allan Hancock Expedition of 1939, by William Campbell Steere 1- 4 2. Land Plants collected by the Velero III, Allan Hancock Pacific Expeditions, 1937-1941, by Howard Scott Gentry, (Plates 1-15, Maps 1-3 ) 5-24^ ^ 3. Plant Ecology of the Channel Islands of California, by Meryl Byron Dunkle, (Figures 1-12, Plates 1-6) 247-386 Index for Land Plants of Allan Hancock Pacific Expeditions 237-246 Index for Plant Ecology of Channel Islands of California 381-386 REPORTS ON THE COLLECTIONS OBTAINED BY ALLAN HANCOCK PACIFIC EXPEDITIONS OF VELERO III OFF THE COAST OF MEXICO, CENTRAL AMERICA, SOUTH AMERICA, AND GALAPAGOS ISLANDS IN 1932, IN 1933, IN 1934, IN 1935, IN 1936, IN 1937, IN 1938, in 1939, in 1940, and in 1941 THE BRYOPHYTA OF THE ALLAN HANCOCK EXPEDITION OF 1939 By WILLIAM CAMPBELL STEERE The University of Southern California Publications Allan Hancock Pacific Expeditions Volume 13, Number 1 Issued May 27, 1946 The University of Southern California Press Los Angeles, California THE BRYOPHYTA OF THE ALLAN HANCOCK EXPEDITION OF 1939 By WILLIAM CAMPBELL STEERE The collection of bryophytes brought back by the Allan Hancock Expedition of 1939 is much smaller than that resulting from the 1934 Expedition, 1 and contains no species new to science. However, since Dr. Wm. Randolph Taylor and Mr. Francis H. Elmore collected the speci- mens in areas little known by botanists, such as the west coast of Costa Rica and Panama, and Socorro Island of the Revillagigedo group (Mexico), it seems worth while to report on them. The five species of Hepaticae listed below were identified by Dr. Margaret Fulford of the University of Cincinnati, and I wish to acknowl- edge here my obligation to her. All specimens are deposited in the her- baria of the Allan Hancock Foundation, The University of Southern California, and the University of Michigan, and a set of the Hepaticae are in the possession of Dr. Fulford. HEPATICAE Lejeuneaceae Brachiolejeunea corticalis (Lehm. & Lindenb.) Schiflfn., Hedwigia 33:180. 1894. Costa Rica (West Coast) : On a large smooth-barked tree in forest, Golfo Duke; 26 March 1939; IV. R. Taylor No. 39-754. Distribution: Florida; West Indies; Central America; northern South America. Caudalejeunea Lehmanniana (Gottsche) Evans, Bull. Torrey Bot. Club 34:554. 1907. Costa Rica (West Coast) : On twigs in the forest, Golfo Dulce; 26 March 1939; W. R. Taylor No. 39-753. Distribution: Florida; West Indies; Central America; northern South America. Rectolejeunea Berteroana (Gottsche) Evans, Bull. Torrey Bot. Club 33: 12. 1906. 1 Steere, Mosses of the G. Allan Hancock Expedition of 1934, collected by Wm. R. Taylor. Hancock Pacific Expeditions 3(1) : 1-12. 1 pi. 1936. C 1 ] 2 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Costa Rica (West Coast) : On bark of trees In the forest, Golfo Dulce; 26 March 1939; W. R. Taylor No, 39-765. Distribution : Florida ; West Indies ; British Honduras. This seems to be the first report of the species from Costa Rica. Stictolejeunea Kunzeana (Gottsche) Schiflfn., in Engler & Prantl, Nat. Pflanzenfam. 1(3): 131. 1895. Costa Rica (West Coast) : On bark of trees in forest, Golfo Dulce; 26 March 1939; PV, R. Taylor No. 39-764. Distribution : Andes of South America. This seems to be one of the first collections to be reported from Central America. Stictolejeunea squamata (Willd.) Schiiifn., in Engler & PrantI, Nat. Pflanzenfam. 1(3): 131. 1895. Costa Rica (West Coast) : On the bark of trees in forest, Golfo Dulce; 26 March 1939; F. H. Elmore No. 39-759, W. R. Taylor No, 39-760. Distribution: West Indies; Central America; northern South America. MUSCI FiSSIDENTACEAE FissiDENs Garberi Lesq. k James, Proc. Amer. Acad. 14: 137. 1879. Mexico (West Coast) : On soil with a small fern, three-quarters of a mile up the canj^on from the landing place, Braithwaite Bay, Socorro Island, Revillagigedo Islands; 18 March 1939; F. H. Elmore No. 39-751. Distribution: Southern United States; West Indies; Mexico; Central America. Calymperaceae Calymperes Donnellii Austin, Bot. Gaz. 4: 151. 1879. Costa Rica (West Coast): On twigs in the forest, Golfo Dulce; 26 March 1939; F. H. Elmore No. 39-757, W. R. Taylor No. 39-762. Trinidad : On branches of Cacao plants by the side of the Manzanilla Beach Road ; 18-20 April 1939 ; JV. R. Taylor No. 39-768. Distribution: Southern United States; West Indies; Central America; northern South America. In spite of its wide distribution in tropical America, this species has not been previously reported from Costa Rica. NO. 1 steere: bryophyta 3 Calymperes Richardi C. Mull., Syn. Muse. 1 : 524. 1849. Costa Rica (West Coast) : Growing high in a tree, among orchids, Port Baker, near Salinas Bay; 24-25 March 1939; F, H. Elmore No. 39-752. Distribution: Southern United States; West Indies; Central America ; northern South America. Although this species is widely dis- tributed in the American tropics, it has not been reported before from Costa Rica. Erpodiaceae Erpodium domingense (Brid.) C. Miill., Bot. Zeit. 1 : 774. 1843. Mexico (West Coast) : On soil with a small fern, mixed with Fissidens Garberi Lesq. & James, three-quarters of a mile up the canyon from the landing place, Braithwaite Bay, Socorro Island, Revillagigedo Islands; 18 March 1939; F. H. Elmore No. 39-7 51a. Only one small stem of this unmistakable moss was found, but it was enough to establish the identification beyond any doubt. Distribution: Cameron County, Texas; Santo Domingo; Haiti; Puerto Rico; Yucatan; Guatemala. This is the first collection and report of what has always been considered to be a typically Caribbean species from the Pacific side of Central America. Orthotrichaceae Micromitrium fragile (Mitt.) Jaeg., Adumb. 1:435. 1872-73. Trinidad : On branches of Cacao plants by the side of the Manzanilla Beach Road; 18-20 April 1939; W. R. Taylor No. 39-767. Distribution: West Indies; Mexico; Central America; tropical South America ; Galapagos Islands. Although this species is known from the West Indies, I do not believe that it has been reported previously from Trinidad. Leucodontaceae Leucodontopsis floridana (Aust.) E. G. Britton, Bryologist 15:28. 1912. Trinidad : On branches of Cacao plants by the side of the Manzanilla Beach Road ; 18-20 April 1939 ; JV. R. Taylor No. 39-769. Distribution: Florida; West Indies; Mexico; Central America; northern South America. ™^ : 4 allan hancock pacific expeditions vol. 13 Pterobryaceae Orthostichopsis tetragona (Hedw.) Broth., in Engler & Prantl, Nat. Pflanzenfam. 1(3) : 805. 1906. Panama (West Coast) : In forest, Bahia Honda; 26 March 1939; F, H, Elmore No. 39-755. Distribution: West Indies; Mexico; Central America; tropical South America. Meteoriaceae Papillaria nigrescens (Hedw.) Jaeg., Adumb. 1:169. 1875-76. Trinidad : On branches of Cacao plants by the side of the Manzanilla Beach Road; 18-20 April 1939; W. R. Taylor No. 39-766. Distribution: Florida; West Indies; Mexico; Central America; tropical South America. This is a common and very widely distributed species. Meteoriopsis patula (Hedw.) Broth., in Engler & Prantl, Nat. Pflanzenfam. 1(3): 825. 1906. Trinidad : On branches of Cacao plants by the side of the Manzanilla Beach Road; 18-20 April 1939; W. R. Taylor No. 39-768. Distribution: Florida; West Indies; Mexico; Central America; tropical South America ; Galapagos Islands. PiLOTRICHACEAE PiLOTRiCHUM AMAZONUM Mitt., Journ. Linn. Soc, Bot., 12:387. 1869. Costa Rica (West Coast) : On delicate twigs in the forest, Golfo Dulce; 26 March 1939; fV. R. Taylor No. 39-756; on the bark of trees, W. R. Taylor No. 39-763. Distribution : Amazon region of South America, extending northward through Panama to Guatemala. This is apparently the first report of the species from Costa Rica. REPORTS OF THE COLLECTIONS OBTAINED BY ALLAN HANCOCK PACIFIC EXPEDITIONS OF VELERO III OFF THE COAST OF MEXICO, CENTRAL AMERICA, SOUTH AMERICA, AND GALAPAGOS ISLANDS IN 1932, IN 1933, IN 1934, IN 1935, IN 1936, IN 1937, IN 1938, IN 1939, IN 1940, AND IN 1941 LAND PLANTS COLLECTED BY THE VELERO III, ALLAN HANCOCK PACIFIC EXPEDITIONS 1937-1941 (Plates 1-15, Maps 1-3) By HOWARD SCOTT GENTRY The UNivERSiTi' of Southern* California Publicatioxs Allan Hancock Pacific Expeditions Volume 13, Number 2 Issued July 22, 1949 The UNivERsiri' of Southern California Press Los Angeles, California TABLE OF CONTENTS PAGE List of Illustrations 9 General Introduction 1 New Names Proposed 13 Introduction to the Catalogues 13 Channel Islands 17 Introduction 17 Catalogue of Collections 20 Literature Cited 42 Cedros and San Benito Islands 43 Introduction 43 Catalogue of Collections 49 Literature Cited 62 Revilla Gigedo Islands 63 Introduction 63 Catalogue of Collections 68 Literature Cited 73 Tres Marias Islands 74 San Juanito 76 Maria Madre 76 Maria Magdalena 77 Maria Cleofa 79 Literature Cited 80 California Gulf Region 81 General Physiography 81 Postinsular Localities 85 Climate 88 Collection Localities 89 Summary of Insular Floras 99 Catalogue of Collections 117 Literature Cited 181 Jalisco and Oaxaca 182 Introduction 182 Catalogue of Collections 183 Costarica 192 Introduction 192 Catalogue of Collections 194 Plates 204 Index 237 [7] LIST OF ILLUSTRATIONS Plates 1. Moving a oletfera Lam. Fig. 1. section of branchlet with leaf x^. Fig. 2. pod x^. 2. Moringa oletfera Lam. Fig. 3. flower at anthesis x3.3. Fig. 4. stamen xll. Fig. 5. bud x5. Fig. 6. sepal x6.7. 3. Echinopepon peninsularis Gentry- Fig. 7. habit x^. Fig. 8. node x4. Fig. 9. fruit prickle x6.5 to compare with. Fig. 10. fruit prickle x6.5 of Echinopepon minimus. 4. Fig. 11. View on the north end of Santa Rosa Island. Grassland covers the more level slopes, while bushy perennials are spotted on the bluff. (Photo from the Los Angeles Museum Channel Island Survey) Fig. 12. Braithwaite Bay, Socorro Island on a calm cloudy day in March. 5. Fig. 13. Punta Frailes, Cape District, Baja California. The arborescent growth is widely dispersed on the steep rocky slopes, dense on the outwash fans. 6. Fig. 14. Vegetation above Frailes Bay, Cape District. On the basic rock slope the vegetation is sparse and stunted. Fig. 15. Dense Thorn Forest vegetation in a broad wash near Frailes Bay, Cape District, Baja California. 7. Fig. 16. View southward near Puerto Escondido, Baja California, over- looking a narrow coastal plain with the scarp of the Sierra Giganta in the background. The foreground shows a xerophytic grass ground cover with scattered tree of Bursera microphylla and Lemaircocereus Thurberi. Fig. 17. Canyon above Escondido, Baja California. The steep slopes are brecciated lavas ; the palm, Erythea Brandegeei. 8. Fig. 18. Angel de la Guardia Island. Typically sparse desert vegetation on washes and fans with a scattered grove of Pachycereus Pringlei. Fig. 19. Angel de la Guardia Island. The effect of wind on the sarcophytic tree Pachycormus discolor puhescens along a rocky crest. Fig. 20. Angel de la Guardia Island. Sparse Desert Shrub on an east exposure with Pachycereus Pringlei on the lower gentler slopes. 9. Fig. 21. Tiburon Island. Low Desert Shrub on the granitic terrain of the southeast coast. Fig. 22. San Pedro Nolasco Island, showing the "raw" rock surfaces, almost no soil, and adventive or pioneering perennials. [9] 10. Fig. 23. San Pedro Nolasco Island. A clump of Echinocereus grandis. Fig. 24. San Pedro Nolasco Island. A succulent vegetation of Agave, Opuntia, and Pachycereus on very rocky terrain. 11. Fig. 25. San Pedro Nolasco Island. Agave chrysoglossa and Lemaireocereus Thurberi in foreground. Fig. 26. San Pedro Nolasco Island. A dense colony of the succulent low shrub, Pedilanthus macrocarpus. 12. Fig. 27. Espiritu Santo Island. A dispersed shrub formation with scattered trees of Pachycereus Pringlei. Fig. 28. Espiritu Santo Island. Detail of branch and fruit of Opuntia cholla. 13. Fig. 29. Tenacatita Bay, Jalisco. The subtropical forest is close upon the beach. Fig. 30. Chacahua Bay, Oaxaca. The scrubby vegetation on the hill in the background shows evidence of having been cut over. 14. Fig. 31. Lowland coastal vegetation of the tierra caliente in Costa Rica near Port Culebra. Fig. 32. Close lowland forest of Costa Rica with a dense tangle of trunks, limbs, and clambering stems, crooked to semi-straight. 15. Fig. 33. Forest of the tierra caliente in the Golfo de Dulce, Costa Rica. The varied tree forms indicate the richness of the flora. [10] LAND PLANTS COLLECTED BY THE VELERO III, ALLAN HANCOCK PACIFIC EXPEDITIONS 1937-1941 (Plates 1-15, Maps 1-3) By Howard Scott Gentry General Introduction The complete itinerary of the voyages of the Velero III on the Allan Hancock Pacific Expeditions is given by Eraser (1943). An outline of the land plants secured are given in Table 1. Since the expeditions were primarily concerned with marine biology and especially the collection of marine faunas, the land plant collections were only incidental and are not large. They are samplings of several distinct floral elements of west- ern North America. ( 1 ) Those from the Channel Islands belong to the unique California flora, in one of the five regions of the world having a Mediterranean type of climate. This climate is characterized by winter rainfall, dry summers, and maritime influence conducive to equable temperatures, on- shore winds, and regular seasonal fogs. Cedros Island, off the west coast of middle Baja California, contains in its high elevations a southern out- post of the California flora. (2) The Sonoran Desert flora is generally peripheral to the Gulf of California. It extends farther south on the peninsula (to the Cape Dis- trict) than it does on the mainland (to about Guaymas, Sonora). In its area an arid contintental type of climate competes with an arid maritime one, the former most evident in the northern part of the region around the lower Colorado River basins, while the latter is particularly stead- fast in the middle and southern outer coastal part of the peninsula. While both types have low and irregular rainfalls, the maritime desert differs in having more equable annual and daily temperatures, higher relative humidity, and summer rainfall is more common in the southern latitudes. This latter feature is a tropical factor in the situation. (3) The Sinaloan subtropical flora is the great transitional element between deserts and tropics. Rainfall is about 90% summer. The high temperatures are ameliorated, particularly through the spring, by on-shore westerlies. The dominating life form is the tree of short to medium [11] OGIC^I^ 12 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Stature, densely spaced, with epiphytes common in the more humid locali- ties. On the mainland of Mexico, the Sinaloan flora occupies the low and middle elevations from southern Sonora south at least to Nayarit. Farther south it is local or transitional with the Central American floral element. The Velero III collections from the Tres Marias Islands, the Revilla Gigedo Islands, and the Cape District of Baja California are referable to the Sinaloan flora. (4) The Central American flora is distinctly tropical under medmm to high rainfalls. It contains a very great number of species and is domi- nated by both evergreen and deciduous trees of medium to tall stature. It is further characterized by broad leaf blades and an abundance of epi- phytes. The Pacific part of the area is distinctly arid as compared to the Atlantic, and its coastal forests are mostly deciduous, of medium stature, and are interspersed with areas of savanna. There is a long dry season from November to May. From the standpoint of vegetation there is little to separate the Pacific coast of Costa Rica from the Pacific coast of southern Mexico, since the dominating plant forms and associations apparently extend throughout with but little modification as far north as Nayarit. The Velero III collections from coastal Jalisco, Oaxaca, and Costa Rica reflect this general relationship. The report has been organized according to geographic regions under the following headings: The Channel Islands of California, Cedros and San Benitos Islands, Revilla Gigedo Islands, Tres Marias Islands, The California Gulf Region, Jalisco and Oaxaca, Mexico, and Costa Rica. The desirability of so enumerating these scattered collections, representa- tive of several regions and several floras, is so obvious it is unnecessary to recount them here. Each section is introduced by a general discussion of the physiography, the climate, and the plant geography with special emphasis on the historical or developmental aspects of the. flora involved. I have also attempted to evaluate the botanizing that has been done to date on the respective areas, the islands in particular. The discussions are opinionated summaries, according to my experiences and observa- tions in the fields or to reports read, rather than documented conclusions. It is hoped that they will stimulate interest and activity in the richly re- warding botanical field of northwestern Mexico. As is usual in works of this kind, the author is indebted to many people for their willing assistance in making the report possible. To all of them he extends his sincere thanks. Ira L. Wiggins of Stanford Uni- NO. 2 GENTRY : LAND PLANTS 13 versity, E. Yale Dawson and Kenneth O. Emery of the Allan Hancock Foundation and The University of Southern California, and P. A. Munz of the Santa Ana Botanical Garden all read various sections of the manuscript and made valuable criticisms and suggestions. The following taxonomists made determinations in certain genera or families: S. F. Blake National Arboretum Compositae Elzada Clover University of Michigan Cactaceae L. Constance University of California Hydrophyllaceae E. Y. Dawson Allan Hancock Foundation Cactaceae L. H. Harvey Montana State University Gramineae C. V. Morton National Herbarium Pteridophytes Hugh O'Neill Catholica University of America Cyperaceae R. C. Rollins Stanford University Cruciferae L. C. Wheeler University of Southern California Euphorbiaceae I. L. Wiggins Stanford University Malvaceae New Names Proposed in this Publication Agave costaricana Gentry sp. nov. Agave Shawii sebastiana (Greene) Gentry new comb. Lyrocarpa linearifolia Rollins sp. nov. Calliandra Brandegeei (Brit. & Rose) Gentry new comb. Tephrosia hamata (Rydb.) Gentry new comb. Dalea variegata (Rydb.) Gentry new comb. P achy cormus discolor Veatchiana (Kell.) Gentry new comb. Pachycormus discolor pubescens (Wats.) Gentry new comb. Echinopepon peninsularis Gentry sp. nov. Vaseyanthus Palmeri (Wats.) Gentry new comb. Introduction to the Catalogues In the plan of the catalogues of species collected by members of the Allan Hancock Expeditions, families are listed according to the sequence of the Engler and Prantl system. The genera and species are listed alpha- betically under each family. The first entry under the species heading is the citation of collection together with any note that the collector may have left upon his field label. It will be noted also that I have included the date of collection, which I estimate of some importance because it gives a record of the time of flowering or fruiting. In cases where the specimen is sterile, it is either so indicated in parentheses along with speci- men citation or is noted in the following paragraph. 14 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 TABLE 1 LAND PLANTS COLLECTED BY THE VELERO III Collector and Locality and Date his numbers 1937 Baja California (outer coast) Manuela Lagoon near Lagoon Head Anchorage, March 1 Rempel 1-22 San Juanico Bay, March 2 Rempel 24-52 Cabeza Ballena, March 3 Rempel 54-74 Islands in Pond Lagoon, July 11 Rempel 350-355 Gulf of California Ensenado de los Muertos, Cape District, Baja California, March 5 Rempel 76-78 San Gabriel Bay, Espiritu Santo Island, March 7 - Rempel 80-1 00 San Francisco Island, March 9 Rempel 101-112 Agua Verde Bay, Baja California, March 10 Rempel 114-140 Puerto Escondido, Baja California, March 13 Rempel 141-170 Ildefonso Island, March 15 Rempel 170a-170b West Cove in Concepcion Bay, Baja California, March 15 Rempel 171-186 Island in Concepcion Bay, Baja California, March 16 Rem.pel 188-209 Tortuga Island, March 17 Rempel 210-230 Los Angeles Bay, Baja California, March 19, 20 Rempel 231-250 North end of Los Angeles Bay, Baja California, March 20 Rempel 251a-262 Puerto Refugio, Angel de la Guardia Island, March 20 Rempel 262-280 Willard Point, Gonzaga Bay, Baja California, March 23 Rempel 283-285 North of Point Lobos, Sonora, March 26 Rem.pel 287-288 Patos Island, March 26 Rempel 290 San Esteban Island, March 27 Rempel 291-294 South end of Tiburon Island, March 27 Rempel 295-298 San Pedro Nolasco Island, March 29 Rempel 300-307 Ensenado de San Francisco, Sonora, March 30 Rempel 311-315 Fraile Bay, Cape District, Baja California, April 4 Rempel 317-328 Cedros and San Benito Islands East side of Cedros Island, July 10 Rempel 330-348 San Benito Islands, July 14 Rempel 356-362 West San Benito Islands, July 15 Rempel 364-371 1938 Channel Islands Middle Island of Anacapa Group, August 1 Elmore 220-252 Becher Bay, Santa Rosa Island, August 2 Elmore 170-215 Santa Barbara Island, August 12 Elmore 295-310 San Miguel Island, Tylers Bight, August 3 and Point Bennett, September 12 Elmore 312-341 San Clemente Island, February 18, 19 Elmore 381-428 Santa Catalina Island, February 28 Elmore ^No. Sp. Col. 25 50 26 6 30 16 34 40 2 17 25 23 21 12 24 3 2 1 6 6 10 8 12 19 15 8 40 50 20 47 53 3 NO. 2 GENTRY : LAND PLANTS 15 TABLE 1— Continued LAND PLANTS COLLECTED BY THE VELERO III Collector and *No Sp. Locality and Date his numbers Col. 1939 San Cleraente Island, February 18, 19 Elmore 312-341 47 Santa Catalina Island, February 28 Elmore 430-439 15 1941 Santa Cruz Island: Between Pelican Bay and Prisoners Harbor, April 17 Elmore 254-293 45 Hill west of Prisoners Harbor, April 17 Elmore 440-468 30 1939 Mexico Cedros Island, "Cannery Bay," March 14 Elmore A1-A37 80 Revilla Gigedo Islands Sulphur Bay, Clarion Island, March 16 Elmore B1-B15 50 Braithwaite Bay, Socorro Island, March 18 Elmore Cl-Cll 40 Chacagua Bay, Oaxaca, March 21 Elmore D1-D25 130 Tenacatita Bay, Jalisco, May 8 Elmore 1A1-1A24 90 Magdalena Island, Tres Marias Group, May 9 Elmore 1B1-1B3 10 1939 Costa Rica Port Parker, Salinas Bay, March 24, 25 Elmore El-E-25 115 Golfo de Duke, March 26 Elmore F1-F28 150 Southwest island of Secas Group, March 27 Elmore G1-G2 10 1940 Gulf of California, Mexico Guaymas, Sonora, January 23 „ Dawson 1000-1009 45 Tiburon Island, January 25 Dawson 1010-1020 28 Puerto Refugio, Angel de la Guardia Island, January 26 Dawson 1021-1032 50 San Pedro Nolasco Island, February 6 Dawson 1033-1036 20 Pond Island, February Dawson 4 San Esteban Island, February Dawson 6 San Carlos Bay, Sonora, February 8 Dawson 1050-1073 90 Near Guaymas, Sonora, February 9 „ Dawson 1074-1084 60 Puerto Escondido, Baja California, February 11 Dawson 1085-1109 50 Punta Frailes, Baja California, February 16 Dawson 1111-1149 100 San Jose del Cabo, Baja California, February 17 Dawson 1150-1225 180 Table 1. The land plant collections made by members of the Allan Hancock Foundation Expeditions in the Eastern Pacific on the Velero III, 1937 to 1941. *No. Sp. Col.=number of specimens obtained, approximate. 16 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 In the following paragraph I have noted the distribution of each species generally, drawing upon my notes and travels, upon notes and records furnished me by Dr. Forrest Shreve of the recent Carnegie Desert Laboratory, upon manuscript copy of the "Flora of the Sonoran Desert" under preparation by Dr. Ira Wiggins of Stanford University and which he was generous enough to loan, upon information furnished by Dr. Dawson of the Allan Hancock Foundation regarding cactus dis- tributions, and upon information in publications. The type locality is given if it is known or specific enough to mean anything. Some taxonomic notes on the species or specimen at hand are also offered and in this I have often tried to clarify some of the important characters for specific recognition. The list, however, makes no attempt to be a descriptive flora. I have tried to add to, rather than merely duplicate, information already carried in earlier publications. The abbreviations used are those that are well established in taxo- nomic literature with but two exceptions. The first is C.N.H. for Con- tributions from the United States National Herbarium, rather than Contr. U. S. Nat. Herb., as has been frequently used. In employing this abbreviation I have followed Riley of the Royal Botanic Gardens at Kew (cf. Flora of Sinalao, Kew Bull. 1923-1924). Proposed here is the abbreviation Brge. for Townsend Seth Brande- gee, who did so much pioneer work on the Flora of Baja California, and K. Brge. for Katherine Brandegee, his coworker and wife. T. S. Brande- gee has been abbreviated in various ways, viz., T. S. Brandeg., Brandg., and Brand. The first is too long to function as an efficient abbreviation, two digits only having been stricken out, one of which is replaced by the period, leaving a net gain of only one digit. The third is easily confused with Brand. According to the recommendations in article 49 of the In- ternational Rules of Nomenclature, Brandegee should be contracted to Brande. However, as an abbreviation this is ineffective since it has elimi- nated only two digits of a long name. Since none of the above abbrevia- tions are short enough to carry the functional advantage of brevity, Brge. stands for T. S. Brandegee in the following pages. Trinomials have been briefed to the citation of the varietal author, or authors, only, and are not designated as to subspecies, variety, or form. With very few exceptions the authors have regarded them as varieties. The criteria used to differentiate subspecific entities is quite variably subjective, and particularly meaningless in dealing generally with a wild and incompletely known flora. Subspecific entities can be used effectively NO. 2 gentry: land plants 17 when genetic values are obtained. Hence, pending the genetic stages of taxonomic inquiry into the plants of the regions considered, the nomen- clature has been kept simple. Synonomy is given when required in proposing new names and also in a few cases of special significance. CHANNEL ISLANDS Introduction The southern California coast from Point Arguello southeastward describes a long shallow irregular curve. The outer margin of the con- tinental shelf, trending more nearly southward, accordingly broadens. Its submarine surface is irregular with submarine valleys and ridges; the topography suggesting land surface rather than sea bottom. The greatest heights of the shelf rise above the sea and form the Channel Islands, bearing the appellations cast upon them by the early Spaniards, San Miguel, Santa Rosa, Santa Cruz, Anacapa, Santa Catalina, Santa Barbara, San Nicolas, and San Clemente. They comprise a residual area of considerable antiquity, which, according to Reed (1933) and other geologists, dates from the Cretaceous. The ratio of sea to land area over the shelf has varied greatly, but during much of the Tertiary this shelf segment actually formed a large land body, known as Catalinia. In the early Miocene and again in the Pleistocene, if not also at other times, Catalinia appears to have been bridged to the continent. Structurally the islands are separated into two groups. The northern consists of San Miguel, Santa Rosa, Santa Cruz, and Anacapa, repre- senting disjunct segments of the Santa Monica Mountains, which ap- pear to have been more variably emerged and submerged in the sea than has the southern group. The southern group, consisting of Santa Bar- bara, Santa Catalina, San Nicolas, and San Clemente, have been part of a more consistent land area and are structurally tied to the San Pedro peninsula. The present islandic configuration is the result of geologically recent submergence of all but the higher elevations. The biota of such a land area can be expected to difler considerably from that of the near mainland and actually the evolution of the plants and animals is to be correlated with that of the land. Behind the present configuration of flora and fauna the elusive steps of evolution can be discerned. The climate of the Channel Islands is Mediterranean in type, semi- arid, and, of course, maritime. The average annual rainfall is around 12 inches, about 90% of which is precipitated in winter (November 18 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 through March). The dominating factor is the northwest winds, which pour around the islands for the majority of the days of the year. This increases aridity over what might be expected on the basis of rainfall and temperature (60°F. mean annual) only. However, transpiration and general aridity are tempered by high humidity and frequent fogs. The climate borders on a fog or maritime desert over the southern por- tion of San Clemente Island, the southern-most of the group. There is considerable variation in climate locally, both on individual islands and between different islands. This is explainable primarily in terms of the predominating direction of air flow. Physical conditions vary greatly whether a given area is on a windward or leeward side. This is expressed by the vegetation. Windward slopes commonly are covered with grass and forbs ( PI. 4, fig. 11), or stunted wind-tolerant shrubs, while protected slopes support shrub and tree communities. Some of the smaller islands that have been greatly disturbed by man have been nearly denuded of cover and top soil by the erosive action of wind. This is true of San Miguel, which catches the full force of the north- westers, and its sterile mobile sands are being discharged into the lee sea. The vegetation of the islands has been greatly altered by man. Both the aborigine, who burned and cut the native plants, and the modern European segregates, who burned, cut, cleared, and pastured, have left only remnants of the virgin cover. Many of the native plants are intoler- ant to grazing, and as they were destroyed, the aggressive weeds were introduced and have persistently spread. While considerable has been published about the flora of the islands, very little is on record regarding the vegetation. The only study of the plant ecology, that I have been able to find, which gives an adequate account of the vegetation is a doc- toral thesis by M. B. Dunkle (1944). Between 1939 and 1943 he made repeated visits to the islands and his introductory statement (1944:128- 129) is a concise general picture of the island vegetation. ^'Varied as are the different islands in topography and climate, they possess certain basic similarities. The western slope of all the islands, except San Nicolas which consists of barren sand dunes in this area, are covered with grasses, low forbs, suffrutescent perennials, and a few low wind-tolerant shrubs. The eastern slopes, except on the smaller islands, are quite generally covered with chaparral, or shrub and tree savannas. The canyons, which afford protection from the wind, usually have more or less shrubby growth on their slopes. This varies from an Opuntia lit- toralis association, through various facies of the coastal sage brush associa- tion, to chaparral on the larger islands. On these larger islands, and on NO. 2 GENTRY : LAND PLANTS 19 Anacapa, there may be occasional trees in the canyons, and where there is a continuous supply of running water there may be a riparian com- munity with trees or arborescent shrubs. The seaward bluffs have a varied growth of succulents, forbs, suffrutescent perennials, and occa- sional shrubs. Sand dune vegetation, very similar to that of mainland areas, is present on low dunes back of the few sandy beach areas. On the larger islands the protected north and northeast slopes support scattered groves of trees, while Santa Cruz has an extensive area of woodland." It is to be expected that such a residual area, representing a con- siderable area of long though interrupted isolation, would have many endemics in its flora. Of the approximately 950 species and varieties that have been catalogued from the ilsands, 80 of them are endemic. The original endemic element was undoubtedly reduced during times of the land bridges and migratory exchanges appear to have been made in both directions. Hence, if the apparent migrants from the islands to the main- land were included, the total island endemics would be about 100 species and varieties. Many of the endemics make up the unique plant com- munities known only on the islands, as the Pinus and the Lyonothamnus associations among the trees, the shrubby or suiifrutescent communities, as the Coreopsis-Artemisia association, the Atriplex-Hemizonia-Lotus- Astragalus community, the Echevaria-Eriogonum-0 puntia community, and the Eriogonum-Eriophyllum association forming a low tangle of suffrutescents. As on the mainland, the grasslands are dominantly com- posed of aggressive introductions and indicate little of the natural virgin climax. The most exhaustive flora published on the islands is that of Mills- paugh and Nuttall (1923) on Santa Catalina Island. Subsequent papers have stressed other islands, as Hoffman (1932) and one has recently catalogued the plants from all of the islands (Eastwood 1941). The fol- lowing annotated catalogue is based upon the collections secured by Mr. Francis H. Elmore on voyages of the Velero III of the Allan Hancock Foundation in 1938, 1939, and 1941, as outlined in Table 1. The com- plete itinerary of the Velero III voyages is given by Fraser (1943). On these and several other works as well, the author has drawn in formu- lating the collections. Philip A. Munz, an authority on southern Cali- fornia botany (1935), of the Rancho Santa Ana Botanical Garden, has read considerately and criticized the manuscript. No new plants are re- ported but the collections confirm and add to the distributional knowl- edge of plants in and about the Channel Islands. 20 allan hancock pacific expeditions vol. 13 Catalogue of Collections polypodiaceae Pellaea andromedae folia (Kaulf.) Fee, Gen. Fil. 129. 1850-52. San Clemente Island, February 18, 19, Elmore 383, 413. In maritime climate up to 4000 feet elevation from Oregon to northern Baja California, and known from the adjacent islands of Santa Catalina, Santa Rosa, and Santa Cruz. Selaginellaceae Selaginella Bigelovii Underw., Bull. Torr. Bot. Club 25:130. 1898. San Clemente Island, February 18, 19, Elmore 398. Southern half of California west of the deserts and adjacent islands of Santa Rosa, Santa Cruz, and Catalina. Probably also in northern Baja California. PiNACEAE PiNUS RADIATA Don, Trans. Linn. Soc. 17:441. 1836. Santa Cruz Island, Prisoners Harbor, August 2, Elmore 284. Maritime of central coastal California from Pescadero to Santa Cruz; type from Monterey. The Santa Cruz Island pines have recently been reviewed by Howell (Leafl.West. Bot. 3 :l-7. 1941 ), who recognized only P. remorata Mason in varying forms. Of the authors of California floras, Jepson and McMinn & Maino attribute P. radiata to Santa Cruz Island, while Abrams and Munz do not, excepting the doubtful status of P. radiata binnata (Engelm.) Lemmon. Hence the Elmore collection was considered criti- cally. The 3-needled leaf fascicles and the light brown, slightly assj^me- tric, open cone specimens appear to belong definitely to P. radiata. Regarding his Santa Cruz Island pine collection, Mr. Elmore has re- cently written, "I remember well collecting the pine specimens, but as the trees were in a regular grove and near a fence line, I remember thinking at the time that they were probably exotics, having been planted there." (Letter dated February 2, 1948). PiNUS REMORATA Mason, IMadrono 2:8-10. 1930. Santa Cruz Island, Prisoners Harbor, August 2, Elmore 290. Known only from Santa Cruz and Santa Rosa Islands. no. 2 gentry : land plants 21 Gramineae DiSTiCHLis SPICATA (L.) Greene, Bull. Calif. Acad. Sci. 2:415. 1887. Santa Cruz Island, Prisoners Harbor, September 14, Elmore 289. On both coasts of North America through the temperate and sub- tropical regions in saline soils and marshes ; type from the north Atlantic coast. Also known on Catalina, San Nicolas, San Miguel, and Santa Rosa Islands. Elymus triticoides Buckl., Proc. Acad. Phil. 1862:99. San Miguel Island, Tyler Bight, August 3, Elmore 317. Widely distributed in the western United States; type from the Rocky Mountains. Known from San Miguel, Santa Cruz, Santa Rosa, and Santa Catalina Islands. Festuca megalura Nutt., Jour. Acad. Phil. II, 1:188. 1848. Santa Cruz, hill west of Prisoners Harbor, April 17, Elmore 444. Mostly maritime from British Columbia and Idaho south to Baja California; type locality, Santa Barbara, California. On the islands it is known from Santa Cruz, Santa Rosa, San Miguel, and Catalina. Festuca pacifica Piper, C.N.H. 10:12. 1906. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 454. Maritime from British Columbia to Baja California and the Chan- nel Islands of San Miguel (Hoffman ??) and San Nicolas. Monanthocloe littoralis Engelm., Trans. Acad. St. Louis 1 :437. 1859. San Miguel, Tyler Bight, August 3, Elmore 318. Littoral in salt marshes, tidal flats, and strands throughout tropical America and north on the Pacific coast to Santa Barbara; type from Texas. The only other records from the Channel Islands are from Catalina. Polypogon monspeliensis (L.) Desf., Fl. Atlant. 1:67. 1798. Santa Rosa, Becher Bay, August 3, Elmore 193. Santa Cruz, Pris- oners Harbor Canyon, September 14, Elmore 267 . Introduced from Europe, common now from Alaska to Mexico along the coast. Also on Santa Catalina, San Miguel, San Nicolas, and Santa Cruz Islands. 22 allan hancock pacific expeditions vol. 13 Cyperaceae SciRPUS CALiFORNicus (Meyer) Brit., Trans. N. Y. Acad. 2:80. 1892. Santa Cruz Island, Prisoners Harbor, September 14, Elmore 293. From California to Florida across southern United States and south through tropical America to South America; type from California. It is not known from any other of the Channel Islands. LiLIACEAE Brodiaea capitata Benth., PI. Hartvv. 339. 1857. San Clemente Island, February 18, 19, Ebnore 391. Common through the coast ranges from southern Oregon to northern Baja California. On all the Channel Islands. Fagaceae QuERCUS AGRiFOLiA Nee, Anal. Cien. Nat. 3:271. 1801. Santa Cruz, Prisoners Harbor, September 14, Elmore 273. Maritime climate of the coast ranges from Mendocino County, Cali- fornia south to San Pedro Martir Mountains of northern Baja Cali- fornia. Also known from Santa Rosa Island. QuERCUs DUMOSA Nutt., Silva 1:7. 1842. Santa Cruz, Prisoners Harbor Canyon, September 14, Ebnore 283. Mostly in the chaparral of the Lower Sonoran Life Zone from northern California south into Baja California; type from Santa Bar- bara. Also on Santa Rosa and Catalina Islands. Urticaceae Urtica holosericea Nutt., Jour. Acad. Phil. II, 1 :183. 1847. Santa Cruz, Prisoners Harbor Canyon, August 2, Elmore 258, dry river bed. From Washington and Idaho south to Baja California; type locality, Monterey, California. Also on Santa Catalina Island. POLYGONACEAE Eriogonum arborescens Greene, Bull. Calif. Acad. Sci. 1:11. 1884. Middle island of Anacapa group, August 1, Elmore 234, steep rocky slope. Canyon walls and steep slopes on the California Islands of Santa Rosa, Santa Cruz, and Anacapa; type from Santa Cruz Island. Eriogonum giganteum Wats., Proc. Am. Acad. Sci. 20:371. 1885. Santa Barbara Island, August 12, Elmore 306, dry hillside. NO. 2 GENTRY : LAND PLANTS 23 In canyons and on bluffs on the islands of Santa Cruz, San Clemente, and Santa Catalfna; type from the latter. Elmore's collection from Santa Barbara is apparently the first citation for that island. Eriogonum GRANDE Greene, Pittonia 1 :38. 1887. Middle island of Anacapa group, August 1, Elmore 232, steep rocky slope. Southern California and the Channel Islands; the type from Santa Cruz. It appears to be lacking only on San Nicolas Island. Eriogonum rubescens Greene, Pittonia 1:39. 1887. San Miguel Island, Tyler Bight, August 3, Elmore 321. Point Ben- nett, September 12, Elmore 334. San Miguel Island, August 10, Elmore 325. Known only from the Channel Islands of Santa Cruz, Santa Rosa, Santa Catalina, San Clemente, and San Miguel. This plant is closely related to E. grande Greene, to which it has been referred as a variety by some authors. Numbers 325 and 334 are atypical in the densely flocculose tomentum of the involucres and the compact cymose head-like inflorescences. They may represent a hybrid form, a condition also indicated by the tendency of the mature involucres to double their number of teeth. The interesting genus Eriogonum shows considerable specific and subspecific variation and in this it appears to be expressive of environ- mental differences. Hence, since environments have varied in the geolo- gic past, the present segregates of Eriogonum would be reactive products. A cytogenetic study of the genus might find physiographic correlatives in insular evolution. Who w^ill make it? POLYOGONUM AVICULARE L., Sp. PL 362. 1753. Santa Rosa Island, Becher Bay, August 2, Ebnore 213. A cosmopolitan weed native of Eurasia; on Santa Rosa and Santa Catalina Islands also. Chenopodiaceae Atriplex leucophylla (Moq.) Dietr. in DC, Prodr. 132:109. 1849. Middle island of Anacapa group, August 1, Elmore 237, steep rocky slope. A littoral halophyte from Plumboldt Bay, California to Vizcaino Bay, Baja California; type from California. 24 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Salicornia subtermixalis Parish, Erythea 6:87. 11 San Clemente, February 18, 19, Elmore 402. Salt marshes along the coast from San Francisco Bay, California to Sinaloa, Mexico; occasional in saline soils of the interior valle3^s; type from the San Jacinto River, California. On the Channel Islands of San Miguel, Santa Rosa, Santa Cruz, Santa Catalina, and San Clemente. Also collected at Avila, Port San Luis, near San Luis Obispo, August 4, Ehnore 402. SuAEDA CALiFORNiCA Wats., Proc. Am. Acad. Sci. 9:89. 1874. San Clemente Island, February 18, 19, Ehnore 403. Salt marshes from San Francisco, California to northern Baja Cali- fornia; type from San Francisco Bay. On Santa Cruz, Santa Catalina, Anacapa, and San Nicolas Islands. SuAEDA TAXiFOLiA Standi., N. Am. Fl. 21 :91. 1916. Santa Barbara Island, August 12, Elmore 302, dry hillsides. Salt marshes along the coast from Santa Barbara County to Los Angeles County; t3^pe from Playa del Rey. Some of the collections reported from the islands as S. californica may rightly belong to this species. It has not previously been listed from the islands. Elmore's collection compares favorably with mainland ma- terial reviewed by the author. Chenopodium murale L., Sp. PI. 219. 1753. Santa Rosa Island, Becher Ba^^, August 2, Elmore 178. A cosmopolitan weed naturalized from Europe. On all of the Chan- nel Islands. Nyctaginaceae Abronia alba Eastw., Proc. Calif. Acad. Sci. Ill, 1 :97. 1898. San Miguel Island, Tyler Bight, August 2, Ehnore 232. Point Ben- nett, September 12, Elmore 339. San Miguel Island, August 10, Elmore 330a. Insular. In addition to San jMiguel Island it is also known from San Nicolas, Santa Rosa, and San Clemente Islands. Abronia maritima Nutt. ex Wats., Bot. Calif. 2:4. 1880. Middle island of Anacapa group, August 1, Elmore 243. San Cle- mente Island, February 18, 19, Elmore 381. San Miguel Island, Point Bennett, September 12, Ehnore 340. Sandy sea strands from San Luis Obispo County south to Baja Cali- fornia and Sinaloa, IVIexico; type from San Pedro, California. NO. 2 GENTRY : LAND PLANTS 25 Hesperoxla laevis (Benth.) Standi., C.N.H. 12:363. 1909. San Clemente Island, February 18, 19, Elmore 41 5 y 328. Coast ranges and adjacent islands from Monterey County south to central Baja California; type from Magdalena Bay, Baja California. Aizoaceae Mesembryanthemum chilense Molina, Sagg. Chile ed. 2:133. 1810. San Miguel, Tyler Bight, August 3, Elmore 316. Coastal, common on bluffs and sandy soils from Oregon to northern Baja California; type from Chile. Also reported from Australia and Tasmania. Mesembryaxthemum nodifloruisi L., Sp. PL 480. 1753. San Clemente Island, February 18, 19, Elmore 406. Santa Barbara Island, August 12, Elmore 310. Native of South Africa and the Alediterranean region; introduced and now common along the shores of southern California and northern Baja California. It has been collected on all the Channel Islands except Santa Cruz. This small mat plant is the least conspicuous and showy of the adventive Mesembryanthemu?n. PORTULACACEAE Montia perfoliata (Donn.) Howell, Erythea 1:38. 1893. San Clemente Island, February 18, 19, Elmore 418. A hydrophytic shade-tolerant herb widely distributed through western North America north of Mexico. Caryophyllaceae SiLENE GALLICA L., Sp. PI. 417. 1753. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 456. A common weed naturalized from Europe; known on all the Chan- nel Islands. SiLENE laciniata Cav., Ic. PI. 6:44. 1801. Santa Rosa Island, Becher Bay, August 2, Elmore 206. Middle island of Anacapa group, August 1, Elmore 239. Widely distributed in the mountains of western North America from central California and western Texas south to southern Mexico; type from Mexico. In addition to the above cited islands, it is also known on Santa Cruz and San Miguel Islands. 26 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Spergularia macrotheca (Hornem.) Heynh., Nomen. 2:689. 1840. Santa Rosa Island, Becher Bay, August 2, Elmore 189. Saline soils near the coast from Washington to Baja California; type from California. Papaveraceae Eschscholtzia californica Cham, var., in Nees Hor. Phys. Ber. 73. 1920. Santa Rosa, Becher Bay, August 2, Elmore 187, on dry hillsides. The material is not sufficient for certain identification. It may be the same as the collection annotated by Hoffman (Bull. So. Calif. Acad. Sci. 31:102. 1932) as Eschscholtzia sp. He stated that it is an annual with clear yellow flowers on Santa Cruz Island. On the basis of fruit and re- ceptacle, however, I have no hesitancy in assigning it to the E. californica complex, which is so well known for its variability. It differs from E. californica mariti?na, the common variety of the islands, in the non- glaucous and longer-lobed leaves, and the reduced stature. Eschscholtzia californica maritima Jep., Man. Fl. PI. Calif. 402. 1925. San Miguel Island, August 10, Elmore 329. Tyler Bight, August 3, Elmore 320. Point Bennett, September 12, Elmore 337. Endemic to the Channel Islands where it is known from Santa Cruz, Santa Rosa, and San Miguel, the latter island, according to Hoffman, having the most widely dispersed population. Platystemon californicus Benth., Trans. Hort. Soc. Lond. II, 1:405. 1835. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 458. Widely distributed in western United States; also in northern Baja California. Cruciferae Lepidium lasiocarpum Nutt. in Torn & Gray, N. Am. Fl. 1 :115. 1838. San Clemente Island, February 18, 19, Elmore 401. Widely distributed in the southwestern United States and northern Mexico; type from near Santa Barbara, California. It has been collected on nearly all of the other Channel Islands. The above cited collection consists of several depauperate specimens, 6-8 cm high, fruiting, and are the first taken from San Clemente. NO. 2 GENTRY : LAND PLANTS 27 Caulanthus inflatus Wats., Proc. Am. Acad. Sci. 17:364. 1882. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 453. The known range of this plant is from the Mojave Desert in Cali- fornia and Nevada and thence westward to the more arid localities in the San Joaquin Valley and Monterey County; type from the Mojave Desert, California. It has never before been reported from the Channel Islands, and in so far as it is a desert plant, Elmore's collection from the maritime habitat is open to question. There was no accompanying note of locality in the field sheet, as there was in the majority of them, but it was in with the sheets in the pacquet of plants marked from Santa Cruz Island. Crassulaceae Dudleya Greenei Rose, Bull. N. Y. Bot. Gard. 3:17. 1903. Santa Rosa Island, Becher Bay, August 2, Elmore 180. Rocks and cliffs near the sea in southern California and on Santa Cruz, Santa Rosa, and San Miguel Islands. Dudleya farinosa (Lindl.) Brit. & Rose, Bull. N. Y. Bot. Gard. 3:27. 1903. San Miguel Island, Tyler Bight, August 3, Elmore 314, dry hill- side. Along the coast of northern and central California. The above cited collection is the first known record of the species in the Channel Islands. It agrees well with mainland material. Crossosomataceae Crossosoma californicum Nutt., Jour. Acad. Phil. II, 1 :150. 1847. Santa Catalina Island, February 29, Elmore 437. Southern California, Baja California and adjacent islands; type from Santa Catalina Island. Also on San Clemente. Rosaceae Lyonothamnus floribundus asplenifolius (Greene) Brge., Zoe 1:136. 1890. Santa Cruz Island, ranch yard in Central Valley, August 2, Elmore 255. Endemic to the islands, Santa Cruz, Santa Rosa, and San Clemente. Photinia arbutifolia (Ait.) Lindl., Trans. Linn. Soc. 13:103.,.,^^^^ 182L /1n^^^^^^^ 28 ALLAN' HAXCOCK PACIFIC EXPEDITIONS VOL. 13 Santa Cruz, Prisoners Harbor Canyon, September 1-r, Elmore 269. Santa Rosa Island, Becher Bay, August 2, Elmore 210. Widely distributed in the Upper Sonoran Zone of California; type from ^lonterey, California. Known also from the islands San ^Miguel, Santa Catalina, San Nicolas, and San Clemente. Elmore 269 from Santa Cruz Island has small, subentire, proximate leaves on short diffuse twigs making a compact crown of foliage indicative of an arid situation. Photixia arbutifolia macrocarpa IMunz, Bull. So. Calif. Acad. Sci. 31:64. 1932. San Clemente Island, Februan,- 18, 19, Elmore 421. Known previously only from Santa Catalina Island. Pruxus Lyoxi (Eastw.) Sarg., PL Wilson. 74. 1911. Santa Cruz Island, Prisoners Harbor Canyon, September 14, Elmore 27 L rocky hillside. Endemic to the islands, Santa Cruz. Santa Rosa, Santa Catalina, and San Clemente; the type from Santa Catalina. Rosa gr-ATISSIMA Greene, Fl. Fran. 73. 1891. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 447. Central Sierra Nevada to southern California; type from the moun- tains of Kern County. Not previously credited to the Channel Islands, but the few weak straight spines, the spatulate-tipped sepals, and the nature of the stipules appear to relate the above-cited collection with the R. gratiss'nna complex rather than with R. calif ornica. Legumixosae AsTR-AGALUS leucopsis Torr. & Gray, N. Am. Fl. 1 :344. 1838. Middle island of Anacapa group, August 1, Elmore 252, steep rocky slopes. San ^liguel Island, Tyler Bight, August 3, Elmore 319. Upper and Lower Sonoran Life Zones from Santa Barbara County south into northern Baja California; type from Santa Barbara, Cali- fornia. Also reported from Santa Barbara and Santa Catalina Islands. AsTR.\GALUs Nevixii Wats., Proc. Am. Acad. Sci. 21:412. 1886. San Clemente Island, February- 18, 19, Elmore 395. Known only from the Channel Islands of Santa Catalina, Santa Barbara, Anacapa, and the t}-pe localit}', San Clemente. Lotus DENDROIDEUS (Greene) Greene, PIttonIa 2 :148. 1890. Santa Rosa Island, August 2, Elmore 196, dry canyon wall. Endemic to the Channel Islands of Santa Catalina, Santa Cruz, Santa Rosa, and Anacapa; t}'pe from Santa Cruz. NO. 2 GENTRY : LAND PLANTS 29 Lotus niveus (Greene) Greene, Pittonia 2:148. 1890. San Clemente Island, February 18, 19, Ebnore 409. Known only from Santa Cruz and San Clemente Islands. Referred here is Elmore 410, a sterile shrubby perennial 2-4 dm high with silvery sericeus twigs, spikoid foliage, also densely silvery tomentose, with 3 ovate-lanceolate leaflets 10-15 mm long. Lotus ornithopus Greene, Bull. Calif. Acad. Sci. 1:185. 1885. San Clemente Island, February 18, 19, Elmore 392. Southern California, northern Baja California and adjacent islands; type from Guadelupe Island off Baja California. Lupinus sparsiflorus Benth., PI. Hartw. 303. 1848. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 446. Widely scattered in southern California and in adjacent Baja Cali- fornia. Trifolium microcephalum Pursh, Fl. Am. Sept. 2:478. 1814. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 441. Open grassy slopes from British Columbia to Baja California and east to Nevada; type from Bitter Root River, Montana. On San Miguel, Santa Rosa, Santa Catalina, and San Clemente Islands. Trifolium tridentatum aciculare (Nutt.) McDermott, N. Am. Trifol. 26. 1910. Santa Cruz Island, rock slide between Pelican Bay and Prisoners Harbor, April 17, Elmore 465. Central valley of California, cismontane southern California, and adjacent islands; type from Santa Barbara. Also on Santa Rosa and San Clemente Islands. Oxalidaceae OxALis CERNUA Thunb., Diss. Oxal. 14. 1781. San Clemente Island, February 18, 19, Elmore 423. Naturalized from southern Africa and now common in cismontane southern California ; on the Channel Islands it has previously been listed from Santa Catalina only. 30 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 EUPHORBIACEAE Eremocarpus setigerus (Hook.) Benth., Bot. Voy. Sulph. 53. 1844. Santa Cruz Island, Prisoners Harbor Canyon, August 2, Elmore 259, dr>^ river bed. Common in wasteland and fallow fields from Washington to south- ern cismontane California. Known also on Santa Rosa, Santa Catalina, and San Clemente Islands. Anacardiaceae Rhus integrifolia (Nutt.) Benth. & Hook, ex Wats, in Wheeler, Rep. U.S. 100th :\Ierid. 6:84. 1878. Santa Catalina Island, February 29, Elmore 439. San Clemente, February 18, 19, Elmore 394. Santa Cruz Island, Prisoners Harbor Canyon, September 14, Elmore 275. Anacapa Island, August 1, Elmore 224'. Maritime southern California and northern Baja California and adjacent islands from Santa Barbara to Cedros Island; type from San Diego, California. Known from all the Channel Islands except San Nicolas. Toxicodendron diversilobium (Torr. & Gray) Greene, Leafl. 1:119. 1905. Santa Cruz Island, Prisoners Harbor Canyon, September 14, Elmore 265, dry rocky hillside. Common along the Pacific coast below 4000 feet from Washington through California and south in the mountains of northern ^lexico to Sinaloa. Known also from San ^Miguel, Santa Rosa, and Santa Catalina Islands. Malvaceae Lavatera assurgentiflora KelL, Proc. Calif. Acad. Sci. 1:14. 1854. Santa Catalina Island, Indian Rock in Emerald Cove, February 29, Elmore 435. Santa Rosa, Becher Bay, August 2, Elmore 191, dry hill- side. Native of the Channel Islands, apparently originally lacking only on San Nicolas, escaped and cultivated on the adjacent mainland. Sidalcea malvaeflora (DC.) Gray ex Benth., PI. Hartw. 300. 1848. Santa Rosa Island, Becher Bay, August 2, Elmore ISl, wall of wet ravine. Cismontane southern California to northern California and on the islands of Santa Rosa, San Miguel, Santa Cruz, and Catalina. no. 2 gentry : land plants 31 Frankexiaceae Frankenia graxdifolia C. & S., Lfnnaea 1:35. 1826. Middle island of Anacapa group, August 1, Ebnore 250. Along the coast from central California south to northern Baja Cali- fornia and on the adjacent islands of Anacapa, San Miguel, Santa Rosa, Santa Cruz, and Santa Catalina. Cactaceae Opuxtia littoralis (Engelm.) Cockr., Bull. So. Calif. Acad. Sci. 4:15. 1905. San Clemente Island, February 18, 19, Elmore 382. Santa Cruz, Prisoners Harbor Canyon, September 14, Elmore 279, rocky hillside. Middle island of Anacapa group, August 1, Elmore 245, steep rocky slope. Along the coast from Santa Barbara to northern Baja California and on the adjacent islands; the exact type locality is not known. East- wood (1941:67) also reports it from Santa Barbara, San Nicolas, and Santa Rosa Islands. The above series of specimens, particularly Elmore 382 from San Clemente, are atypical of the mainland plants in the more orbicular pads (rather than ovate) and in the straightness of their spines. OXAGRACEAE Oenothera chieranthifolia Hornem. ex Spreng. Syst. 2.228 San Miguel Island, August 10, Elmore 331. Tyler Bight, August 3, Elmore 313. Point Bennett, September 12, Elmore 335. Santa Rosa Island, Becher Bay, August 2, Elmore 209, sandy hill slope. Sea beaches from Oregon to southern California and adjacent islands. Not listed from San Clemente and Anacapa, but apparently common on all the others. Zauschxerla. californica villosa Jeps., Man. Fl. PL Calif. 667. 1925. Santa Rosa Island, Becher Bay, August 2, Elmore 183, dry hillside. Stated by Jepson to be in southern California. It is known from the islands of Santa Cruz, San Clemente, and Santa Catalina, and may originally have been an island endemic. Umbelliferae FoEXicuLUM VULGARE (L.) Gaertn., Fr. Sem. 1:105. 1788. Santa Cruz Island, Prisoners Harbor Canyon, August 2, Elmore 254, dry river bed. A naturalized weed from Europe and now widely dispersed in west- ern North America and in South America. It is also listed from Santa Rosa and Santa Catalina Islands. 32 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Sanicula arguta Greene ex Coult. & Rose, C.N.H. 7:36. 1900. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Ehnore 462. Southern California and the adjacent islands of San Nicolas, Santa Catalina, San Clemente, and Santa Cruz. Primulaceae Anagallis arvensis L., Sp. PL 148. 1753. Santa Rosa Island, Becher Bay, August 2, Elmore 208, grassy hill- side. Weed naturalized from Europe and widely scattered in western North America. Known on the Channel Islands of San Miguel, Santa Cruz, Santa Catalina, and Santa Rosa. Gentianaceae Centaurium venustum (Gray) Robs., Proc. Am. Acad. Sci. 45: 397. 1910. Santa Rosa Island, Becher Bay, August 2, Ehnore 204, grassy hill- slope. From northern Baja California to Butte County, California where it is common along the coast and rarely in the desert. Known also from Santa Cruz and Santa Catalina Islands. Elmore's collection appears to be the first from Santa Rosa Island. CONVOLVULACEAE Convolvulus occidentalis cyclostegius (House) Jeps., Man. Fl. PI. Calif. 776. 1925. Santa Barbara Island, August 12, Elmore 313, climbing on shrubs in dry canyon. San Clemente Island, February 18, 19, Elmore 396. Santa Cruz Island, Prisoners Harbor, August 2, Elmore 282. Coastal from San Francisco Bay to southern California. Its insular distribution has previously been reported only on Santa Catalina Island. POLEMONIACEAE GiLiA AFF. MULTiCAULis Bcnth., Bot. Reg. 19: t. 1622. 1833. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 440. Common in cismontane southern California and reported also from Catalina. Boraginaceae Amsinckia intermedia Fisch. & Mey., Ind. Sem. Hort. Petrop. 2:26. 1836. Santa Rosa Island, Becher Bay, August 2, Elmore 212, dry hillside. Widely scattered in California, northern Baja California, Arizona, and on the Channel Islands. NO. 2 GENTRY : LAND PLANTS 33 Cryptantha Clevelandii Greene, Pittonia 1:117. 1887. San Miguel Island, Tyler Bight, August 3, Elmore 312. Maritime, from northern Baja California to Santa Barbara, Cali- fornia and the Channel Islands of Santa Catalina, Santa Cruz, and San Miguel ; type from the hills above San Diego, California. Heliotropium curassavicum L., Sp. PI. 130. 1753. Middle island of Anacapa group, August 1, Ehnore 222^ steep rocky slope. Santa Cruz Island, Prisoners Harbor Canyon, September 14, El- more 277 y rocky edge of salty pool. San Miguel Island, August 10, El- more 328. Widely dispersed in saline lowland soils of tropical and subtropical America. Pectocarya linearis ferocula Jtn., Contr. Arn. Arb. 3 :95. 1932. Santa Cruz Island, hills west of Prisoners Harbor, April 17, Elmore 463. Cismontane southern California from Ventura south into Baja Cali- fornia and the adjacent islands. Verbenaceae Verbena robusta Greene, Pittonia 3:309. 1898. Santa Cruz Island, Prisoners Harbor, September 14, Elmore 291, along edge of swamp. San Diego County and the Channel Islands of San Miguel, Santa Rosa, Santa Cruz, Santa Catalina, and San Clemente. Apparently origi- nated on Catalinia and migrated to the mainland. Labiatae Marrubium vulgare L., Sp. PI. 583. 1753. San Clemente Island, February 18, 19, Elmore 417. Weed naturalized from Europe, widely scattered across North America and on all the Channel Islands except San Nicolas and Anacapa. Salvia Brandegei Munz, Bull. So. Calif. Acad. Sci. 31:69. 1932. Santa Rosa Island, Becher Bay, August 2, Elmore 194, dry canyon wall. Endemic to Santa Rosa Island. Solanaceae Lycium californicum Nutt. in Gray, Bot. Calif. 1 :542. 1876. Santa Barbara Island, February 12, Elmore 295, forming thickets in dry fields. San Clemente Island, February 18, 19, Elmore 397. 34 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Cismontane southern California and south along the coast to central Baja California, type from San Diego. In addition to the above listed Channel Islands it is also reported from Santa Catalina. Dunkle (1944) reports it as characteristic of the grassland on Santa Barbara Island, where it forms a biome with Suaeda and the gull Larus. Petunia parviflora Juss., Ann. Mus. Paris 2:216, t. 47. 1803. Santa Rosa Island, Becher Bay, August 2, Elmore 170, sandy hills. Widely distributed in the moist sandy soils of southern United States and tropical America. On the Channel Islands it is known only from Santa Rosa. SoLANUM Clokeyi Munz, Bull. So. Calif. Acad. Sci. 31:69. 1932. Santa Cruz Island, rock slide between Pelican Bay and Prisoners Harbor, April 17, Elmore 467. Endemic to Santa Cruz Island. SoLANUM DouGLASii Dunal, in DC, Prodr. 13, 1 :48. 1852. Santa Catalina Island, February 28, Ehnore 438. Cismontane southern California and the Channel Islands of San Miguel, Santa Rosa, Santa Cruz, Santa Catalina, and San Clemente. SoLANUM viLLOSUM ( L. ) Mill., Gard. Diet. ed. 8, n. 2. Santa Rosa Island, Becher Bay, August 2, Elmore 200, dry canyon walls. San Clemente Island, February 18, 19, Elmore 420. A European weed adventive in southern California. Not previously reported from the Channel Islands. Scrophulariaceae Castilleja anacapensis M. B. Dunkle, Bull. So. Calif. Acad. Sci. 41 :135. 1942. Middle island of Anacapa group, August 1, Elmore 230, steep rocky slope. Endemic to Anacapa Island. Castilleja hololeuca Greene, Pittonia 1 :38. 1887. Middle island of Anacapa group, August 1, Elmore 226, steep rocky slope. Endemic to the islands of San Miguel, Santa Rosa, Santa Cruz, and Anacapa. NO. 2 GENTRY : LAND PLANTS 35 Castilleja LATIF0LL4 Hook. & Am., Bot. Beech. Voy. 154. 1839- 40. Santa Rosa Island, Becher Bay, August 2, Elmore 172, 174, sandy hills. From Monterey County to northern California and on the islands of Santa Rosa and Santa Cruz. The flowers are reported as being normally red, but No. 174 was noted by the collector as having yellow flowers. Castilleja mollis Penn., Proc. Acad. Nat. Sci. Phil. 99:185. 1947. San Miguel Island, August 10, Elmore 333. Point Bennett, Septem- ber 12, Elmore 341. Castilleja sp. San Clemente Island, February 18, 19, Elmore 384, 411. The material is too young for certain determination. It is a low shrubby or suffrutescent plant with sordid pubescence, linear attenuate bracts and leaves, both of which are irregularly lobed. DiPLACUs LOXGiFLORUS Nutt. in Taylor's Ann. Nat. Hist. 1, 1 :139. 1838. Santa Cruz Island, Prisoners Harbor, August 2, Elmore 261, rocky hillside. Common on the cismontane and chaparral slopes of southern Cali- fornia and on Santa Rosa, Santa Cruz, and Catalina Islands. Also in northern Baja California. DiPLACus PARVIFLORUS Greene, Pittonia, 1:36. 1887. Santa Rosa Island, Becher Bay, August 2, Elmore 185, dry hillside. Santa Cruz Island, rock slide between Pelican Bay and Prisoners Har- bor, April 17, Elmore 464. Apparently limited to Santa Cruz and Santa Rosa Islands. MiMULUS GUTTATUS DEPAUPERATUS (Gray) Grant, Ann. Mo. Bot. Gard. 11:170. 1924. Santa Cruz Island, Prisoners Harbor Canyon, August 2, Elmore 257, dry stream bed. Widely dispersed in the western United States. This variety has not previously been reported from the islands and the fragmentary^ material is doubtfully referred here. 36 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 LiNARiA CANADENSIS TEXANA (Schcele) Pcnn., Proc. Acad. Phil. 73:502. 1922. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Ehnore 461. Widely distributed in both North and South America. On the Chan- nel Islands it is known from San Miguel, Santa Rosa, Santa Catalina, and Santa Cruz. Orthocarpus purpurascens Benth., Scroph. Ind. Introd. 13. 1835. Santa Cruz Island, hill west of Prisoners Harbor, April 17, Elmore 459, 460. Common along the coast from central California to northern Baja California and the adjacent islands of San Nicolas, San Miguel, Santa Rosa, Santa Cruz, and Santa Catalina. CUCURBITACEAE Marah fabacea (Naud.) Greene, Pittonia 2:129. 1890. San Clemente Island, February 18, 19, Elmore 422, 386. The Channel Islands and mainland from Monterey north to Sonoma County, California. COMPOSITAE Achillea millefolium lanulosa (Nutt.) Piper, Mazama 2:97. 1901. Santa Barbara Island, August 12, Elmore 308, dry hillside. From the mountains to the coast in southern California and northern Baja California and on all the Channel Islands. Artemesia californica Less., Linnaea 6:523. 1831. Middle island of Anacapa group, August 1, Elmore 228, steep rocky slope. Common and widespread on slopes and mesas of the Upper Sonoran from central California to northern Baja California and the adjacent islands. It is one of the strongly successful elements in and about chapar- ral. With a stable population it is quickly adventive on new habitats, and was doubtless migratory on the Tertiary land bridges. Baccharis Douglasii DC, Prodr. 5:400. 1836. Santa Cruz Island, Prisoners Harbor Canyon, September 14, Elmore 281, dry stream bed. NO. 2 gentry: land plants 37 Mostly coastal from San Francisco south to northern Baja Cali- fornia and the adjacent islands of Santa Rosa, Santa Cruz, and Santa Catalina. It is tolerant of new immature soils and apparently has a strong potential as a migrant in new areas of Mediterranean climate type. Baccharis pilularis consanguinea (DC.) O. Kuntze, Rev. Gen. PI. 1:319. 1891. Prisoners Harbor Canyon, Santa Cruz Island, September 14, Elmore 285, dry rocky hillside. Coastal from Oregon to southern California and the adjacent islands of Santa Rosa, Santa Cruz, and Catalina. Baccharis Plummerae Gray, Proc. Am. Acad. Sci. 15:48. 1880. Prisoners Harbor Canyon, September 14, Elmore 287, dry rocky hillside. Santa Cruz Island and the adjacent mainland. Baeria chrysostoma gracilis Hall, U. C. Publ. Bot. 3:170. 1907. Hill west of Prisoners Harbor, Santa Cruz Island, April 17, Ehnore 468. San Clemente Island, February 18, 19, Elmore 400. Common on Pacific slopes from Oregon south to northern Baja Cali- fornia and listed (by Eastwood 1941 :75 under Baeria Palmeri Clemen- tina) from San Nicolas, San Miguel, Santa Rosa, Santa Cruz, and San Clemente Islands. A quick annual in a genus developmentally responsive to the semi-arid climates of both maritime and continental types. Coreopsis gigantea (Kell.) Hall, U. C. Publ. Bot. 3:142. 1907. Santa Barbara Island, August 21, Elmore 297, dry canyon walls and rocky hillside. Catalina Island, February 28, Elmore 436. Coastal bluffs from San Luis Obispo County south to the Santa Monica Mountains and on all the Channel Islands. Apparently evolved on Catalinia and recently migratory to the mainland. Corethrogyne filaginifolia robusta Greene, Pittonia 1 :89. 1887. Middle island of Anacapa group, August 1, Elmore 220, steep rocky slope. The variety is endemic to the islands San Miguel, Santa Rosa, Santa Cruz; not previously known from Anacapa. 38 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Encelia californica Nutt., Trans. Am. Phil. Soc. II, 7 :357. 1841. San Clemente Island, February 18, 19, Elmore 393. Common on the semi-arid coastal slopes from Santa Barbara, Cali- fornia south into northern Baja California and on the adjacent islands of Santa Cruz, Santa Catalina, and San Clemente. An adaptive species of maritime climate which has responsively evolved into varietal popu- lations. Erigeron foliosus Nutt., Trans. Am. Phil. Soc. II, 7:309. 1841. Becher Bay, Santa Rosa Island, August 2, Elmore 177, dry hillside. Mainly coastal from Humboldt County, California south into south- ern California. The tj^pical form, of the species has not been reported from the islands. However, the variety, stenophyllus (Nutt.) Gray, is reported by Eastwood (Leafl. W. Bot. 3:74. 1941) for San Miguel, Santa Rosa, and Santa Cruz Islands. Elmore's collections lack the fili- form leaves of the variety, being up to 5 mm wide. Erigeron glaucus Ken, Bot. Reg. 1: pi. 10. 1815. Point Bennett, San Miguel Island, September 12, Elmore 336. Tyler Bight, San Miguel Island, August 3, Ehnore 315, dry sandy hill slope. San Miguel Island, August 10, Elmore 332. Common along the shores from Oregon south to Monterey County and on the Channel Islands of San Miguel, Santa Rosa, and Santa Cruz. Eriophyllum Nevinii Gra}^, Syn. Fl. I, 2:452. 1886. Santa Barbara Island, August 12, Elmore 300, dry hillside. Apparently endemic to the islands Santa Catalina, San Clemente, and Santa Barbara; not previously reported from the latter. Gnaphalium beneolens Davidson, Bull. So. Calif. Acad. Sci. 17:17. 1918. Becher Bay, Santa Rosa Island, August 2, Elmore 215, dry hillside. Southern California east to Texas and northern Mexico. Gnaphalium bicolor Bioletti, Erythea 1:16. 1893. San Clemente Island, February 18, 19, Elmore 407. Coastal and inland valleys from Monterey and Tulare Counties south to Baja California and on the adjacent islands of Santa Rosa, Santa Catalina, and Santa Cruz; not previously reported from San Clemente. NO. 2 GENTRY : LAND PLANTS 39 Gnaphalium palustre Nutt., Trans. Am. Phil. Soc. II, 7:403. 1841. Hill west of Prisoners Harbor, Santa Cruz Island, April 17, El- more 443. Widely distributed in western North America from British Co- lumbia to Mexico; on the islands Santa Cruz, Santa Catalina, and San Clemente. Haplopappus venetus vernonioides (Nutt.) Munz, Man. So. Calif. Bot. 522, 601. 1935. Middle island of Anacapa group, August 1, Elmore 246, steep rocky slope. Common on dry slopes at low elevations from San Francisco south into Baja California and on all the adjacent islands. Hemizonia CLEMENTINA Brge., Erythea 7:70. 1899. Santa Barbara Island, August 12, Elmore 299, dry hillside. San Clemente Island, February 18, 19, Elmore 408. Endemic to the islands Santa Catalina, San Clemente, Santa Bar- bara, and Anacapa. A perennial herb with dimorphic inflorescence; the early flowers in larger heads on simple leafy peduncles, the later in diffuse corymbose panicles. Elmore's collection from San Clemente is in the early flower- ing stage and differs from typical material also in the long villous pubescence of branches and foliage. It may be in need of varietal or possibly even specific segregation. Hemizonia fasciculata (DC.) Torn & Gray, N. Am. Fl. 2:397. 1841-43. Becher Bay, Santa Rosa Island, August 2, Elmore 175, dry hillside. Common from southern California south through Baja California to Cedros Island ; also on San Miguel, Santa Rosa, Santa Catalina, and San Clemente Islands. It is tolerant of immature soils and apparently is aggressive on local wastelands. Layia platyglossa (F. & M.) Gray, PI. Fendl. 103. 1849. Becher Bay, Santa Rosa Island, August 2, Elmore 202, dry hillside. Hill w^est of Prisoners Harbor, April 17, Elmore 442. Widely scattered in southern California and on San Miguel, Santa Rosa, Santa Cruz, and Santa Catalina Islands. 40 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Malacothrix Clevelandii Gray, Bot. Calif. 1:433. 1876. Santa Barbara Island, August 12, Elmore 370. Aggressive on disturbed areas in cismontane southern California and on the islands Anacapa, Santa Cruz, and on the above new extension of range, Santa Barbara Island. Apparently migratory on the mainland. Malacothrix foliosa Gray, N. Am. Fl. ed. 2, 1, pt. 2, suppl. :455. 1886. Point Bennett, San Miguel Island, September 12, Elmore 338. On the islands Santa Cruz, San Clemente, Santa Barbara, and the above extension of range, San Miguel. Malacothrix incana (Nutt.) Torr. & Gray, N. Am. Fl. 2:486. 1841-43. Tyler Bight, San Miguel Island, August 3, Elmore 322. San Miguel Island, August 10, Elmore 327. Channel Islands of Santa Rosa, Santa Cruz, San Miguel, and the adjacent coast. Malacothrix saxatilis implicata (Eastw.) Hall, U. C. Publ. Bot. 3:269. 1907. Middle island of Anacapa group, August 1, Elmore 248, steep rocky slope. Channel Islands of San Miguel, Santa Rosa, Santa Cruz, San Ni- colas, and Anacapa. The evolution of the genus Malacothrix, judging from the numerous segregates of varying rank that are known from the islands, has ap- parently been highly responsive to the stimulation of successive disjunc- tions and conjunctions of populations as induced by physiographic evo- lution. Perezia microcephala (DC.) Gray, PI. Wright. 1:127. 1852. Prisoners Harbor Canyon, Santa Cruz Island, August 2, Elmore 263, dry rocky hillside. Open arid slopes of the chaparral belt from San Luis Obispo County south to Baja California and the adjacent islands of Catalina, Santa Rosa, and Santa Cruz. The limited island distribution of this strongly successful perennial of the mainland, indicates at least recent appearance on the islands. NO. 2 GENTRY : LAND PLANTS 41 Perityle Emoryi Torr. in Emory, Rep. N. Mex. Bound. Sur. 142. 1848. San Clemente Island, February 18, 19, Elmore 414, 385. Widespread from southern California to Arizona south in Mexico to southern Sonora and throughout Baja California; also on Santa Cata- lina, Santa Rosa, Santa Cruz, San Clemente, and Cedros Islands. It is an aggressive winter annual with white rays in the arid maritime climates. Senecio Lyonii Gray, Syn. Fl. I, 2:456. 1886. San Clemente Island, February 18, 19, Ehnore 404, 399. On the Channel Islands of San Clemente, Santa Barbara, Santa Cruz, and in Baja California. SiLYBUM Marianum (L.) Gaertn., Fruct. 2:378. 1791. Santa Cruz Island, August 2, Ehnore 264. Common weed in California naturalized from Europe; also on Santa Catalina Island. SONCHUS OLERACEUS L., Sp. PI. 794. 1753. Santa Barbara Island, August 12, Elmore 304, dry hillside. Common weed in wasteland, naturalized from Europe; widespread on the islands and probably on all of them. 42 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 LITERATURE CITED DUNKLE, M. B. 1944. Investigations of the plant ecology of the Channel Islands. Ph.D. thesis, 264 pp. Univ. of So. Calif. Eastwood, Alice 1941. The islands of Southern California and a list of the recorded plants. Leaflet West. Bot. 3:27-35, 54-78. Eraser, C. McLean 1943a. General account of the scientific work of the Velero III in the Eastern Pacific, 1931-41, Part I. Historical introduction, Velero III, Personnel. Allan Hancock Pacific Exped., 1(1) :l-48, pis. 1-16. 1943b. General account . . Part II. Geographical and biological associations. Ibid., 1(2) :49-258, pis. 17-128. 1943c. General account . . Part III. A ten-year list of the Velero III collecting stations. Ibid., 1(3) :259-432, charts 1-115. Hoffman, Ralph 1932. Notes on the flora of the Channel Islands off Santa Barbara. Bull. So. Calif. Acad. Sci. 31:46-60, 101-120. ? Flora of San Miguel Island. Typed list of plants in Santa Barbara Museum of National History; ca. 150 species and varieties. MiLLSPAUGH, C. F. AND ThOMAS NuTTALL 1923. Flora of Santa Catalina Island. Field Mus. Pub. Bot. 5:1-413, pis. 1-14, map. MuNZ, Philip A. 1935. Manual of Southern California botany. Published by Claremont College, 642 pp. Reed, Ralph D. 1933. Geology of California. Published by Am. Assoc. Pet. Geol., Tulsa, Okla. 355 pp. (State Minerologist, Ferry Bldg., San Francisco, Calif.) NO. 2 gentry: land plants 43 CEDROS AND SAN BENITO ISLANDS Cedros Island is a partially submerged mountain that stands about midway along the coast of Baja California across the broad Vizcaino Bay. North to south it is about 38 kilometers long, averages about 8 kilometers wide, and contains about 300 square kilometers of rugged land. It has several peaks, the highest of which is about 1200 meters above the level of the Pacific Ocean. Dissected by numerous canyons, some of which dip steeply to the sea, it contains several physiographic habitats; peak, ridge, hill slope, mesa, canyon, cliff, and brief narrow beaches. Alluvial slopes and valleys are notably minor. Salty and other- wise highly mineralized and unpalatable water is reported to stand in the lower reaches of the canyons. Near the south end of the island, about 3 miles from what was formerly known as Bernstein's abalone camp and at about 600 to 900 meters elevation, is a good fresh water spring. Fishing has recently been developed by the construction of a large cannery in the locality and a village has grown up around it. The island is composed mostly of sedimentary rocks. The south end shows fossiliferous marine sediments of Cretaceous shales, Miocene shales and sandstones, and Pliocene sands and conglomerate. Near the middle of the east side of the island, "Grand Canyon" cuts back deeply. G. Dallas Hanna, paleontologist on the California Academy of Science's expedition to the Eastern Pacific islands in 1922 and 1923, wrote in reference to the middle section, as follows (1926:88): "It was found that a fault line crosses the island following approximately the course of the canyon. To the south only, Jurassic cherts, supposedly Franciscan in age, were found. To the north there is a block of Cretaceous shales, 200 or more feet thick, with a generally westward dip of about 30°. — Our studies convinced us that Cedros Island is a zone of intense block faulting and disturbance. At the present time, except for a com- paratively recent post-Pleistocene uplift of little significance, the island is in a period of depression. In other words, at no very distant period geologically, the island was a part of a much higher land mass." He found igneous rocks only at the two extremities of the island. "At the southwestern corner of the island there has been some volcanism and at the north end the land is greatly disturbed with intrusions of serpentine" (1925:268). Since the axis of Cedros Island is aligned with that of Sierra Viz- caino to the south on the western edge of the peninsula, and because soundings show the intervening channel to be only 9 to 12 meters in 44 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 depth, it appears that the two land masses were continuous at times in the geologic past. If this is true, the floras of the two mountains should show much in common and a comparison would contribute evidence for the historical problem. Unfortunately, the flora of Sierra Vizcaino is quite unknown, except b}' what may be inferred from neighboring locali- ties of collections, the nearest being northwest of the mountains on San Bartolome Bay, or, as it is now commonly known, Tortuga Bay. Three collecting parties have visited this locality: Hinds on H.IVI.S. Sulphur in 1839, Pond on the U.S. ship Ranger in 1889, and Brandegee on the Wahlberg in 1897. The latter on his spring visit found, "the region was perfectly dry and seemed not to have been rained upon for years. A few plants w^ere recognized that w^re before known only from Cedros Island, and made it evident that an accurate knowledge of the distribution of neighboring island forms cannot be obtained without a more thorough examination of the adjacent mainland." The isolated position of Sierra Vizcaino, sitting out by itself across a broad low desert plain about 60 miles wide, indicates that it too was for indeterminant periods in the past an insular body and that like Cedros it may possess its own relic biotic elements. There are no meteorological data for Cedros Island nor for the adjacent peninsula. Rainfall is probably under 8 inches annually on the average for the lower slopes and most of the island area, something over that for the highlands. The rainfall incidence is irregular, judging from oral accounts of peninsular natives, and years may pass w^ithout effective precipitation. There is apparently only one source of rainfall and that is in the cyclonic winter storms of the northern latitudes which occasion- ally extend to Cedros. According to the natives, the convectional sumi- mer storms rarely reach Cedros Island. The moist westerlies are reported by the inhabitants of the adjacent peninsula to be the dominant and persistent winds. A fog desert similar to that found along the western border of the peninsula may exist over some of the island. Visitors re- port that fog drips from the trees in the higher elevations to such an extent that it causes little rivulets of water and that these have been mistaken for springs. Few botanists appear to have visited the western margin of Cedros and very little has been published regarding its plant life. According to accounts left us by visitors, most of the island is covered w^ith a dis- persed formation of Desert Shrub. Suftrutescents in Eriogonum, Fran- NO. 2 GENTRY I LAND PLANTS 45 seria, A triplex, Viguiera, and Encelia are common. Succulents are repre- sented in Agave and cacti and semi-succulents by such shrub and tree forms as Euphorbia misera and Pachycormiis discolor. The sclerophyl- lous shrubs of Rhus and Siminondsia appear to be prevalent and Greene mentions (1888:197) two evergreen shrubs, Gilia Veatchii and Harfor- dia fruticosa, ' Vhich grow on these lower hills in sufficient quantities to impart an appearance of verdure." To Greene the most conspicuous tree was the corpulent xerophyte, Pachycormus discolor, endemic to Baja California and adjacent islands. A scrubby juniper, Juniperus cerrosi- anus, he reported as growing throughout all elevations. Towards the summits of the mountains, Pinus muricata cedrosensis is accompanied by Arctostaphylos bicolor. George O. Hale and Lee Haines, as two students of botany at the University of California at Los Angeles, spent six weeks in the early spring of 1939 on the island and made an ecologic study of the vegeta- tion. Hale (1941) reported about 97% of the island is covered through all elevations with Desert Shrub. It has a uniform growth as a regularly dispersed open formation of lov/ bushy shrubs spotted with large shrubs or dwarf tree forms; among the latter, that of Pachycormus discolor being the most ubiquitous. He found differences in composition between the higher elevations and the lower elevations, and divided the Desert Shrub accordingly into ''High Altitude Desert" (650 to 1300 meters elevation) and "Low Altitude Desert" (below 650 meters elevation). As dominants of the former he listed Eriogonmn fasciculatum, Pachy- cormus discolor, Haplopappus propinquus, and Franseria camphorata leptophylla and for the latter he listed as most abundant Harfordia fruticosa, Euphorbia misera, Pachycormus discolor, Franseria chenopodi- folia, and Encelia calif ornica asperifolia. The only other low altitude assocation he reported is ''Maritime Dune Scrub," occupying a small dune area on the southwest coast. Dominated by A triplex julacea and Frankenia Palmeri, it is also charac- terized by Brodiaea, Abronia, Achyronchia Cooperi, Oenothera, and Ly- cium Andersoni. The highland vegetation, other than High Altitude Desert Shrub, occupies limited areas at middle and high elevations. A Closed Cone Pine Forest (250 to 900 meters elev.) occurs in two widely separated areas on western and northern slopes, in the central and northern parts cf the island. The pines usually occur in pure stands and have an average mature stature of about 50 feet. Hale 46 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 concluded that they are coincident with the elevations and tracks of regular fog ingressions. A local Chaparral, dominated by Arctostaphylos, Adenosiema, Quer- cus, Eriogonurn, and Juniperus, occurs on the northern slopes of the highest peaks, above the pine forest. This is a most interesting ecologic find with considerable historical significance. Small patches of Coastal Sagebrush and Juniper Woodland border the pine stands irregularly. These formations show little or no natural transitional grading, but are sharply one or the other. Where the pine forest begins, the desert shrub leaves off. Chaparral occupies higher elevations than the pines and both formations appear largely dependent upon the upsurging fogs from the westward, especially the latter. These are aspects peculiar to the island, and while the vegetations show climatic qualities, they are inter- esting instances of what climate and soil t3'pes have generated out of the limited plant materials available. The flora of the island is now known to consist of about 182 species of vascular plants and has recently been annotated by Eastwood (1929) and Howell (1942). Around 600 numbers have been taken by 14 col- lectors, but little collecting has been done in summer, only one small one made in winter (Pond, 15 numbers), and none at all during the fall (Table 2). Considering this, the fact that little or no collecting has yet been done on the western side, the areal nature of the island, and the irregular rainfall, I would estimate that some, perhaps 10%, of the island flora is still unreported. In addition to the unknowns, more material of many of the known plants, and more field work are needed for a thorough evaluation of the Cedros flora. The plant collections of the Allan Hancock Expeditions to Cedros Island were made in 1937 by P. J. Rempel and in 1939 by Francis H. Elmore. They are enumerated belov/ together with Rempel's collection from the neighboring San Benito Islands. Their collections add three genera and four species to the Cedros Island flora: Aristida adscensionis, Eschscholtzia minutiflora, Fagonia laevis, and Euphorbia bartolomaei. NO. 2 GENTRY : LAND PLANTS 47 TABLE 2 Collector Cedros Island San Benito Islands Year Spring Summer Spring Summer 1859 Veatch 28 1876 Streets 11 10.^ 1882 Belding 10 1885 Greene 82 1889 Pond 15 24 1889 E. Palmer 87 18 1897 T. S. Brandegee 100? 50? 1905 Stewart 7 1911 J.N. Rose 97 44 1922 Hanna 7 1925 Mason 63 1932 Howell 41 1937 Rempel 19 15 1939 Elmore 38 1939 Hale and Haines 150 Table 2. Plant collectors and collections on Cedros and San Benito Islands. Estimated numbers are followed by a question mark. The San Benito Islands consist of three small islands, East, Middle and West, lying about 24 kilometers from the north end of Cedros Island. Although they rest upon the continental shelf, they are separated from Cedros Island by a channel 180 to 190 fathoms deep. This is more than enough to prevent junction during low eustatic sea levels of the Glacial Periods. Whether they were ever land-bridged to Cedros or the peninsula is not presently known. The highest elevation of 200 meters (661 feet) is attained on West Island, the largest of the three. Fraser ( 1943 :65) speaks of them as "all rocky and barren," but Greene (1889 : 261), the first chronicler of the islands, described them in glowing terms. ''Lieutenant Pond judges the San Benito Islands to be of much older formation than the large island of Cedros near by. The surface is not sharply rocky ; the slopes are not abrupt ; there is good depth of soil almost everywhere, and vegetation is abundant, the whole group presenting, on near approach, a picture of freshness and verdure at the showery season of the year, the months from December to February, during which several visits were made (by Pond). At this time sweet flowing water was found in most of the canyons and ravines ; a condition not likely to obtain during the dry summer season." Such opposing impressions in addition to the personal factor of prejudice appear to be conditioned by the lack or presence of seasonal rains. 48 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Before Pond's collection of 1889, Streets had collected, among other plants the endemics Lavatera venosa, and Hemizonia Streetsii. Follow- ing Pond the next important collection is that of the remarkable pereg- rinating botanist, Edward Palmer, who visited the San Benitos in March of the same year that Pond did. His collections w^ere reported by J. N. Rose (1890:20-21). No additions to Greene's first published list of 25 species were made, however, until Brandegee visited the San Benito Islands on the voyage of the Wahlberg in 1897 and published (1900: 22-23) a small supplementary list, bringing the known flora of the San Benitos to 40 species. Rose and Rempel appear to complete the roster of collectors on San Benito Islands (Table 2). Except for the five plants which still appear to be endemic to the San Benito Islands, the flora is almost completely repeated on Cedros Island, the exceptions being common wide-spread species, as Lepidium lasiocarpum, which may have been passed by collectors on Cedros. Be- cause of the striking lack of the genera Astragalus and Eriogonmn in the San Benito flora, Greene saw a straight relationship between the San Benito flora and that of Guadalupe Island 150 miles westward, and not at all between San Benito and Cedros. The absence of these genera, so conspicuous on the peninsula and on Cedros Island, is, of course, interesting, and the explanation of it could easily lead one into many conjectures. However, because so many Cedros genera and species are represented on the San Benito Islands and so few of the Guadalupe Island species are, Greene's statement now appears rather meaningless. The position and character of the San Benito flora, however, does re- main in part anomalous. The collection of 15 numbers by P. J. Rempel from the San Benito Islands during July on the 1937 voyage of the Velero III, is apparently the first to be made in summer. Although no novelties are added, the collection provides fruiting and flowering records and one addition, Mesembryanthemum nodiflorum^ as annotated below in the catalogue of species. NO. 2 GENTRY : LAND PLANTS 49 CATALOGUE OF COLLECTIONS Gnetaceae Ephedra aspera Engelm. ex Wats., Proc. Am. Acad. Scl. 18:157. 1883. Ephedra peninsularis Jtn., U. C. Publ. Bot. 7:437. 1922. East side of Cedros Island, July 10, 1937, Rempel 340, alluvial fan. Widely distributed in the southwestern United States and northern Mexico; type apparently from Frontera, Texas. Rempel's collection is sterile and not certainly determinable, but aspera is the only species known from Cedros Island. Hale, (1941:56) reports that the plants are broad and bushy, scarcely exceeding a foot in stature. Gramineae Aristida adscensionis L., Sp. PI. 82. 1753. Cannery Bay on east side of Cedros Island, March 14, Elmore A 6, in dry wash. A xerophytic grass commonly scattered throughout southwestern United States and southward into Mexico ; also in warmer parts of the Old World. The collection is a single depauperate plant, but it adds another species to the published flora of the island. Amaryllidaceae Agave Shawii Engelm., Trans. Acad. St. Louis 3:314. 1875. West San Benito Island, July 14, 15, 1937, Rempel 364 (sterile), on southerly exposures with wind appearing to prevail from north. Typical A. Shawii is abundant from San Diego, California south to Rosario in northwestern Baja California, where it commonly forms dense stands along the slopes facing the moist sea air. The type was described from Point Loma near San Diego. The short ovate-lanceolate blades with strong crooked, closely set, marginal prickles of the San Benito plant pretty certainly align it with the peninsular plant. Agave Shawii Engelm., var. sebastiana (Greene) Gentry new comb. Agave sebastiana Greene, Bull. Calif. Acad. Sci. 1:214. 1885. East side of Cedros Island, July 10, Rempel 330, general. The variety is known certainly only from Cedros Island. Greene first collected and described it as a distinct species. Trelease (C.N.H. 23:110, 124. 1920) maintains A. Shawii and A. sebastiana as distinct species, separating them in his key on the basis of sinuous (Shawii) or straight (sebastiana) terminal spine. The sebastiana population appears 50 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 also to differ from typical Shawii by its less robust form and by fewer, more remote, and smaller marginal spines, characters which appear to have subspecific rather than specific value. Agave Shawii sebastiana may also occur on the San Benito Islands. Brandegee reported an Agave from the San Benito Islands (1900) which he considered the same as on Cedros. Palmer (Rose 1890:20) spoke of two species of Agaves on San Benito Islands. POLYGONACEAE Eriogonum fasciculatum Benth., Trans. Linn. Soc, 17:417. 1837. Cannery Bay on east side of Cedros Island, March 14, Elmore A24i sandy rocky slope in dry wash, flowers white. Generally distributed on the maritime and desert slopes and valleys from Santa Barbara, California south to central Baja California and on adjacent islands. This collection appears to be related to the variety flavoviride Munz and Johnston, but differs in the pubescent peduncles and larger stature. Eriogonum intricatum Benth., Bot. Voy. Sulph. 46, t. 22. 1844. East side of Cedros Island, July 10, Rempel 337. Known only from Cedros Island and the adjacent peninsula; the type from San Bartolome Bay, Baja California. The above material is very young, but a few young flowers at the base of the plants with wine-red strigulose sepals in sessile involucres and the pubescent leaves limited to the basal node are diagnostic. Eriogonum Pondii Greene, Pittonia 1:267. 1889. Cannery Bay on east side of Cedros Island, March 14, Elmore A16y A 25, dry rocky wash. In and along arroyos in sandy and coarse detrital soils on Cedros Island and the adjacent peninsula near Tortuga Bay (San Bartolome Bay) ; type from Cedros Island. Elmore reports in his notes that the flowers on No. A 16 were pink, and white on No. A 25. A very low wide-spreading suffrutescent bush. Chenopodiaceae Atriplex dilatata Greene, Pittonia 1:264. 1889. West San Benito Island, July 14, 15, Rempel 366, west side of island; small-leaved plant all over island. San Benito Islands, July 14, 15, Rempel 361, general distribution. NO. 2 gentry: land plants 51 Known only from the San Benito Islands to which it is apparently confined. The species is related to A. Bar clay ana, but differs in the woody branches, the thicker fruits with less tuberculation on the sides. The shrubby nature of the plant is apparent in the above cited collec- tions. On the sheet of Rempel 361 even the terminal branches are woody, the leaves reduced, while Rempel 366 shows new herbaceous shoots from a thick woody branch, the leaves uncommonly wide and herbaceous, the broad ovate blades 3-4 cm long, mostly 2-2.5 cm wide; petioles 5-10 mm long. Atriplex pacifica a. Nels., Proc. Biol. Soc. Wash. 17:99. 1904. Cannery Bay on east side of Cedros Island, March 14, Elmore A37, sandy, rocky soils of alluvial fan. Southern California and adjacent islands to Cedros Island; type from San Diego, California. Elmore reports only a few plants observed and the collection consists of a single small plant. I have seen no collec- tions of this species from northern Baja California, but it probably occurs there, since it is a weedy type that is often passed by collectors. SUAEDA SP. San Benito Islands, July 14, 15, Rempel 358 (sterile). Nyctaginaceae Hesperonia cedrosensis Standi., C.N.H. 12:362. 1909. Cannery Bay on east side of Cedros Island, March 14, Elmore A 15, alluvial fan with sand and pebbles. On coarse arid soils from Cedros and San Benito Islands, along the northern coast of Baja California, and on San Clemente Island in the Channel group; type from Cedros Island. Elmore noted but a few plants; flowers purple. AlZOACEAE Mesembryanthemum crystallinum L., Sp. PL 480. 1753. Cannery Bay on east side of Cedros Island, March 14, Elmore A 2 6, gentle slope of pebbly alluvial fan. Naturalized along the coasts of the Californias and adjacent islands from Santa Barbara, California south into northern Baja California; type from Cape of Good Hope. Mesembryanthemum nodiflorum L., Sp. PI. 480. 1753 West San Benito Island, July 14, 15, Rempel 369, 371, north side of island. 52 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 This Mediterranean adventive is abundant on the coast through the northern part of Baja California and is also common farther north as far as Oregon. Rempel's collections are the first records from San Benito Islands. PORTULACACEAE Calandrinia maritima Nutt. in Torr. & Gray, N. Am. Fl. 1 :197. 1838. Cannery Bay on east side of Cedros Island, March 14, Elmore AZS^ gentle slope of a rocky alluvial fan; flowers purple. From Santa Barbara Count}^, California south along the coast to central Baja California. Locally on Cedros the plant is known as "ver- dolaga" and was regarded by Elmore's informant as edible. EscHSCHQLTZiA MiNUTiFLORA Wats., Proc. Am. Acad. Sci. 11 :122. 1876. Cannery Bay on east side of Cedros Island, March 14, Elmore J17y gentle slope of rocky alluvial fan. From southern Utah south through the Sonoran Desert to middle Baja California. The above collection is a considerable extension of range for the species and an addition to the known flora of the island. Cruciferae SiBARA PECTINATA Greene, Pittonia 3:11. 1896. Cannery Bay on east side of Cedros Island, March 14, Elmore A32f on gentle rocky slope of alluvial fan; flowers light purple. Central Baja California and Cedros Island; type from San Bart- olome Bay (Tortuga Bay). A delicate and rarely collected winter an- nual, slender, erect, with dissected leaves 4-6 cm long, the lobes remote and narrowly linear, spreading. Capparidaceae IsoMERis ARBOREA Nutt. in Torr. & Gray, N. Am. Fl. 1 :124. 1838. Cannery Bay on east side of Cedros Island, March 14, Elmore A14, dry wash and fans on exposed, sandy, pebbly, gentle slopes; flowers yellow. Widely but irregularly distributed in the deserts of California, So- nora, and Baja California; type from San Diego, California. Forms a low bushy shrub on the island and flowers through February and March after rainy winters. NO. 2 GENTRY : LAND PLANTS 53 Leguminosae Errazurizia Benthami (Brge.) Jtn., Proc. Calif. Acad. Sci. IV, 12:1043. 1924. Dalea Benthami Brge., Proc. Calif. Acad. Sci. II, 2:148. 1890. Cannery Bay on east side of Cedros Island, March 14, Elmore A20, on rocky, sandy, exposed, gentle slope of alluvial fan. East side of Cedros Island, July 10, Rempel 341, alluvial fan and hill side. Known only from the Sierra Vizcaino area and the neighboring islands; type from Santa Margarita Island. This is a low stiff shrub with short intricate branches, gray twigs, and conspicuously gland-dotted light gray leaves, the whole often matted into a low rounded crown. Both Johnston and Rydberg have indicated the advisability of separating this species (together with the related E, megacarpa from the California Gulf Region), with its regular non- papilionaceous corolla, from the genus Dalea, having a typical papili- onaceous corolla; a view with which the writer thoroughly agrees. Un- fortunately, Phillip's awkward name, Errazurizia, (Ann. Univ. Chile 1872:688) has priority over Rydberg's Psorobatus (N. Am. Fl. 24:41. 1919). It was reported to Elmore by local informant as edible to human beings, but caused insanity among animals. The informant may have been confusing the plant with members of the genus Astragalus. Lotus humilis Greene, Pittonia 2:140. 1890. Cannery Bay on east side of Cedros Island, March 14, Elmore A27. A low prostrate winter annual common to the sands of central Baja California and Cedros Island; type from San Bartolome Bay. LupiNUS PoNDii Greene, Pittonia 1:288. 1889. Cannery Bay on east side of Cedros Island, March 14, Elmore Al, a few plants in a dry wash. Central and northern Baja California and Cedros Island; type from San Bartolome Bay, Baja California. It is a low annual herb with spreading stems, sparse, coarse, long pubescence, and the flowers glomer- ate along the rachis of the inflorescence, according to the specimen at hand. Zygophyllaceae Fagonia laevis Standi., Proc. Biol. Soc. Wash. 24:249. 1911. Cannery Bay on east side of Cedros Island, March 14, Elmore A18, gentle slope with rocky sandy soil on alluvial fan; flowers light purple. 54 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Widely distributed around the upper half of the Gulf of California in California, Arizona, Sonora, Baja California, and on intervening and adjacent islands. This is a rather dense form of the usuall)^ more open habit of the species, which is characterized by the 3 small very narrow leaflets and the short spreading stipules. EUPHORBIACEAE Euphorbia bartolomaei Greene, Pittonia 1:290. 1889. Cannery Bay on east side of Cedros Island, March 14, Elmore 129^ rocky sandy soil in dry wash ; flowers white. One of the small finely cut prostrate spurges inhabiting the arid soils on Cedros Island and adjacent Baja California; type from San Bartolome Bay. Euphorbia misera Benth., Bot. Voy. Sulph. 51. 1844. Euphorbia Benedicta Greene, Pittonia 1:263. 1889. Cannery Bay on east side of Cedros Island, March 14, Elmore A19, gentle slope on sandy rocky alluvial fan. San Benito Islands, July 14, 15, Rempel 370, north side of West Island (flowering). Irregularly distributed through the deserts of southern California, Baja California, and perhaps in the Thorn Forest of central Sinaloa (Gentry 7001, near Culiacan) ; type from San Diego, California. Eastwood (1929:432) reported it tentatively as collected by Mason, his material being too poor for certain identification. Elmore's number agrees well with typical material from the peninsula. The San Benito Island plants are apparently more succulent and dwarfed in stature, and are perhaps worthy of varietal distinction. Buxaceae SiMMONDSiA CHiNENSis (Link) Schneider, 111. Handb. Laubholzk 2:141.1907. Simmondsia californica Nutt., Lond. Jour. Bot. 3:400, t. 16. 1844. Simmondsia pabulosa Kell., Proc. Calif. Acad. Sci. 2:21. 1860. Cannery Bay on east side of Cedros Island, March 14, Elmore A3, dry wash. East side of Cedros Island, July 10, Rempel 331 (sterile). Widely scattered upon the arid slopes on the mountains and in the canyons of the Sonoran Desert from southern California and southern Arizona south into Sonora and through Baja California to the Cape District; the type is probably from San Diego, California, although the describer attributed it to China, due to error in labeling or sorting by the collector who visited both California and China on the same voyage. NO. GENTRY : LAND PLANTS 55 It is a leathery-leaved evergreen shrub vv^ith acorn-like fruits, which, although it may assume a dominant role in the vegetation, is more often w^idely scattered as individual plants. Although somewhat bitter the seeds are eaten by the desert peoples as they ripen, either raw or roasted, and a kind of coifee has been made of them by the Mexicans in times of coffee shortages. Elmore reports that it is employed locally as a pomade by women for their eyelashes and by men for their moustaches. The family affinities of the plant are uncertain. Anacardiaceae Pachycormus discolor Veatchiana (Kell.) Gentry new comb. Rhus Veatchiana Kell., Proc. Calif. Acad. Sci. 2:24. 1860. Veatchia discolor Veatchiana (Kell.) Jtn., Proc. Calif. Acad. Sci. IV, 12:1081. 1924. Cannery Bay on east side of Cedros Island, March 14, Elmore A13, on wash walls and fans. East side of Cedros Island, July 10, Rempel 344 (with young leaves and budding inflorescence). This variety, which according to Johnston is distinguished by its larger, more colored, coarser, and more conspicuously pubescent flowers, is known definitely only from Cedros Island. However, it likely will be found to occur also on the adjacent island of Natividad and the Sierra Vizcaino part of the mainland. Elmore collected it at 75 feet elevation where he noted a few trees or shrubs up to 10 feet high, trunks as much as two feet in diameter and with peeling papery bark, known locally as "Copalquin." Hale (1941:68) reports it as "Abundant throughout the desert formation, and reaches its best development as to size in Grand Canyon, where on the lateral alluvial fans it reaches tree size. — On the west coast near the ocean it is often of low stature, not two feet tall." This remarkable plant with its massive, smooth, round, paper-barked trunk and branches together with the pinnate leaves make it appear much like the Burseras, which abound in Mexico. The massive stems accommodate the enlarged water-storing tissues, w^hich serve to carry the plant over the extended and regular drought periods of the peninsu- lar desert, and which in extreme times are known to last for several years. Bentham first described it under Schinus (Bot. Voy. Sulph. 11, pi. 9. 1844), while later authors have assigned it to Veatchia and Rhus, Were not the name Veatchia preoccupied, it would have priority over Pachycormus of Coville. 56 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Rhus Lentii KelL, Proc. Calif. Acad. Sci. 2:16. 1863. Cannery Ba}^ on east side of Cedros Island, March 14, Elmore A4, on sandy gentle slope with pebbles on alluvial fan. On outer coast of middle Baja California and adjacent islands; type from Cedros Island. It forms a dense green shrub or even a small tree, flowering in the early spring. Elmore noted that it was used locally for wood and that the fruits are three to four times larger than our "lemon- ade berry." Malvaceae Lavatera venosa Wats., Proc. Am. Acad. Sci. 12:249. 1877. Middle island of the San Benito Islands, July 14, 15, Rempel 362, near shore on south side. It is known only from the San Benito Islands. The collection is in full leaf, flower, and fruiting. Sphaeralcea fulva Greene, Pittonia 1:201. 1888. Cannery Bay on east side of Cedros Island, March 14, Elmore All, gentle slope of alluvial fan. Commonly in washes along the outer coast of middle Baja Cali- fornia and on Cedros Island, the type locality. Elmore noted that the flowers are orange-colored, that it is known locally as ''malva silvestre," and a decoction of the herbage is made and applied to the back for fevers. LOASACEAE Mentzelia hirsutissima Wats., Proc. Am. Acad. Sci. 12:252. 1877. Cannery Bay on east side of Cedros Island, March 14, Elmore AS sandy pebbly soil of alluvial fan ; flowers light yellow. Infrequently scattered in the mountains of the mid Baja California Desert and on some of the adjacent islands; tj^pe from Angel de la Guardia Island. The single collection contains two depauperate plants 6-7 cm high with the dry corollas 12-15 mm long. Frankeniaceae Frankenia Palmeri Wats., Proc. Am. Acad. Sci. 11:124. 1876. West San Benito Island, July 14, 15, Rempel 368. San Benito Islands, July 14, 15, Rempel 359 (sterile). A halophyte on either salty or "sweet" soils of the arid coastal slopes and flats throughout northern Baja California and adjacent southern California. It forms low brittle woody bushes 2-6 dm high, rather dense, NO. 2 GENTRY : LAND PLANTS 57 and often forming dispersed pure stands over considerable area in some localities. Cactaceae CocHEMiEA PoNDii (Greene) Walton, Cact. Jour. 2:51. 1894. East side of Cedros Island, July 10, Rempel 332 (sterile). Known only from Natividad and Cedros Islands, the type locality being the latter. EcHiNOCEREUS MARITIMUS (Jones) Schuman, Gesambt. Kakteen 27. 1898. East side of Cedros Island, July 10, Rempel 333 (sterile). West coast of Baja California and adjacent islands; type from En- senada. Ferocactus chrysanthus (Orcutt) Brit. & Rose, Cactaceae 3: 127. 1922. East side of Cedros Island, July 10, Rempel 331 (sterile). Northwestern Baja California and the adjacent islands; the type from Cedros Island. The numerous, curved, light brown to gray, an- nular spines in the closely set areoles identify this insular plant. Mammillaria Goodridgei Scheer in Salm-Dyck, Cact. Hort. Dyck. 1849:91. 1850. West San Benito Island, July 14, 15, Rempel 364. West coast of middle Baja California and the adjacent islands of Cedros, San Benito, and Guadalupe; type from Cedros Island. Some of the Rempel specimens are fruiting. The series show all low plants tend- ing to be obovate in outline, 5-10 cm high, woody at base, and one plant is branched. Opuntia sp. San Benito Islands, July 14, 15, Rempel 357, infrequent — mostly top and north sides. This is a sterile Cylindr opuntia, showing affinities to the section Imhricatae, The inflated straw-like sheaths of the spines suggest O. cal- malliana, but the joints are too thick, the spines too few, too long, and the areoles too remote for that species. It is neither O. tesselata nor O. prolifera, the only two Opuntia which have been reported (Brandegee 1900:20) for the San Benito Islands. The collection probably repre- sents an undescribed species, but is insufficient for diagnosis. 58 allan hancock pacific expeditions vol. 13 Hydrophyllaceae Phacelia IXODES KclL, Bull. Calif. Acad. Sci. 1:6. 1884. East side of Cedros Island, July 10, Rempel 335, damp place in canyon. Known only from Cedros Island, the species is readily recognized by its coarse stems, long scorpioid cymes, and heavy glandular pubescence. BORAGINACEAE Cryptantha maritima Greene, Pittonia 1:117. 1887. Cannery Bay on east side of Cedros Island, March 14, Elmore A9, sandy soils of wash and alluvial fans. In coarse sandy soils apparently throughout the Sonoran Desert in California, Arizona, Baja California, and on adjacent islands. Labiatae Salvia cedrosensis Greene, Bull. Calif. Acad. Sci. 1:212. 1885. Cannery Bay on east side of Cedros Island, March 14, Elmore A33, dry wash. Known certainly only from Cedros Island, but possibly also at Mag- dalena Bay, Baja California. It is a low sufErutescent plant with blue flowers. The collector reported but few plants observed and the one col- lected is depauperate. Teucrium glandulosum Kell., Proc. Calif. Acad. Sci. 2:23. 1863. Cannery Bay on east side of Cedros Island, March 14, Elmore AlO, a few plants in a wash in the bottom of a small canyon in pebbly sand, corolla white tinged with lavender. Ranges sparingly through middle Baja California and the adjacent islands, the type from Cedros Island. Scrophulariaceae Galvezia juncea (Benth.) Gray, Proc. Am. Acad. Sci. 22:311. 1887. Cannery Bay on east side of Cedros Island, March 14, Elmore A 30, on alluvial fan with gentle, rocky, sandy slope. Widely scattered through central Baja California and adjacent islands; type locality, west coast of Baja California, probably at San Quentin. It is a perennial herb with round, smooth, strictly ascending, green branches and few small linear-lanceolate, glabrous, ephemeral leaves about 1 cm long; flowers red, the sepals and peduncles strongly glandular pubescent. NO. 2 GENTRY : LAND PLANTS 59 CUCURBITACEAE EcHiNOPEPON MINIMUS (Kell.) Wats., Proc. Am. Acad. Scl. 24: 52. 1889. Marah minima Kell., Proc. Calif. Acad. Sci. 2:18. 1863. Cannery Bay on east side of Cedros Island, March 14, Elmore A12y climbing on shrubs on the sides of dry washes. "Guisapol." Central and southern Baja California and the outer adjacent islands; type from Cedros Island. It is a small scabrous vine, the leaves mostly broadly 3-lobed, small white flowers, and solitary fruits rather strongly but flatly prickled. COMPOSITAE Baccharis sarathroides Gray, Proc. Am. Acad. Sci. 17 :211. 1882. East side of Cedros Island, July 10, Rempel 338 (sterile). Desert washes of the southwestern United States and northwestern Mexico ; type from near Old Mission Station, San Diego County, Cali- fornia. Bebbia juncea (Benth.) Greene, Bull. Calif. Acad. Sci. 1:180. 1885. East side of Cedros Island, July 10, Rempel 345, fan. Cannery Bay on east side of Cedros Island, March 14, Elmore J35, steep rocky slope with sandy soil. This is a bushy broom-like shrub 1-2 m tall, long-flowering through spring and summer, but with ephemeral reduced leaves, and found throughout the Sonoran Desert in California, Arizona, Sonora, Baja California, and accompanying islands. It is common to the banks and bottoms of the desert arroyos. Type locality: Magdalena Bay, Baja California. Encelia californica asperifolia Blake, Proc. Am. Acad. Sci. 49:368. 1914. Cannery Bay on east side of Cedros Island, March 14, Elmore A22, dry wash. Northern Baja California and adjacent islands; type from Cedros Island. This forms a low spreading bush with many radiating stems from the base. It is a more xerophytic edition of the species, characterized by the light brittle branches and the smaller paler leaves. 60 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Encelia stenophylla Greene, Bull. Torr. Bot. Club 10:41. 1883. Cannery Bay on east side of Cedros Island, March 14, Elmore A7 y dry wash and fans in coarse rocky soil, flowers yellow, leaves resinous and odorous. East side of Cedros Island, July 10, Rempel 342, fan. Known only from Cedros Island. It is a low shrubby plant with very narrow glutinous leaves (1-3 mm wide) and Senecio-like flowering heads. Local name reported by Elmore is "y^rba de venado." Ericameria diffusa Benth., Bot. Voy. Sulph. 23. 1844. Haplopappus sonoriensis (Gray) Blake, C.N.H. 23:1490. 1926. East side of Cedros Island, July 10, Rempel 346, fan. On both coasts of middle Baja California and the adjacent islands; type from Magdalena Bay, Baja California. The material is sterile and doubtfully referred here. It differs from typical peninsular material in having larger linear-lanceolate leaves as well as the typical narrow linear ones. This may be due to the late vegetative stage in which it was col- lected. Not previously listed from Cedros Island. Franseria chenopodifolia Benth., Bot. Voy. Sulph. 20. 1844. Cannery Bay on east side of Cedros Island, March 14, Elmore A 31, alluvial fan. From southern California south through Baja California to the Cape District, Cedros Island. Type from Magdalena Bay, Baja Cali- fornia. A low shrubby bush with crowded whitish to yellowish leaves and rich brown stems with glutinous golden pubescence. Haplopappus tridentatus (Greene) Blake, C.N.H. 23:1493. 1926. East side of Cedros Island, July 10, Rempel 348 (its flowering peri- od is done, only a few achenes remain in the persistent involucres). Known from Cedros Island and adjacent peninsula; type from Cedros Island. It is a low suffrutescent plant with linear leaves termi- nally tridentate. Hemizonia fasciculata (DC.) Torr. & Gray, N. Am. FI. 2:397. 1841-43. Cannery Bay on east side of Cedros Island, March 14, Elmore AS, on gentle slope of alluvial fan. Southern California and northern Baja California. NO. 2 GENTRY : LAND PLANTS 61 Hemizonia Streetsii Gray, Proc. Am. Acad. Sci. 12:162. West San Benito Island, July 14, 15, Rempel 367 ^ north side of island. Known only from the San Benito Islands. It is a low sufifrutescent plant with yellow flowers, the base definitely woody and as much as a half inch in diameter. Perityle Grayi Rose, in Coulter, Bot. Gaz. 15:117. 1890. Cannery Bay on east side of Cedros Island, March 14, Elmore A21^ dry wash. East side of Cedros Island, July 10, Rempel 336, wash. Southern California, Baja California and adjacent islands; type from Guadelupe Island. Porophyllum gracile Benth., Bot. Voy. Sulph. 29. 1844. Cannery Bay on east side of Cedros Island, March 14, Elmore A34, gentle slope with rocky soil on alluvial fan. Found nearly throughout and limited to the Sonoran Desert in California, Arizona, Sonora, and Baja California; type from Magda- lena Bay, Baja California. It is a small highly ramified suffruticose plant with small linear leaves and long pedunculate reddish involucres with white flowers. The herbage is glandular and emits a pungent aromatic odor when plucked or crushed. Never really abundant, it is common in arid situations and may often be found growing up through the branches of low shrubbery. Viguiera lanata (Kell.) Gray, Proc. Am. Acad. Sci. 17:218. 1881-82. East side of Cedros Island, July 10, Rempel 339, wash (fruiting). Cannery Bay on east side of Cedros Island, March 14, Elmore A 22a, dry wash. Known only from Cedros Island. A scapose herb frutescent at the base, white-woolly pubescent on stems, leaves, and involucres. The older scapes shed their woolliness and show the brown-colored stems. It is a handsome plant and worthy of cultivation. 62 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 LITERATURE CITED Brandegee, T.S. 1900. Voyage of the Wahlberg. Zoe 5 :19-28. Eastwood, Alice 1929. Studies in the flora of Lower California and adjacent islands. Proc. Calif. Acad. Sci. IV, 18:420-441. Eraser, C. McLean 1943a. General account of the scientific work of the Velero III in the Eastern Pacific, 1931-1941, Part I. Historical introduction, Velero III, Personnel. Allan Hancock Pacific Exped., 1(1) :l-48, pis. 1-16. 1943b. General account . . Part 11. Geographical and biological associations. Ibid., l(2):49-258, pis. 17-128. 1943c. General acccount . . Part III. A ten-year list of the Velero III collecting stations. Ibid., 1(3) :259-432, charts 1-115. Greene, Edward L. 1888. Botany of Cedros Island. Pittonia 1:194-208. 1889. Vegetation of the San Benito Islands and a list of San Benito Island plants. Pittonia 1:261-269. Hale, George O. 1941. A survey of the vegetation of Cedros Island. Master of Arts thesis, Univ. of Calif, at Los Angeles. Hanna, G. Dallas 1925. Expedition to Guadalupe Island, Mexico. Proc. Calif. Acad. Sci. IV, 14:217-275. 1926. Expedition to the Revilla Gigedo Islands, Mexico. Proc. Calif. Acad. Sci. IV, 15:87-89. Howell, John Thomas 1942. A short list of plants from Cedros Island. Leaflet West. Bot. 3:180-185. Rose, J. N. 1890. San Benito Island plants. Contributions of the National Herbarium, 1:20-21. NO. 2 GENTRY : LAND PLANTS 63 REVILLA GIGEDO ISLANDS The Revilla Gigedo Islands are small and widely scattered far off the Mexican west coast. Southward about 380 kilometers from the tip of Baja California is San Benedicto Island. Just south of it is the largest, Isla Socorro, w^hich is about 520 kilometers west by south of Jalisco. Three hundred miles westward of Socorro is Clarion. Rocca Partida is a barren double rock westward of San Benedicto and Socorro. All these islands are volcanic, their geologic history obscure, and their relationship vague. They may be oceanic islands, but Johnston (1931:45) considers them as continental. On the basis of bathymetrics and flora he regards them as peaks of a submerged land mass that was part of the Mexican mainland. The islands are in general alignment with the Tarascanahuan Cordillera, a volcanic massif forming the southern border of the Mexican central plateau, which breaks off in Jalisco at the sea. Because the flora of the Revilla Gigedos is related to both that of the Cape District of Baja California and southern Mexico, his theory presupposes that the whole area of the California peninsula and its periphery was once a part of the continental land mass, more or less outlined by the bathymetric contour of 4000 meters. This could only have been true previous to the Upper Tertiary, because Miocene and Pliocene formations on Baja California, the Mexican west coast, and adjacent islands show that their respective areas were covered by salt water. Johnston's argument is brilliantly developed, but the entire structure of the area is in need of detailed field study before credence of his theory can be assured. The climate of the islands is semi-arid and maritime tropical with a dry spring season. There are no meteorological records, but in these latitudes rainfall should be about 80% summer and fall. The geographic position places it in Koppen's classification of Savanna Climate, which is characterized by a dry winter and wet summer. The long dry spring and the foggy montane forest of Socorro, however, make it atypical. That the spring is dry and the winters in part wet is attested by the five expeditions which have visited the islands between 1889 and 1939. This is strongly indicated both by their reports and the dry quality of the specimens collected during the spring months. Though the northern anticyclonic storms may cause some mild precipitation during the winter and local temperatures be productive of winter and spring fogs, the rainfall regime belongs definitely with the tropical and appears quite comparable to that which prevails on the west coast mainland in these 64 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 latitudes. The members of the California Academy of Sciences Expedi- tion to the islands in 1922 found the vegetation dripping with early morning fog around the summit of Mount Evermann. These fogs ameliorate the spring aridity and help support the epiphytes of the mon- tane forest, and even the tree growth by reducing transpiration. The total known flora of this archipelago consists of only 120 vas- cular plants (Johnston 1931) ; Socorro 102 species, Clarion 43 species, San Benedicto 11 species, and is based upon about 325 numbers col- lected by five men. Table 3. All of these collections but one were made in the spring dry season. Barkelew only collected during the summer months (May 15 to July 8) and he apparently failed to exploit fully the opportunity to obtain representatives of the rich upper canyon flora. It is not known if the summer rains preceded him. I have not seen his specimens, which would indicate the condition of the vegetation at the time of his visit. As on the Mexican mainland, collecting on the islands should be best from August through November and the first botanist to engage the flora at this season will undoubtedly be richly rewarded with plants and information. Johnston (1931) has given good geo- graphic descriptions of the islands, and with his and Hanna's reports (1926), I have drawn up the following brief summaries of the islands. TABLE 3 Socorro Clarion San Benedicto "Vcsr Collector ■ " , ' ' ■ , ^ spring summer sprmg summer sprmg summer 10? Total known species 102 43 11 Table 3. Plant collectors and their collections from the Revilla Gigedo Islands. San Benedicto Island. 48 kilometers north of Socorro, is about 5 kilometers long, averages about 1 kilometer wide, and contains in the neighborhood of 5 square kilometers. The southern half is an ash cone about 296 meters high, the northern half a lava plateau. Physiographi- cally it is young and little developed with a poor diversification of habitats. The known flora consists of but 1 1 species of land plants. Clarion Island, the extreme western outpost, is roughly rectangu- lar, about 8 kilometers long, 3 kilometers wide, and 24 square kilo- meters in area. It is surrounded by an immature coral reef (Hanna, 1887 Stockton (Anthony) 41 25 1889 Townsend 19 12 1903 Barkelew 70 1925 Mason 83 44 1939 Elmore 11 15 NO. 2 GENTRY : LAND PLANTS 65 I.e. p. 36). It has three distinct hills, the eastern being the highest and reaching about 321 meters above sea level. "The hills are rounded by erosion and show few canyons of any size. Over most of the island there is a deep reddish brown soil that suggests antiquity. Undoubtedly erosion now takes place very slowly; the rainfall is slight but the island is w^ell covered with vegetation" (Johnston 1931 :25). The few habitats of two sand beaches on the south side, rocky sea cliffs, mesa, slopes, and hill tops are not indicative of a rich flora. There is no permanent source of fresh water. Hanna (1926:39) mentions a pool of brackish water at the east end behind a sand beach, which he thought might be potable during the rainy season. The island appeared to Hanna to have been very little altered by human interference. Such spots are so rare on the extra-polar portions of our little earth that his remarks on this point are quoted in full (1926:33). "No mammals of any kind were found on the island. Fortunately, the place has never been inhabited, even by temporary residents; hence those curses of the isles to the northward, mice, cats, and goats, have not become established. In fact, Clarion Island is one of the few places remaining which has not been modified in some way through the agency of man. The original 'balance of nature' still obtains. We know of only one case of the introduction of any kind of life. In 1903 the California Academy of Sciences sent an expedition to these islands and during the course of the work on Socorro Island some paroquets were captured alive. Mr. E. W. Gifford, a member of the expedition, told us that some of these birds were liberated on Clarion Island. We saw no sign of them during our stay and it is supposed that they perished through lack of fruit which constitutes their chief food on their native island." The natural vegetation is dominated by thorny shrub under 15 feet in stature. Hanna noted "One species of plant — , a shrub about 15 feet high and the nearest approach to a tree found" (1926:38). One of the most abundant shrubs was Opuntia occidentalis, which "grows very luxuriantly in a broad zone around the shore line and more or less in patches to the top of the island. Intertwined in it everywhere are dense growths of vines" (1926:32). For ingress to the interior it was neces- sary to cut trails through the dense spiny thickets. Other dominant thorn shrubs include Zantroxylon fagara, and Euphorbia anthonyi is probably the shrub, the fruits of which stained the parties' clothing. On the tops of the island they found large areas covered with grass. 66 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 The 43 species of vascular plants known from Clarion Island have been discussed in some detail by Johnston. The collection and notes obtained by Elm.ore on the Allan Hancock Pacific voyage to that island are listed below together with his collections from Socorro Island. As with other spring collections from the island, their condition reflects the dryness of that season. Socorro Island at about latitude 18°50' N and longitude 111° 00' W, stands as a rough quadrangular hulk (Plate 4, fig. 12). Fourteen kilometers in length by about 11 kilometers in width, and with about 154 square kilometers of area, it is the largest of the Revilla Gigedo group. Near the center is Mount Evermann, a little over 1138 meters in elevation, a volcanic structure which Hanna (1926:56) and party found mildly active in 1922. ''The whitish mud flows out in the side of the gulch and makes a marker which Is visible for a long distance — . The fissures are very active. We had no means of measuring the tem- perature of the steam Issuing therefrom but the rush due to the high pressure produced a great roar. Around the vents there was much crystalized sulphur and the odor of hydrogen sulphide was very ap- parent. The largest vents were about eight inches in diameter. In the upper part of the gulch some of the fumaroles contained water but this was found to be highly acid and entirely undrlnkable." The land form of the island is mountain broken up into ridges, slopes, and radiating canyons. A small dry lake exists in an eastern canyon. Recent and ancient lavas have contrasting soils and vegetations. Permanent water is available at sea level at Gra5^son Cove on the south side of the island, where a spring of fresh w^ater comes out of a crack in the lava rocks. Sheep were introduced on the island in 1896 and were still running freely about the island In considerable numbers in 1922. Grayson Intro- duced hogs at about this same time, but they have not been observed since and it is most likely that they have expired. One species of lizzard is known from Socorro and It is the home of numerous sea and land birds. The vegetation apparently consists of two types determined by the topography, soil, and the upland fogs; a maritime subtropical Thorn Forest on the lower slopes and ridges and a maritime tropical montane drought deciduous forest in the upper reaches of the canyons. Of the brushy Thorn Forest, the ornithologist Anthony wrote (Auk II, 15 :312. 1898) ''The greater part of the island is covered with a very dense growth of underbrush, the weather side [north and west exposure] being especially thickly covered, making travel, except in favored spots, NO. 2 GENTRY : LAND PLANTS 67 well nigh impossible. Trees are abundant on the weather side of the island but on the south and east sides they are mostly confined to can- yons, and were smaller than on the north slopes. They were nowhere seen over forty or fifty feet in height, though usually covering consider- able area with their broad spreading branches." Hanna and party found an upland section of red hills with flattened and denuded vegetation, which they attributed to a cataclysmic wash- out by a tropical cloud burst. On the south slopes of the mountain above Grayson Cove, they found grass, cactus, and some shrub adventive over the area that Grayson (1871:295) had fired 53 years earlier. The montane forest is found only in the upper reaches of the canyons — "In the canyon were many strange trees, flowers, epiphytic plants and orchids." Birds were excessively abundant and droves of sheep were met with here and there all the way. ''The forests in the canyons were so dense that the sunlight rarely penetrated to the ground ; hence mosses, lichens, ferns, and orchids, were abundant on the trees and branches. 'Trom the top we were able to study the best means of approaching the mountain and found it unquestionably to be from Grayson Cove. But that route does not pass through any such interesting country as we had traversed on the ascent. Wooded canyons are absent on the south side but are abundant on the north, east and west. Between them brush covered ridges radiate outward like spokes of a wheel" (Hanna, 1926:48, 54, 57). The flora of Socorro is the richest of the Revilla Gigedo Islands. The larger and higher area of more diversified terrain is accompanied by a more highly evolved indigenous flora. The five collectors who have visited the island have given us records of 102 species of vascular plants. Table 3. Barkelew's collection of 70 numbers is the only one made in the summer and he apparently failed to reach much of the higher richer flora of the interior. To Mason goes credit for first hav- ing brought the rich potentialities of the montane forest to our atten- tion by his collections. Late summer and fall collections would add more species and genera to the island flora. And, as Johnston wrote (1931 :15) "the most important botanical work now awaiting attention on the islands concerns not species so much as the vegetation and the living plant. The past collectors on the islands have been quite satisfied in making a single collection of each species found on each of the islands. No attempt has been made to make repeated collections either to show variation of the plants or their distribution on particular islands. There is almost nothing on record regarding the abundance, habits, stature, 68 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 habitats, associates, flower-color, etc., of the various plants of the islands. Few, if any, notes have been made which would permit the botanists who have not visited the islands to visualize the living plant and see it in relation to its environment. The plant ecology of the islands is an untouched subject." That is as true today as it was when Johnston wrote it. Elmore's collection of 11 numbers is annotated below along with those of Clarion. Catalogue of Collections polypodiaceae Cheilanthes peninsularis insularis Weatherby, Am. Fern Jour. 21:25. 1931. Braithwaite Bay, Socorro Island, March 18, Elmore C8, elev. 100 feet. This variety is endemic to the Revilla Gigedo Islands, where it has been collected from Socorro and Clarion. Elmore reports it growing in shaded crevices of lava rocks in the side of a canyon. Gramineae Cenchrus mysuroides H.B.K., Nov. Gen. & Sp. 1:115, t. 35. 1816. Braithwaite Bay, Socorro Island, March 18, Elmore CI, among boulders on upper beach. Widely distributed in the warmer parts of North America. JouvEA PiLOSA (Presl) Scribn., Bull. Torn Bot. Club 23 :143. 1913. Braithwaite Bay, Socorro Island, March 18, Elmore C4, among lava rocks on steep slopes, elev. 10 feet. Widely scattered along the coast from southern Baja California to Nicaragua. Sporobolus argutus (Nees) Kunth., Enum. PI. 1.215. 1833. Sulphur Bay, Clarion Island, March 16, Elmore B12, on edge of fresh water lagoon. From Kansas to Arizona and south through tropical America. MORACEAE Ficus cotinifolia H.B.K., Nov. Gen. & Sp. 2:49. 1817. Braithwaite Bay, Socorro Island, March 18, Elmore C6, elev. 20 feet. NO. 2 GENTRY : LAND PLANTS 69 In low and middle elevations from southern Baja California and central Sonora south into Central America, especially along canyon water courses. It is the most common and widespread fig tree in western Mexico. Elmore notes it as growing on the island ''in good red dry volcanic soil with aerial roots from overhead branches." Johnston ( 1931 ) reports that these broad large trees are a favorite haunt of the untended sheep so numerous on Socorro Island. He attributes this to their desire for shade. They may also be attracted by the edible fruits dropping upon the ground and which are a favorite source of food for both wild and domestic animals in Mexico. People eat them in times of hunger, but generally as a source of food they are not desirable. This island collection shows some minor differences in leaf vena- tion from that of typical mainland material, but due to lack of flores- cence it cannot be more critically compared. Its affinities are obviously close to that of F. cotinifolia, and in spite of the apparently long isola- tion of the Socorro Island population, it is unlikely that it would show more than varietal distinction. Ficus is an old genus, known from the Cretaceous. Nyctaginaceae BoERHAAViA CARiBEA Jacq., Obs. Bot. 4:5, t. 84. 1771. Braithwaite Bay, Socorro Island, March 18, Elmore C7y Cll, elev. 100 feet, in crevices of basaltic rocks where there was sufficient dis- integrated rock to form soil. A low viscid perennial herb with small dark purple flowers widely distributed in the American tropics. The specimens are fragmentary remains of plants indicating an earlier period of florescence. It normally flowers in the summer and fall. AlZOACEAE Sesuvium portulacastrum L., Syst. Nat. ed. 10, 2: 1058. 1753. Sulphur Bay, Clarion Island, March 16, Elmore BIO, clay soil around old lagoon, flowers pink. Widely distributed in brackish soils in the American tropics, but not abundant on the adjacent Mexican coast and not known from other Revilla Gigedo Islands. Leguminosae Phaseolus atropurpureus sericeus Gray, Proc. Am. Acad. Sci. 5:156. 1861. Sulphur Bay, Clarion Island, March 16, Elmore B2, many plants on upper beach, flowers dark red. 70 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 On coarse arid soils in southern Arizona and western Mexico. The densely sericeus variety is here recognized for the race inhabiting north- western Mexico. The species ranges from southwestern United States to Costa Rica. SOPHORA TOMENTOSA L., Sp. PI. 373. 1753. Sulphur Bay, Clarion Island, March 16, Elmore B4, upper beach. Although widely distributed in the tropical littoral of both hemi- spheres, it is not known from the Mexican mainland, so that its pres- ence on Clarion Island is a special problem in distribution. Native peoples make use of the plant for supposed or actual medicinal proper- ties, and it is possible that early seafaring people introduced the plant on Clarion Island. Zygophyllaceae Tribullus cistoides L., Sp. PL 387. 1753. Sulphur Bay, Clarion Island, March 16, Elmore Bl, beach sand. Widely distributed through the w^armer parts of America. Euphorbiaceae Euphorbia Anthonyi Brge., Erythea 7:7. 1898. Braithwaite Bay, Socorro Island, March 18, Elmore C9, a few plants growing in crevices of lava rocks in side of canyon wall. A low shrubby plant endemic to the Revilla Gigedo Islands," the type described from San Benedicto. In Standley's "Trees and Shrubs of Mexico" (C.N.H. 23:602) the t^^pe locality is given erroneously as San Benito Island. Dr. Wheeler, who determined the above collec- tions, states that it is atypical, perhaps juvenile. Euphorbia cf. californica Benth., Bot. Voy. Sulph. 49, pi. 23B. 1844. Sulphur Bay, Clarion Island, March 16, Elmore B8, rocky loam on gradual slope, flowers about one-quarter inch across, yellow green. The material is insufficient for certain identification. Typical E. cali- fornica has not been reported from these islands and is known only in the California Gulf Region. Euphorbia clarionensis Brge., Erythea 7:7. 1898 and Zoe 5:27. 1900. Sulphur Bay, Clarion Island, March 16, Elmore B6, upper beach in exposed dry sandy soil, flowers white, the white tips with red brown centers. Endemic to the Revilla Gigedo Islands. NO. 2 gentry: land plants 71 Sterculiaceae Waltheria aisiericana L., Sp. PI. 673. 1753. Sulphur Bay, Clarion Island, March 16, Elmore B9, among lava boulders on dry slope. Braithwaite Bay, Socorro Island, March 18, El- more C2, elev. 5 feet. Widely distributed in the warmer parts of the world, common through Mexico. A low perennial herb, often pol^-podial, 5-8 dm high, very aggressive on disturbed areas, and one of the most collected weeds of tropical America. Its presence on even the remote Revilla Gigedo Islands is not surprising. Elmore reports that only a few plants were observed. Cactaceae Opuntia occidentalis Engelm. & Big., Proc. Am. Acad. Sci. 3:291. 1856. Braithwaite Bay, Socorro Island, March 18, Elmore CIO, in ba- saltic canyon in between boulders where there is a little soil formed by disintegrating lava. Southwestern California, northern Baja California, and the adjacent islands along the maritime slopes; type from "western slopes of the California Mountains." Elmore's collection is the first identifiable material of the Platy- opuntia which has long been noted upon the slopes of Socorro. Although sterile, his specimens exhibit only minor differences from California plants, as reviewed by Dr. Clover, who made the determination. The presence of this species appears to be another bit of evidence supporting the theory of the dissemination of migrules by oceanic currents. The relationship of the Revilla Gigedo plants to the flora of California and means of dispersal has been discussed by Johnston (1931). COXVOLVULACEAE Ipomoea cathartica Poir., Encycl. Suppl. 4:633. 1816. Sulphur Bay, Clarion Island, March 16, Elmore B5, upper beach in exposed sandy soil, flowers blue or purple. Common to the American tropical lowlands; among the Revilla Gigedo it is known only from Clarion Island. BORAGINACEAE Heliotropium curassavicuini L., Sp. PI. 130. 1753. Sulphur Bay, Clarion Island, March 16, Elmore Bll, on edge of fresh water lagoon. Common on saline or alkaline soils throughout most of tropical and subtropical America. 72 allan hancock pacific expeditions vol. 13 Labiatae Teucrium Townsendii Vasey & Rose, Proc. U.S. Nat. Mus. 13: 146. 1890. Sulphur Bay, Clarion Island, March 16, Elmore B14, on 40° slope in lava rocks with a little soil between, flowers white. A low succulent appearing herb with crowded ovate leaves, remotely and irregularly crenate, thickish, inflorescence crowded at the ends of the branches. Known only from Clarion Island. Solanaceae NicoTiANA NESOPHiLA Jtn., Proc. Calif. Acad. Sci. IV, 20:93. 1931. Braithwaite Bay, Socorro Island, March 18, Elmore C3, in crevices of basaltic rocks where sufficient soil has formed. Known only from Socorro Island. The lower leaves of the Elmore specimens are undulate, tending to be lobed toward the base, not cren- ate as described for the species by Johnston, but otherwise agreeing. The flowers are reported as cream colored. COMPOSITAE Brickellia peninsularis amphithalassa Rob., Proc. Calif. Acad. Sci. IV, 20:93. 1931. Sulphur Bay, Clarion Island, March 16, Elmore B7. Known only from Clarion and Socorro Islands, this variety is a low brittle shrub with pale yellowish branches and prominent nodes. Mason (in Johnston 1931 : 100) reported it the most dominant cover on Clarion Island. Perityle socorrosensis Rose, Bot. Gaz. 15:118, t. 13, f. 9. 1890. Sulphur Bay, Clarion Island, March 16, Elmore BIS^ B3, growing in crevices of lava boulders, partially shaded on upper beach. Herbaceous, said to be perennial. Known only from Socorro, San Benedicto, and Clarion Islands. Elmore B3 is a more vigorous form with larger leaves and without ray flowers. ViGUiERA DELTOiDEA TowNSENDii Vasey & Rose, Proc. U.S. Nat. Mus. 13:148. 1890. Braithwaite Bay, Socorro Island, March 18, Elmore C5, growing among lava boulders in decomposed lava soil at 15 feet elev. Known only from Socorro Island. NO. 2 GENTRY : LAND PLANTS 73 LITERATURE CITED Grayson, A. J. 1871. Exploring expedition to the island of Socorro. Proc. Boston Soc. Nat. Hist. 14:287-302. Hanna, G. Dallas 1926. Expedition to the Revillagigedo Islands, Mexico, in 1925. Proc. Calif. Acad. Sci. IV, 15:1-113. Johnston, I. M. 1931. Flora of the Revillagigedo Islands. Proc. Calif. Acad. Sci. IV, 20:9-104. 74 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 TRES MARIAS ISLANDS The Tres Marias Islands are out on the steep edge of the con- tinental shelf about sixty to seventy miles off the mainland shore of Nayarit, opposite the port of San Bias. They appear to have been in existence since sometime in the Pliocene and in part much earlier. They are composed of several kinds of both igneous and sedimentary rocks. The Miocene and Pliocene marine formations on Maria Madre and the Mesozoic rocks on Maria Magdalena indicate local differential faulting. To Nelson (1899:9-11) the position of the islands upon the continental shelf and the similarity of the vertebrate animals to those of the main- land indicated a Quaternary land bridge. However, the seven species of endemic mammals, the 24 species and subspecies of endemic birds and an endemic reptile, as well as the 21 species of endemic plants known from the islands, strongly indicate a considerable period of insular de- velopment. A local insular biota was apparently well developed before the Quaternary land bridge, so that aggressive mainland adventives would have had to compete to establish themselves. Soundings show a rather narrow submarine ridge extending out from Punta Mita, Nayarit, toward the Tres Marias and in line with their axis. The deepest sounding along this line shows 70 fathoms (ca, 30 m), with greater depths to port and starboard. Glaciation is thought to have lowered sea levels from 80-100 m (Zeuner, 1945:248), which is suf- ficient to cause peninsulation in this case. Probability of the land bridge hypothesis is also dependent upon local epirogenic movements and a study of geology still in waiting. The climate of the Tres Marias is subhumid, tropical, maritime, and equable with a binary pattern of seasons; rainy summer, and arid spring. The maritime influence appears to ameliorate the long spring dry season. (Grayson 1871:267) *'In the dry season heavy dews are frequent, the drops of which I have often seen the birds sipping, for want of other means of quenching their thirst, there being but few ojas de agua (springs)." Summer rains are convectional in type, the "chubasco" storms of wind-driven rain being common and making navi- gation for small boats hazardous. Hurricanes in the fall rarely swing in across the islands from the southwest (Schiaffino, 1939). No rain- fall data are available for the islands but precipitation should be ap- proximately that of San Bias on the mainland, which has an annual average of 58.5 inches. The annual average temperature at San Bias is 24° C, with no record of frost (1939). The natural vegetation of the islands is a subhumid tropical drought NO. 2 GENTRY : LAND PLANTS 75 deciduous forest with a multiple dominance of tree species. One of the most conspicuous dominants originally was Spanish cedar, Cedrela^ which oddly enough, although still living upon the islands, has never been collected. The first naturalist to see these forests was Grayson in 1865, 1866, 1867, who wrote of them (1871:264). ''The immense cedar (Cedrela odorata) grows in great abundance on this island (Maria Magdalena), not having been disturbed by the wood cutters. This tree makes the finest lumber in the world. It is also common on the coast of tierra caliente. Cleofa, the smallest of the three islands, is also well wooded and has a good little port. All these islands, except Juanito, are covered with a dense forest from the water's edge to the top of the highest hills. The shape of the trees (of which there is a great variety), is generally straight or straighter and taller than upon the main. There is but little thorny underbrush, so characteristic of the tierra caliente." Cutting of Cedrela was in progress on Maria Madre at the time of Grayson's visits. Later a penal colony was established on Maria Madre by the Mexican government and disturbance of the native vege- tation has continued. Ferris in 1925 (1927:64) found the area about the penal colony cut over and a weedy aggressive cover advancing, many species of which appeared to be newcomers from the mainland and characteristic of disturbed areas about Mazatlan and San Bias. Little agriculture has been practiced on this island. The islands are reputed to have been uninhabited by man until the coming of Europeans, the first of whom to inhabit the islands were buccaneers (cf. Dampier, 1703). If it is true that the Amerindian never inhabited the islands, it is very remarkable indeed, for traces of man indicate that he has over-run almost every square mile of North America at some time during his long residence on this continent. Sixty miles of sea water is a small barrier and although the Spaniards may have found the islands uninhabited, it is still likely that some of the prehistoric peoples who inhabited the adjacent mamland, and some of whom had advanced cultures, must have known of, visited, and even inhabited the Tres Marias for some periods. Maria Magdalena has apparently been only partially disturbed by man, and should this frag- ment of primeval vegetation (never affected by early man) still exist, it would be a singular boone to all students of natural vegetation. The known flora of the islands consists of 11 species of ferns and 313 spermatophytes. These have been catalogued in 3 separate papers. Rose (1899), Ferris (1925), Eastwood (1929). Over 90% of these 76 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 are known only from Maria Madre. In addition to a species of Cedrela, a giant Agave, and a native cotton mentioned by authors and still un- identified, are many other plants still awaiting collection and study. Altogether these islands offer a most interesting subject for field research. The job calls for a botanist of ethnographic and geologic background in insular residence throughout the summer and fall, at least. San Juanito IsIand.The Tres Marias Islands are actually four. The odd one is San Juanito, the northwesternmost of the group, lying across a narrow channel from Maria Madre. It is the smallest, being about 5 kilometers long, 3 kilometers wide, 15 square kilometers in area, and about 300 meters in elevation. Nelson and Goldman are apparently the only biologists who have visited it and the former gives the follow- ing short description (1899:10, 12), "San Juanito which is nearly flat with a narrow border of low bluffs along the northern shore — . On San Juanito the vegetation is largely made up of bushes and scrubby trees 8 to 15 feet high, with many Agaves on the sandy southern end. Agaves are very numerous also on the northern end of Maria Madre." There are no known plant collections and it apparently has never been visited by botanists, the accounts of other expeditions do not mention any land- ings there. Maria Madre Island is the largest of the Tres Marias group, 21°35' N. by 106°40' W. It is approximately 20 kilometers long, 10 kiolmeters wide, and 200 square kilometers in area. The peak of the island is 616 meters above sea level. The land forms consist of narrow beaches, cliffs, canyons, mesas, hill slope and ridge. Arroyo Hondo is a notable intermittent water course originating near the central peak and descending northwestward through a deep canyon to discharge water into the sea during the rainy season. In the past it has been difficult of access by land. Hanna (1926:67-71) found the island to consist of a central pediment of granite with Plio- cene marine sediments of chert, limestone, and sandstone lapped upon it to near the top of the peak. ''During a portion of Pliocene time large coral reefs existed around this old land mass and large blocks of the fossiliferous material, firmly cemented, having fallen from the exposures and have rolled indiscriminately far out into the forest." The island appears to have had its inception in Miocene times, if not before, and grew in area during the Pliocene. Maximum size was probably attained during one of the glacial periods when sea levels were universally lower. What is known of the vegetation has been described above. As with nearly all of the Eastern Pacific islands, very little attention has been NO. 2 gentry: land plants 77 paid to the life histories and the ecology of the plants that live upon them. We are still ignorant of the community make-ups and the inter- relationships of the biota. Certainly the terrain is sufficiently diversified to support various associations, although the elevation is not spectacular and the present island area is not an ancient one. The area of the island, its tropical nature, the observations of visitors, and the lack of summer and winter collecting all indicate that the known flora of about 300 plants is probably little over half of what actually exists. Many of the plants listed, due to inadequate material, have been determined to genus only, others are tentative. The Euphorbiaceous genus, Calaenodendron, appears to be endemic. Should more be found, we would have to revise our opinion regarding the age of the island, which seems to have been little more than a large rock until some time in the Pliocene. Until the amount and nature of endemism and the identity of most of the species are known, all inferences regarding the development and relations of the flora must be very tentative. There is no group of islands in the Eastern Pacific more worthy of thorough investigation. Maria Magdalena Island lies about 11 kilometers across chan- nel from Maria Madre. It is approximately 150 square kilometers in area, 16 kilometers long, 10 kilometers wide, and with a central peak of about 450 meters elevation. Physiographic habitats include beaches, sea cliffs, canyons, rocky slopes and ridges. Hanna (1926:73) mentions various canyons and a water hole in the next canyon west of the one in which they were camped, near the center of the north side of the island. They were there in May, when the dry season is well advanced, so that the water must be nearly or actually perma- nent. Concerning the geology he wrote (1926:72) "Maria Magdalena has had an entirely different history from Maria Madre. Basement rocks are volcanic and are overlain by a great series of cherts, sand- stones, and mud shales. These we took to be Cretaceous in age but definite paleontologic proof was not found. Miocene appeared to be absent and Pliocene was not positively identified. Pleistocene, however, is exposed near the sea and on the beach at the creek mouth and the flat eastern end of the island is probably an elevated terrace of this age. The dangerous reefs projecting from the north side of the island are com- posed of resistant layers of the supposed Cretaceous rocks, the softer shale layers having been eroded away. Many of these resistant layers weather out as huge flagstones. The high western end of the island, the Pacific side, with its enormous sea cliffs, is composed of highly altered 78 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 cherts with volcanic rocks in many places. No evidence of granite, such as compose the central core of Maria Madre Island was found." In 1865-67 Grayson found Maria Madre "unoccupied and covered with a grand forest of fine timber" (1871 :264). Nelson in 1897 found the vegetation similar to that on Maria Madre. He stated that most of the Spanish cedar was gone, but that a large percentage of the origi- nal forest remained intact. Hanna and party "all agreed that the fauna and flora of Maria Madre and Maria Magdalena were almost identi- cal." This is what one would expect of the vegetation, but not of the flora, if Hanna's supposition regarding the geology of the two islands is correct. If Miocene and Pliocene rocks are absent on Magdalena and present on Madre, one would infer that the former is the older, larger area and its flora with a consequent difference in speciation of related groups, which might in part be represented on the younger Maria Madre. The known flora of Maria Magdalena consists of only 31 species of vascular plants collected principally by Nelson in 1897 and Mason (35 numbers) in 1925. Francis H. Elmore made a small collection in May of 1939 during the Allan Hancock Pacific voyage. They are enumerated below, Jacquinea aurantiaca being a new addition to the island flora. CATALOGUE OF COLLECTIONS Polygon AC EAE CoccoLOBA ScHiEDEANA Lindau, Bot. Jahrb. Engler 13:187. 1890. Magdalena Island, May 9, Elmore lB3y a few plants in boulders and sand along a dry stream at 15 feet elev. Apparently along both coasts of Mexico from Sinaloa and Vera Cruz south to Guatemala; type from Papantla, Vera Cruz. Theophrastaceae Jacquinea aurantiaca Ait., Hort. Kew. ed. 2, 2:6. 1811. Magdalena Island, May 9, Elmore IBl, dry rocky partially shaded stream bank, elev. 15 feet. From Sinaloa to southern Mexico, Central America, and the West Indies; also on Maria Madre Island. Verbenaceae Avicennia nitida Jacq., Enum. PL Carib. 25. 1760. Magdalena Island, May 9, Elmore 1B2, dry sandy soil of the upper beach. Widespread along the coasts of tropical America. NO. 2 GENTRY : LAND PLANTS 79 Maria Gleofa Island. The last and southern-most of the group is Maria Cleofa. In dimensions this island is approximately 6 kilometers long, 5 kilometers wide with an area of about 25 square kilometers. The one central peak is given as 402 meters in elevation. Except for the brief remarks of Grayson and Nelson, the only naturalists who have visited the island, there is little known of the island. Nelson noted that canyons descend from the central peak in all directions. At least one of them carries an intermittent stream, which sinks in its bed before reaching the sea during the dry season. In com- parison to the other islands Nelson states (1899:12), *'Maria Cleofa is more rocky and sterile, and the trees are bushy and stunted." The total known flora of land plants consists of Nelson's collection of these four species: Zamia loddigesii (?), Arundo donax, Cyperus ligularisy and Trixis Wrightii, TABLE 4 Year Spring Summer Fall Winter 221 Maria Madre Island 1897 E. W. Nelson 95 1925 H. L. Mason 128 1925 R. S. Ferris Maria Magdalena Island 1897 E. W. Nelson 14 1925 H. L. Mason 35 1939 F. H. Elmore 3 Maria Cleofa Island 1897 E. W. Nelson Totals 279 221 Table 4. Plant collectors and their respective numbers collected on the Tres Marias Islands. 80 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 LITERATURE CITED Atlas Climatologico de Mexico 1939. Seer. Agr. y Fom., Mexico, D. F. Dampier, William 1703. A new voyage around the world. 5th ed. Eastwood, Alice 1929. Studies in the flora of Lower California and adjacent islands. Proc. Calif. Acad. Sci. IV, 18:442-468. Ferris, Roxana S. 1925. Preliminary report on the flora of the Tres Marias Islands. Contr. Dud. Herb. Stanford Univ. 1 :65-90. Grayson, A. J. 1871. On the physical geography and natural history of the islands of Tres Marias, etc. Proc. Boston Soc. Nat. Hist. 14:261-302. Hanna, G. Dallas 1926. Expedition to the Revillagigedo Islands, Mexico, in 1925. Proc. Calif. Acad. Sci. IV, 15:66-76. Nelson, E. W. 1899. General description of Tres Marias Islands. No. Am. Fauna 14:7-13. Rose, J. N. 1899. Plants of the Tres Marias Islands. No. Ara, Fauna 14:77-91. Zeuner, F. E. 1945. The Pleistocene period. Ray Society, Pub., pp. 1-322. NO. 2 gentry: land plants 81 CALIFORNIA GULF REGION General Physiography The California Gulf Region possesses considerable physiographic unity, although its boundaries are not fast. For the purpose of this study it includes, besides the great gulf nucleus itself and its chromosomic islands, the surrounding coastal plains, mostly narrow and in part lack- ing, the coastal mountains, and the Cape District of the peninsula. It forms a long narrow area nearly 1000 miles long containing about 150,000-175,000 square miles. It is nearly co-extensive with the Sonoran Desert, the latter comprising additional area in northern Sonora, south- eastern California, southwestern Arizona, and on the peninsula except in the Cape District. The western middle portion of the peninsula is being considered in another study. Nearly half the area is occupied by the sea. Its western shores are generally precipitous and without rivers, while its eastern are mostly low coastal shores with several intermittent rivers. From the mouth of the Colorado River in the apex, the gulf gradually deepens to 2600 meters under its 225 kilometer-wide mouth. The upper part of the gulf is generally less than 300 meters in depth with low gradients on the eastern and northern margins. South of Angel de la Guardia and Tiburon Islands it rapidly deepens to over 1000 meters and the 1500 meter contour comes well up into the gulf. In this water are 25 to 30 larger islands and many smaller ones, some of which are no more than jagged rocks set upon by the tides and wind-driven waves. The topo- graphy of the northern half is terrestrial in type rather than marine. It is the water relations, both of the sea and of the air, that give the country its distinctive quality. The gulf water is changeful. It can be as quiet as a forest pool at dawn or as choppy as Lake Erie. High seas often run under the strong and recurrent winds. In the late sum- mer of the convectional storm cycle, hard squalls suddenly appear and have upset many a light ship, and in the early colonial days made the passage of the gulf and the colonization of the peninsula a hazardous undertaking. The waters teem with life ; fish, shrimp, whale, and many other animals run in and out in their seasons, as also do a vast assem- blage of birds over the surface; others, from plankton and crabs to sharks and turtles, are permanent residents. The sublittoral zone is covered with an abundance of varied algae. Under the brilliant sun- light, the ever mobile waters, blue or green or vermilion or gray, are set sharp against the rocky, uneven, and disconsolate shores. The water of the air is rare, because it seldom falls as rain, although 82 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 actually the hot air contains more moisture than usually exists over temperate regions. Hence, in contrast to the dense biota in the sea, the land is relatively barren. The land biota, though seemingly dead, is nevertheless there, and with many special adaptations for survival of drought. When rains fall, the plant life suddenly becomes intensely active and the land becomes green and flower-colored. To the man who may be there the desolate then becomes a garden of well-spaced forms. And it, too, in brilliant sunshine is set sharp against the inanimate rocks beside the waters of the gulf. Such is the impression of the gulf region on the senses today. Geologically, the perception has quite a different quality, because the eye and mind are removed from the object by many milleniums. In spite of changeful orogeny, the California Gulf Region has had a persistent character since the early Mesozoic. The gulf itself is a de- pressed block, which Schuchert (1935), has discussed as a part of the southern Pacific geosyncline. This sea-invaded trough has had a strik- ing physiographic evolution, the remarkable events and character of which are geologically revealed at every great turn. The biota of the lands has had a restless place and has endured displacements, inunda- tions, extinctions, and has been forced into migrations with the coming and going of the sea, with the submergence or emergence of mountains, and with the concomitant changes of local climate. Close study of the plant and animal life, when directed by a correlating intelligence, should reveal a course of evolution, expressed jointly by plant and rock, hardly equaled in plant geography. There is some lack of agreement among geologists regarding the history of the gulf region, primarily because investigations are still in preliminary stages. From the reconnaissance work that has been done, however, the general history can be outlined. Schuchert has been foremost in synthesizing available knowledge and I have drawn heavily upon his great work in the following paragraphs. Judging from the world-wide deposits of fossil plants, the modern angiosperms had their inception in the Upper Mesozoic, following the biota dominated by such groups as the seed ferns, the cycads, and the saurians. When the Tertiary opened, the majority of the modern plant families and genera were pretty well established. Hence, in seeking the origin of the desert flora of the gulf region we are concerned with events in the Cretaceous and onward. Of all the major floral elements, the desert floras are the most obscure in origin. They are almost un- represented in the fossil record. We must seek other sources of evidence for determining their origin, their migrations, and their growth. A NO. 2 GENTRY : LAND PLANTS 83 knowledge of the evolution of land forms is primary. The withdrawal of the Mesozoic seas from the Mexican continent was accompanied by widespread uplift. Large areas became land. They included not only what is now the Gulf of California, but also a margi- nal area. During the Cretaceous and early Tertiary the shore lines of Baja California and northwestern Mexico reached 50 or more miles westward of present limits, including some of the islands. This land mass might even have included the Channel Islands on the northwest, Guadelupe on the west, and the Revilla Gigedos to the southwest, al- though there is no real geologic evidence to support such a presumption (cf. Jtn. 1931). However, irrespective of the exact boundaries, it is evident that during the Cretaceous and early Tertiary, Baja California and its gulf were a part of a land mass areally different than exists today. The progenitors of our modern flora had a broad base for develop- ment in arid latitudes; arid because the land lay in a rain shadow of the Mexican continent, barred from the trade winds, and because the weak westerlies blowing onshore are warmed and dried by the radia- tion of these latitudes. The climate of the Eocene appears to have been somewhat wetter than the late Mesozoic, which, because of the extensive sandstone and gypsiferous deposits, is judged to have been relatively arid. However, the granite batholithic and pyroclastic intrusions of the late Cretaceous and Eocene must have raised some mountains, which in turn localized climate abetting drought on the one side and decreasing it on the other, according to the mountain orientation to air flow. However, the interiors of the western land areas have apparently been relatively arid since the middle Mesozoic. Since then, if not before, there have been deserts, though the boundaries of them have been modified or shifted according to climatic cycles and to the raising and lowering of land masses. So, at the beginning of the Tertiary in the California Gulf Region, the environment was already diversified and the evolution of modern seed plants well begun. In the Oligocene the sea began to invade the old downwarped block of the Southern Pacific Geosyncline for the first time since the Triassic. As the bend deepened the sea invaded the gulf and by the middle Mio- cene reached half way up the present gulf to about Angel de la Guardia Island and Tiburon Island. By middle or late Miocene it was maximal and covered all of the present gulf, the Colorado Desert, and a section of adjacent Arizona and northwestern Sonora besides. Some of the gulf islands apparently date from that time and still persist as the tops of 84 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 submerged mountains, i.e., Angel de la Guardia, San Lorenzo and Ceralbo. Others, as Tortuga may not have appeared until the marked diastrophism of the Upper Pliocene and Pleistocene, when the peninsula gained its present elevation and general outline. San Jose, Carmen, San Marcos, and other islands close along the peninsular shore, may be fault splinters, and their developments intimately related to the dynamics of this compensatory zone. Too little is known geologically to time events in detail. It was not until the Pleistocene that the modern peninsula arrived. During the late Miocene and the early Pliocene most of the modern southern half of the peninsula was covered by sea, judging from the sedimentary beds. The marine formation described by Darton as "the yellow beds" (Jour. Geol. 24:720-748. 1921), is particularly significant. The beds appear to have covered all but the several old central igneous masses along the Sierra Giganta axis south of San Ignacio and a couple of western outposts. Hence it would appear that for a portion of the Upper Tertiary, the peninsula south of Sierra Calmalli was represented only by a series of islands, where Tertiary pediments crested the invad- ing sea, and which are still represented by locally exposed schists and granites about the bases of modern mountains (Darton I.e. fig. 3). The more important of such postinsular and prepeninsular masses appear to be represented by the Sierra Vizcaino, the Magdalena Island area. Sierra Zacatecas, a segment of the Sierra Giganta about Cerro Giganta east of Comondu, and the Cape District. A shallow portal across the mid-peninsula about the latitude of San Ignacio appears to be clearly defined and to have been contemporaneous with the yellow beds. As the Giganta fault subsequently became active, the modern pen- insula grew southward by anticlinal uplift and accompanying pyroclas- tics, which now form the greater part of the higher land. Today we see that the whole eastern side of the peninsula from San Ignacio south was tilted upward, the great Giganta scarp reared, the Pliocene waters retreated from the southern and western borders and from the trans- peninsular portals. The mountains of the Cape District and the Sierra Vizcaino complex were tied onto the peninsula. This was not accom- plished as one gradual sea recession, but according to local orogeny and to the eustatic periods of the Pleistocene. Hence such lowlands as the Vizcaino Depression were alternately opened and closed. The expanse of sand in remotely serried dunes appears to have been formed by suc- cessive beaches. In summary, there is little or nothing in the stratigraphy to indicate that the southern half of the peninsula was anything more NO. 2 GENTRY : LAND PLANTS 85 than an archipelago during a considerable portion of the Upper Ter- tiary. The evolution of the biota, the distributions and speciations, are linked with the physiography. It was inevitable that the evolution of the fauna and flora developed synchronously with the radical physiographic stages, but even though we have both spacial and temporal yardsticks, we have yet to measure these events in terms of plant development and to make specific correlations between organic and inorganic evolution. Plant populations were re- peatedly restricted or provided with new areas and divergent habitats. This effected lines of descent with swamping, with infrequent crossing opportunities, with new placements for variants and chance natural selec- tions, and also with entire eliminations. However, it appears that plant evolution can be evaluated in accordance with rates of divergence in many groups, in so far as isolation has fostered it, or as migration has interrupted it, and as speciation has expressed it. The California Gulf Region was split by the sea invasion, creating disjunct populations on the peninsula, the mainland, and the islands. Specific divergence due to isolation should be greatest about the mouth of the gulf because the sea invaded that end first and the distances sepa- rating mainland and peninsula populations are greatest. The Cape Dis- trict appears to have been isolated, except for its Quaternary union with the peninsula, since early Tertiary times and its high ratio of en- demism is to be expected and correlates nicely with a tempo-spacial yardstick. Endemism in the disjunct segments of the upper part of the gulf dates generally from the Upper Miocene or Lower Pliocene. Cor- respondingly speciation is not so clearly developed and we encounter difficulties in separating the entities to our taxonomic satisfaction. The Pleistocene disjuncts are even less mature and here we often engage the aggravating problem of choosing between species, subspecies, or varieties. Nevertheless we have to deal with them, since they represent stages in a natural rate of evolution. Were it not for the irrepressible tendency of life to vary, the units of life would be fixed and we would be denied many fascinating problems. By a study of the events that demark the periods of California Gulf history in relation to the evolving organisms, the origins of our desert flora should be less obscure. The Postinsular Localities The Cape District, consisting primarily of a granitic batholithic block and volcanic intrusives, was not until recently a continuous part of the peninsula, as has been stated above. Westward and southward of La Paz are a series of fossiliferous beach deposits underlaid by marine 86 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 deposited arkose. In this locality also, trending north and south, may be seen a southern segment of the Sierra GIganta fault, which Itself appears to be a continuation of the San Leandro fault zone, so well known In '*Alta California." The sediments of this locality show that this low section of the peninsula did not arrive above the sea until the Pleistocene. The adjacent part of the peninsula, the Sierra GIganta anticline, is composed largely of Pliocene marine sediments. This all Indicates that the block of Igneous rock In the Cape District had a long existence In the Tertiary as an island. This Is significant in the con- sideration of the biota and especially the flora. Many plants are known only from the Cape District and the adjacent peninsula, to which latter area they have migrated In Quaternary times. I propose to call such land bodies as the Cape District, postinsular. Besides the Cape District there are many coastal "cerros" (a word which the Mexicans apply to their craggy hills), whose positions and the adjacent landward strata indicate are also postinsular mountains. It Is not within the scope of this paper to present detailed evidence for this Interpretation of the physiography, but the facts : that post-Pliocene detrital and alluvial materials from the Sierra Madre Occidental have built up the coastal plains from 100 to 500 feet; that the Pacific coast is rising; that higher sea levels existed universally in the Interglaclal periods; and that several Quaternary estuarlne deposits and sea caves exist several miles Inland, all provide excellent grounds for this theory. More detailed evidence accumulated during my several years of paleon- tologlcal reconnaissance In the area, will, I hope, appear in a later study. The larger coastal mountains which may be considered postinsular are; in Sonora, Sierra Coloral, Sierra Serl, the monadnock north and west of Guaymas, probably Sierra Bacatete, Sierra Bojihuaqueme, and in Sinaloa, Sierra Navachlste, Sierra Tecomate, and possibly Sierra Ta- culchamona. The floras of these postinsular localities have not been Individually studied. On most of them few if any collections have been made. More collections have been made on the Guaymas monadnock than on any of the others, but they are vv^Idely scattered In herbaria and not available for detailed study. Nor do they represent all of the postinsular land body lying northwest of Guaymas. Beginning with the collections of Thomas Coulter in 1829-30, the plants taken at Guaj^mas and vicinity have disclosed a surprising number of novelties, some of which presum- ably had their specific origin on the Guaymas monadnock. The post- insular Sierra Coloral has not been studied by botanists. McDougal, NO. 2 GENTRY : LAND PLANTS 87 Shreve, Gentry, and perhaps a few others have made some fleeting col- lections on the north end. No collections are known from Sierra Seri; except a few Mammillaria, there are none from Sierra Bacatete and Sierra Bojihuaqueme; none from Sierra de Navachiste except Edward Palmer's collections at the harbor of Topolobampo ; from Cerro Teco- mate, Gentry 100 numbers; from Sierra Tacuichamona, Gentry 192 numbers. These postinsular floras should show differences from that of the mainland, more or less correlated with the ages of the respective cerros or with the duration of their islandic isolations. Some of them may show considerable endemism, if not of species rank, then of lesser degree. The rare plants that are known from only one or two restricted areas along the Mexican west coast may be postinsular endemics that have persisted or migrated locally since their original habitats joined the mainland. Some of the cases that attract the attention at this point are the fol- lowing : Adelia obovata Wiggins & Rollins Jatropha purpurea Rose Desmodium Wigginsii Schubert Lippia P aimer i Wats. Phrygilanthus sonorae (Wats.) Standi. Caesalpinia gracilis Benth. Karwinskia latifolia Rose Indigofera laevis Rydb. Ruellia leucantha postinsularis Gentry Physalis purpurea Wiggins Sesbania sonorae Rydb. Physalis sonorensis Standi. Prosopis reticulata Wats. Aloysia nahuire Gentry & Mold. Holographis pallida Leonard & Gentry Porophylluf?! pausodynum Rob. & Greenm. Each of these postinsular localities possesses its own floristic problems in relation to the land flora. A detailed knowledge of their floras would provide considerable evidence regarding their geologic histories and the rate of evolution, in so far as species divergence can be chronologized with physiographic developments. While it may be possible to determine which plants have insular origin, it will be harder in many cases to ascertain what species are of mainland origin and whose distributions now represent migrations upon postinsular mountains. All such problems 88 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 offer a fascinating invitation to the student of plant evolution and geography and their solution would contribute to the great latent story of the California Gulf Region. Climate The California Gulf Region in spite of its large water surface, is extremely arid around its upper portion. In so far as the continental air circulation dominates, its climate is continental in type. Theoretically, the prevailing air currents are those of (1) the western Pacific anti- cyclone, the maritime character of which is eliminated by its passage over the west coast land surface northwest of the gulf region; (2) the continental anticyclone, dry, and with marked annual and daily tem- perature extremes; (3) the mild Pacific westerlies prevailing through the spring as on-shore breezes are consistently dry because of the warmer land surface. The conspicuous fog desert along the outer coast of the peninsula is lacking in the gulf region proper. Near the mouth of the gulf epiphytes occur locally on the mainland, as at Cerro Tecomate. (4) A maritime air current appears to swing in from the near south Pacific as a western limb of the Caribbean anticyclone. In the fall, storms develop along this track (Schiaffino, 1939) and may reach into the gulf. However, unstable humid air masses are of two sources and two seasons; the winter rains from the northwest Pacific storms which are uncertain and do not fall every year; the summer rains of the tropical convections, which occur regularly in the southern portion of the region, but lighten materially in the interior of the gulf. Averages of yearly rainfalls, as far as regional records are available and vegetative growth indicate, range from 20 to 30 inches for the mountains on either side of the mouth of the gulf to about 3 inches for the lowlands in the upper portion of the gulf. Data are carried in the accompanying table (Table 5). Average winter rainfall is similar for all stations except the high of Alamos in southern Sonora, which is explained by the proximity of that station to Sierra de Alamos (height 1800 m) and its obvious precipitative effect on humid air masses. The same phenomenon is operative through- out the region on other mountains having comparable mass and height. Average sum.mer rainfall generally decreases from south to north. The average for the five southern stations (Muleje, La Paz, Guaymas, Alamos, Topolobampo) is 9.5 inches, while for the northern stations of Brawley, Lechuguilla, Tule Tank, Cirio Point, and Libertad, it is only 1.6 inches. This greater summer rainfall in the southern part of the NO. 2 GENTRY I LAND PLANTS 89 California Gulf Region is most responsible among the climatic factors for the tropical elements in the southern flora. It fosters the existence of many plants requiring or responding only to summer rains, when higher temperatures, longer daylight periods, periods of shortening day- light (fall), higher humidities, etc., prevail. The lack of synchrony of these factors in the northern area excludes many subtropical plants. TABLE 5 p, ^ Length Average Average Average Station r.^* of record summer winter annual (years) rainfall rainfall rainfall Brawley, California -100 29 0.63 2.01 2.64 Mexican, Baja California 3 15 1.0 2.0 3.0 Muleje, Baja California 110 15 2.7 1.3 4.0 La Paz, Baja California 60 15 4.4 2.7 7.0 Lechuguilla, Arizona 700 5 1.7 2.6 4.3 Tule Tank, Arizona 1100 9 2.2 1.9 4.1 Cirio Point (Sierra Coloral), Sonora 180 10 2.1 1.5 3.6 Libertad, Sonora 100 10 2.1 1.8 3.7 Hermosillo, Sonora 700 15 10.1 2.1 12.2 Guaymas, Sonora 13 15 7.6 2.5 10.1 Alamos, Sonora 1200 15 22.1 6.2 28.3 Topolobampo, Sinaloa 10 15 11.0 3.2 14.2 Table 5: Average rainfall in inches for some representative stations in the Cali- fornia Gulf Region. Data extracted from Turnage and Mallery (1941) and Atlas Climatologico de Mexico (1939). COLLECTION LOCALITIES The Cape District since its beginnings in the Cretaceous (cf. Schu- chert 1935:132-133), apparently has had a relatively long and stable geologic history as an island. Due to its age, its high elevations, and its tropical and wetter climate, it has the best developed vegetation and the richest flora of any other part of the gulf region of comparable size. The complex physiography and variety of rock also contribute to its floral richness by providing a wide range of habitats. Present are val- leys, hillsides, cliffs, mountain tops, sandy beaches, detrital slopes, springs, short intermittent streams, and on top of Sierra Laguna a small meadow- like basin holds shallow water during the summer rains. The arroyo beds, although with steep gradients, carry deep lenses of granitic sands and support a rich assemblage of mesophytic trees and shrubs. The principal vegetation formations are: Desert Shrub, Thorn Forest, Short-tree Forest (a subtropical, mixed, drought-deciduous for- 90 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 est), and Pine-oak Forest. Grasslands are lacking except for a few small mountain meadow-lands, as *'La Laguna" on Sierra Laguna and an obvious Increase in grasses in the Pine-oak Forest generally. Thorn Forest is not as well developed as It is on the adjacent mainland, being but a brief transitional element between Desert Shrub and the more prevalent Short-tree Forest of the slopes and canyons which comprise the greater part of the area. Except for the minor role of Thorn Forest and the lack of a well defined Oak Grassland, the vegetations show a comparable alignment to that of southern Sonora. The variance may be attributable to the islandic origin. Views of Cape District vegetation are shown In Plate 6. The flora of the Cape District is more tropical than temperate. Brandegee's floristic survey of the eighteen nineties (1891, 1892, 1894, 1901), listed 104 species of Leguminosae, 103 species of Composltae, and 52 species of Gramineae. While this is incomplete, he did collect generally through all elevations and his collections appear fairly repre- sentative of these three major groups of angiosperms. The respective ratios of these groups per given area are often instructive about floristic relations. The light showing of grasses is comparable to that found In Thorn Forest and Short-tree Forest on the adjacent mainland. Generally, the Composltae are more numerous through temperate regions, while Leguminosae dominate the floras of tropical regions. The fact that le- gumes equal or exceed the composites in numbers of species in the Cape District attests Its tropical affinities. The number of genera common to the cape and the adjacent mainland is far greater than those common to the cape and the northern part of the gulf region, as may be expected. In North America the following genera of the Leguminosae in the Cape District have their centers of area in southern Mexico and Central America. Acacia, Mimosa, Pithecolobium, Lysiloma, Desmanihus, Albiz- zia, Caesalpinia, Cassia, Bauhiniaj Indigofera, Tephrosia, Coursetia, Benthamantha, Sesbania, Nissolia, Aeschynomene, Erythrina, Galactia, Phaseolus, Haematoxylon, and Leucaena. This list is Incomplete, but much longer than we would cite for those having northern centers of areas, as Astragalus, Lupinus, Lotus, and Trifolium, and which are not strongly represented In species in the Cape District. Three endemic genera {Coulter ella, Clevelandia, and Faxonia) are known from the Cape District, and over 100 endemic species and varieties have been described (Interpreting their occurrences along the Sierra GIganta as postinsular migrations). Actually, except for the flora of the Desert Shrub formation, the NO. 2 gentry: land plants 91 flora of the Cape District is still in isolation. The climatic and edaphic conditions of the peninsula, particularly north of the Sierra Giganta, are not tolerable to the majority of the cape plants. Comparable habitats on the mainland are still over 100 miles of water and coast away. Equally significant to differences in the make-up of the cape flora and the adjacent mainland are the numbers of species the two areas have in common. Considering the long tenure of isolation for the cape flora and the vicissitudes besetting diaspores across salt water, a higher endem- ism could reasonably be predicted. It may be that the Cape District was bridged to the peninsula during the middle Tertiary, allowing in- gress of the aggresive Sinaloa element, which displaced some of the insular population. However, whether cape Isolation dates from early or late Tertiary, it is clear that genera have generally been conserva- tive in species generation. The genesis of species does not appear to have progressed nearly as rapidly in land plants as it has in land mammals (cf. Zeuner, 1945:253-269, and 1946). In April of 1937 and in February of 1938 the Velero III made stops in the Cape District. P. J. Rempel and E. Yale Dawson made collections of land plants at San Jose del Cabo, and at Punta Frailes and vicinity (Tables 1, 6). Their collections are enumerated below in the catalogue of species for the California Gulf Region. Punta Frailes is among the least known of the localities visited by the Velero III, the collections from there being the first. Punta Frailes is a granite cerro that juts out into the mouth of the gulf on the southeastern tip of the peninsula (Plate 5, fig. 13 ; pi. 6, figs. 14, 15). Southwestward of the cerro is a sandy beach, where members of the expedition landed and made collections. For causes which are still ob- scure, this southeastern tip of the peninsula is the driest portion of the Cape District. North of Punta Frailes is a mountain, known by the natives as Sierra Victoria. This has been confounded by cartographers (who have followed an early error) with Sierra Laguna which is the main central mountain mass of the Cape District. There are no known collections from Sierra Victoria. This mountain and the adjacent area around Punta Frailes form a locality in need of detailed field work. Puerto Escondido is a small harbor with a neighboring rancho lying at the foot of the precipitous Sierra Giganta scarp. Here the mas- sive sedimentary formation is exposed and variously over-lapped by volcanic lavas and breccias. A narrow plain bounds part of the shore, in part the scarp rises high and spectacular out of the water (Plate 7, figs. 16, 17). The climate is hot and arid, being on the lee side of the 92 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Sierra GIganta, which blocks the westerlies that cool the opposite side of the peninsula. The vegetation is dominated by a dispersed, succulent, microphyllous Desert Shrub with scattered small trees which increase in stature and in number of species up in the steep canyon. Among these latter are Bursera sp., Lemaireocereus Thurberiy Cercidium molle, and the palm Erythea Brandegeei, Four collections have been made from Puerto Escondido and vicinity, mainly in the canyons cutting steeply back into the mountains of Sierra Giganta. I. M. Johnston May, June 1921 50 num.bers P. J, Rempel March 1937 30 numbers H. S. Gentry April 1938 50 numbers E. Y. Dawson February 1940 25 numbers It should be noted that all collections were made in the spring and that the late summer flora fostered by the summer rains has not been collected. Due to the precipitous slopes, the arid climate, and the tor- rential type of rains, the soils are rankly immature. The humic soils are confined to pockets in rocks, or to riparian embayments, and in them are commonly found the small succulents (e.g. Mammillaria) and mesophytes of erratic dispersion. In the high, narrow, and rocky canyons, where shade and run-ofiF conserve and augment soil moisture, there are found many subtropical species of the wetter Cape District. Pachycormus discolor here forms one of the largest trees with erect straight trunks 30 feet high or more, presenting quite a different habit from the dwarf contorted members of the species found in more arid habitats on the peninsula. Quercus idonea occurs in the mesic saddles and slopes near the tops of the Sierra Giganta. In the sierran area about Puerto Escondido, including the collect- ing localities of Agua Verde Bay 25 miles to the south and Comondu on the other side of the mountains, there are several plants having local distributions. These include Verbesina oligocephala, Polygala apopetala, Ruellia cordata, Cercidium molle, Dalea vetula, Mimulus sp., Perityle aurea, Agave sohria, and Vallesia laciniata. Their limited distributions corroborate the little geologic evidence gathered to date that indicates insular periods during the Tertiary for the Sierra Giganta area. Angel de la Guardia Island, lying close along the peninsula, is the second largest of the gulf islands. It has an approximate area of 975 square kilometers and the highest peak is 1315 meters above sea level. As indicated in Plate 8, figs. 18-20, the island is rugged, "barren," with precipitous slopes, and canyons shortly discharging into the sea. NO. GENTRY : LAND PLANTS 93 No permanent source of fresh water is known. Physiographlcally it is young and the arid climate tends to perpetuate the youthful appearance. Sedimentary rocks appear to overlie basic igneous and are in part over- lapped by extensive lavas. Very little appears to be on record describ- ing the island. W. H. Burt of the University of Michigan in his notes generously loaned to me described the island as follows: "A range of mountains, attaining a height of 4315 feet in the northern part, traverses the entire length of the island. This mountain range is highest at the two ends and there is a low pass near the center of the island. The west shore is for the most part precipitous but there are several landing places on the east shore and at the north and south ends. The mammalian fauna which is represented by three species, a pocket mouse {Perog- naihus)j rock mouse (Peromyscus) , and a wood rat (Neotoma), seems very small for an island of this size which is only eight miles from the mainland shore at its nearest point." Vegetation is sparse, particularly at the northern end of the island. Both Slevin (1923:69) and Johnston (1924) state that they found more vegetation at the southern end. It is densest on the low gentle slopes in the valleys, and along washes. Judging from the flora (Table 7 and Plate 8), it appears to be a microphyllous, succulent tree and suf- frutescent shrub desert, not essentially different from that common to the low and middle elevations of the adjacent mid-peninsula. The suc- culent or sarcophytic tree forms consist of Pachycereus Pringlei, Le- maireocereus Thurberi, Pachycormus discolor pubescens, Bursera ?nicro- phylla, and are accompanied, at least in the washes and valleys, by such nonsucculent microphyllous trees and shrubs as Prosopis juli flora, Acacia Greggii, Cercidium micro phy Hum, and Olneya tesota. A similar mixture of succulent and microphyllous species occurs in the shrub populations also. Low shrubby suffrutescents are well represented in Dalea, Erra- zurizia, Frankenia, A triplex, Petalonyx, Franseria, and Encelia. The flora also shows a strong relationship to the central peninsula, although some of the more striking peninsular species are apparently lacking, e.g., Idria, palms and yuccas. Many of the plants of the upper gulf region are here near their southern limits. The 96 species known to the island are enumerated in Table 7. The known endemics are only 6, surprisingly few for the extent and nature of the area. But here again only the spring flora is known. Additional field work is necessary before we can be satisfied of an adequate showing of the flora and before prob- lems of distribution and speciation can be evaluated. The Rempel and Dawson collections are enumerated below in the general catalogue of 94 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 species. They collected on the north end of the island, chiefly at Puerto Refugio. Dawson has 4 numbers from Pond Island, a small rocky body connected with the east side of the Angel de la Guardia at low tide. Their collections add 6 species to the known flora of Angel de la Guardia Island, one of which is new to science, Lyrocarpa linearifolia. Tiburon Island is the largest in the gulf. It is separated from the Sonoran mainland by a shallow narrow channel, *'el infiernillo," two to five kilometers wide and only three to four meters deep. Roughly quadrangular in shape, the island contains about 1170 square kilometers. Although mountainous, there are extensive valleys and several "aguajes," where fresh water is available for indeterminate periods following rains. On the north end, where Bahia Agua Dulce roundly indents the shore line, there is a permanent fresh water spring. In former times it was regularly used as a base settlement for the seminomadic Seri Indians, who still occasionally roam over the island hunting, fishing, and forag- ing upon the native wild plants and animals. There are two igneous ranges of mountains trending north and south and paralleling the ad- jacent ranges of Sonora. The western is Sierra Menor, the eastern and higher is Sierra Kunkaak (the Seri name for it) with a middle peak elevation of 1218 meters. A porphyritic granite occupies the southeast part of the island, ac- cording to Jones (1910), who explored the island. He traversed the island north to south along the east side, east to west across the south end, and went into the interior around the highest peak. The greater area of parent rock is volcanic. ''The types are profuse and belong to the effusive class." Also present are andesite, rhyolite, *'the latter pass- ing into the extreme phases of obsidian and pumice." No limestone or other sedimentaries were observed by him and his party. The volcanics may be a part of the Upper Miocene pyroclastics (Comondu formation), which predominate much of the California gulf area, although Jones took them to be much younger. The vegetation and the flora, so far as known, appear closely re- lated to that of the adjacent mainland (Plate 9, fig. 21). This is to be expected, since the shallow infiernillo channel was emerged repeatedly during the low sea levels of the glacial periods. The land bridge would then have allowed plant migration to or from the island, excepting those plants restricted to the rocky slopes and which find the sandy lowlands intolerable. The dry rocky slopes support a dispersed Desert Shrub for- mation, while the bottomlands and aggrading surfaces are thinly forested NO. 2 GENTRY : LAND PLANTS 95 with the wide spread mesquite, Olneya tesota, Cercidium, and several of the cactus trees. The known flora consists of only 77 species, which is considerably less than one would expect on the basis of area and elevation of the island. Obviously, Tiburon Island has been little collected and like other islands in the California gulf, the summer-fall flora is unknown. When the floras of Tiburon and adjacent Sierra Seri are better known, a com- mon origin will probably become apparent that may show some degree of independence from the surrounding region. Both Dawson and Rempel visited Tiburon Island on the voyages of the Velero III. Their collec- tions, consisting of 36 numbers were obtained on the southeast corner of the island and add 1 1 plants to the recorded flora of Tiburon Island. Tiburon Island has the richest insular vertebrate fauna known in the gulf. The following mammals are represented: coyote (Canis), rock squirrel (Citellus), two species of pocket mice (Perognathiis) y kangaroo rat (Dipodomys) , rock mouse (Peromyscus), wood rat (Neotoma), jack rabbit (Lepus), and the burro deer (Odocoileus). Bird life is well rep- resented and the following varieties appear to be endemic to the island : Tiburon quail {Lophortyx gambeli pembertoni) , Tiburon woodpecker (Centurus uropygialis tiburonensis) ^ Tiburon gnatcatcher (Polioptila melanura curtata)^ Tiburon cardinal {Richmondena cardinalis town- sendi)y Tiburon towhee (Pipilo fuscus jamesi). San Esteban Island lies in mid-channel \n the middle of the gulf region o£E Tiburon Island. Quadrangular in shape, it embraces about 35 square kilometers of rugged land with a top elevation of 540 meters at the south end. Volcanic in origin, it has ''scoriae-covered slopes and much breccia" (Johnston 1924:954). It is shored mostly with high vertical cliffs, but on the southeast is a pebble beach and a broad valley above the beach provides ready ingress. No source of fresh water has been reported. The ubiquitous Desert Shrub forms a fairly close cover over the gentle valley gradients, where growth is fostered by run-off, and a widely dispersed cover upon the open rocky slopes. The flora is very similar to that of adjacent islands and there are no known endemics other than the small number of plants it shares with neighboring islands, as Echino- cereus grandis which is common to San Pedro Nolasco, San Lorenzo, and San Esteban Islands. The mammal fauna consists of one endemic species of Peromyscus and a colony of introduced rats, Rattus, The Velero III land plant collections from San Esteban Island, Rempel 6 numbers, Dawson 6 numbers, consist mainly of cactus. The 96 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 early springs 1937 and 1940, when their collections were made, appar- ently were too dry for general floral response. The summer-fall flora of San Esteban Island is not known, all collections having been made in the spring, Table 6. The known flora consists of 48 species. The Rempel and Dawson collections are annotated in the following catalogue of species. San Pedro Nolasco is a rocky island 8 to 10 miles off the Sonoran coast along the latitude of Guaymas (N 28°). About 3.5 kilometers long and 1 kilometer wide, it is approximately 3.5 square kilometers in area, and reaches a height of 315 meters. Like most of the gulf islands, it rises sharply w^th steep rocky slopes out of the restless sea. Landings have been made on the southeast side, where a narrow rocky defile leads up into the interior. No source of fresh water has been reported. It is reported as both volcanic and granitic (Fraser 1943:137, 149). Judging from its stage of weathering (Plate 9, fig. 22), it is a youthful island, but its age is unknown. In accordance with the rocky slopes there is a heavy succulent ele- ment in the vegetation represented by Pachycereus Pringlei, Lemaireo- cereus Thurberi, Fouquieria peninsularis, Agave chrysoglossa, Pedilan- thus macrocarpus^ and several smaller inconspicuous cacti (Plates 10, 11, figs. 23-26). In addition to the tree cactus listed above there are Bursera micro phylla, the peninsular Ficus Pahneri, and the Sonoran Acacia Wil- lardiana. Unique also is the extensive growth of a bunch grass, Setaria macrostachya, which Johnston reports (1924:987), "extremely abun- dant on north-facing slopes on San Pedro Nolasco Island (4397) where it makes some hillsides appear like hay fields." This odd assortment of flora is dominantly composed of rock-inhabiting species and suggests for- tuitous occupancy characteristic of an infant island. The island appears too young for soils and maturely balanced plant communities to have developed. Comparable pioneer societies have been noted on the recent volcanics of the adjacent mainland, as on the southwestern out-lyers of the Bacatete range along the Sonoran coast. The known flora (Table 7) consists of 27 species, of which 8 have been added by the Velero III collections, represented by 15 numbers collected by Rempel and Dawson as annotated below. The iguana {Ctenosaura hemilopha) is a conspicuous resident. The Guaymas monadnock consists of a discontinuous range of cer- ros with intervening and bordering valleys and plains. The area has a certain physiographic unity and comprises about 3000 square kilometers with a peak elevation of 1316 meters. Clifl and rocky slopes form an NO. 2 gentry: land plants 97 extensive habitat, which show considerable variance according to the degree of sun and wind exposure. Such rare plants as Perityle Pabneriy Asclepias leptopus, and Dcsmodium Wigginsii have been found along the cliffs, as well as many cacti, Agave, and Ficus Palmeri. Conglomer- ates are apparent and the Recent alluvial deposition of clay and silt has tended to lap upon the mountain sides. The present cycle of degrada- tion at work in the area, however, exposes and sorts the coarse and fine gravels along the shallow arroyo channels, or hurries the coarse frac- tured rocks down the mountain slopes and canyons. Plants that attain true tree stature are limited to the run-off chan- nels or the margins thereof. Saline littoral flats are extensive and an unusually rich assortment of halophytes make up a complex association, and among which are commonly found the following genera: A triplex^ Lycium, Atamisquaea, Zizyphus, IVisUzenia, Salicornia, Prosopis, Steg- nospennum, Rhizophora, and many others. The sandy beach lines are widely and intricately interrupted by the steep rocky slopes and cliffs that are hammered by the palpitating sea. On the landward side of the area, alluvial materials have been banked high upon the mountain pedi- ments forming broad plains of Pleistocene aggradation. The strata are dominated by volcanics which have intruded and overlaid sedimentaries and are in turn partly overlaid by Quaternary beach and littoral deposits. The process of deposition on the bajadas and a structural uplift have raised the area above the confines of the sea. "The hills about Guaymas and for about 80 kilometers to the north are volcanic, consisting of basalt, tufa, and agglomerate. In the hills behind Guaymas a number of old sea caves were noted which, though now over 50 meters above the ocean, contained unconsolidated sands and modern shells. This indicates recent movement at least in one sec- tion of the coast" (Johnston 1924:953). The implications of the area being postinsular have been noted above, and while we are not yet cer- tain of this, it is obvious that it has grown in area during the recent period. The predominating plant formation of the Guaymas locality is Desert Shrub. However, because it is near the southern limits of the Sonoran Desert (see map in Shreve, !Mallen', and Turnage 1936:215) and per- haps because of conditions imposed by insulation, there are at}'pical ele- ments in the floral composition. The Sonoran Desert Region is bounded on the south by the Sinaloan Thorn Forest and species from this latter formation find their northern limits in or near Guaymas. This com- ponent includes such prominent Thorn Forest species as Acacia cymbi- 98 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Spina, Lysiloma divaricatu?n, Melochia tomentosa, Bursera laxifiora, Pithecolobium sonorae, Zizyphus sonoriensis^ and Hinionia pterosperma, which here are either reduced in stature and incidence, or confined to the more favorable situations of soil moisture, as along valley drainways. Certain shrubs achieve local dominance in the Guaymas area that are quite secondary in vegetational weight elsewhere. Among these are Cordia parvifolia and Lippia Palmeri. With low to medium shrub stat- ure they act a strong part locally in the dispersed shrub formation. There is present also an arborescent element typical of the Sonoran Desert, including species of Prosopis and CercidiuTu, which have normal growth in the valleys but are stunted on the dry rocky slopes. Pachy- cereus Prtnglei is abundant upon the rocky slopes and like Ficus Palmeri presents a special problem in distribution. These two plants are found widely over the peninsula and on the adjacent islands. Why is their incidence on the Sonoran coast so restricted? The Guaymas flora was referred by Axelrod (1939) to the Sierra Madre element of the American flora. Chaney's (1944) inclusion of it in the ''Southwest American Element" is more appropriate, since the relation of the Guaymas flora to that of the Sierra Madre is remote and at best but general. It is, however, an integral part of the great floral complex that has a long but obscure Tertiary history and a large spacial occupancy in the arid and semiarid southwest of North America. As sug- gested above, the Guaj^mas flora has as yet an indeterminate uniqueness in so far as insular isolation may have effected local evolution. Another striking feature of the flora is the number of apparent natural erratics. Some of the species whose occurrences at Gua^^mas and vicinity appear extralimital are Ficus Palmeri, Pachycereus P ringlet, Hermannia pauci- fiora, Lysiloma Candida, Indigofera mucronata, Acacia cymbispina, Vin- cetoxicum petiolare, Colubrina glabra, Boerhaavia Xantii, Lobelia splen- dens, Bouchea dissecta, and Vitex mollis. There is, of course, always the possibility that early man made such local displacements, but in any case they now belong to the native flora. Guaymas and San Carlos Bays are well known collection localities. Edward Palmer in 1877 was the first to do detailed botanizing in their vicinities. According to Watson's report (1888:36-87), the only paper that has itemized the Guaymas flora. Palmer collected 299 species of flowering plants during the summer months from mid-June to mid- November. Dawson's collections (43 numbers) from there and Rempel's 8 numbers from neighboring Ensenada de San Francisco represent the land plant samplings of the Velero III, catalogued below. NO. 2 GENTRY : LAND PLANTS 99 Summary of the Insular Floras Table 6 is a summary of plant collections that have been made on the islands in the California Gulf Region. They represent species or varieties collected, rather than numbers of each collector. In most cases, species collected equals numbers collected, the only notable exception TABLE 6 §o ti? .^ |. .2^ I^. 1^ Total Coll. Islands M^ o|^ c^ ^^ g- g- ^^ T" " J'^lS'oSP^-^-^-^f^-H oS rtS summer wmter ^ ZO^ ^^«Q fall spring Northern Gulf San Luis 9 9 Mejia 9 9 Angel de la Guardia 78 20 16 114 Pond 1 4 5 Raza 11 7 18 Sal si puedes 6 6 Las Animas 3 3 San Lorenzo 19 19 Patos 5 1 6 Tiburon 40 60 4 21 125 Turner's 3 3 San Esteban 37 4 6 47 San Pedro Martir 23 1 14 21 17 San Pedro Nolasco 13 10 6 29 Tortuga 19 21 40 Southern Gulf San Marcos 3131 62 Santa Inez 4 4 Ildefonso 11 2 13 Coronados 10 10 Carmen 70 32 49 70 81 Danzante 2 2 Monserrate 4 4 Catalina 10 5 15 Santa Cruz 7 7 San Diego _ 9 9 San Jose. 9 9 San Francisco 12 12 24 Partida Sur 9 9 Espiritu Santo 22 19 48 20 109 Ceralbo 2 42 44 Table 6: Collections of plants from the California Gulf Islands, according to collector and to island. 100 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 being I. M. Johnston, who frequently made more than one collection of a species at one locality or on one island. The list of collectors is not complete, but as far as I have been able to determine from literature, it includes all of the important ones. Others that may have collected on the islands, as Diguet and George Lindsay who had special interests in the Cactaceae, have provided few, if any, cited records. From the table it is clear that botanical exploration has been fleet- ing. On all except San Pedro Martir and Carmen Islands the summer flora has not been collected and is therefore imperfectly known. While many of the xerophytes produce leaf and bloom following both winter and summer rains, there are also many which respond only in the sum- mer-fall period. These latter are abundant in the southern portion of the region, where summer rains are heavier. Except for Johnston's pub- lished notes of a general nature, the ecology of the islands has been given only the most cursory attention. The development of the vegetation, the nature of plant communities, the amount of endemism, and the many diverse problems facing the phytogeographer cannot be determined until further careful and less nomadic field work has been systematically done. In Table 7 are listed all the plants that have been recorded from the larger California gulf islands. A few of the smaller islands have been omitted from the table for reasons of space, smallness of flora, or because their floras are unknown. Among the numerous small islands, the floras of which have not been collected, are Consag Rock, Smith, Tassne, and Montague Islands. The known plants on Mejia and Granite Islands are included in the list of Angel de la Guardia Island. Omitted are some small islands found in the bays of Guaymas, Concepcion, and La Paz (Pichilinque). Islands having small lists of spring flora are the following: Georges Island Chenopodium murale L. Patos Island* Bouteloua barbata Lag. Atriplex Barclayana typica H. & C. Amaranthus fimbriatus (Torr.) Atriplex Barclayana Palmeri Benth. (Wats.) H. & C. Carnegiea gigantea (Engelm.) Machaereocereus gummosus (En- Britt. & Rose gelm.) Brit. & Rose Opuntia sp. Encelia farinosa phenicodonta (Blake) Jtn. * In 1945 all the sporophytes were stripped off Patos Island by a Mexican guano company and the guano producing Peruvian cormorant was introduced. NO. 2 GENTRY: LAND PLANTS 101 San Luis Island Eriogonum galioides Jtn. (type loc.) Hoffmanseggia microphylla Torn Lotus tomentellus Greene Larrea divaricata Cav. Euphorbia polycarpa hirtella Bofss. Cryptantha marltima pilosa Jtn. Pond Island Mammillaria angelensis Craig Aster frutescens Wats. Raza Island Monanothocloe littoralis Engelm. Batis maritima L. Sesuvium sessile Pers. Lemaireocereus Thurberi (En- gelm.) Brit. & Rose Opuntia tunicata Lehm. Lycium brevipes Benth. Turner's Island Mentzelia adhaerens Benth. Santa Inez Island Atriplex Barclayana Palmeri (Wats.) Atriplex Barclayana sonorae (Standi.) Lophocereus Schottii (Engelm.) Brit. & Rose Isla Partida** Antigonon leptopus H. & A. Atriplex Barclayana typica H. & C. Atriplex Barclayana Palmeri (Wats.) H. &C. Dalea mollis Benth. (o) Bursera Hindsiana (Benth.) Engl. Eucnide cordata Kell. Vaseyanthus insularis Rose Atriplex Barclayana typica H. & C. Atriplex Barclayana Palmeri (Wats.) H.&C. Fouquieria peninsularis Nash Pachycereus Pringlei (Wats.) Brit. & Rose Cressa truxillensis HBK. Opuntia Bigelovii Engelm. Amaranthus Watsoni Standi. Lemaireocereus Thurberi (En- gelm.) Brit. & Rose Cressa truxillensis HBK. Atriplex Barclayana typica H. & C. ** There are two Isla Partidas in the Gulf of California; one in the northern area between Isla Raza anci Isla Angel de la Guardia, the other in southern waters on the northern tip of Espiritu Santo Island, It is here suggested that these should be designated respectively as Isla Partida del Norte and Isla Partida del Sur. From the accounts of the California Academy of Sciences Expedition to the Gulf of California (Slevin 1923, Johnston 1924), it is not possible to determine whether the plants listed from "Isla Partida" belong to the north or to the south island. 102 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Amaranthus Watsoni Standi. Lemaireocereus Thurberi (En- gelm.) Brit. & Rose Datura discolor Bernh. Nicotiana trigonophylla Dunal Bebbia juncea (Benth.) Greene Las Animas Island Ficus Palmeri Wats. Amaranthus Watsoni Standi. Solanum Hindsianum Benth. Atriplex Barclayana Palmeri (Wats.) H.&C. Cuscuta corymbosa stylosa ( Choi- sey) Engelm. Lycium brevipes Benth. Hofmeisteria fasciculata (Benth.) Walp. Atriplex Barclayana sonorae (Standi.) H. &C. Echinocereus grandis Brit & Rose Vaseyanthus insularis Rose Some synonomy in names of the California gulf islands are : Las Animas=North San Lorenzo (Proc. Calif. Acad. Sci. IV, 12. 1923-24) Pichilinque=San Juan Nepomucens (according to Leon Diguet) San Jose=San Josef (Proc. Calif. Acad. Sci. IV, 12. 1923-24) San Pedro Martir==San Pedro Martin (Rose, C.N.H. 1 :78-79. 1890) San Pedro Martir=San Pedro Martin (Rose, C.N.H. 1 :78-79. 1890) Tassne=Pelican Turner's^El Datil (local idiom) NO. 2 GENTRY: LAND PLANTS 103 Ceralbo Espiritu Santo. San Francisco.., San Jose San Diego... Santa Cruz. Catalina Monserrate. Danzante.... Carmen Coronados. Ildefonso... XXXXO XXX X X X X > San Marcos. Tortuga S. P. Nolasco. X S. p. Martin. San Esteban. Tiburon X X XX X San Lorenzo... Sal si puedes. Angel de Guardia. X X o fi O M ^ -a a; 0. g :: X o ^ a * 8i c o X : «^ w ^ W OS's W c S U o -r; : <^3 -^ i ^ "rt rt : o ca c-e cu c: b aj >? O P (U c w <: u h S I < £ O o- Pi -a n C« CD (Zl cfi C r; _ >^ S 5 == O O Ui !/3 Oj "^ ^ 'i^ _ ti ti y o-= ^_^ pq >. ^.^ hJ rt ^^ ■ > V) — 3 w Ok-G X5 -Q j=" _, ti c c 12 C3 C (u < Qj qj Qj qj ^ > t> > > > aj <<< S rt « f^ >> i7 G a rt C3 = O *^ ^ 2 'c "o, ex c 8 'E 'ill ^ ;r; C/D «.2 S- < s s << 3 . -J . Oh ^S-^ 073 Vi ^ C 'i^ 'C *n Tr P^ 2 « « « b£ <;gqq o rr] 1-1 106 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Ceralbo Espiritu Santo San Francisco San Jose San Diego Santa Cruz Catalina Monserrate Danzante Carmen Coronados Ildefonso San Marcos Tortuga S. P. Nolasco S. P. Marti7-7Z7~ San Esteban Tiburon San Lorenzo Sal si puedes Angel de Guardia O X o X o o X o c* h X O X o X X O X X XX X X o * X* O X si ^ 'i ^- ■C T3 ^ -C 2i o c > C g UU D X ii'- cn O C g c «« .2 .2 g E CO er, 3 ZI C C I ^ ^ cs re 4j 5j O C UUffiffi NO. 2 GENTRY: LAND PLANTS 107 o c o o o c o o O X X o X X X o o o XX X X X X X X X X X XXX X X X X X O X X X X X v tJD I- — "^-2^ rt « g o 0(5 ijs.> C c3 Wo ^ P3 C3 rt -5 .^ f3 ^ tw r' <» - b 2 o as ^ Sg s •= .-, o 3 cj C CJ «-> !-i •50^ c3 n: N 2 g rt S 22 <) u: re '^ i-J ^re _re hx)^- c c re Z A^' 108 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Ceralbo O X X XX X Espiritu Santo X C X X X X X X * C San Francisco O XX X 1 c c X San Diego O X X X c X o Santa Cruz Catalina X Vfonserrate o o o o o Danzante - Carmen - X X X H X XX X Coronados X o Ildefonso San Marcos X XX X X XX XXX Tortuga S. P. Nolasco ^ X X i S. P. Martir I- San E^teban X X X Tiburon ^. XX XXX XX X X X X Sal si puedes X X X X X X X Angel de Guardia. X fci: x! u TS rr Vr C/D '-^ -^ 'x ft r 5^ ~ "c < 2.5 u u a. u ^ ■— « ^ >• ~ '" >!"— — V", •"•"•— •"•~'~'— — -J-. 'X '~ '^ 'X '~ r" '~1 "" < "^ >i^ — r u 2i .^ u 'i HI'H.r ^ " -.^' ^-.^-'S'H'^'^'H'^T -rz'^'i'^'S'E-^ H £ X C < S ^ X o ^ c * :^ f I If 1 1 1 1 iiiiiiif Siiiii f 1 1 S £ < < w w ;^ —. c c ;^ ;i: u: ;^ '^ ►^ ',^ '^ '^ u; '>^ >— ^« CA. cc U C-l xvo. 2 gentry: land plants 109 * * >< X X ><^ X X X X X c X o o X c c c o c X X XXX X o c o c X X X O XX X O X c c c c c X X XXX X X X X X XX X XX X X c X X XX X c c el li 8s : : ! : >^. \^ r: 5 c «- <^ < = r ^ <^ c i 1 r: z J < r; y. C o c c - X "G u 5 r: c r: -2 v. z £ c < = < ^ > = < < cr r C c < > i 2 "C .^ u u z < > r: < y ■J- : y y • y : : u i >;£ ; u u >>^^ I- R ^ O iiz - " __ ^ r: c: 3 ; il L y. y. — r: • A* "x v-/ y It = 'x j < : ! : : : i i • J c c k 2 Z E <^ < « O 5 ar R 2 i 1 c K on J — C - »: 1 no ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Ceralbo Espiritu Santo San Francisco San Jose San Diego Santa Cruz Catalina Monserrate Danzante Carmen Coronados Ildefonso San Marcos Tortuga S. P. Nolasco S. P. Martir San Esteban Tiburon _ San Lorenzo Sal si puedes Angel de Guardia. o o o o X o o o o XX X O O X O X o X * o o o o * * X X XX* X * o o X XX 8 ^ _^ a . -c -^ S ^ g •O -C3 ^ -a ?> o o < — i3 >^ u a> o o C -^ 1- c C « c c o w "5 c c S S (^ OD C/3' < rt U fcfl pq o «« '^ T ^- 2^ rt r: w iJ pq t£j>.^ CO en cr Q ^ ^ 1 : c 'i: , ^ .*^ ' iP^M s P5 : i^^ ^ ; C — fciC a ^—^ ■' i c 5 D 13 1 J3 1 1 3 ^ i : CD en O t) 3 .' O >^ ^C3 ^2 .«^ «^ *n *v- 'u, 'C rt c3 rt C3 esse s 6 £ s rt w cs cs pq ::3 ^ .eq rt P3 pq rt « rt £ rt o fctpq S o C VI &.i pq c3 c3 rt ^ pqpq ^ '13 T3 g C C ^ a. a OO O U o rt c3 rt «a C 03 U fcjcpqc .2 ^ g-S: "5 -2 « o £•35 ^^ a c , a. D, c ooo ='p^ rt c3 rt flJ •r .t; 'z ^ 3 3 3 0, C- C O. C3 f tJD . o .t:P5 6 :cq7 C|V c pq C- ^ C C3 C 3pq < , U a; C c 5.2 < c <^ O P^ = -?^ (n a p ^ IB 15 > = o H i2 < ^ W u rn < c Q n < Oh rt < C/J h-J rt u D- C/) !U < ?^ c« O (u O g 3 O O < o o t:. « 2- -e s >. o rt .-. « 3 rt > *C *c rt 2 2^ o o o i io a rt c3 r3 rt c c c a f3 rt rt rt c a. ex o, uuuo NO. 2 GENTRY : LAND PLANTS 113 X X o o X o X X X X X X X X X X X * H * X X X X H X X xH rt ^ 5 .2 .- .2 2 _2 "^ a, a,H.fh "> ffiEffi pSj C fc- rt t7 cj jg a, =3 '^ s^ W o < ^ C/3 rt a a >^ .22 ^ 3 £« M-i O 5 I- c IS C C3 en 1/3 C rt C3 fS M i c O rt *: -a "^ « ?i rt S^ oj G 1) „• -I rt rt o .-; WO < C' c c o ^ C3 C3 Ph X X X X Catalina Monserrate. Danzante.... Carmen Coronados. Ildefonso... X X X X XX San Marcos.... Tortuga S. P. Nolasco. S. P. Martin. San Esteban. Tiburon San Lorenzo. Sal si puedes. Angel de Guardia si ^1- •t; "" -a -a -a ^ -o O O (U (U ^ S bjc « ■^f§ (U ^ 3^ 5 fcX S a b r^ en -^ -^ "^ — J2 -- pq ?^ c i; fcJD'^ -; en en « opq O rt .is 2 3 tx r=. ^ f? y O CO ^pq rt -O ^ C/3 en c/3 en en 3 3 3 3 &, D, a, CX ,T! 9- 5 u u fs cs rt 'J o o o o o . C NO. 2 GENTRY : LAND PLANTS 115 X X X h X XH m C3 I- o fc, O U PQM.S o w "C-r <«^ rt o D en ^ O .S c pq c pj !3 o M-i PQ 3 •^ Oh 3 D 03 c« t« en tn w fJ 3 H! 3 3 3 P ^-. _^ cxaHCxci,ciia.n) „:^coooo^^.';: o- cu^ o O •r -;;; t> ^ 8 en tn (U gj (U CJ ^- "S O, Cl, CI, a, O cfl «! C3 rt .5 £ oj o *33 'S w *5 M-( V-( "^ V4_| O O^ O XX X en en V3 en 6 S o o XX .i2 o OJ c §1 o >, en <^ , O 1-1 QiJ -G — ' I S 2 o ^= ^ fc- >^ o ^ 17, Dazi'son 1213. Puerto Escondido, February 10, Dawson 1085, March 13, Rempd 156. Tortuga Island, IMarch 17, Rempd 223. A common small vine throughout Baja California below the chapar- ral and on adjacent islands; type from Magdalena Bay, Baja California. The leaves are dimorphic; flowers pink. Stylosanthes viscosa Sw., Prod. Veg. Ind. Occ. 108. 1788 vel aflF. San Jose del Cabo, February 17, Dawson 1154. Swartz' species is widely distributed through tropical America. The genus is in need of revision and Dawson's collection is tentatively as- signed. Tephrosia hamata (Rydb.) Gentry new comb. Tephrosia Pahneri Brge., Proc. Calif. Acad. Sci. II, 3:126. 1891. Not T. Pahneri W^ts. 1889. Cracca hamata Rydb., N. Am. Fl. 24:177. 1923. Tephrosia hamata Brge. ex. Rydb., N. Am. Fl. 24:177. 1924, a MS name in synonomy only. San Jose del Cabo, February 17, Dawson 1153. Endemic to the postinsular Cape District; type from San Jose del Cabo. It is a low silvery sericeus perennial with sagging stems. Tephrosia tenella Gray, PI. Wright. 2:36. 1853. Near Guaymas, Sonora, February 9, Dawson 1076. Punta Frailes, February 16, Dawson 1144. Tortuga Island, March 17, Rempel 206, in wash. Coarse arid soils from Texas to Baja California and Sinaloa; type from San Pedro, Sonora. A low bushy perennial herb with slender branches from near the base, and purple flowers. Apparently long flower- ing following favorable rains, from October to March. The Cape Dis- trict specimen has longer stipules, larger and more pubescent flowers. Zygophyllaceae Fagonia californica Barclayana Benth., Bot. Voy. Sulph. 10. 1844. San Juanico Bay, Baja California, March 2, Rempel 42, wash. West side of the peninsula from San Juanico Bay south to the Cape District; type from Magdalena Bay. Fagonla. californica glandulosa Vail, Bull. Torr. Bot. Club 22:229. 1895. 136 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Puerto Escondido, February 11, Dawson 1102. Los Angeles Bay, March 19, 20, Rempel 237. Island in Concepcion Bay, March 16, Rempel 203, on fan. Arid desert slopes around the Gulf of California on the peninsula and on the mainland. Larrea tridentata (DC.) Cov., C.N.H. 4:75. 1893. Los Angeles Bay, March 19, 20, Rempel 242. Found in nearly all the hot American deserts. Viscainoa geniculata (Kell.) Greene, Pittonia 1:163. 1888. Los Angeles Bay, March 19, 20, Rempel 241. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 277. An irregular branching shrub along rocky arroyos of the southern two-thirds of the peninsula, along the mid-Sonoran coast, and known also from Tiburon Island. Rempel 277 appears to represent the first record from Angel de la Guardia Island. Simarubaceae Castela peninsularis Rose, C.N.H. 12:278. 1909. San Gabriel Bay, Espiritu Santo Island, March, Rempel 85 (sterile) A low stiff spiny shrub with sclerophyllous drought-deciduous and rather ephemeral leaves, the branches varying from yellowish green to reddish purple, the whole pubescent except the spines. It is known from the southern part of the peninsula from Magdalena Bay and near Mu- leje southward; the type from San Jose del Cabo. Johnston (Proc. Calif. Acad. Sci. IV, 12:1056) collected it on several of the adjacent gulf islands from Catalina south. BURSERACEAE Bursera Hindsiana (Benth.) Engl, in DC, Monogr. 4:58. 1883. Bursera rhoifolia (Benth.) Jtn., Proc. Calif. Acad. Sci. IV, 12:1058. 1924. West cove in Concepcion Bay, March 15, Rempel 181, low shrub or tree on rocky slopes. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 267a. Los Angeles Bay, March 19, 20, Rempel 234a. Tor- tuga Island, March 17, Rempel 229. A thick-butted and usually dwarf tree with thick reddish or gray branches and spur branchlets. The leaves are quite variable; simple, or partially trifoliate, or distinctly trifoliate, as on the Tortuga specimen (up to 7 cm long) or sm.all, as on the Concepcion Bay sheet (10 to 15 mm long). As Bullock (Kew Bull. 1936:366) has pointed out, the NO. 2 gentry: LAND PLANTS 137 name B. Hindsiana takes precedence over B. rhoifolia, because Art. 56 of the International Rules makes binding the first selection of a name out of two or more applicable to the same species under the same date. In this case the rule invalidates later combinations made apparently on the basis of page priority. B. Hmdsiana appears to be common to the rocky slopes and washes from Los Angeles Bay south through the penin- sula and has also been collected on the north central Sonoran coast (Pringle, sine no. in 1884). BuRSERA MICROPHYLLA Gray, Proc. Am. Acad. Sci. 5:155. 1861. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 93. Ca- beza Ballena, Cape District, March 3, Rernpel 57. On rocky slopes and draws it is scattered through the Sonoran Desert. A closely related plant, B. morelensis from the state of Morelos, Mexico, has doubtfully been referred to synonomy under this species (Bullock I.e. p. 371). Malpighiaceae Thryalis angustifolia (Benth.) Kuntze, Rev. Gen. PI. 1:89. 1891. San Jose del Cabo, February 17, Dawson 1209. Widely distributed in northwestern Mexico; the type from Cape San Lucas, Baja California. A suffrutescent plant, in the above speci- men 2 to 3 dm tall, the ovary glabrous except for a few minute hairs on the apex and angles. Euphorbiaceae Adelia virgata Brge., Zoe 4:406. 1894. Cabeza Ballena, Cape District, March 3, Rempel 63. Agua Verde Bay, Baja California, March 10, Rempel 137 , in wash. Arroyos along the mountains of the Cape District and the Sierra Giganta as far north as Comondu, and reported by Johnston on the adjacent gulf islands of San Jose, Espiritu Santo, and Ceralvo; type from the Sierra Laguna, Cape District. It is an erect slender shrub with several branches bearing fascicles of obovate leaves. The Rempel collec- tions are sterile. DiTAXis Brandegei (Millsp.) Rose & Standi., C.N.H. 16:13. 1912. Puerto Escondido, February 11, Dawson 1108. Agua Verde Bay, Baja California, March 10, Rempel 123. Widely but infrequently scattered in the California Gulf Region; type from San Gregorio, Baja California. A rather succulent polypodial 138 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 subshrub with purplish stems 1 to 2 m tall with regularly lanceolate, remote, finely serrate leaves. DiTAXis LANCEOLATA (Benth.) Pax & Hoffm. in Pflanzenr. IV, 147:64. 1912. Puerto Refugio, Angel de la Guardia Island, January 26, Dawson 1032. San Jose del Cabo, February 17, Daivson 1215. Open situations in the California Gulf Region, Sonora and Baja California; ty^t from Magdalena Bay, Baja California. A low suffru- tescent plant with pale, linear-lanceolate, cinereus leaves. Also related to, but not conspecific, is Dawson 1150, from San Jose del Cabo. DiTAXis NEOiMEXiCANA (Muell. Arg. ) Heller, Cat. N. Am. PI. 5. 1898. North end of Los Angeles Bay, March 19, 20, Re?npel 257, washes and sandy benches. Arid and semi-arid climates from Texas to southern Arizona, Sonora, and middle Baja California. Euphorbia californica Benth., Bot. Voy. Sulph. 49, pi. 23B. 1844. San Jose del Cabo, February 17, Dawson 1170. Southern part of the gulf region in Baja California, Sonora, and Sinaloa; type from Magdalena Bay, Baja California. The specimen is unusually robust, reflecting better soil moisture con- ditions than commonly obtained in other parts of its range. Questionably referred here also is Rejnpel 70 from Cabeza Ballena, Cape District, a shrub having unusually thin, long petiolate, obcordate leaves, and exfoli- ating bark. Euphorbia eriantha Benth., Bot. Voy. Sulph. 51. 1844. Los Angeles Bay, March 19, 20, Rempel 248. San Juanico Bay, outer coast, March 2, Re?npel 34. San Jose del Cabo, February 17, Daw- son 1221. Throughout the deserts of the California Gulf Region in sandy soils, east to Texas; type from Magdalena Bay, Baja California. An erect winter annual with single stem and ascending branches, long linear leaves, and terminal capitate inflorescences, the erect gray- green fruits usually conspicuous. Commonly scattered in dispersed colo- nies over sandy areas. Euphorbia leucophylla Benth., Bot. Voy. Sulph. 50. 1844. Punta Frailes, February 16, Dawson 1127. NO. 2 gentry: land plants 139 Known from the southern part of Baja California and the adjacent islands of Tiburon and Ceralbo ; type from Cape San Lucas. Euphorbia magdalenae Benth., Bot. Voy. Sulph. 50. 1844. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 97 y west end of pass on shell and coral. San Francisco Island, March 9, Rempel 107, shrub 2 feet high on south-facing hill slope. On both coasts and adjacent islands of the southern part of the peninsula; type from Magdalena Bay, Baja California. Johnston ob- served what he took to be this plant at San Pedro Bay on the coast above Guaymas, but apparently failed to collect it. It is a low slender- stemmed, densely and intricately branched shrub 4 to 10 dm tall, form- ing low globose bushes with minute flowers in the spring. Euphorbia pediculifera Engelm. in Torr., U.S. & Mex. Bound. Bot. 186. 1859. Tortuga Island, March 17, Rempel 222. North end of Los Angeles Bay, March 19, 20, Rempel 259. Island in Concepcion Bay, March 16, Rempel 199, in wash. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 83, on beach dunes. Widely scattered on the deserts of southwestern United States and northern Mexico and throughout most of the California Gulf Region. Euphorbia pediculifera linearifolia Wats., Proc. Am. Acad. Sci. 24:76. 1889. Guaymas, Sonora, January 23, Dawson 1005 (topotype). Southern Sonora; type from Guaymas. It is distinguished from the typical in the species by its erect habit and linear leaves. Both are winter annuals flowering in the early spring or late winter. Euphorbia petrina Wats., Proc. Am. Acad. Sci. 24:75. 1889. Puerto Refugio, Angel de la Guardia Island, January 26, Dawson 1022. Known from southern Sonora, Angel de la Guardia and San Pedro Martir Islands, the type from the latter. Perennial prostrate herb form- ing dense mats; spring flowering. Euphorbia polycarpa Benth.?, Bot. Voy. Sulph. 50. 1844. Tortuga Island, March 17, Rempel 204a. Widely scattered apparently throughout the California Gulf Region ; type from Magdalena Bay, Baja California. A perennial, spring-flower- 140 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 ing, prostrate herb. Since the inflorescence of the Rempel collection has been modified by galls, Wheeler has questionably referred the collection to this species. Euphorbia tomentulosa Wats., Proc. Am. Acad. Sci. 22:476. 1887. Near Guaymas, Sonora, February 9, Dawson 1080. San Carlos Bay, February 8, Dawson 1057. Common on coarse rocky dry soils nearly throughout the gulf region as far south as southern Sonora and the Cape District. It is a low suf- frutescent, bushy, flat-topped herb. Euphorbia Xantii Engelm. in Boiss., DC, Prodr. 152:62. 1862. Puerto Escondido, February 11, Dawson 1092, 1099. Canyon above Puerto Escondido, March 13, Rempel 149. Southern Baja California and northern coastal Sinaloa; type from Cape San Lucas. It forms a slender, erect, flat-crowned shrub 1 to 2 m tall with dichotomous virgate branching; the leaves are ephemeral fol- lowing the summer rains. The plant may bloom in a leafless condition in the spring. It has the largest flowers, 6 to 9 m in diameter, of any of the gulf region euphorbs and is attractive in bloom. It was cultivated successfully in the greenhouse of the Carnegie Desert Laboratory in Tucson for many years. Jatropha cinerea (Ort.) Lluell. Arg. in DC, Prodr. 152:1078. 1866. Questionably referred here is Rempel 119 from Agua Verde Bay collected March 10, 1937. The specimen consists of a sterile and what was once a turgescent shoot, bearing large cordate, ternately lobed, long- petiolate leaves, 9 to 14 cm broad, the emarginate sinuses very narrow and deep. Other collections seen have been referred to either this species or to /. canescens of Bentham, but neither name seems satisfactorily ap- plicable. Flower and fruiting material are needed for taxonomic place- ment of the plant. It occurs in the deep can3^ons of the Sierra Giganta. Jatropha cuneata Wiggins & Rollins ?, Cont. Dud. Herb. 3 :272. 1943. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 92, small thick-trunked tree throughout pass except on alkali. The succulent v/oody branches with numerous short spur-branchlets of the sterile Rempel collection are doubtfully referred here. The species is widely distributed in the California Gulf Region; tj^pe from near Kino Bay, Sonora. NO. 2 GENTRY : LAND PLANTS 141 BUXACEAE SiMMONDSiA CHiNENSis (Link) Schneid., Handb. Laubholzk. 2: 141. 1907. San Carlos Bay, Sonora, February 8, Dawson 1056. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 95. Southern Sonora, southern California, Sonora, and Baja California. It is a characteristic shrub of the arid rocky slopes in the California Gulf Region, occurring usually in widely scattered stands. It is not known south of the Rio Yaqui and the above collection from Sonora is near the southern limits of the species on the mainland. Anacardiaceae Cyrtocarpa edulis (Brge.) Standi., C.N.H. 23:659. 1923. Tapirira edulis Brge., Zoe 5:78. 1900. Cabeza Ballena, March 3, Rempel 56. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 89, up to 12 feet high in Cactus scrub. Known only from the Cape District and adjacent islands. This is a thick-stemmed Bursera-like tree with smooth yellowish bark that peels ofi perennially in thin sheets. The above cited specimens are sterile and the leaves are the small drought-pauperized ones of the arid spring. It is widely and naturally spread in the Cape District and also occasion- ally planted by the inhabitants for its edible plum-like fruits. It appears to be a postinsular endemic, since it is quite distinct from its only near relative, C. procera of the Mexican mainland. However, little is known of the ecology of the plant. There is the possibility that it was intro- duced to the peninsula by early man. Pachycormus discolor var. pubescens (Wats.) Gentry new comb. Schinus discolor Benth., Bot. Voy. Sulph. 11, pi. 9. 1844. Pachycormus discolor (Benth.) Cov., Cent. Diet. rev. ed. 6708. 1911. Bursera pubescens Wats., Proc. Am. Acad Sci. 24:44. 1889. Veatchia discolor Brge., Proc. Calif. Acad. Sci. II, 2:140. 1889. Veatchia discolor pubescens (Wats.) Jtn., Proc. Calif. Acad. Sci. IV, 12:1079. 1924. Puerto Refugio, Angel de la Guardia Island, January 26, Dawson 1025 (sterile). March 20, Rempel 273. A low, sprawling, openly crowned, sarcophytic tree with massive trunk and branches, endemic to and characteristic of the arid peninsula about Los Angeles Bay and the adjacent Angel de la Guardia Island, the type locality. The above cited specimens are sterile, but the smaU 142 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 pubescent leaves differ materially from leaves of other varietal popula- tions of the species studied on the peninsula. According to Johnston, this variety is characterized by "its very loose deltoid inflorescence, by its rather small leaves, and perhaps also by its brownish sap." Celastraceae Maytenus phyllanthoides Benth., Bot. Voy. Sulph. 54. 1844. Puerto Escondido, February 11, Dawson 1088. Cabeza Ballena, Cape District, March 3, Rempel 74 (sterile). San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 87 (sterile). Island in Concepcion Bay, March 16, Rempel 188 (mature fruit). Coastal from Baja California to Cuba and Florida; type from Mag- dalena Bay, Baja California. Schaefferia cuneifolia Gray, PI, Wright. 1 :35. 1852. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 86, shrub 3-4 feet high scattered in nonalkaline soil. Semiarid localities of northern Mexico and western Texas ; type from "high prairies of the San Felipe and on the San Pedro." In Baja Cali- fornia it has been collected previously only on the slopes of Tres Virgines in middle peninsula, at Muleje, and in the Cape District. It is a short stiflE Condalia-like shrub with spinescent branchlets and small, obovate, thick- ish, drought-deciduous, caducous leaves. Sapindaceae Cardiospermum halicacabum L., Sp. PI. 366. 1753. Guaymas, January 23, Dawson 1000. San Carlos Bay, Sonora, Feb- ruary 8, Dawson 1069. San Jose del Cabo, February 17, Dawson 1198. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 96. Agua Verde Bay, March 10, Rempel 128. A scandent diffuse shrub or self-supporting subshrub forming a low bush with interlocking stems and recurved branches. The common Mexi- can name is "bolsilla," aptly referring to the inflated papery fruits, hang- ing showily like unlit jack-o-lanterns through the fruiting season of late fall and again perhaps in the spring. The species in the larger sense is variable and subspecific names have been assigned. Rhamnaceae CoNDALiA globosa Jtn., Proc. Calif. Acad. Sci. IV, 12:1086. 1924. Los Angeles Bay, March 19, 20, Rempel 234c (sterile). Known in middle and southern Baja California, on adjacent islands in the gulf, and near Guaymas, Sonora ; type from La Paz. Though not important in the aggregate of vegetation, this stiif low shrub with spine- tipped branchlets appears to be widely scattered through the desert of the California Gulf Region. NO. 2 gentry; land J^lants 143 Karwinskia Humboltiana (Zucc.) R. & S., Abh. Akad. Wfss. Muchen 2:351. 1832. Cabeza Ballena, Cape District, March 3, Rempel 69. The species as now known includes several variants which range over Mexico and southwestern Texas. The peninsular population exists in the Cape District and along the Sierra Giganta scarp, mainly in the wetter canyons and northern slopes, and is not readily separable from some of the populations of northwestern Mexico. Malvaceae Abutilon californicum Benth., Bot. Voy. Sulph. 8. 1844. Puerto Escondido, February 11, Dawson 1086. West coast of Mexico from Baja California and Sonora south to Oaxaca; type from Magdalena Bay, Baja California. Abutilon crispum (L.) Sweet, Hort. Brit. 53. 1827. San Jose del Cabo, February 17, Dawson 1226. In open situations from southern Arizona south to Central America. A decumbent perennial herb with inflated fruits; the present specimen with remote, cordate, acuminate, and characteristically gray canescent leaves, the pale petals marked with bright carmine in the base. Abutilon incanum (Link) Sweet, Hort. Brit. SZ. 1827. San Jose del Cabo, February 17, Dawson 1210. San Carlos Bay, Sonora, February 8, Dawson 1066. Widely distributed as a wayside and fallow-land weed in north- western Mexico; also reported in Hawaii. It is a bushy perennial 1 to 2 m high with rather strictly ascending branches. The yellow petals are carmine spotted within and reflexed at anthesis, closing at night. GossYPiUM Davidsonii Kell., Proc. Calif. Acad. Sci. I, 5:82. 1873. San Jose del Cabo, February 17, Dawson 1197. Southern Baja California; type from San Jose del Cabo. Rarely col- lected. Hibiscus denudatus Benth., Bot. Voy. Sulph. 7, pi. 3. 1844. West Cove in Concepcion Bay, March 15, Rempel 176, hill slope of west exposure. Guaymas, Sonora, January 23, Dawson 1006. The more arid situations in northwestern Mexico and western Texas to southern Arizona in the United States. It appears to be quite generally scattered through the California Gulf Region. At best it forms a low open suffrutescent bush, the young branches are densely yellowish tomen- 144 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 tose, the old glabrous. The flowers are lavender and with only light winter rains the plant will persist in flowering from late winter months through the spring. HoRSFORDiA ALATA (Wats.) Gray, Proc. Am. Acad. Sci. 22:297. 1887. Puerto Escondido, February 11, Dawson 1086. California Gulf Region on both the peninsula and the mainland; type from northwestern Sonora, Mexico. The genus is distinguished by its winged carpels ; this species is distinguished from H. Newberryi, the only other Sonoran Desert species, by its relatively large, thickish, ovate leaves and pink petals drying purplish or blue. HoRSFORDiA Newberryi (Wats.) Gray, Proc. Am. Acad. Sci. 22: 297. 1887. West Cove in Concepcion Bay, March 15, Rempel 179, foot of rocky slope. Island in Concepcion Bay, March 16, Rempel 191, in wash. Rather infrequently scattered in the lower elevations of the Cali- fornia Gulf Region. It is uncommon or lacking along the outer coast of the peninsula and has a smaller range than H. data. SiDA Xantii Gray, Proc. Am. Acad. Sci. 22:296. 1887. San Jose del Cabo, February 17, Dawson 1190. Known in Baja California and listed also by Standley (C.N.H. 23: 765. 1920-26) as occurring in Sinaloa. The type locality is from Cape San Lucas, Baja California. The pale flowers are unusually large for the genus, the petals may be as much as 1.5 to 2 cm long. Sphaeralcea Coulteri (Wats.) Gray, Proc. Am. Acad. Sci. 22: 291. 1887. San Carlos Bay, Sonora, February 8, Dawson 1050. In sandy and coarse detrital soils of the Sonoran Desert from the Colorado Desert, California south and in the Sinaloa Thorn Forest south to Mazatlan; type apparently from southwestern Arizona or ad- jacent Sonora. It is a leafy annual, single-stemmed or polj^podial, rather showy with salmon-colored flowers through the arid spring. Sphaeralcea Coulteri californica (Rose) Kearney, U. C. Publ. Bot. 19:32. 1935. Puerto Escondido, February 11, Dawson 1091. Southern half of Baja California; type from La Paz. NO. 2 gentry: land plants 145 Sphaeralcea Hainesii Brge., Proc. Calif. Acad. Sci. II, 2:136. 1889. Tortuga Island, March 17, Rempel 224. Los Angeles Bay, Baja Cali- fornia, March 19, 20, Rempd 244. Middle part of the California Gulf Region along the eastern coast of the peninsula and the islands of San Pedro Martir, San Marcos, Tor- tuga; type from Jesus Maria, Baja California. On the peninsula it fre- quents waste lands like a weed, but Johnston reported "it is the most abundant herbaceous perennial on the island" of San Pedro Martir. Sterculiaceae Hermannia Palmeri Rose, C.N.H. 1 :67. 1890. San Jose del Cabo, February 17, Dawson 1220. Apparently endemic to southern Baja California; type from La Paz. This plant is rare in collections. The above cited specimen is in flower only, the characteristic Solanum-like stamens conspicuous. The cordate- triangular crenate leaves are unusually ample, the blades up to 4 cm wide and 4 cm long, indicating good moisture conditions or shade during the weeks preceding collection. Melochia tomentosa L., Syst. Nat. ed. 10:1140. 1759. San Carlos Bay, Sonora, February 8, Dawson 1062. San Jose del Cabo, February 17, Dawson 1191. Lower elevations nearly throughout Mexico and south into Central America. It is a common shrub in the environs of Guaymas, where, in the fall after the summer rains, it forms a rather showy slender shrub with light purplish flowers. WaLTHERIA AMERICANA L., Sp. PI. 673. 1753. San Jose del Cabo, February 17, Dawson 1158. Widely distributed through the American tropics and subtropics; type from the Bahama Islands. Frankeniaceae Frankenia Palmeri Wats., Proc. Am. Acad. Sci. 11:124. 1876. Los Angeles Bay, March 19, 20, Rempel 262, 238. Littoral flats of the upper gulf region, where it forms low rounded brittle-stemmed bushes. It is one of the common halophytes of the region, but Johnston (Proc. Calif. Acad. Sci. IV, 12:1097) attributes its appar- ent halotropism to the salt air father than to saline soils. It is also found on nonalkaline soils and locally occurs in extended stands making a low suffrutescent vegetation with or without associates. 146 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 FOUQUIERIACEAE FouQUiERiA PENiNSULARis Nash, Bull, Torn Bot. Club 30:454. 1903. Near Guaymas, Sonora, February 9, Dawson 1079. Baja California: Punto Frailes, February 16, Dawson 1136; San Jose del Cabo, Feb- ruary 17, Dawson 1219; Puerto Escondido, March 13, Rempel 167, on alluvial fan; Frailes Bay, April 4, Rempel 328. Mainly coastal in the southern part of the California Gulf Region in southern Baja California, Sonora, and northern Sinaloa; type from La Paz, Baja California. These collections all show the thickish pedicels and narrow panicles typical of the species. FouQUiERiA SPLENDENS Engelm. in Wisliz., Mem. Tour. Mex., 30th Cong. 1st Sess. misc. rep. No. 26:98. 1848. Los Angeles Bay, March 19, 20, Rempel 249. Northern part of the California Gulf Region in Sonora, on the pen- insula, on Tiburon Island, north into the deserts of California, Arizona, and Nevada, and eastward in northern Mexico. TURNERACEAE TuRNERA DIFFUSA Willd., Schult. Syst. Vcg. 6:679. 1820. San Jose del Cabo, February 17, Dawson 1152. Widely distributed in Mexico, Central America^ and South America, mainly in arid soils on hill slopes. Passifloraceae Passiflora arida (Mast. & Rose) Killip, Jour. Wash. Acad. Sci. 12:256. 1922. Island in Concepcion Bay, March 16, Rempel 190, on fan. Widely scattered along the coasts in the California Gulf Region as far south as Mazatlan, Sinaloa; type from Guaymas, Sonora. It is par- ticularly abundant along the inner shore of the mid-peninsula, infrequent on the outer peninsular coast and the Sinaloa coast. It is a densely pu- bescent vine flowering and fruiting in the spring and related to P. Pal- meri and P. fruticosa, but distinguished from those two species by the lack of oily stipitate glands. Passiflora Palmeri Rose, C.N.H. 1:131, pi. 14. 1892. West Cove in Concepcion Bay, March 15, Rempel 175, on hillside. Island in Concepcion Bay, March 16, Rempel 195, generally distributed on fans and lower hillsides. On the gulf side of the peninsula and adjacent islands in the gulf NO. 2 gentry: LAND PLANTS 147 from Angel de la Guardia south to San Jose del Cabo; type from Car- men Island. A low spreading flat-topped viscous shrub with showy flowers in the spring. LOASACEAE EucNiDE CORDATA Kell. in Curran, Bull. Calif. Acad. Sci. 1:137. 1885. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 99, in alkaline flats not far from beach dunes. Agua Verde Bay, Baja Cali- fornia, March 10, Rempel 121, wash. Tortuga Island, March 17, Rem- pel 215. Common to the cliffs, rocky slopes, and occasional in washes through the mid-section of the peninsula and the adjacent islands; type from Cedros Island. It is a coarse rather succulent or turgescent perennial herb with bright green, cordate, crenate, lobate leaves, and creamy white tubular flowers. Mentzelia adhaerens Benth., Bot. Voy. Sulph. 15. 1844. Turner's Island, January 25, Dawson 1012. Near Guaymas, Feb- ruary 9, Dawson 1096. Tortuga Island, March 17, Rempel 226. Sonora and Baja California; type from Magdalena Bay, Baja Cali- fornia. Johnston also reports it from Tortuga and Tiburon Islands. The Dawson numbers show the pauperate foliage of a dry winter season, the blades being mostly less than 2 cm long. It is a brittle herb sometimes found clambering among the low branches of shrubs. The Mexicans have been heard to call it "pega pega." Petalonyx linearis Greene, Bull. Calif. Acad. Sci. 1:188. 1885. Tortuga Island, March 17, Rempel 214. A low bushy subshrub with rough scabrous leaves, large floral bracts, and small white flowers. Widely scattered over the northern half of the peninsula and the adjacent islands; into southern California and Ari- zona. Type from Cedros Island. Sympetaleia rupestris (Baill.) Gray in Wats., Proc. Am. Acad. Sci. 24:50. 1889. Tiburon Island, January 25, Dawson 1013. Puerto Escondido, March 13, Rempel 162. Island in Concepcion Bay, March 16, Rempel 202. On cliffs and dry rocky situations in the mid-gulf region in Sonora, Baja California, and the gulf islands of Sal si puedes, Tiburon, and Tortuga. A low sticky herb forming rounded clumps ; spring flowering. 148 allan hancock pacific expeditions vol. 13 Cactaceae CocHEMiEA PosELGERi (Hildm.) Brit. & Rose, Cactaceae 4:22. 1923. Punta Frailes, Cape District, February 16, Daiuson 1111. Agua Verde Bay, Baja California, March 10, Rempel 115, in wash. Frailes Bay, Cape District, April 4, Re?npel 327. Cape District of Baja California and northward along the Sierra Giganta scarp ; exact locality of type collection unknown. EcHiNOCEREUS GRANDis Brit. & Rose, Cactaceae 3:18. 1922. San Pedro Nolasco Island, February 6, Dawson 1037; San Esteban Island, February, Dawson 1042. South end of San Esteban Island, March 27, Rempel 294. Known only from San Pedro Nolasco, Las Animas, San Lorenzo, and San Esteban Islands in the Gulf of Baja California; the ty^t from San Esteban. Flowers March and April. ECHINOCEREUS MAMILLATUS (Engelm.) Brit. & Rose, Cactaceae 3:41. 1922. West Cove in Concepcion Bay, March 15, Rempel 172 (sterile), hillside of east exposure. Frailes Bay, April 4, Rempel 323a (sterile). A short-bodied, long-spined species, the stems tapered at the base. Ranges along the Sierra Giganta scarp and through the mountains of the Cape District; type from "mountain sides south of Muleje, Lower California." EcHiNOCEREUs sciURUS (K. Brge.) Purpus, Monatsschr. Kakteenk. 14:130. 1904. San Jose del Cabo, February 17, Dawson 1222. Known only from the Cape District of the peninsula. The tj^pe locality is the hills near San Jose del Cabo, Dawson 1222 being a topo- type. EcHiNOCEREus scopuLORUM Brit. & Rose, Cactaceae 3:31. 1922. Ensenada de San Francisco, Sonora, March 30, Rempel 315. Guay- mas, January 23, Dawson 1008. Native of the coastal mountains of southern Sonora and northern Sinaloa; type from near Guaymas, Sonora. EcHiNOCEREUs Websterianus G. Lindsa}^ Cact. Succ. Jour. 19: 153. 1947. San Pedro Nolasco Island, March 29, Rempel 301, 303; February 6, Dawson 1040. NO. 2 gentry: land plants 149 Known only from San Pedro Nolasco Island where it commonly grows in close association with Mamjnillaria jnultidigitata. Lindsay (I.e.) has written that "its large golden clumps [contrast] pleasantly with the white masses of the latter." Ferocactus acanthodes (Lem.) Brit. & Rose, Cactaceae 3:129. 1922. North of Point Lobos, Sonora, I\Iarch 26, Rempel 287 (sterile). This is the common Ferocactus or bisnaga on the arid slopes and mesas in the northern part of the gulf region ; type locality, "California." Ferocactus Covillei Brit, k Rose, Cactaceae 3:133. 1922. Guaymas, Sonora, January 23, Dawson 1009 (in part). From southern Arizona to southern Sonora; the type from near Altar, northern Sonora. Guaymas is near the southern limit of the species. Ferocactus Townsendianus Brit. & Rose, Cactaceae 3 :127. 1922. Frailes Bay, April 4, Rempel 318, hillside. A poorly known species from the southern part of the peninsula and adjacent islands; the type from Isla San Jose. Continuation of this name is questionable, since it appears applicable to the species described earlier by Weber as Echinocactus peninsulae, Bull. Mus. Nat. Paris 1 :320. 1895. Britton and Rose failed to resolve the question, apparently be- cause they did not have access to Weber's material collected by Leon Diguet in 1894 near Santa Rosalia, and based their concepts of the species on a fragmentary collection of Gabb. This they state to be with- out radial spines, the main character segregating F. Townsendianus from F. peninsulae (Weber) Brit. & Rose, Cactaceae 3 :133. 1922, but Weber describes "Aiguillons rougeatres, a pointe jaune; exterieurs 11, ." Taxonomically the question can be cleared by field work and ample herbarium material, and more solidly by access to Diguet's original notes and photos. Britton and Rose's name is retained here solely because it is the one in common use for the southern peninsular plexus. Ferocactus sp. Tetas de Cabra near Guaymas, Sonora, January 23, Dawson 1009a. Referred here is a specimen strongly suggestive of Ferocactus ala- mosanus platygonus G. Lindsay, Cact. Succ. Jour. 14:139. 1942, but differing materially in the strongly ascending 7-8 lateral spines and as reported by the collector in the large size; up to 1.5 m. This is the plant 150 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 which Johnston referred to F. alamosanus, apparently, Proc. Calif. Acad. Sci. 13:1110. 1924. Although the present collection, the species, and its variety appear to be related, further field work and study may reveal that we are dealing with three specific entities. Lemaireocereus Thurberi (Engelm.) Brit. & Rose, C.N.H. 12: 426. 1909. South end of Tiburon Island, March 27, Rempel 297. San Pedro Nolasco Island, March 29, Rempel 305. Ensenada de San Francisco, Sonora, Rempel 312a. Frailes Bay, Re?npel 322. The pitaya (traditionally spelled pitahaya, but poorly so from the phonetic standpoint) is widely distributed in the central and southern part of the California Gulf Region; from southern Arizona to central Sinaloa. In central Baja California it favors the rocky slopes where run- off increases soil moisture. In southern Sonora, the area of greatest abundance, it is found both on rocky slopes and alluvial plains, as on the plain south of Navojoa, where it abounds in dense nearly pure stands over many square miles. Flowers in late spring, fruits ripen in late May and June. All the above collections are sterile. Lemaireocereus Thurberi littoralis (K. Brge.) G. Lindsay, Desert PI. Life 12:186. 1940. Cereus Thurberi Engelm. var. littoralis K. Brge., Zoe 5:191. 1904. Canyon above Puerto Escondido, March 13, Rempel 145 (imma- ture). Fraile Bay, April 4, Rempel 326. Cape District and north along the Sierra Giganta scarp to Con- cepcion Bay; type from "on steep seacoast bluffs between San Jose del Cabo and Cabo San Lucas, Baja California." This low slender plant appears to be a good variety. The collections show only 6 to 7 ribs and weaker spines than is typical of the species. Rempel 209 from an island in Concepcion Bay is probably also referrable here. All specimens are sterile. LoPHOCEREUs ScHOTTii (Engelm.) Brit. & Rose, C.N.H. 12:427. 1919. Los Angeles Bay, Baja California, March 19, 20, Rempel 232, ir^ wash and on fan. Fraile Bay, April 4, Rempel 324. Widely distributed in the gulf area; type from Magdalena, Sonora. A widely spreading plant commonly 2 to 3 m tall with many stems from the decumbently spreading base, recognizable by the sordid "beard" of long bristles of the terminal 1 to 3 feet of the flowering branches. NO. 2 gentry: land plants 151 Machaerocereus gummosus (Engelm.) Brit. & Rose, Cactaceae 2:116. 1920. Puerto Escondido, Dawson 1109, abundant over all brush-land. Fraile Bay, Rempel 323, Tortuga Island, Rempel 210. Common and widespread through Baja California south of Ensen- ada and on the adjacent islands; type from northwestern Baja Cali- fornia. In places it is abundant and is a codominant with other desert shrubs. Mammillaria albicans (Brit. & Rose) Berger, Kakteen, 308. 1929. San Pedro Nolasco Island, February 6, 1940, Dawson 1039. Recorded only from Santa Cruz, San Jose, and San Pedro Nolasco Islands, the above cited specimen being new to the known flora of the latter island. Referred here doubtfully is Rempel 302 also from San Pedro Nolasco Island. It is a smaller, slenderer plant with an evident tendency to be cespitose. Mammillaria angelensis Craig, Mam. Handb. 165. 1945. Angel de la Guardia Island (probably Puerto Refugio), January 26, Dawson sine no. Pond Island, February 5, Dawson sine no. Punta Frailes, February 16, Dawson 1112. Known previously only from Angel de la Guardia Island, the species is now certainly known from Pond Island, but Punta Frailes collection (sterile) is doubtfully referred to this species. The long reflexed petals are singularly characteristic of this species. Mammillaria dioica K. Brge., Erythea 5:115. 1897. Tiburon Island, January 25, Dawson 1018. Apparently ranges throughout Baja California, but hitherto not re- ported for any of the California Gulf Islands. Mammillaria Evermanniana (Brit. & Rose) Orcutt, Cactogra- phy 7. 1926. Canyon above Puerto Escondido, Sierra Giganta scarp, Rempel 144j sides of canyon. Along the Sierra Giganta scarp in southern Baja California and adjacent islands; type from Ceralbo Island. Craig (Mam. Handb. 82. 1945) states, but without citation, that it is also reported from San Pedro Nolasco Island. It is a small globose plant 6 to 10 cm high preferring humic soils in the detrital pockets of rocky terrain. It is rare in both living and herbarium collections. 152 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Mammillaria frailiana (Brit, k Rose) Boedeker, Mammil.Vergl. Schluss. 30. 1933. South end of San Esteban Island, Rempel 293 (sterile). It is also known from Ceralbo, Catalina, and Pichilinque Islands, the latter being the type localit3\ It will probably be found also on the adjacent part of the peninsula. A closely related species is il/. Verhaerti- ana Boedeker, reported from Los Angeles Bay. IVlA^miLLARiA iNSULARis Gates, Cact. Succ. Jour. 10:25. 1938. Ildefonso Island, March 15, Rempel 17 Oh, on rocky ridge. Ensenada de San Francisco, Sonora, ]\Iarch 10, Rempel 316. Central part of the Gulf of California; type from easternmost islet of Smith Island, Los Angeles Bay, Baja California. A low cespitose plant with purplish hooked spines. Doubtfully referred here also is Rempel 325 (in part). All of his collections, made in Alarch, are sterile. Mammillaria multidigitata G. Lindsay, Cact. Succ. Jour. 19: 152. 1947. San Pedro Nolasco Island, ^larch 29, Rempel 302 \ February 6, Dawson 1038. Apparently endemic to San Pedro Nolasco Island, this plant is re- lated to M. albicans, but difiEers in its cespitose habit, its slender cylindri- cal stems (not globose), and the appressed fine white radial spines. Lindsay reports it as being abundant on the steep slopes of the island. Mammillaria Slevinii (Brit. Sc Rose) Boedeker, Mammil. Vergl. Schluss. 44. 1933. San Francisco Island, i\Iarch 9, Rempel 101 (sterile). A rare and little known species related to AI. albicans and may be as Craig (I.e. p. 262) surmised, conspecific with it. Known only from San Francisco and San Jose Islands in the Gulf of California, Rose having collected the type on the latter island. Opuntia Bigelovii Engelm., Proc. Am. Acad. Sci. 3:307. 1856. Angel de la Guardia Island, Puerto Refugio, January 26, Dawson sine no. Turner's Island, Dawson sine no. Tiburon Island, January 25, Dawson 1019. Tortuga Island, Rempel 210a. South end of San Esteban Island, March 27, Rempel 292. A common cactus shrub of the northern part of the California Gulf Region in IMexico and the United States; type from Bill Williams River, Arizona. The Tortuga Island collection is the southernmost record. All the above collections are sterile. NO. 2 gentry: LAND PLANTS 153 Opuntia Burrageana Brit. & Rose, Cactaceae 1 :70. 1919. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 80, round-jointed Opuntia in cactus-scrub. Fraile Bay, April 4, Rempel 319 (flowering). Puerto Escondido, ]\Iarch 13 (flowering), Rempel 142, alluvial fan; February" 11, Daivson sine no. (sterile). The collections from Puerto Escondido are doubtfully referred here. They are atypical of the species in the elongate slender joints and with more prominent tubercles. The ovaries have numerous, acicular, yellow- ish spines about 1 cm long that deciduate soon after anthesis on the growing ovar}'. The specimens also suggest O. ciribe, but are scarcely consepecific. Typical O. Burrageana is recognized as ranging through- out the Cape District and on the adjacent islands as far north as Car- men. It may be a postinsular endemic of the Cape District. O. ciribe distribution is thought to be only in the central part of the peninsula from the vicinity of Santa Rosalia south to the vicinity of Comondu. It is apparent that there is a complex of cylindropuntia species in the central part of the peninsula in marked need of field work and taxonomic study. Opuntia aff. clavellina Engelm., in Coulter, C.N.H. 3:444. 1896. Patos Island, IVIarch 26, Re?npel 290. San Pedro Nolasco Island. March 29, Rempel 304. South end of Tiburon Island, March 27, Rem- pel 292. All specimens sterile. Typical O. clavellina is known only along the outer peninsula coast; type from near Purissima (see Map 2). The rather short thickish joints with prominent paper>^-sheathed spines of the above collections indicate a close but scarcely specific relationship. Opuntia comonduensis (Coult.) Brit, k Rose, Smith. ]Misc. Coll. 50:519. 1908. San Gabriel Bay, Espiritu Santo Island, March 7, Rempel 82 (sterile). Fraile Bay, Baja California, April 4, Rempel 320 (sterile). Southern Baja California and the adjacent island of Espiritu Santo and probably others; type from Comondu, Baja California. Little is known of the actual habitat and range of this platyopuntia and the above collections extend the known range southward and to Espiritu Santo Island. Johnston (Proc. Calif. Acad. Sci. IV, 12:1116. 1924) reported observing it on all of the southern gulf islands except Catalina, stating that it was the only plat3'opuntia observed on these islands. The small oval joints, the remote prominent areoles, and the 1 or 2 long porrect 154 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Spines, either whitish or yellow, characterize this species among the platyopuntia of the peninsula. Opuntia aff. fulgida Engelm., Proc. Am. Acad. Sci. 3:306. 1856. San Pedro Nolasco Island, March 29, Rempel 304a. Doubtfully referred here, the specimen also shows relationship to O. ciribe Engelm., a peninsular species (see Maps 1 and 2). Opuntia Gosseliniana Weber, Bull. Soc. Acclim. France 49:83. 1902. Ensenada de San Francisco, Sonora, March 30, Rempel 313. Reported by Britton and Rose from *'Sonora and Lower California, Mexico;" type locality, ''Coast of Sonora on the Gulf of California." Rempel's collection is flowering and typical except that the joints are suborbicular being broader than long. Opuntia invicta Brge., Proc. Calif. Acad. Sci. II, 2:163. 1889. Los Angeles Bay, March 19, 20, Rempel 233 (fruiting), in wash on fan. Central part of the peninsula, the t^^pe from San Juanico Bay. A low plant with trailing stems, short thick joints with high tubercles, large, 3-cornered, whitish, sheathless spines, and bristly fruits. Opuntia leptocaulis DC, Mem. Mus. Hist. Nat. Paris 17:118. 1828. Tiburon Island, January 25, Dawson 1020. Widely distributed in both the high and low deserts of the south- western United States and Mexico ; type locality in Mexico. Opuntia ramosissima Engelm., Am. Jour. Sci. II, 14:339. 1852. Los Angeles Bay, March 19, 20, Rejnpel 231 (flowering), in wash. Widely distributed in the southwestern United States and adjacent Mexico west of the continental divide; type from near the Colorado River in California. This appears to be the first cited collection from Baja California, although the species has long been known to occur there. Opuntia versicolor Engelm. in Coulter, C.N.H. 3:452. 1896. Ensenada de San Francisco, Sonora, March 30, Rew.pel 312. South end of San Esteban Island, Rempel 291. Widely distributed in the northern part of the gulf region on the mainland from southern Arizona to southern Sonora ; type from Tucson, Arizona. I have seen no specimens nor records referrable to the peninsula and the above citation from San Esteban Island is the first for the gulf islands ; fruiting, it compares favorably with mainland material. NO. 2 GENTRY : LAND PLANTS MAPI 155 ?u NTIA IN CALIFORNIA GULF AREA 1. 0. versicolor Engelm. 2. molesta Brge. 3. fulgida Engelm. 4. spinosior (Engelm.) Toumey 5. prolifera Engelm. 6. alcahes Weber 7. Burrageana B. & R. 8. invicta Brge. 9. cineracea Wiggins 0. reflexispina Wig. & Rollins A. rosarica G. Lindsay 156 ALLAN HANCOCK PACIFIC EXPEDITIONS MAP 2 VOL. 13 CYLINDROPUNTIA IN CALIFORNIA GULF AREA 1. 0. mortolensis B. & R. 2. 3. leptocaulis DC. Thurberi Engelm. 4. echinocarpa Engelm. 5. 6. 7. 8. 9. serpentina Engelm. Parryi Engelm. Bigelovii Engelm. ciribe Engelm. cholla Weber 0. calmalliana Coulter X. clavellina Engelm. NO. 2 GENTRY : LAND PLANTS MAPS 157 PLATYOPUNTIA IN CALIFORNIA GULF AREA 1. O. basilaris Engelm. 2. pycnantha Engelra. 3. comonduensis (Coult. 4. Gosseliniana Weber 5. occidentalis Engelm. 6. discata Griffiths 7. chlorotica Engelm. 8. tapona Engelm. 9. Wilcoxii B. & R. ? species undescribed O. Bravoana Baxter B. &R. 158 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Opuntia aff. Wilcoxii Brit. & Rose, Cactaceae 1 :172. 1919. San Pedro Nolasco Island, March 29, Rempel 306, West coast of Mexico in southern Sonora and northern Sinaloa; type from Fuerte in northern Sinaloa. This first insular collection ap- pears to be in good character with the species. It was just coming into flower when collected. Unusual are the areoles around the edge of the joints; they are larger and with longer glochids than those on the flat faces. Opuntia sp. Dawson 1091 from Puerto Escondido, Baja California, collected in a sterile condition February 11, 1940, is a flat-jointed Opuntia resemb- ling O. tardispina Griffiths from eastern Texas, as illustrated in Brit. & Rose (I.e. p. 141). Doubtless, this is one of the unknown Opuntiae men- tioned by Johnston (Proc. Calif. Acad. Sci. IV, 12:1117). Pachycereus Pringlei (Wats.) Brit. & Rose, C.N.H. 12:422. 1909. South end of Tiburon Island, March 27, Rempel 298 (sterile). Abundant over wide areas through middle and southern Baja Cali- fornia, on the gulf islands, on the cerros southwest of Altar in north- western Sonora, and in the hills about Guaymas, Sonora; type from the Altar River, Sonora. Its western limit is Cedros Island, where it was recently discovered by Howell (Leafl. West. Bot. 3:183. 1942). Sterile nubbins represented by Rempel 211, 307, 317, from Tortuga Island, San Pedro Nolasco Island, and Fraile Bay respectively, are either referrable to this plant or to Pachycereus pecten-ahoriginum, the other giant cactus abundant through the southern part of the gulf region. Rathbunia alamosensis (Coult.) Brit. & Rose, C.N.H. 12:415. 1909. Ensenada de San Francisco, Sonora, March 30, Rempel 314. Near Guaymas, Sonora, January 23, Dawson 1074. Coastal lowlands from southern Sonora to Nayarit; type from Alamos, Sonora. It commonly forms colonies several yards in diameter, spreading by declining or broken stems taking root. It is also employed locally and effectively for making fences, the cuttings taking root readily and eventually forming a dense hedge row. Rhizophora mangle L., Sp. PI. 443. 1753. West Cove in Concepcion Bay, March 15, Rempel 174. NO. 2 GENTRY : LAND PLANTS 159 The above cited locality is about the northern limit of this wide- spread littoral plant on the east side of the peninsula. On the Mexican mainland it reaches Tiburon Island. Onagraceae Oenothera angelorum Wats., Proc. Am. Acad. Sci. 24:29. 1889. Los Angeles Bay, March 19, 20, Rempel 236a. Known only from the eastern shore of the peninsula near the type locality, Los Angeles Bay. Oenothera cardiophylla Torr., Pacif. R.R. Rep. 5:360. 1856. North end of Los Angeles Bay, March 19, 20, Rempel 253a. Widely distributed around the north and west sides of the Gulf of California, south to central Baja California, and known from the gulf islands of San Luis, Angel de la Guardia, San Pedro Martir, and San Marcos, Type locality, near Fort Yuma, Arizona. A winter annual or possibly a short-lived perennial, it blooms in the early spring months. Sapotaceae BuMELiA occidentalis Hemsl., Biol. Cent. Am. Bot. 2: 298. 1881. Canyon above Puerto Escondido, March 13, Rempel 153. Los En- cinos. Sierra Giganta, Gentry 4264. Usually a small rather bushy tree scattered in the canyons of the southern part of the peninsula and northern Sonora on the lower moun- tain slopes; type locality, "Sonora alta." it flowers and fruits in the spring. Apocynaceae Vallesia glabra (Cav.) Link, Enum. PI. 1:207. 1821. Island in Concepcion Bay, March 16, Rempel 196. An evergreen shrub with white flowers and small pyriform trans- lucent fruits adapted to the mesophytic bottomlands, particularly the river margins from southern Baja California and southern Sonora south along the coastal tierra caliente to South America and the West Indies. ASCLEPIADACEAE AscLEPiAS ALBICANS Wats., Proc. Am. Acad. Sci. 24:59. 1889. Puerto Escondido, February 11, Dawson 1107. Island in Concepcion Bay, March 16, Rempel 194, on fan and low on north exposure. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 268. Nearly throughout and confined to the Sonoran Desert; type from ravine near Los Angeles Bay, Baja California. 160 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 ASCLEPIAS CURASSAVICA L., Sp. PL 215. 1753. Puerto Escondido, February 11, Dawson 1087. Widely distributed in the tropics and subtropics of America. It is a mesophyte found only along stream banks or wastelands of irrigated fields in Baja California. It is an erect perennial herb or strictly branched bush with showy heads of reddish orange flowers. AscLEPiAS SUBULATA Decne. in DC, Prodr. 8:571. 1844. Guaymas, January 23, Dawson 1101. San Jose del Cabo, February 17, Dazvson 1156. Found throughout the desert of the California Gulf Region, mostly along arroyos and canyon bottoms in sandy or gravelly soil ; type from "Nova Hispania." This and A. albicans are perennial herbs with several to many rush-like stems 1 to 2 m tall, bearing remote filiform leaves for short periods in the rainy seasons, the terminal pani- culate white and yellowish flowers often outlasting the ephemeral leaves. The flowers are greedily visited by many species of wasps and flies. CONVOLVULACEAE EvoLVULUs LiNiFOLiA L., Sp. PI. ed. II :392. San Jose del Cabo, Baja California, February 17, Dawson 1188. Sonora and Baja California. Distinguished from the more widely spread E. aisinoides by the narrower leaves and smaller flowers. Jacquemontia abutiloides Benth., Bot. Voy. Sulph. 34. 1844. Puerto Escondido, February 11, Dawson 1093; March 13, Rempel 148. Punta Frailes, February 16, Dawson 1126. San Jose del Cabo, February 17, Dawson 1162, 1186. Southern Sonora and middle and southern Baja California; type from Magdalena Bay, Baja California. Hydrophyllaceae Phacelia scariosa Brge., Proc. Calif. Acad. Sci. II, 2:185. 1889. Puerto Escondido, February 11, Dawson 1103; March 13, Rempel 150. Tortuga Island, March 17, Rempel 205. Southern Baja California and adjacent islands including Tortuga and Carmen in the gulf ; type from Magdalena Island on the outer coast. Boraginaceae Bourreria sonorae S. Wats., Proc. Am. Acad. Sci. 24:62. 1889. San Jose del Cabo, February 17, Dawson 1208. Baja California and Sonora, type from Guaj^mas, Sonora. It is a shrub 1 to 2 m tall scattered in the coarse soils of the less arid localities. NO. 2 gentry: land plants 161 CoLDENiA Palmeri Gray, Proc. Am. Acad. Sci. 8:292. 1870. Los Angeles Bay, March 19, 20, Rempel 235, 255. In sandy soils through the northern part of the gulf region, where it forms a low suffrutescent mat or mound with parts of the older stems exfoliating a white cortical layer. Heliotropium curassavicum L., Sp. PI. 130. 1753. Punta Frailes, Cape District, February 16, Dawson 1138. Coastal lowlands throughout most of tropical and subtropical America. Heliotropium Hintonii Jtn., Jour. Arn. Arb. 21 :50. 1940. Punta Frailes, February 16, Dawson 1120, 1176. San Jose del Cabo, February 17, Dawson 1185, 1216. Known from the mountains of west central Mexico and from the Cape District of Baja California. It is a low suffrutscent 3 to 4 dm high with elongating racemes of small white flowers, the stems leafy with strongly pubescent linear-lanceolate leaves 15 to 25 mm long. Labiatae Hyptis Emoryi Torr. in Ives, Rep. Col. River 20. 1861. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 276. Near Guaymas, Sonora, February 9, Dawson 1081. Found throughout the California Gulf Region ; type from the Colo- rado River country. It is one of the regular opportunists among the shrubs along rocky arroyos and is able to withstand, or at least to endure as a species, the grinding flash floods that arise with torrential desert rains. Guaymas appears to be about its southern limit on the Mexican mainland. A variant of the species occurs about Guaymas and was des- cribed by Watson as a species, H. Palmeri, but later was reduced to a variety by Johnston, H. Emoryi Palmeri. It may well be a postinsular endemic. Standley (C.N.H. 23:1276. 1924) lists Hyptis Emoryi from Tepic, but his statement is in part based on the related H. albida Kunth., which is known to range through Nayarit. Hyptis laniflora Benth., Bot. Voy. Sulph. 42, pi. 20. 1844. Punta Frailes, February 16, Dawson 1124. San Jose del Cabo, Feb- ruary 17, Dawson 1183. Southern Baja California and adjacent islands; type from Cape San Lucas. The material from Punta Frailes has calyx lobes shorter than usual, scarcely half as long as the tube, and the lanate pubescence of the calyx is much denser than on the San Jose del Cabo specimen. 162 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Stachys coccixea Jacq., PL Hort. Schoenbr. 3:18. 1798. Canyon above Puerto Escondido in the Sierra Giganta, March 13, Rempel 170, canyon bottom. Widely distributed in the warmer mountains of southwestern United States and northwestern Mexico. Verbenaceae AviCENNiA nitida Jacq., Enum. PI. Carib. 25. 1760. West cove in Concepcion Bay, March 15, Rempel 173. As a regular associate of the mangrove marshes, it is common along the coasts of tropical and subtropical America. Concepcion Bay is near its northwestern limit. As residue of sea mist, salt cr>'stals are commonly apparent on its leaves. SOLANACEAE Datura discolor Bernh., Prommed. N. Jour. Pharm. 26:149. 1838. North end of Los Angeles Bay, March 19, 20, Rempel 251a. Ti- buron Island, January 25, Dawson 1010. San Jose del Cabo, February 17, Dawson 1224. From southeastern California and southern Arizona south through the California Gulf Region to Central America; type from the West Indies. A relatively small Datura with small leaves and narrow flowers with a purple flush in the throat. The Tiburon material has smaller flowers and heavier fruiting spines than the typical peninsular material. Lycium aff. Andersoni Gray, Proc. Am. Acad Sci. 7:388. 1868. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 276a. Common through the deserts from southern Utah and southern Ne- vada south throughout the California Gulf Region, more common along the coasts than inland in the latter area. Lycium brevipes Benth., Bot. Voy. Sulph. 40. 1844. Lycium Richii Gray, Proc. Am. Acad. Sci. 8:292. 1870. Island in Concepcion Bay, March 16, Rempel 189. Agua Verde Bay, March 10, Rempel 118. Sandy slopes, washes, and alluvial and saline soils along the coast throughout the gulf region and south into Sinaloa. It is one of the larger-leaved, bushy Lyciums, making dense plants and dense thickets locally. NO. 2 GENTRY : LAND PLANTS 163 NicoTiANA Greeneanum Rosc, C.N.H. 1:18. 1890. Agua Verde Bay, March 10, Rempel 127, wash. Known previously only from Cedros Island and the adjacent western part of the peninsula, this appears to be the first record of the plant from the gulf side of the peninsula. The annual habit, nonclasping and non- auriculate leaves with ovate to lanceolate blades, the small corolla with very narrow limb (3 mm wide in the dried specimen), and the dull light brown muriculate seeds relate it pretty definitely to Rose's plant. NicoTiANA TRiGONOPHYLLA Dunal in DC, Prodr. 131:562. 1852. San Carlos Bay, Sonora, February 8, Dawson 1023, 1059, Near Guaymas, February 9, Dawson 1078. Tortuga Island, March 17, Rempel 217. Mostly in the coarse alluvial soils in arroyos and valleys in Desert Shrub and Thorn Forest from Texas to California and south to Nayarit. This is a common member of the tobacco genus in the arid gulf region. It appears to be perennial in the lower latitudes of its range, as shown by the woody base of Dawson 1023. Another sheet, Dawson 1078, repre- sents a young plant in its first season of spring flower, indicating it as having germinated during the preceding early fall or late summer rains. Petunia parviflora Juss., Ann. Mus. Hist. Nat. Paris 2:216. 1803. San Jose del Cabo, February 17, Dawson 1159. Southern Florida to California and south into tropical America. Through the deserts it is chiefly confined to the moist sands of perma- nent or intermittent streams. Low herb with minute lavender flowers. Physalis crassifolia Benth., Bot. Voy. Sulph. 40. 1844. Willard Point, Gonzaga Bay, Baja California, March 23, Rempel 283, wash. Wide-spread in the southwestern United States and adjacent Mexico in arid climates on sandy and rocky soils. Type locality, Magdalena Bay, Baja California. Physalis crassifolia infundibularis Jtn., Proc. Calif. Acad. Sci. IV, 12:1156. 1924. Puerto Refugio, Angel de la Guardia Island, January 26, Dawson 1026. Known only from San Esteban Island, Angel de la Guardia Island, and the adjacent coast of the peninsula; type from Angel de la Guardia 164 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Island. It differs from typical P. crassifolia in having a funnelform co- rolla as long as or longer than wide, rather than a rotate corolla. Physalis GLABRA Bcnth., Bot. Voy. Sulph. 39. 1844. Punta Frailes, February 16, Dawson 1128. Known only from the Cape District of Baja California. Physalis pubescens L., Sp. PI. 183. 1753. Canyon above Puerto Escondido, March 13, Rempel 157; April 22, Gentry 3764, riparian in canyon bottom. A mesophytic to hydrophytic annual with pale yellow flowers, purple stamens, and pubescent viscid herbage, widely distributed across southern United States and southward in Mexico ; Galapagos Islands. Physalis purpurea Wiggins, Cont. Dud. Herb. 3;74. 1940. San Carlos Bay, Sonora, February 8, Dawson 1063. Known only from the vicinity of Guaymas, Sonora. This perennial Physalis with its bright purple corolla and rather open habit with remote leaves is not easily confused with any other members of the genus in the gulf area. Its apparent very limited distribution, which kept it from being discovered so long, probably marks it as an endemic of the coastal mountains near Guaymas, and which for a period in the Tertiary may have been insular. SoLANUM HiNDSiANUM Bcnth., Bot. Voy. Sulph. 39. 1844. Tiburon Island, January 25, Dawson 1011. San Carlos Bay, Sonora, February 8, Dawson 1052, Punta Frailes, Baja California, February 16, Dawson 1142. San Jose del Cabo, February 17, Dawson 1192. Baja California and Sonora; type from Magdalena Bay. This plant is a low openly branched shrub rather closely related to S. elaeagnifolium, but in the field is at once distinguished by its larger size and larger corollas. It becomes abundant locally in the southern part of the gulf region, but is infrequent in the northern part. Scrophulariaceae Antirrhinum cyathiferum Benth., Bot. Voy. Sulph. 40. 1844. Near Guaymas, Sonora, February 9, Dawson 1084. Southwestern Arizona, Sonora, and Baja California; type from Mag- dalena Bay. Guaymas appears to be about the southern limit for the species on the mainland. MiMULUS SP. Canyon above Puerto Escondido, February 11, Dawson 1104, seep- NO. 2 GENTRY : LAND PLANTS 165 age in palm canyon. Canyon above Puerto Escondido, March 13, Rempel 161 J on damp rocks. These collections represent two species, neither of which I can place satisfactorily. Dawson 1104 is an erect herb about 15 cm high with rela- tively large orbicular coarsely dentate leaves, strongly 3-veined from the base, the calyces prominently red-spotted, but the spots fading on the fruiting calyces. The plant represented by the Rempel 161 is a diminutive, procum- bent, finely cut herb with yellow flowers forming mats in wet or moist sand or on rocks by seeps and pools in the canyon bottom. It is to be expected in other localities of the Sierra Giganta. It has also been col- lected in the same locality; Gentry 3772 and Johnston 4113. The latter collection was referred by Grant (Ann. Mo. Bot. Gard. 11 :186. 1924) to Mbnuliis dentilobus, described from Nacari, Sonora by Robinson and Fernald. While I have not seen Johnston's collection, those of both Rempel and Gentry fail to show the laciniatelj^-lobed corollas character- istic of M. dentilobus. Though the Puerto Escondido plants are clearly related to M. dentilobus, they appear worthy of taxonomic recognition. Unfortunately, none of the material at hand is worthy of type designa- tion. Future collectors should make it a point to secure a large series of good material of both of these rare Mimulus, not otherwise known. MoHAVEA CONFERTIFLORA (Benth.) Heller, Muhl. 4:48. 1912. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 270. Widely but infrequently scattered through the deserts from southern Nevada south to Angel de la Guardia Island ; not known from Sonora. The type locality is uncertain. A diminutive winter annual with large showy flowers. Stemodia arizonica Penn., Notul. Nat. Acad. Sci. Phil. 43:1-10. 1940. Canyon above Puerto Escondido in the Sierra Giganta, March 13, Rempel 158, in canyon bottom. Hydrophytic herb, often with the older branches decumbent. Along the perennial streams from low to middle elevations in the mountains and foothills of northwestern Mexico and adjacent United States. Martyniaceae Martynia altheaefolia Benth., Bot. Voy. Sulph. 37. 1844. Punta Frailes, February 16, Dawson 1134. Widely scattered in the sandy deserts from Texas to California, 166 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 Sonora and Baja California; type from Magdalena Bay, Baja California. A procumbent spreading viscid herb 1 m or more in diameter and locally forming extensive dispersed colonies; the sticky stem often with adher- ing sand particles, the long petiolate leaves cordate to oricular. The Papago Indians still use the tough fibers of the fruits in weaving baskets. BiGNONIACEAE Tecoma stans (L.) HBK., Nov. Gen. & Sp. 3:144. 1789. Canyon above Puerto Escondido, Sierra Giganta, March 13, Rempel 166, hillside. Widely distributed in tropical and subtropical America. It is often cultivated locally for its showy yellow blossoms and one of its most common names is ''lluvia de oro." In the Cape District it is commonly found in the rocky swales where run-off adds to the precipitated soil moisture. Acanthaceae Beloperone californica Benth., Bot. Voy. Sulph. 38. 1844. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 275 (sterile) ; January 27, Dawson 1029. Puerto Escondido, March 13, Rempel 165. Punta Frailes, February 16, Dawson 1141. San Jose del Cabo, February 17, Dawson 1217. Cabeza Ballena, Rempel 65. San Carlos Bay, Sonora, February 8, Dawson 1070. A subshrub with long flowering branches bearing orange-red flowers in the spring through the California Gulf Region and adjacent areas, and south to central Sinaloa along the sandy coast (Isla Tachechilte, January 20, Gentry 7127). Beloperone Purpusii Brge., Zoe 5:172. 1903. San Jose del Cabo, February 17, Dawson 1175. Known only from the Cape District of Baja California; type from San Felipe. The lower anther only is mucronate, the upper lip is bicus- pidate. It is suffrutescent or grows to a small shrub. Rarely collected. Berginia Palmeri Rose, C.N.H. 1:86. 1890. Island in Concepcion Bay, March 16, Rempel 197, 208. Puerto Es- condido, February 11, Dawson 1098. Known only from the southern part of the peninsula and adjacent Carmen Island ; type from Santa Rosalia. This is a small usually nondescript shrub with brittle shiny branches, sparse foliage, and lavender or pink flowers. It is rare in collections and has often been confused with B. virgata. Though very similar in appear- ance to the latter it is distinguished by the stalked glands of the inflor- NO. 2 GENTRY: LAND PLANTS 167 escence, and by the cordate bases of the uppermost leaves. Standley's key (C.N.H. 23:1337. 1926) does not make use of the important gland character and the foliage feature he uses is applicable only to the reduced leaves of the inflorescence; the stems and leaves of both species being narrow^ly lanceolate and acute at the base. Although he reports B. vir- gata as occurring in Baja California, it is doubtful if it actually occurs there. Others in following his key have also assigned Baja California to B.virgata, as Johnston (I.e. p. 1168) who attributed his collections from San Nicolas Bay and Carmen Island to B. virgata, but states that they have glandular calyces, which identifies them as B. Pahneri. Though these two species are closely related they appear to be geo- graphically distinct and no intergrading forms have been seen by the author. For the time being they should be kept distinct, or until such time as adequate collections can be brought together for study. Berginia virgata Harv. in Benth. & Hook, Rev. Gen. PI. 2:1097. 1873. Guaymas, Sonora, January 23, Dawson 1004 ; February 9, Dawson 1075. Coastal and foothill regions in Desert Shrub and Thorn Forest of Sonora from Puerto Libertad to the Rio Mayo country. A small slender strictly or openly branched shrub about 1 m tall with lavender flowers. Carlowrightia californica Brge., Zoe 5:172. 1903. Punta Frailes, February 16, Dawson 1146. San Jose del Cabo, Feb- ruary 17, Dawson 1196. Baja California, Sonora, and Sinaloa; type locality, southern Baja California. Doubtfully distinct from C. cordifolia Gray, described from southwestern Chihuahua and with a similar range on the mainland. Dicliptera resupinata (Vahl) Juss., Ann. Mus. Hist. Nat. Paris 9:268. 1807. San Carlos Bay, Sonora, February 8, Dawson 1071. Southern part of the California Gulf Region and south to tropical America. Jacobinia candicans (Nees) Benth. & Hook, in Hook. & Jacks., Ind. Kew. 1:1246. 1893. San Carlos Bay, Sonora, February 8, Dawson 1065. Common through the Thorn Forest of southern Sonora and Sinaloa, thence southeast along the Pacific coast to southern Mexico ; type from 168 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 the mountains of Oaxaca. It forms a low irregularly branched shrub with rather thin ovate to lanceolate acuminate leaves and short racemes of bilabiate red flowers in the spring. Closely related to /. mexicana, which occurs in the same region, but distinguished from it by the floral bracts which equal or exceed the calyx, while in /. rnexicana they are shorter than the calyx. Both plants belong with the Thorn Forest rather than with the desert, and their occurrence in the latter is marginal and restricted to bottomlands of overflow or the more moist canyon slopes. JusTiciA HIANS Brge., U. C. Publ. Bot. 6:194. 1915 and Proc. Calif. Acad. Sci. II, 2:194. 1889. San Jose del Cabo, February 17, Daivson 1200. Low suffrutescent herb rarely collected and apparently a postinsular endemic of the Cape District of Baja California. RuELLiA CALIFORNICA (Rosc) Jtn., Proc. Calif. Acad. Sci. IV, 12:1171. 1924. Guaymas, January 23, Dawson 1002. San Carlos Bay, February 8, Dawson 1073a. Agua Verde Bay, Baja California, March 10, Rempel 133, 138, rocky hillside. Island in Concepcion Bay, March 16, Rempel 193. Southern part of the gulf region, mainly on the rocky slopes of the coastal cerros; type from Santa Rosalia, Baja California. It forms a low shrub with twiggy branches often in dispersed small colonies with showy, lavender, campanulate, caducous flowers about 3 cm long. The foliage is vernicose or glutinous with dull, sparse, blunt or capitate hairs, erect or impacted in the surface excretion. The species is closely related to R. peninsularis, from which it is dis- tinguished by the longer, more attenuate, calyx lobes bearing clavate glandular "hairs." The indument on R. peninsularis although glandular is not clavate. Johnston's attempt to separate these two species on foliage characters alone is not altogether satisfactory ( Proc. Calif. Acad. Sci. IV, 12:1172), since both species are glandular and glutinous, nor is the foliage of R. peninsularis glabrate. The varnish tends to accumulate on the leaf surface with age and in time may submerge the ''hairs" in a glutinous film, a condition apparently mistaken by Johnston for glabrate. It is common to both species. R. peninsularis appears to be limited to the peninsula, since all the collections I have reviewed from the mainland are referrable to R. calif ornica. Johnston assigned some Guaymas collections to R. peninsularis, but in view of the criteria used in separating the two species, I believe he NO. 2 gentry: LAND PLANTS 169 was In error, although I have not seen the specimens he cites. This is not surprising since Rose's original descriptions are inadequate. Leonard of the United States National Herbarium, who has access to the type col- lections, has kindly determined Dawson's Guaymas collections. With his assistance I believe I am correct in restricting R. peninsularis from the Sonora flora until such time as it may be found in typical form. RuELLiA LEUCANTHA Brge., Zoe 5:109. 1901. Punta Frailes, February 16, Dawson 1135. Endemic to the postinsular Cape District. It is a suffrutescent peren- nial with white flowers and densely tomentose leaves. It is rare in col- lections. RuELLiA PENINSULARIS (Rose) Jtn., Proc. Calif. Acad. Sci. IV, 12:1172.1924. San Jose del Cabo, February 17, Dawson 1163. Southern part of Baja California; the type from mesas about La Paz. Plantaginaceae Plantago insularis fastigiata (Morris) Jeps., Man. Fl. PL Calif. 956. 1925. Puerto Refugio, Angel de la Guardia Island, March 20, Rempel 269. A low villous annual common to the deserts of the California Gulf Region. RUBIACEAE HousTONiA ASPERULOiDES (Benth.) Gray, Proc. Am. Acad. Sci. 5:158. 1860. Houstonia Brandegeana Rose, C.N.H. 1:70. 1890. Punta Frailes, February 16, Dawson 1113, 1132. San Jose del Cabo, February 17, Dawson 1187. Apparently limited to southern Baja California. A finely cut annual with purple flowers. Although the flowers on the specimens from Punta Frailes are smaller than those described for the species (3 to 4 mm high), the calyx is sparsely strigose and the lobes acute, two characters agreeing with H. asperuloides, and there are no other significant differences. Houstonia mucronata (Benth.) Rob., Proc. Am. Acad. Sci. 45: 401. 1910. San Francisco Island, March 9, Rempel 110, beach dunes. Southern Baja California and the adjacent islands; type from Mag- dalena Bay. It forms a low shrubby bush 2 to 9 dm high with white corollas which turn black on drying. Reported by Johnston to be abun- dant on some of the islands in the southern part of the gulf. 170 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 13 MiTRACARPUs PORTORICENSIS Urb., Symb. Antill. 4:609. 1911. Doubtfully referred to this Caribbean species is Dawson 1184 from San Jose del Cabo, Februar}^ 17 (flowering). Standley, who determined the collection, reported that he had not been able to place the specimen satisfactorily. CUCURBITACEAE CucuMis DIPSACEUS Ehrenb. in Sprach, Hist. Veg. Phan. 6:211. 1838. San Jose del Cabo, February 17, Dawson 1194. Apparently originally from the north African highlands, it is now widely but discontinuously dispersed in both the Old and New Worlds. In North America it has been collected from such widely separated regions as Oregon, Baja California, and the West Indies. EcHiNOPEPON MINIMUS (Kell.) Wats., Proc. A.m. Acad. Sci. 24: 52. 1889. Marah mini?na Kell., Proc. Calif. Acad. Sci. 2:18. Canyon above Puerto Escondido, Sierra Giganta, ]\larch 13, Rempel 752; April 20-22, 1938, Gentry 3747 (fruiting). Rather widely scattered in central and southern Baja California and the adjacent islands on the outer coast; type from Cedros Island. Appar- ently annual, it has very slender and rather short stems, about the smallest of the cucurbit vines in the region, with densely echinate fruits, dehiscing irregularlj'^, the prickles rather short and somewhat flattened. Echinopepon peninsularis Gentry sp. nov. Herba annua; caulis ramique graciles, striati, ad nodos pilis longis albis praediti; petioli 2-3 cm iongi; folia membranacea, 4-6 cm longa, 6-8 cm lata, lobis 5, late lanceolatis, serrulatis, aristatis, sparse scabro- pubescentibus ; cirrhi graciles 1-2-fidi;