PROCEEDINGS OF THE California Academy of Sciences FOURTH SERIES Vol. XL SAN FRANCISCO PUBLISHED BY THE ACADEMY 1974-1975 COMMITTEE ON PUBLICATION George E. Lindsay, Chairman Diana R. Young, Editor Paul H. Arnaud, Jr. William N. Eschmeyer Alan E. Leviton CONTENTS OF VOLUME XL Pages No. 1. Williams, Stanley C. A new genus of North American scorpions with a key to the North American genera of Vaejovidae (Scorpionida: Vaejovidae). Published October 30, 1974 1-16 No. 2. Carter, Annetta M. The genus Cercidium (Leguminosae: Caesalpinioideae) in the Sonoran Desert of Mexico and the United States. Published October 30, 1974 17-57 No. 3. Hallacher, Leon E. The comparative morphology of ex- trinsic gasbladder musculature in the scorpionfish genus Sebastes (Pisces: Scorpaenidae). Published October 30, 1974 59-86 No. 4. Murphy, Robert W. A new genus and species of euble- pharine gecko (Sauria: Gekkonidae) from Baja California, Mexico. Published October 30, 1974 ........ 87-92 No. 5. Murphy, Robert W. Two new blind snakes (Serpentes: Leptotyphlopidae ) from Baja California, Mexico with a contribution to the biogeography of peninsular and insular herpetofauna. Published February 18, 1975 93-107 No. 6. Chen, Lo-chai. The rockfishes, genus Sebastes (Scor- paenidae), of the Gulf of California, including three new species, with a discussion of their origin. Published Feb- ruary 18, 1975 . 109-141 No. 7. Hayes, Alan H. The larger moths of the Galapagos Islands (Geometroidea: Sphingoidea & Noctuoidea). Published August 8, 1975 __„ 145-208i No. 8. Smith, Alan Reid. New species and new combinations of ferns from Chiapas, Mexico. Published August 8, 1975 209-230 No. 9. Zavortink, Thomas J. A new genus and species of eucerine bee from North America (Hymenoptera: Anthophoridae). Published August 8, 1975 231-242 No. 10. Gold, J. R., and G. A. E. Gall. The taxonomic structure of six golden trout (Salrno aguabonita) populations from the Sierra Nevada, California (Pisces: Salmonidae). Pub- lished August 8, 1975 _ 243-263 NO. 11. ESCHMEYER, WlLLIAM E., AND JOHN E. RANDALL. The scorpaenid fishes of the Hawaiian Islands, including new species and new records (Pisces: Scorpaenidae). Published October 3, 1975 265-334 No. 12. Howard, L. D. Muscular anatomy of the hind limb of the sea otter (Enhydra lutris) . Published October 3, 1975 335-416 No. 13. McCosker, John E., and Richard H. Rosenblatt. The moray eels (Pisces: Muraenidae) of the Galapagos Islands, with new records and synonymies of extralimital species. Published October 3, 1975 _. _ 417-427 No. 14. Davis, John, and W. Z. Lidicker, Jr. The taxonomic status of the southern sea otter. Published October 3, 1975 _ 429-437 No. 15. Paxton, John R. Heraldia nocturna, a new genus and species of pipefish (family Syngnathidae) from eastern Aus- tralia, with comments on Maroubra perserrata Whitley. Published October 3, 1975 .. 439-447 Index to Volume XL 449-460 PROCEEDINGS OF THE ! Marine Biological Laborator LIBRARY NOV 1 1 CALIFORNIA ACADEMY OF SCIENCES Woods Hole, Mass. FOURTH SERIES Vol. XL, No. 1, pp. 1-16; 5 figs.; 1 table. October 30, 1974 A NEW GENUS OF NORTH AMERICAN SCORPIONS WITH A KEY TO THE NORTH AMERICAN GENERA OF VAEJOVIDAE (SCORPIONIDA: VAEJOVIDAE) By Stanley C. Williams Research Associate in Entomology, California Academy of Sciences San Francisco 941 18 Abstract: A new genus of North American vaejovid scorpions is described and named Nullibrotheas. Nullibrotheas is a monotypic genus, the only species being Niillibrotheas allenii (Wood). The systematic relationships of N. allenii are analyzed and it is concluded that the closest known relatives are in the vaejovid genera Uroctonus and Vaejovis. A key to the North American genera of Vaejovidae is included. Introduction In 1863, Horatio Wood described a new species of North American scorpion which he named "Scorpius allenii" (Wood, 1863a). This new taxon was based on specimens collected by L. J. Xantus de Vesey from Cabo San Lucas, in Baja California, Mexico, presumably between 1859 and 1861. This new species had the distinction of being one of the first species of North American scorpions to be described and named, and it has remained one of the most misunderstood taxa in North America. The actual identity and phylogenetic relationships of this species remained unknown until just recently. The main purposes of this study are to more clearly define the species 'allenii' to determine its geographical distribution, to study its systematic affinities, and to erect a new genus to permit appropriate taxonomic placement of 'allenii: Most of the specimens cited as new records are deposited in the collections of the California Academy of Sciences. [l] 2 CALIFORNIA ACADEMY OF SCIENCES (Proc. 4th Ser. Acknowledgments Much appreciation is due the following individuals and their respective institutions for cooperation and loan of specimens which materially aided this study: M. A. Cazier, Arizona State University; J. A. Chemsack, California Insect Survey, University of California, Berkeley; W. J. Gertsch, American Museum of Natural History; C. F. Harbison, San Diego Museum of Natural History; H. W. Levi, Harvard Museum of Comparative Zoology; P. H. Arnaud, Jr., California Academy of Sciences. Thanks are due Richard and Mary Lou Adcock for providing transportation to the islands in the Gulf of California aboard their boat Marisla in 1968, to Thomas and Doris Hearne, who sponsored an expedition to study the Gulf Islands aboard their boat Muy Pronto in 1969, and to Richard and Elenore Dwyer who sponsored another research expedition to the Gulf Islands aboard their boat Sea Quest in 1970. Appreciation is owed to Carolyn Mullinex for making technical drawings, to V. F. Lee for technical and field assistance, and to C. F. Williams for clerical assistance. This study was partially supported by the National Science Foundation through research grants GB 7679 and GB 23674 and the California Academy of Sciences through the use of the research facilities of the Department of Entomology. Nullibrotheas Williams, new genus Type-species: Scor plus allenii Wood, 1863. Description. Pedipalp palms greatly swollen, fingers shorter than palm or carapace; brachium with posterior border of inferior surface with six long trichobothria; fingers with supernumerary denticles flanking single principal row of denticles medially. Pectines with fulcra sub-triangular in shape; middle lamellae with elongate basal piece and one row of subcircular to ovate sclerites; males with longer comb, longer teeth, and greater number of teeth than females. Males with large distinct genital papillae on inner margins of genital operculum. Sternum pentagonal, with almost parallel sides; length approximates width or is slightly longer. Stigma of book lungs minute, circular in shape. Last tarsomere of walking legs with one row of short spines on ventral surface; two pedal spurs. Two median eyes on low ocular tubercule; median eyes entirely on anterior half of carapace ; lateral eyes three per group, third eye reduced to degenerate, often giving appearance of two-eyed condition. Nullibrotheas allenii (Wood) . (Figures 1-4; table 1.) Scorpius allenii Wood, 1863a, p. 360 (original description) ; 1863b, p. 107. Broteas allenii, Marx, 1887, p. 91. Uroctonns privus Karscii, 1879, p. 103 (original description); Hjelle, 1972, p. 28-29. Vol. XL] WILLIAMS: NEW SCORPION GENUS Table 1. Measurements (in millimeters) of Nullibrotheas allenii Wood, SCW #116 (1), topotypes from Cabo San Lucas, Baja California Sur, Mexico. Male Female Total length Carapace, length width (at median eyes) Metasoma, length segment I (length/width) segment II (length width) segment III (length /width) segment IV (length width) segment V (length width) Telson length Vesicle (length Width) depth Aculeus, length Pedipalp Humerus (length Width) Brachium (length width) Chela (length/width) depth movable finger, length fixed finger, length Pectines, teeth (left /right) Sternum (length width) Stigma (circular), diameter Number middle lamellae Genital operculum (length width) 42 5.0 4.1 2.2/3.1 2.6/2.7 2.9/2.6 3.5/2.4 5.7/2.3 6.0 4.4/2.4 2.0 1.6 3.4/1.7 3.9/1.6 7.7/2.7 4.8 4.3 2.4 11/11 1.1/1.1 0.2 6 0.8/1.6 30 4.5 3.7 1.5/2.4 1.9/2.2 2.0/2.0 2.6/1.8 4.3/1.8 4.2 2.8/1.8 1.4 1.4 3.0/1.4 3.5/1.4 6.7/2.5 3.6 3.5 2.2 8/9 1.1/1.2 0.2 0.6/1.7 Uroctonus mordax (part), Kraepelin, 1894, p. 194 (synonymy of U. privus) . Broteas jormosus Marx, 1889, p. 211. Broteas alleni, Ewing, 1928, pp. 3, 6 (part) ; Hoffmann, 1931, pp. 332-333; Kraepelin, 1899, p. 176 (lists as doubtful species of chactid) ; Gertsch, 1958, pp. 2-5; Gertsch and Soleglad, 1966, p. 1; Diaz Najera, 1970, pp. 113-114. Broteochactas allenii (part), Banks, 1910, p. 188. Type data. Scorpius allenii Wood, two cotypes (one male, one female); collected at Cabo San Lucas, Baja California Sur, Mexico; male cotype deposited in U. S. National Museum (Type number S-5, jar 2), the location of the female cotype is unknown. The male cotype is intact, badly yellowed, and not well preserved. This specimen was designated as a lectotype by Gertsch ( 1958). It appears to be a subadult. The type of Uroctonus privus Karsch is a holotype female with the following CALIFORNIA ACADEMY OF SCIENCES IProc. 4th Ser. Figure 1. Nullibrotheas allenii (Wood), dorsal and ventral views of male topotype from Cabo San Lucas, Baja California Sur, Mexico. Vol. XL] WILLIAMS: NEW SCORPION GENUS 5 data attached "Californien; leg. graber." It is deposited in the Zoological Museum of Berlin, DDR; catalog number 3036. The holotype is a small juvenile that does not differ significantly from the juveniles of A 7 , allenii collected recently in the Cape region of Baja California, Mexico. The holotype is somewhat faded with age but the color pattern is still apparent; the metasoma is severed between segments 2 and 3. Redescription based on topotypes from Cabo San Lucas. Variable species throughout range in coloration, cuticular granulation, hirsuteness, and body size; apparently forming many more or less isolated regional races. Typically with base color of cuticle golden yellow with underlying dusky to dark markings on carapace and mesosomal dorsum; pedipalps with dark reddish brown fingers; metasoma with inferior keels with more or less distinct dusky stripes; other parts of cuticle may or may not show underlying dusky markings. Carapace. Anterior margin usually with deep median emargination and rounded lateral ends; lateral eyes generally three per group (sometimes appar- ently two per group); median eyes two in number, located on elevated ocular tubercule; median diad completely on anterior half of carapace. Sternum large, pentagonal; lateral sides almost parallel; length approximates width or is slightly longer; with deep median longitudinal furrow along posterior half. Mesosoma. Terga 1 to 6 usually lustrous; seventh tergum with one pair short, obsolescent lateral keels, these granular; terga 1 to 6 lacking distinctly developed dorsal keels; sterna smooth and lustrous; stigma small, circular in shape; all sterna lack keels. Metasoma. Dorsal and dorsolateral keels present and granular on segments I to IV; inferior lateral keels smooth to crenulate on I, irregularly crenulate on II and III, serrate on IV and V. Inferior median keels smooth to obsolescent on I, crenulate to obsolescent on II, crenular on III, serrate on IV and V. Telson. Aculeus short, less than one-half length of vesicle; larger individuals normally with laterally swollen vesicle. Pectines. Male with middle lamellae consisting of elongate basal piece and about five or six subcircular to ovate sclerites in single row; fulcra subtriangular; pectinal teeth six to nine in females, 10 to 14 in males; female comb much smaller than male, proximal one-fourth of female comb lacking teeth. Genital operculum. In males completely divided longitudinally; in females divided only in posterior region of furrow; males with very large, conspicuous genital papillae, these lacking in females. Chelicerae. Fixed finger with one simple tooth and one greatly elevated bicuspid tooth; movable finger with superior border armed with three or four simple teeth in addition to terminus; inferior border with three or four small denticles, these not conspicuous. CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. Figure 2. Nullibrotheas allenii (Wood) figured from topotype male from Cabo San Lucas, Baja California Sur, Mexico. Top. Ventral view of pedipalp brachium. Bottom. Mesosomal sternum showing openings to book lungs. Pedipalps. Fingers very short, palm very deep and swollen; keels obsolete to smooth except for pronounced inferior keel; palm surface lustrous, smooth to finely granular. Movable finger distinctly shorter than carapace; fixed finger shorter than metasomal segment III. Each finger armed with one longitudinal row of small denticles; no distinct gap between fingers when chela closed. Fingers very hirsute. Brachium with six trichobothria on posterior border of inferior surface. Walking legs. Two pedal spurs; one row of minute, longitudinally arranged hairs on ventral surface of last tarsomere. Vol. XL] WILLIAMS: NEW SCORPION GENUS Figure 3. Nullibrotheas allenii (Wood), figured from topotypes from Cabo San Lucas, Baja California Sur, Mexico. Top left. Ventral view of male showing prosomal sternum, genital opercula, and pectines. Top right. Ventral view of female showing prosomal sternum, genital opercula, and pectines. Bottom left. Right anterolateral aspect of carapace showing three lateral eyes. Bottom right. Right anterolateral aspect of carapace showing two lateral eyes. Geographical distribution. Known only from the southern region of the Baja California peninsula and associated islands. Specimens have been collected from Cabo San Lucas north to El Coyote on Bahia Concepcion. Reports of this species in the United States appear to be due to errors in locality records and to mistaken identification of this species. New records. Known from the following localities in Baja California Sur, Mexico: 1 mi. SW. Rancho Canipole, elevation 800 ft., 16 May 1969, (S. C. Williams), 2 juveniles; 4 mi. SW. San Miguel de Comondu, elevation 900 ft., 15 May 1969, (S. C. Williams), 2 juveniles; 5 mi. SW. San Miguel de Comondu, CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. Figure 4. Nullibrotheas allenii (Wood), figured from topotypes from Cabo San Lucas, Baja California Sur, Mexico. Top left. Ventral view of male metasoma. Top right. Movable pedipalp finger of male. Bottom left. Female telson. Bottom right. Male telson. Vol. XL] WILLIAMS: NEW SCORPION GENUS 9 elevation 1000 ft., 2 July 1968, (S. C. Williams, M. A. Cazier), 1 male, 1 female, 1 juvenile; 5 to 10 mi. SW. San Miguel de Comondu, elevation 1000 ft., 3 July 1968, (S. C. Williams, M. A. Cazier), 1 female, 3 juveniles; San Jose de Comondu, IS February 1966, (V. Roth), 1 male, 1 female; 4 mi. W. La Purisima, elevation 375 ft., 1 July 1968, (S. C. Williams, M. A. Cazier), 1 female; Rancho Las Parras, 26 May 1970, (S. C. Williams, V. F. Lee), 1 juvenile; 9.9 mi. N. Loreto, 27 May 1970, (S. C. Williams, V. F. Lee), 1 juvenile; Loreto city dump, just N. Loreto, 20 June 1964, (Chris Parrish), 1 female; 8 mi. S. Loreto, base of La Gigantia, 27 January 1965, (V. Roth), 1 male juvenile; Puerto Escondido, 17 mi. S. Loreto, 27 May 1970, (S. C. Williams, V. F. Lee), 1 juvenile; 22 mi. NE. Santo Domingo, 15 February 1966, (V. Roth), 1 female, 1 juvenile; 51 mi. N. El Crucero, elevation 400 ft., 14 May 1969, (S. C. Williams), 1 male, 3 females, 7 juveniles; 10 mi. N. El Crucero, 15 February 1966, (V. Roth), 1 male juvenile, 1 female juvenile; Bahia Concepcion near El Coyote, 26.40 N., 111.50 W., 17 February 1966, (V. Roth), 3 males; 10.3 mi. SE. Santa Rita, 27 July 1968, (S. C. Williams, M. A. Cazier), 1 male, 1 female, 1 juvenile; 49.3 mi. SE. Santa Rita, 2 7 July 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; 49.3 mi. SE. Santa Rita, 27 July 1968, (S. C. Williams, M. A. Cazier), 1 male, 4 juveniles; 31.0 mi. W. Los Aripes, elevation 800 ft., 25 July 1968, (S. C. Williams, M. A. Cazier), 3 juveniles; 1 mi. E. Los Aripes, 8 July 1968, (S. C. Williams, M. A. Cazier), 1 male; 1.4 mi. W. El Coyote, 30 December 1958, (H. B. Leech), 1 male, 3 juveniles; La Paz, 1 to 3 February 1965, (V. Roth), 1 female, 3 juveniles; 2 mi. E. La Paz, elevation 50 ft., 5 July 1968, (S. C. Williams, M. A. Cazier), 1 male; 12.4 mi. E. La Paz on road to Las Cruces, Arroyo Agua de los Pozos, 4 January 1959 (A. Leviton), 1 juvenile; 2 mi. NW. Los Pozos, 23 July 1968, (S. C. Williams, M. Bentzien, W. Fox), 8 juveniles; 14.5 mi. E. La Paz on road to Las Cruces, 4 January 1959, (H. B. Leech), 1 juvenile; .25 mi. N. La Paz airport, 12 July 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; 6 mi. N. La Paz on road to Pichilingue, 6 January 1959, (H. B. Leech), 1 juvenile; 12.5 mi. N. La Paz on road to Pichilingue, 29 December 1958, (H. B. Leech), 1 female; Puerto Balandra, approximately 13 mi. N. La Paz, 3 September 1963, (P. R. & D. L. Craig), 1 juvenile female; 14 mi. NE. La Paz along A. Balandra road, 14 July 1968, (S. C. Williams, M. A. Cazier), 2 males, 1 female, 1 juvenile; 14 mi. NE. La Paz along A. Balandra road, 9 July 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; 75 mi. NW. La Paz, elevation 200 ft., 4 July 1968, (S. C. Williams, M. A. Cazier), 1 female; 5 mi. SW. La Paz, 2 August 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; Las Cruces, 5.0 mi. SW. Las Cruces, 30 July 1968, (S. C. Williams, M. A. Cazier), 2 females, 5 males, 4 juveniles; 5 to 6 mi. SW. La Paz, elevation 25 ft., 5 July 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; 14.8 mi. N. Todos Santos, elevation 500 ft., 24 July 1968, 10 CALIFORNIA ACADEMY OF SCIENCES LProc. 4th Ser. Fic.ure 5. Distribution of Xullibrothcas allenii (Wood) in Baja California Sur, Mexico. Vol. XL] WILLIAMS: NEW SCORPION GENUS 11 (S. C. Williams, M. A. Cazier), 1 female, 6 juveniles; 5.5 mi. NW. Todos Santos, road to La Pastora, 13 January 1959, (H. B. Leech), 1 juvenile; 5.9 mi. N. Todos Santos, elevation 500 ft., 24 July 1968, (S. C. Williams, M. A. Cazier), 14 males, 18 females; 9 mi. N. Todos Santos, elevation 200 ft., 4 May 1969, (S. C. Williams), 1 female; 3.5 mi. S. El Pescadero, elevation 20 ft., 23 July 1968, (S. C. Williams, M. A. Cazier), 1 male, 1 female, 4 juveniles; 4 mi. N. Tinaja, 23 July 1968, (S. C. Williams, M. Bentzien, W. Fox), 1 female, 1 juvenile; Santiago, 19 August 1964, (H. W. Campbell), 1 male, 1 female; Boca de la Sierra, near Miraflores, 10 February 1966, (V. Roth), 1 female; Bahia de los Frailes, 9 March 1947, (I. La Rivers), 1 male; 1.5 mi. NE. Punta Palmilla, elevation 50 ft., 16 July 1968, (S. C. Williams, M. A. Cazier), 4 males, 13 females; 1.5 mi. NE. Punta Palmilla, elevation 50 ft., 17 July 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; 2.2 mi. SW. Punta Palmilla, 17 July 1968, (S. C. Williams, M. A. Cazier), 3 juveniles; 3.5 mi. SW. Punta Palmilla, 17 July 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; 3.9 mi. SW. Punta Palmilla, 17 July 1968, (S. C. Williams, M. A. Cazier), 4 males, 6 females; 16.5 mi. NW. Cabo San Lucas, 23 July 1968, (S. C. Williams, M. Bentzien, W. Fox), 5 juveniles; 5 mi. N. Cabo San Lucas, 21 July 1968, (S. C. Williams, M. A. Cazier), 1 juvenile; 4 mi. N. Cabo San Lucas, 21 July 1968, (S. C. Williams, M. A. Cazier), 2 females, 1 juvenile; 2 mi. N. Cabo San Lucas, 19 July 1968, (S. C. Williams, M. A. Cazier), 1 male, 1 female, 1 juvenile; 3 mi. E. Cabo San Lucas, 18 July 1968, (S. C. Williams, M. A. Cazier), 2 females; Cabo San Lucas, 20 July 1968, (S. C. Williams, M. A. Cazier), 3 males, 9 females, topotypes; Isla Carmen, Puerto Ballandra, 24 May 1970, (S. C. Williams, V. F. Lee), 2 juveniles; Isla Carmen, Puerto Ballandra, 23 March 1971, (V. F. Lee), 1 female, 1 juvenile; Isla Carmen, Marquer Bar, 22 June 1964, (R. C. Banks), 1 juvenile; Isla Santa Catalina, 27 April 1964, (C. H. Croulet), 3 females, 6 juveniles; Isla Santa Catalina, NE. side, 10 April 1962, (Members Belvedere Expedition), 1 female, 1 juvenile; Isla Santa Catalina, NW. end, Punta Arena, 24 June 1964, (C. Parrish), 1 male; Isla Santa Catalina, cave on S. end, 25 June 1964, (R. C. Banks), 1 juvenile; Isla Santa Catalina, SW. end, 26 April 1964, (A. J. Sloan), 1 male, 1 female, 4 juveniles; Isla Santa Catalina, W. side, 9 April 1962, (Members Belvedere Expedition), 5 juveniles; Isla Coyote, 19 April 1962, (George Lindsay), 1 female, 2 juveniles; Isla San Diego, WNW. side, 18 April 1962, (C. Parrish, M. Soule), 1 female; Isla San Diego, 19 April 1962, (George Lindsay), 1 juvenile; Isla San Jose, N. side, 26 March 1971, (V. F. Lee), 3 juveniles; Isla San Jose, NE. side of Arroyo de Aguada, 11 April 1962, (Members Belvedere Expedition), 1 female, 3 juveniles; Isla San Jose, S. end near lagoon mouth, 27 June 1964, (C. Parrish), 1 juvenile; Isla San Jose, SW. end, near cattle ranch, 24 April 1964, (A. J. Sloan), 2 females; Isla San Jose, Los Ostiones, 12 April 1962, (Members Belvedere Expedition), 2 females, 1 juvenile; Isla Cayo, 28 June 1964, (C. Parrish, G. Lindsay, 12 CALIFORNIA ACADEMY OF SCIENCES IProc. 4th Ser. I. Wiggins), 1 male, 2 females, 2 juveniles; Isla San Francisco, SW. side, 25 May 1969, (S. C. Williams), 1 juvenile; Isla San Francisco, S. end, 17 April 1962, (R. C. Banks), 1 male; Isleta Rock, N. Partida Island, 20 April 1962, (M. Soule), 1 male, 1 juvenile; Isla Partida, central valley, elevation 25 ft., 9 July 1968, (S. C. Williams, M. Bentzien, W. Fox), 1 male, 2 juveniles; Isla Partida, dry lake, 29 June 1964, (I. Wiggins), 1 female; Isla Ballena, 21 April 1962, (C. Parrish), 3 juveniles; Isla Espiritu Santo, SW. shore, 7 July 1968, (S. C. Williams, M. A. Cazier), 1 male, 2 females, 10 juveniles, 1 female with 2i in litter; Isla Espiritu Santo, Bahia San Gabriel, 24 May 1969, (S. C. Williams), 5 juveniles; Isla Espiritu Santo, Bahia Canon, 29 June 1964, (I. Wiggins), 2 juveniles; Isla Espiritu Santo, Bahia Candelaro, 30 June 1964, (I. Wiggins), 1 juvenile; Isla Cerralvo, Piedras Gordas, 17 May 1970, (S. C. Williams, V. F. Lee), 5 juveniles; Isla Cerralvo, sand dune area on SW. side, 16 April 1962, (C. Parrish), 1 juvenile; Isla Cerralvo, Arroyo Aguaje, 15 April 1962, (G. Lindsay), 2 juveniles; Isla Cerralvo, Rancho Ruffo, 16 April 1962, (I. Wiggins, G. Lindsay), 1 juvenile; Isla Magdalena, Puerto Magdalena, 17 March 1957, (R. Zweifel), 1 male; Santa Margarita Island, June 1970, (USNM), 1 specimen. Remarks. Nullibrotheas allenii is an extremely variable species in coloration, cuticular granulation, hirsuteness of metasoma and telson, and body size. Throughout its distribution it appears to form series of local races with some evidence of genetic continuity between many of the races. For this reason it appears best to treat this as one variable species rather than as a complex of many poorly defineable, separate ones. In regard to size, adults in most populations range from 30 to 45 millimeters in length. A few local populations have been found in which adults have attained much larger sizes, in the range of 50 to 60 millimeters. Such giant races have been found in the area of La Paz and on some of the islands of the gulf, notably on Isla Catalina. In regard to color, most populations have a golden yellow base color with underlying dark coloration on the carapace and mesosomal dorsum. Some populations have the dark markings reduced to faint dusky markings, notably on the Magdalena Plain, while others may have these markings very dense and dark, notably in the volcanic habitats, such as around Comondu. South of Todos Santos the dark coloration of the mesosomal dorsum is divided longitu- dinally by a thin unpigmented stripe thus creating the appearance of one pair of dark longitudinal stripes. North of the Todos Santos area the metasomal dorsum is usually completely pigmented giving no indication of stripes. Around the Todos Santos and Las Cruces areas individuals are variable and show all forms of intermediate conditions. The granulation of the cuticle, especially on the carapace, vesicle, and Vol. XL] WILLIAMS: NEW SCORPION GENUS 13 pedipalp palms is highly variable and difficult to interpret. It appears that the amount of granulation is determined by age, sex, length of time since last moult, and individual genetics. Most individuals with more distinctive granulation are larger than average. The hirsuteness of the vesicle and metasoma is another variable character. In the Cabo San Lucas area the metasoma and telson are usually only moderately hirsute. The northern and island populations generally had both vesicle and metasomal segments IV and V considerably more hirsute than the southern populations. Samples taken from intermediate areas indicate the variation in hirsuteness is to some extent individual, and developmental, and over the range is somewhat clinal. The collection records of this species indicate that individuals probably spend most of their life cycle in burrows in the ground. They do, however, appear to leave the burrow and seek shelter under suitable surface rocks during part of the year. Of the hundreds of specimens collected during this study only two females were observed with litters on their backs. Both of these were discovered under the protection of moderate-size flat rocks during July. The females appear normally not to leave their protective shelters at time of birth or while carrying young on their backs. Both females had only small to moderate numbers of young in their litters. The first litter was found 4 miles N. of Tinaja on 23 July 1968 and consisted of the mother and 8 young. The second litter was found on Isla Espiritu Santo on 7 July 1968 and consisted of the mother and 23 young. During the summer of 1970 the male cotype in the U. S. National Museum was examined and compared with topotypes recently collected at Cabo San Lucas. The type was almost identical to a sub-adult male topotype except for being more yellowish, less hirsute, and lacking the characteristic color pattern (three characteristics commonly altered in this way by poor preservation and age). It was interesting that the cotype was labeled H Broteas alleni Wood" by an old hand-written label included in the specimen jar. This was apparently a later addition or alteration of the original data since this species was originally named Scorpius allenii. It is notable that when Gertsch (1958) studied and redescribed the types two cotypes were present. In 1970 only one of the cotypes could be located, this presumably is the lectotype designated by Gertsch in 1958. Taxonomic relationships of Nvllibrotheas. Nullibrotheas allenii was first placed within the genus Scorpius by Wood in 1863 where it remained until 1887 at which time George Marx placed it within the chactid genus Broteas. Since this species has been relatively unstudied it has essentially remained placed in the genus Broteas and family Chactidae until now. However, Banks (1910) apparently realized that 'allenii' was not properly placed within the genus Broteas because he assigned this species to the genus Broteochactas thus 14 CALIFORNIA ACADEMY OF SCIENCES IProc. 4th Ser. retaining it within the family Chactidae. Banks' generic assignment was, however, not followed by others working with North American scorpions. It is interesting that European workers did not agree with the systematic placement of 'aUemi' by American workers. In 1879, Karsch studied a juvenile specimen of N. allenii and named it Uroctonus privus, thus placing it within the family Vaejovidae. Later Kraepelin (1899), considered U. privus (a junior synonym of N. allenii) to be a junior synonym of Uroctonus mordax, thus also retaining N. allenii within the family Vaejovidae. Kraepelin (1899) did not see the relationship between u Broteas allenii" and Uroctonus privus, but he did indicate that he believed U B. allenii" to be a good species, and doubted that it should be placed in the Chactidae. The family placement of the genus Nullibrotheas is made somewhat difficult by the obscure relationship between the families Chactidae and Vaejovidae. The primary characteristic that has been used to differentiate these two families is the presence of two lateral ocelli at each anterolateral corner of the carapace in the Chactidae. The family Vaejovidae is contrasted by having three or more lateral ocelli at each anterolateral corner of the carapace. Numbers of lateral ocelli are not satisfactory criteria alone for placement of scorpions within a family. Several species which clearly belong to families other than the Chactidae have secondarily lost lateral ocelli. For example, in the Diplocentridae several species belonging to the Didymocentrus group show only two lateral ocelli at each anterolateral corner of the carapace. These were even considered by Stahnke (1968) to represent a new diplocentrid genus which he called Bioculus. Gertsch (1972) has found several species of Vaejovis and Uroctonus which either have external signs of the third lateral ocellus completely absent or in an obsolescent state. Actually, most species of Vaejovis and Uroctonus have the third lateral ocellus showing signs of becoming reduced in size or approaching obsolescence. Nullibrotheas allenii clearly does not belong in the genus Broteas because it has the openings to the book lungs completely circular rather than linear, and the last tarsomere of the walking legs has one row of ventral hairs (not two parallel rows of spine-like hairs). Nullibrotheas allenii also clearly does not belong in the genus Broteochactas because it has six trichobothria in a posterior row on the ventral surface of the pedipalp brachium (not seven such trichobothria), and also because the walking legs have the last tarsomere with one row of ventral hairs (not with abundant irregular hairs). Study of the lateral ocelli of Nullibrotheas allenii reveals two distinct eyes at each anterolateral corner of the carapace in some specimens. However, individuals usually show evidence of a third reduced or degenerate ocellus. This line of evidence indicates that this genus should be placed within the family Vaejovidae. Furthermore, in other characteristics it appears somewhat related to the genera Uroctonus and Vaejovis (both members of the family Vaejovidae). Vol. XL1 WILLIAMS: NEW SCORPION GENUS 15 Nullibrotheas can be clearly distinguished from Vaejovis and Uroctonus in that it has the ventral surface of the pedipalp tibia (brachium) with six trichobothria in a posterior row (not with two or three such trichobothria) and has minute, round openings to the book lungs (not slit-like or oval) . Considering all the morphological evidence available it appears that Nullibrotheas has no known close living relatives, but that it is more closely related to the vaejovid genera Uroctonus and Vaejovis than to any other living ones known. Nullibrotheas is therefore here considered to be a member of the family Vaejovidae. Key to the North American genera of the family Vaejovidae 1. Book lung openings round ._ _ _ Nullibrotheas, new genus Book lung openings not round, but oval or elongate __ 2 2. Metasoma with single, unpaired, inferior median keel on segments I to IV genus Syntropis Metasoma with inferior median keels of segments I to IV paired or obsolescent 3 3. Lateral ocelli four per group; pectines with middle lamellae angular and not circular genus Anuroctonus Lateral eyes three or fewer per group ; pectines with middle lamellae subcircular 4 4. Inferior margin of movable cheliceral finger with one long, dark, conspicuous tooth ; pedipalps with ventral surface of brachium with over 20 trichobothria ._. genus Hadrurus Inferior margin of movable cheliceral finger with small teeth, minor denticles, denticle-like crenulation, or completely lacking denticles; pedipalps with ventral surface of brachium with two or three trichobothria along posterior border 5 5. Pectines with middle lamellae usually composed of 9 or fewer circular sclerites; openings to book lungs usually oval or long oval in shape and generally with length less than three times width; pedipalp with ratio of palm width to palm depth generally greater than 1.05; pedipalps with ventral surface of brachium with three trichobothria genus Uroctonus Pectines with middle lamellae usually composed of 10 or more circular sclerites; openings to book lungs usually elongate or slit-like and generally with length greater than three times width ; pedipalp with ratio of palm width to palm depth generally less than 1.05; pedipalps with ventral surface of brachium always with 2 trichobothria (never with three) 6 6. Chelicerae with inferior margin of movable finger with one or several small denticles or crenulations __. __ _ genus Paruroctonus Chelicerae with inferior margin of movable finger smooth and completely lacking crenulations genus Vaejovis References Banks, N. 1909-1910. The scorpions of California. Pomona College Journal of Entomology, vol. 2, no. 2, pp. 185-190. Diaz-Najera, A. 1970. Contribucion al conocimiento de los alacranes de Mexico (Scorpionida) . Revista de Investigation en Salud Publica, vol. 30, no. 2, pp. 111-122. 16 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. Ewing, H. 1928. The scorpions of the western part of the United States with notes on those occurring in northern Mexico. Proceedings of the United States National Museum, vol. 73, no. 9, pp. 1-24. Gertsch, W. J. 1958. Results of the Puritan-American Museum Expedition to western Mexico. 4. The scorpions. American Museum Novitates, no. 1903, pp. 1-20. Gertsch, W. J., and M. Soleglad 1966. The scorpions of the Vejovis boreus group (sub-genus Paruroctonus) in North America (Scorpionida: Vejovidae). American Museum Novitates, no. 2278, pp. 1-54. 1972. Studies of North American scorpions of the genera Uroctonus and Vejovis (Scorpionida, Vejovidae). Bulletin of the American Museum of Natural History, vol. 148, pp. 547-608. HjELLE, J. T. 1972. Scorpions of the Northern California Coast ranges (Arachnida: Scorpionida). California Academy of Sciences, Occasional Paper, no. 92, pp. 1-59. Hoffmann, C. C. 1931. Monografias para la entomologia medica de Mexico. Monografia num. 2, los scorpiones de Mexico. Primera parte: Diplocentridae, Chactidae, Vejovidae. Anales del Instituto de Biologica, Universidad de Mexico, vol. 2, no. 4, pp. 291-408. Karsch, R. 1879. Scorpionologische Beitrage. Ueber scorpione. Mittheilungen des Munchener Entomologischen Vereins, vol. 3, pp. 97-136. Kraepelin, K. 1894. Scorpiones und Pedipalpi. Das Tierreich, vol. 8, p. 182. Marx, G. 1887. Remarks on the types of Scorpionidae described by Wood. Proceedings of the Entomological Society of Washington, vol. 1, pp. 90-94. 1889. Arachnida in the scientific results of explorations by the U.S. Fish Commission steamer Albatross. Proceedings of the United States National Museum, vol. 12, pp. 207-211. Stahnke, H. L. 1968. Some diplocentrid scorpions from Baja California del Sur, Mexico. Proceedings of the California Academy of Sciences, fourth series, vol. 35, no. 14, pp. 273-320. Wood, H. C. 1863a. Descriptions of new species of north American pedipalpi. Proceedings of the Academy of Natural Sciences of Philadelphia, April, pp. 107-112. 1863b. On the pedipalpi of North America. Journal of the Philadelphia Academy of Natural Sciences, series 2, vol. 5, pp. 357-376. LIBRARY PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FOURTH SERIES Vol. XL, No. 2, pp. 17-57; 15 figs.; 1 table. October 30, 1974 THE GENUS CERCIDIUM (LEGUMINOSAE: CAESALPINIOIDEAE) IN THE SONORAN DESERT OF MEXICO AND THE UNITED STATES By Annetta M. Carter Herbarium, Department of Botany, 2017 LSB, University of California, Berkeley, California 94720 Cercidium comprises nine taxa, five of which occur in the Sonoran Desert of Mexico and southwestern United States (figs. 1, 2). Of these five, only one, C. praecox (Ruiz and Pavon) Harms, has a widespread distribution, occurring from western and southern Mexico south to Peru. Of the remaining taxa, two occur in Texas and northern Mexico and two in Argentina. Thus, it can be seen that the greatest concentration of species occurs in the Sonoran Desert (Shreve, 1951, 1964). Cercidiums, with their green branches which at a distance appear leafless much of the year, and are laden from March into June with a profusion of yellow, caesalpinoid flowers, are among the most conspicuous and characteristic trees of the Sonoran Desert. The newcomer to this region tends to refer to all these green-barked trees as palo verde, but if he goes into the field with a paisano, he soon learns that there is not only palo verde, but there are also dipua (dipuga), palo brea, and palo estribo (fig. 3). One other green-branched tree, closely related to Cercidium, also occurs in the Sonoran Desert. It is Parkinsonia aculeata Linnaeus, known locally in Baja California as junco. Continued experience with palos verdes in the field, especially in the Sierra de la Giganta of Baja California Sur, made it evident that they merited special attention in order to understand the relationships within the group. [17] 18 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. As with so many Sonoran Desert plants, most species of Cercidium present an entirely different aspect in the dry season than they do following the rains. Speaking generally, in the southern part of the Sonoran Desert, the heaviest rains fall in brief, hard-hitting storms during the hot months of July through September, sometimes extending into October (Hastings and Humphrey, 1969). With this moisture most plants come into full leaf, but with the end of the rains, many are soon leafless again. In 'good' years, gentle winter rains also come to the southern part of the Sonoran Desert and many species repeat the leaf cycle. Farther north, in south-central Arizona, where the total rainfall is about equally divided between summer and winter, Cercidium usually has two leafy seasons. All of the cercidiums flower at the height of the dry season, March through June, when for the most part, the trees are leafless or nearly so. Sometimes there is minor 'off-season' fall flowering. Keys to the taxa are included herein for both the vegetative and the flowering and fruiting stages. Although no attempt is made here to resolve the long-standing controversy as to whether Cercidium should be regarded as distinct from Parkinsonia, some points brought out in this paper may add fuel to the fire. Watson (1876) argued for union of these two genera; Sargent (1889), principally on the basis of the legume, maintained them as distinct; Johnston (1924a), on other grounds and after some realignment of the taxa within the two genera, also argued for their maintenance as distinct genera; Britton and Rose in 1930 erected the genus Cercidiopsis for Cercidium micro phyllum; Brenan (1963), on consideration of additional African taxa which he placed in Parkinsonia, felt that it is not possible to maintain Cercidium as distinct from Parkinsonia, but he refrained from making any nomenclatural changes in the American material. In accordance with Johnston's delimitation, Cercidium in America is easily distinguished from Parkinsonia and is kept separate in recent floras. The question of generic relationships cannot be resolved without first making extensive comparative studies of these trees (their biology, morphology, cytology, genetics, etc.) from seedling stage to maturity. Lack of evidence for resolution of problems at the generic level, however, does not preclude our attaining a better understanding of relationships at the subgeneric level. For the purposes of this paper, I am following Johnston's generic delimitation. The following key summarizes the differences between the species of Cercidium and Parkinsonia aculeata Linnaeus. Armature comprised of the first leaf developing at a node having a long-persistent, indurate petiole and rachis which terminates in a sharp, stout spine; leaves with pinnae 1 8 to 30 (-60) cm. long, the rachis 1 to 3.6 mm. wide, flattened, phyllodial, persistent; leaflets falling with drought, alternate and/or opposite, 10 to 40 on one 1 As used herein, the term "pinnae" refers to the secondary rachises plus leaflets of a bipinnate leaf. In many representatives of Caesalpinioideae the pinnae are reduced to a single pair. Vol. XL] CARTER: CERCIDIUM 19 32° — 28°— Cercidium 24° — R C.microphyllum .^ C.praecox C. X sonorae Figure 1. Distribution of Cercidium microphyllum, C praecox and C. X sonorae in the Sonoran Desert of northwestern Mexico and southwestern United States. side of a given rachis; axillary leaf-bearing shoots usually 2 to 12 (-23) mm. long; inflorescences developing with the leaves, the racemes much shorter than the pinnae; pedicels 10 to 21 mm. long (mean 14.4 mm.) Parkinsonia aculeata Armature of one or two axillary thorns, or lacking; leaves with pinnae 0.3 to 6.5 cm. 20 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. 32° — 28° — 24 Cercidium floridum ,_ ] subsp. floridum subsp. peninsulare 114° 110° Figure 2. Distribution of Cercidium floridum subsp. floridum and C. floridum subsp. peninsulare in the Sonoran Desert of northwestern Mexico and southwestern United States. long, the rachis less than 1 mm. wide, terete or sub-terete, not phyllodial, often falling with drought; leaflets opposite, 2 to 9 (-17) pairs; axillary leaf-bearing shoots so reduced that the leaves appear to arise in the primary leaf axil; inflorescences developing before the leaves or if with them, the racemes subequal to or exceeding the pinnae; pedicels 2.S to 14 mm. long (mean 7.5 mm.) ._ species of Cercidium Vol. XL] CARTER: CERCIDIUM 21 Herbarium material of Cercidium belonging to the following institutions has been studied: British Museum (Natural History), London (BM) ; California Academy of Sciences, San Francisco, California (CAS); Dudley Herbarium, Stanford University, California (DS); Escuela Nacional de Ciencias Biologicas, Instituto Politecnico Nacional, Mexico, D. F. (ENCB); Field Museum of Natural History, Chicago, Illinois (F) ; Gray Herbarium of Harvard University, Cambridge, Massachusetts (GH); Herbario Nacional del Instituto de Biologia, Universidad Nacional Autonoma de Mexico, Mexico, D. F. (MEXU) ; Instituto de Botanica, ' : Antonio Jose Cavanilles," Madrid (MA); Missouri Botanical Garden, St. Louis (MO); New York Botanical Garden (NY); Rancho Santa Ana Botanic Garden, Claremont, California (RSA); The Herbarium and Library, Royal Botanic Gardens, Kew (K) ; San Diego Museum of Natural History, California (SD) ; United States National Museum, Washington, D.C. (US) ; University of Arizona, Tucson (ARIZ) ; University of California, Berkeley (UC). Abbreviations for these institutions are those given by Lanjouw and Stafleu (1964). Collections made by the author will be widely distributed. Specimens cited include, in addition to those indicating limits of distribution, representative flowering and fruiting specimens as well as all specimens cited in illustrative material. I wish to express my appreciation to the directors and curators of the herbaria at these institutions for enabling me to study their collections. In citing collections from the peninsula of Baja California, political sub- divisions are indicated as follows: Baja California Sur for the area from 28°N. southward to the tip of the peninsula, a territory of Mexico; Baja California [Norte] for the area from 28°N. northward to the Mexico-United States border, a state of Mexico. Cercidium Tulasne Cercidium Tulasne, 1844, Arch. Mus. Paris, vol. 4, p. 133. Trees or large shrubs with smooth greenish bark, the branches usually armed with axillary thorns, or having short, spine-tipped branches. Stipules slender, foliaceous, caducous. Clusters of dry axillary bud scales usually present. Leaves usually not long persistent, sessile or petiolate, bipinnate with one to three pairs of pinnae, the leaflets small, opposite; leaf rachis terete or sub-terete, 1 mm. or less in diameter and terminating in a slender foliaceous or semi-indurate bract. Flowers borne in condensed to open axillary racemes. Pedicels jointed, the joint being in the distal half. Calyx tube green, broadly campanulate, 1.5-2 mm. long, 3-4 mm. wide, the lobes 5, narrowly ovate, valvate or subvalvate, 4-7 mm. long, 2 mm. wide, yellow or greenish yellow. Petals 5, yellow or creamy white, pubescent at base, the upper (posterior) with a conspicuous claw which usually elevates the limb of this petal above the others, the base of the limb auriculate, the auricles often curved over the apex of the claw to form a channel ; other four 22 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. p eninsu lare •-• '> -A A 4k\ i J 16 / 4134 ^T — i — i — niy Figure 3. Comparison of leaves, legumes, and seeds of Cercidium: A, C. micro phyllum, the upper figure is a single pinna of a bipinnate, sessile leaf; B, C. X sonorae, the leaves show variation in petiole length from subsessile to short-petiolate (one pinna has fallen from a leaf in 4945) ; C, C. praecox; D, C. jloridum subsp. peninsulare. Numbers refer to Carter, and/or Carter et al. collections ; data for these are given under "Representative collections cited" for the respective taxa. petals with shorter claws. Stamens 10, distinct; filaments pilose near the base; anthers versatile. Pods oblong to linear-oblong, or linear and torulose, coriaceous or papery. Seeds 1 to 4, flattened or subglobose. Based on Cercidium spinosum Tulasne, 1844, p. 136: "Regionem Amazonum (Bonplandi herb, propr. nunc in herb. Mus. Paris), Colombiam prope Maracaibo (Plee, herb. No. 73), nee non prov. Oaxaca Novae Hispaniae propter Tehuacan (in Cordillera alt. 1700 metr. — Galeotti herb. No. 3212) habitat." Key to Cercidium in the Sonoran Desert (Based on vegetative characters) Branches spinescent at tip; bark yellow-green, horizontally striate on older branches; axillary thorns lacking; leaves sessile, the pinnae of the primary 2 leaves borne on a scale-like, semi-indurate rachis which is tardily deciduous, the leaflets minute, O.S-S mm. long 1. C. micro phyllum 2 The term "primary" leaf refers herein to the first leaf developing at a node in contrast to those developing subsequently from a succession of buds produced on a reduced shoot in the axil of the primary leaf or primary leaf scar. Vol. XL] CARTER: CERCIDIUM 23 Branches not spinescent at tip; axillary thorns usually present; leaves petiolate (sometimes subsessile in C. X sonorae), the leaflets larger, usually 3-12 mm. long. Pinnae usually bearing 4-12 (-17) pairs of leaflets. Nodes bearing one or two stout axillary thorns 2-18 (-25) mm. long and having conspicuous clusters of dry, dark bud scales in the axils; bark waxy-coated on older branches and having a fine, quadrate-pustulate pattern; petioles 4-11 mm. long. 2. C. praecox Nodes bearing a single slender axillary thorn 2-11 (-20 mm.) long (or sometimes none) ; dry bud scales present at nodes but usually not conspicuous; bark on older branches smooth or inconspicuously horizontally-striate, the waxy coat inconspicuous or lacking; petioles 0-3 (-12) mm. long 3. C. X sonorae Pinnae usually bearing 2-4 pairs of leaflets; nodes without conspicuous clusters of dark axillary bud scales; bark of older branches horizontally striate ... 4. C. jloridum Branchlets of mature trees glabrous to glabrate ; leaflets 4-8 mm. long, mostly 3 pairs per pinna (Sonoran Desert except Baja California). 4a. C. jloridum subsp. jloridum Branchlets villous or pilose, leaflets 6-15 mm. long, mostly 2 pairs per pinna; clusters of reddish brown glandular hairs in axil of primary leaves (confined to southern Baja California). 4b. C. jloridum subsp. peninsulare Key to Cercidium in the Sonoran Desert (Based on flowering and fruiting characters) Ovary glabrous at anthesis; legumes not strongly long- tapered at ends. Inflorescences borne in clusters along older branches; petals deep yellow, the upper often orange-dotted near base of limb ; mature legumes papery, flat, conspicuously net-veined; seeds flattened, gray-brown with brown mottling. ._.. 2. C. praecox Inflorescences on terminal or subterminal branches ; legumes when immature bearing conspicuous white dots (stomata) but these obscure in mature legumes, coriaceous, the veins inconspicuous ; seeds oblong to ovate, flattened, brown with marginal area lighter and sometimes faintly mottled. 4. C. jloridum Inflorescences open, the racemes with rachis 1.0-4.5 (-7.0) cm. long, pedicels 6-12 (-20) mm. long; petals bright yellow, the upper sometimes orange- dotted (Sonoran Desert except Baja California). .... 4a. C. jloridum subsp. jloridum Inflorescences somewhat congested, the racemes with rachis mostly 0.3-1.0 (-2.0) cm. long; pedicels 4-9 (-12) mm. long; petals deep or bright yellow, the upper not orange-dotted (confined to southern Baja California) 4b. C. jloridum subsp. peninsulare Ovary lightly pubescent to strigose at anthesis ; legumes strongly tapered at both ends. Legumes torulose, lightly striate-veined ; petals light yellow 3 except for the upper one which is white or creamy-yellow but never orange-dotted; seeds sub-globose, brown. 1. C. microphyllum Legumes sinuate, conspicuously striate-veined; petals light yellow 3 except for the upper one which is variable (whitish or pale yellow and sometimes orange- dotted) ; seeds oblong to ovate, flattened, dark mottled brown. 3. C. X sonorae 3 Petal color is useful only with fresh material; use vegetative key to distinguish these two species if only dry material is available. 24 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. 1. Cercidium microphyllum (Torrey) Rose and Johnston. Cercidhim microphyllum (Torrey) Rose and Johnston in I. M. Johnston, 1924, Contr. Gray Herb., vol. 70, p. 66. Parkinsonia microphylla Torrey, 1857, Pac. R. R. Rep., vol. 4, p. 82; 1859, Bot. Mex. Bound, p. 59. Cercidiopsis microphylla Britton and Rose, 1930, N. Amer. Fl., vol. 23, p. 306. Shrub or tree 3-4 m. tall and nearly as broad, occasionally up to 8 m., the branches usually ascending and often presenting a broom-like appearance; bark yellow-green with fine horizontal striations; branches pubescent (strongly strigose in Baja California), puberulent or pilose to glabrate, the branchlets rigidly divaricate, spinose-tipped ; axillary thorns lacking, the nodes bearing few to many bud scales; leaves 1-2 per node, sessile [i.e. no petiole evident on the common rachis except in seedlings (fig. 7b)], the single pair of pinnae (3-65 mm. long) of the primary leaves borne on a scale-like, indurate rachis which is some- times tardily deciduous after the pinnae fall; leaflets 1-7 (-13) pairs, broadly elliptical, obtuse to emarginate at apex, 0.5-5 mm. long, 0.75-1.5 (-2) mm. broad, puberulent when young; rachis of the raceme 0.2-4.5 cm. long, puberulent, bearing 1-6 (-10) flowers on pedicels 4-8 (-15) mm. long; calyx pubescent to glabrate; petals yellow except for the upper (posterior) whose limb is usually white but occasionally creamy or pale yellow; upper petal 7-10 mm. long, the limb 4-5 mm. long, 3.5-6.0 mm. wide, rhomboidal or broadly ovate, the claw 3-4 mm. long, subequal to the limb in length, the other four petals slightly shorter than the upper, the limb rhomboidal or lanceolate, the margins not overlapping one another at anthesis, the claw short, 1-2 mm. long; ovary at anthesis strigose to puberulent; legumes 3.5-11.0 cm. long, 7-9 mm. wide, strongly tapered basally and apically, torulose, circular in cross-section where the seeds are borne, longi- tudinally striate, not coriaceous, dehiscence irregular, the valves often breaking as well as the sutures opening; seeds 1-4, brown, sub-globose, 8 (-10) mm. long, 5-7 mm. wide. Type. Mexican Boundary Survey: Diluvial banks of the Colorado, Ft. Yuma, 13 January 1854. A. F. Schott (NY, holotype; F!, date of 1855 is in error) [flowering specimen]; Williams' River [Arizona], 12-22 February [nSA],J.M.Bigelow (NY.) [in fruit]. Representative specimens. MEXICO. Gulf of California Islands, Baja California Sur: S. end of Amortajada Bay, San Jose Island, 11 April 1952, R. Moran 3772 (DS, UC) ; Puerto Balandra, Isla Carmen, 23 March 1971, /. R. Hastings 71-131 (ARIZ, SD); Danzante Island, 7 April 1962, Moran 9228 (CAS, SD) ; Arroyo de los Chivos, NE. side of San Marcos Island, 29 March 1962, Moran 8981 (CAS, RSA, SD). Baja California [Norte]: main arroyo, San Esteban Island, 26 April 1966, Moran 13054 (DS) ; ca. M.> mi. S. of Refugio Vol. XL] CARTER: CERCIDIUM 25 Bay, Isla Angel de la Guarda, 26 March 1963, Moran 10427 (DS, UC). Sonora: Tiburon Island, 5 May 1952, Moran 4064 (DS, SD, UC, US), 23 April 1966, Moran 12992 (ARIZ, CAS, RSA, SD). Peninsula, Baja California Sur: Rancho El Salto, Arroyo Coyote (ca. 24°47'N., 110°50'W.), altitude 525 m., 5 November 1971, Moran 19020 (14 mi. [22.4 km.] (by road) W. of San Luis Gonzaga, 21 October 1964, Turner & Hastings 64-380 (ARIZ); Agua Verde, 1 April 1911, /. A 7 . Rose 16578 (US); Portus Escondido, 2 February 1842, Wosnessensky (GH); 55 km. E. of Villa Insurgentes on highway to Loreto, 3 May 1972, Carter 5670: 1 km. W. of Las Parras summit, road from Loreto to San Javier, 6 May 1972, Carter 5672; Rancho La Venta, ca. 16 km. westerly from Loreto on road to San Javier, 21 April 1962, Carter 4415 \ Rancho Aguajito, Arroyo Gua between Loreto and Rancho Sauce, 24 April 1955, Carter & Ferris 3446; Arroyo Gua, N. of Loreto, 7 November 1960, Carter 4110; Rancho Naucajoa (26°16'N., 111°36.5'W.), W. of Llanos de San Juan, Carter & Reese 4531; Coyote Bay [Cove], Conception Bay, 18 June 1921, Johnston 4172 (CAS, F, K, MO, UC, US); wash 25 mi. [40 km.] S. of San Ignacio, 19 April 1931, Wiggins 5434 (CAS, DS, GH, RSA) ; eastern bajada of Sierra Calvario, Systema de Sierra Viscaino, 10-15 March 1947, H. S. Gentry 7501 (DS). Baja California [Norte]: ca. 5 mi. [8 km.] N. of Mision de San Borja along road to Bahia de Los Angeles, 17 May 1959, /. L. & D. B. Wiggins 14857 (CAS, DS); Cajon de Santa Maria, 12 May 1889, T. S. Brandegee s.n. (DS, UC) ; San Luis Gonzales [Gonzaga] Bay, 29 April 1921, Johnston 3348 (CAS, US) ; San Felipe Desert between El Cajon and Algodones, along eastern foot of Sierra San Pedro Martir, altitude 2500 ft., [800 m.] , 10 May 1941, Wiggins 9845 (DS, UC, US). Sonora: island in harbor, Guaymas, 14 April 1921, Johnston 3084 (CAS, US); hills NW. of shrimp cannery, Guaymas, 6 April 1962, Carter 4363; 26 km. S. of Hermosillo on Guaymas road, 21 March 1934, Ferris 87 59 A (DS); between Hermosillo and Kino Bay, 21 December 1968, V. Rudd 3038, with P. Bauer & A. C. Fox (ARIZ, SD, US); Sonora Alta, 1830, Coulter 490 (K, 2 sheets, one lacking collection number); N. of bay, within 5 mi. [8 km.] of coast, vicinity of Libertad, 2 May 1928, E. H. Graham 3822 (DS) ; 7 mi. [11 km.] S. of Altar, 4 May 1928, Graham 3906 (DS) ; Pasa de San Luis [Poso de Luis on US sheet] , 4 June 1894, E. A. Mearns 2697 [Int. Bound. Commission] (DS, US). UNITED STATES. Arizona: Tumamoc Hill, Tucson Mts., Pima Co., 8 June 1938, H. S. Gentry 3777 (CAS) ; 5 mi. [8 km.] S. of Florence, Pinal Co., 26 July 1927, H. W. Graham s.n. (DS) ; Tempe Butte, altitude ca. 1300 ft. [415 m.], Maricopa Co., 4 May 1952, E. P. Killip 42150 (US) ; Government Springs, 3.5 mi. [5.6 km.] N. of Bumblebee, State Highway 69, Yavapai Co., 3 July 1940, Ferris 9902 (CAS, DS, GH, RSA, UC); 14 mi. [22 km.] NE. of Topock, 26 CALIFORNIA ACADEMY OF SCIENCES LProc. 4th Ser. Mohave Co., 10 June 1967, Carter 5243. California: 1 mi. [1.6 km.] S. of Copper Basin Lake, Whipple Mts., San Bernardino Co., 22 April 1940, A. M. Alexander & L. Kellogg 1200 (DS, GH, RSA, US); base of Whipple Mts. adjacent to Colorado River 11 mi. [18 km.] above Earp on road to Parker Dam, San Bernardino Co., 15 March 1940, Wolf 9721 (RSA). Coarse soil of plains and hillslopes, mostly below 600 m., but also between 700 and 800 m. in north central Sonora and southeastern Arizona. The trees leaf out following the rainy seasons, but the leaves are ephemeral. Flowering is mostly from March through May, the height of the Sonoran Desert dry season. Leaves, if present during flowering, are confined to non-floriferous branches. This is the most widespread of the taxa in Baja California where it is known as dipua and extends from a little south of Latitude 25°N. northward to the United States border. It is also abundant in west central Sonora and southwestern Arizona where it extends to Latitude 35°N. Even when leafless, the branches are an important source of food for stock. Trees with their upper branches lopped off by machetes indicate that a traveller rested nearby and provided food for his mule, burro, or horse. In times of drought, the branches are also cut to feed cattle. The seeds are edible. As an historical aside, it is interesting to note that Thomas Coulter collected both Cercidium microphyllum and C. floridum in "Sonora Alta" [vicinity of Hermosillo, Sonora] in 1830. According to the correspondence between Harvey and Bentham [Kew archives], some years after the packages of Coulter's col- lections were received at Trinity College, Dublin, Harvey separated the legumes and sent them to Bentham at Kew for identification. Specimens of both taxa are annotated in Bentham's hand as new species. At about that time, Asa Gray visited Kew, presumably saw all of the Coulter material, and subsequently published Cercidium floridum Benth. ex Gray. Five years later, John Torrey, in describing material collected on the Mexican Boundary Survey, named Parkins onia micro phylla, chose a Schott specimen as type, and made no mention of the much earlier Coulter collection to which Bentham had applied the same specific epithet. The following collections are putative hybrids between Cercidium micro- phyllum and C. floridum subsp. floridum: P. Kamb 2014. Bottom of NE. caldera of Molina Crater, altitude ca. 950 ft. [310 m.], crater region NW. of Sierra Pinacate, Sonora, Mexico, 29 April 1951 (DS, UC). The leaves are petiolate. The leaflets are larger than typical for C. microphyllum and there are too many leaflets for it to be C. floridum. It has the stem pubescence and the strigose ovary of C. microphyllum and thorns similar to those of C. floridum. On both of the above cited sheets of this collection, the flowering branch is the putative hybrid and the separate fruiting branch — presumably from a different tree — is typical C. floridum subsp. floridum. Both of the putative parents are Vol. XL] CARTER: CERCIDIUM 27 known to occur in this area (Hastings, Turner and Warren, 1972). A sheet at University of Arizona bears only a fruiting branch and this is typical C. jloridum subsp. jloridum. C. B. Wolf 9722. Base of Whipple Mts., adjacent to Colorado River, 11 mi. [18 km.] above Earp on road to Parker Dam, altitude ca. 400 ft. [130 m.], San Bernardino Co., California, 15 July 1940 (DS, RSA). Collector's note: "This lone tree looks like a typical C. microphyllum in olive-green color and shape, but the pods are rich brown, somewhat flattened, and not constricted between the seeds, which are larger and flattened. Leaflets somewhat larger than typical C. microphyllum and C. jloridum. Both species grow here in abundance." The valves of the pods are coriaceous, striate, and white-dotted. The stems are sparsely appressed-pubescent and bear short spines. Papers treating these and other putative hybrids in Cercidium are published elsewhere (Carter, 1974; Carter and Rem, 1974). 2. Cercidium praecox (Ruiz and Pavon) Harms. (Figure 4.) Cercidium praecox (Ruiz and Pavon) Harms, 1908, Bot. Jahrb., vol. 42, p. 91. Caesalpinia [Sappania] praecox Ruiz and Pavon, 1802, Fl. Peruv., vol. 4, pi. 376, plate only [Entire volume published by Consejo Superior de Investigaciones Cientificas, Madrid, 1957]. Caesalpinia praecox Ruiz and Pavon, 1833, in Hooker and Arnott, Bot. Misc., vol. 3, p. 208. Cercidium spinosum Tulasne, 1844, Arch. Mus. Paris, vol. 4, p. 134. Rhetinophloem viride Karsten, 1862, Fl. Columb., vol. 2, p. 25, pi. 113. Cercidium viride Karsten, 1887, Bot. Jahrb., vol. 8, p. 346. Cercidium plurijoliolatum M. Micheli, 1903, Mem. Soc. Phys. et Hist. Nat. Geneve, vol. 34, p. 269, pi. 18. Cercidium unijuga Rose, 1905, Contrib. U. S. Nat. Herb., vol. 8, p. 301, 1905. Cercidium Goldmanii Rose, 1905, Contrib. U.S. Nat. Herb., vol. 8, p. 301. [Goldman 735, type (US !) : leaves glabrous; waxy coating of stems only slightly visible.] Shrub or tree usually 2-A m. tall, but up to 9 m. in forested areas of north- eastern Sonora, the crown usually rounded, or flat-topped and spreading in exposed habitats, erect and less branched in sheltered habitats; branches and trunk bright green to the base, the bark with a minute quadrate-pustulate pattern and bearing a heavy coating of wax, the branchlets glabrate to pubescent, the hairs usually appressed; axillary bud scales prominent; thorns axillary, stout, usually one but sometimes two per node, 2-18 (-25) mm. long (mean 8.6 mm.), often dark brown; leaves 1-3 (-6) per node, pubescent, the petiole (1-) 4-11 (-21) mm. long, bearing 1 (occasionally 2-3) pair of pinnae 0.4-4.5 cm. (-5.0, Wiggins 6473) cm. long; leaflets (3-) 5-9 (-17, Wiggins 6473) pairs, oblong, rounded at apex, 3-10 mm. long, 1.4-3.8 mm. wide; inflorescences borne in 28 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. Figure 4. Cercidium praecox, palo brea, showing typical growth form in open habitat (Carter, Hastings & Turner 5577, northeastern Sonora between Moctezuma and La Noria). clusters along mature branches, usually compact, 1-3 (4) per node, the rachis (1.5-) 2-11 (-15) mm. long, sparingly pilose, bearing 1-6 (-9) flowers on pedicels 7-10 mm. long; petals deep yellow, the upper often orange-dotted near base of limb; upper petal 9-11 mm. long, the limb 6-7 mm. long, 6-8 mm. wide, broadly ovate, the claw 3-5 mm. long, shorter than the limb, the other four petals slightly shorter than the upper, the limb broadly ovate and sometimes auriculate at base, the claw 1-2 mm. long; ovary glabrous at anthesis; legumes 3-6 (-8) cm. long, 0.6-1.0 cm. wide, flat and papery, not narrowed between the seeds, the veins conspicuous, forming an elongate-reticulate pattern; seeds 1-2 per pod, oblong, flattened, gray-brown with dark brown mottling, up to 1 cm. long, 3-4 mm. wide. Type. Middle western Peru. Ruiz and Pavon plate, [cf. Johnston, 1924, p. 67.] (Ruiz and Pavon collections, MA! ). Representative specimens. MEXICO. Gulf of California Islands, Baja California Sur: summit, Ildefonso Island (26°37'N., 111°27'W.), 17 May 1921, Johnston 3753 (CAS), 2 April 1962, Moran 9066 (RSA, SD, US); Tortuga Island (27°26'N., 111°54'W.), 24 April 1952, Moran 4007 (DS), 30 March 1962, Moran 9016 (SD), 11 May 1921, Johnston 3592 (CAS, GH, K, UC, US). Peninsula, Baja California Sur: San Jose del Cabo, 26 March 1911, /. N. Rose 14466 (US) ; La Paz, 10 August 1944, Maximino Martinez s.n. Vol. XL] CARTER: CERCIDIUM 29 (US) ; along dry washes between Medano and Venancio, 29 April 1931, Wiggins 5532 (DS, RSA, UC, US); between Rancho Segundo Paso and San Javier, altitude 300 m., 21 April 1962, Carter 4412 ; between Canon de Las Calaveras and La Tinaja, western side of Mesa de San Alejo, altitude 690 m. (ca. 25°51'N., 111°36'W.), 11 November 1961, Carter 4306; along old mission trail SE. of Comondu, altitude 420 m., 20 April 1955, Carter & Ferris 3424 (ARIZ, DS, GH, SD, UC, US); Cuesta de Los Encinos, SE. of Cerro Giganta, altitude 500 m., 29 March 1960, Carter & Ferris 4046 (UC), same tree, 9 November 1960, Carter 4146 (UC); 4 mi. [6 km.] (by road) E. of San Lina (suburb of San Ignacio), 29 October 1963, Turner & Hastings 63-294 (ARIZ, DS, SD). Sinaloa: La Constancia, Munic. El Fuerte, December 1924 [sic! on original label at DS; US copied label is 1926], Jesus Gonzales Ortega 6200 (DS, GH, MENU, US). Sonora: Alamos study area, 7.5 mi. [12 km.] W. of Alamos, 28 April 1967, R. D. Krizman 16 (ARIZ); Agua Caliente N. of Alamos, 2 November 1939, Gentry 4839 (ARIZ, MO, distributed as C. torreyanum and so cited by Gentry, 1942, p. 131); San Bernardo, Rio Mayo, 1 March 1935, Gentry 1377 (ARIZ, GH, MEXU, UC, distributed as C. torreyanum and so cited by Gentry, 1942, p. 131); Cerro de Bayajori, 12 mi. [19 km.] W. of Navajoa, 11 April 1948, Gentry 7947 (UC, US, distributed as C. jloridum) ; island in Bay, Guaymas, 14 April 1921, Johnston 3078 (CAS, US); low hills and flats near tannery E. of Guaymas, 28 February 1933, Wiggins 6348 (DS, RSA, US); small valley 20 mi. [32 km.] N. of Guaymas, 8 March 1933, Wiggins 6473 [spines up to 30 mm. long, pinnae 45-50 mm. long, leaflets up to 15 pairs] (DS); 20 km. S. of Carbo junction, 3 May 1971, Carter, Hastings & Turner 5597; Horcasitas, 17 April 1932, Abrams 13360 (DS) ; 2 mi. [3 km.] S. of Los Hoyos, altitude 810 m., 23 April 1971, Carter, Hastings & Turner 5576; 4.2 km. S. of Los Hoyos, 4 July 1971, Hastings 71-199 (SD); Colonia Oaxaca, 24 July 1938, Stephen S. White 663 (ARIZ, GH). Ranging from coastal plains up to bajadas, mesas, hills, and mountains at elevations up to 825 m. in Baja California and 1115 m. in the mountains of northeastern Sonora. Flowering is from March through May with the peak in April, before the leaves develop. In Baja California, C. praecox extends from near the tip of the peninsula (23°N.) northward to the vicinity of San Ignacio (27° 25' N.). On the main- land, it occurs in southern Mexico and is abundant in Sonora from a little south of Guaymas (27° 56' N.) almost to the United States border (30° 54' N.). Occasional collections have been made as far south as 26°N. in Sinaloa. This is the only one of our Sonoran Desert species of Cercidium having a disjunct distribution. According to Johnston (1924a), in South America it occurs from extreme middle-western Peru to northern Venezuela. 30 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. During the months when the trees are leafless, C. praecox may be recognized readily by the bright green bark extending down to ground level, and the conspicuous accumulation of dark bud scales at the nodes. Development of new branch growth appears to be limited, and most seasonal growth (production of leaves and flowers) is axillary year after year on the old branches. In addition, the stout, often dark brown spines, which may reach a length of 2.6 cm., set it apart from the other taxa. It is the only one of our Cercidium species with bark having a quadrate-pustulate pattern. When in flower, the species may be recognized easily by the usually deep yellow flowers borne in close clusters along the length of the old branches. The common name, palo brea, is derived from the fact that the waxy substance coating the bark, after being scraped from the branches and melted by heat, is used as a 'gum' for gluing together leather objects and furniture; thus it is used just like la verdadera brea. The following Mexican collections are putative hybrids between Cercidium praecox and Parkinsonia aculeata: edge of town, Coyuca, Mina, Guerrero, 11 May 1934, G. B. Hinton 6040 (BM, NY), 25 March 1937, Hinton 9968; (ARIZ, BM, K, MEXU, NY, RSA, TEX) ; near Los Hoyos, northeastern Sonora, 23 April 1971, Carter, Hastings & Turner 5575 (to be distributed). This material is discussed in a separate paper (Carter and Rem, 1974) . 3. Cercidium X sonorae Rose and Johnston. (Figure 5.) Cercidium X sonorae Rose and Johnston, 1924 (pro sp.) stat. nov., in I. M. Johnston, Contr. Gray Herb., vol. 70, p. 66. (April). [C. micro phyllum (Torrey) Rose and Johnston X C. praecox (Ruiz and Pavon) Harms]. Cercidium molle I. M. Johnston, 1924, Proc. Calif. Acad. Sci., IV, vol. 12, p. 1038. (May). Spreading tree 4-8 m. tall, usually with lax branches (but sometimes the branches short and stiff), the bark smooth or sometimes faintly horizontally striate, green to yellow-green, sometimes with an inconspicuous, thin, waxy coating, branches short-villous, pilose, or sometimes glabrate; axillary thorns present or lacking, slender (2-) 5-11 (-20) mm. long (mean 6.5 mm.), variable as to presence on a given tree; bud scale clusters prominent to inconspicuous; foliage bright yellow-green, the leaves 1 to 3 per node, the petioles usually 1-3 mm. long, but varying from to 12 mm.; pinnae 1 to 2 pairs, 0.7-5.0 cm. long, the leaflets (3-) 6-8 (-12) pairs elliptical-oblong, (1.5-) 2-3 (-6.0) mm. long, (1.0-) 1.5 (-3.0) mm. wide; inflorescence usually open, the rachis of the racemes 0.5-4.0 cm. long, bearing 1 to 10 flowers on pedicels 6-10 (-14) mm. long; petals light yellow, except for the upper whose limb may be whitish, creamy, or creamy and yellow and often orange-dotted near the base of the limb; upper petal 10-12 mm. long, the limb 5-7 mm. long, 5-8 mm. wide, broadly ovate but with Vol. >>L] CARTER: CERCIDIUM 31 Figure 5. Cercidium X sonorae, palo estribo, showing typical habit of a mature tree (Carter & Reese 4554, Baja California, Sierra de la Giganta west of Loreto). pointed apex, the claw 4-5 mm. long, shorter than the limb; the other four petals slightly shorter than the upper, the limb ovate, the claw short, 1.5-3 mm. long; ovary strigose at an thesis; legumes 3-8 cm. long, 7-10 mm. wide, slightly flattened and somewhat coriaceous, sinuate if more than one-seeded, long- tapering to each end, the surface with conspicuous longitudinal striations, dehiscing along the sutures or irregularly; seeds 1 to 2 (4), oblong to oblong- ovate or ovoid, dark mottled brown, flattened or thick, 9-10 (-12) mm. long, 4-5 mm. wide. Type. Dry hills in the vicinity of Guaymas, Sonora, Mexico, 1910, Rose, Standley & Russell 12586 (US! ). Representative specimens. MEXICO. Baja California Sur: Arroyo San Ramon just W. of Rancho San Ramon (25°14.5'N., 111°17'W.), 21 October 1964, Carter 4818; Rancho Tasajera, ca. 3.5 km. NE. of San Jose de Agua Verde, 3 June 1965, Carter & Shar smith 4936; Agua Verde Bay, 26 May 1921, Johnston 3877 (CAS!, type of C. molle, DS, GH, K, MO, UC, US); Puerto Agua Verde [Bahia Agua Verde], 5 June 1965, Carter & Sharsmith 4945, 23 August 1971, Carter 5610; Mision San Javier, 36 km. SW. of Loreto, 6 May 1972, Carter 5679; Arroyo Ranchito, ca. 9 km. SE. of Llanos de San Julio on road from San Javier to Comondu (26°02'N., 111°39'W.), 5 June 1963, Carter & Reese 4554; Cuesta de Los Encinos, SE. of Cerro Giganta, altitude ca. 500 m., 32 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. 29 March 1960, Carter & Ferris 4045, and from same tree 9 November 1960, Carter 4145; La Higuera, NE. base of Cerro Giganta, 18 October 1966, Carter & Soma 5212 ;Arroyo Hondo, N. side of Cerro Giganta, altitude 540 m., 17 October 1966, Carter & Sousa 5199; Tortuga Island, 11 May 1921, Johnston 4409 (CAS, K, UC, US), 16 mi. [26 km.] from San Ignacio on road to Santa Rosalia, 10 March 1934, Ferris 8626 (US); 10.5 [17 km.] mi. ENE. of San Ignacio, 27.4°N., 112.8°W., altitude 170 m., 17 October 1971, Turner & Hastings 71-134 (ARIZ). Sonora: N. of Guaymas (0.6 mi. [.96 km.] N. of junction with Bahia San Carlos road), 2 May 1971, Carter, Hastings & Turner 5595; 25 mi. [40 km.] N. of Guaymas, 2 April 1935, Shreve 7310 (MO; sheets at ARIZ and F bear both C. X sonorae and C. praecox); 31 mi. [50> km.] N. of Guaymas, 2 April 1935, Shreve 7313 (ARIZ, F, MO); El Pozo, 26 km. S. of Hermosillo on road to Guaymas, 21 March 1934, Ferris 8761 (DS, US) ; Torres, 10 February 1903, F. V. Coville 1664 (US); Sierra Lopez Rancho [37 mi., 59 km. NW. of Hermosillo], 13 April 1932, Abrams 13327 (DS,F). Plains and hillslopes below 600 m. Flowering March to June with the peak in April; leaves usually present when trees in flower. In Baja California, where it is known as palo estribo, Cercidium X sonorae occurs principally in the Sierra de la Giganta from about 25°15'N. (opposite the northern end of Isla San Jose) northward to the northern slopes of Cerro Giganta (26°10'N.). One outlying collection (Turner & Hastings 71-134) is from east of San Ignacio (2 7.4°N.). In Sonora, C. X sonorae is most abundant from near Guaymas northward to the vicinity of Carbo (i.e., from about 28° to 30° N.). On neither side of the Gulf of California is C X sonorae as abundant as C. micro phy Hum and C. praecox, one or the other or both of which appear always to occur in the vicinity of trees of C. X sonorae. Throughout its range, but especially noticeable in Baja California, is the fact that C. X sonorae occurs only within the limits of the distributional overlap of C. microphyllum and C. praecox. In char- acters of bark surface, spininess, pubescence, flower color, and seed shape, C. X sonorae appears to be more variable than any of the other taxa in the group. In several morphological characters, C. X sonorae falls between C. praecox and C. microphyllum; furthermore, pollen of C. X sonorae stained with 'cotton blue' (aniline-blue-lactophenol) shows a much lower percentage of presumably viable grains than does the pollen of either of its putative parents. These facts lead to the hypothesis that C. X sonorae is of hybrid origin, the putative parents being C. microphyllum and C. praecox. Discussion of the hybrid origin of C. X sonorae and of other hybrids involving species of Cercidium is published else- where (Carter, 1974; Carter & Rem, 1974). In 1965 while on a field trip to Agua Verde Bay in search of Johnston's Cercidium molle, I was told that a large palo estribo tree comparing favorably to that described by Johnston (1924b) had been cut down some twenty years Vol. XL] CARTER: CERCIDIUM 33 before. On a subsequent trip I explored the "huge amphitheater-like canyon" (Johnston, 1924b, p. 1056) south of the puerto of Agua Verde Bay and found only a few trees of pain estribo. Collections from one of these (Carter 5610) closely approach the type specimen of C. molle in its lack of spines, its pubescence, flower size, and leaflet size, but differ in the pinnae being shorter, and the leaflets more closely spaced than in the type. Considering the amount of variation occurring in C. X sonorae, however, C. molle falls well within the limits of that taxon. Had Johnston been able to explore further in the Sierra de la Giganta (instead of being limited to short sallies from aboard ship), he undoubtedly would have encountered trees of palo estribo similar to that which he described from across the Gulf near Guaymas as C. sonorae and would have realized the close affinity of the trees as they occur on both sides of the Gulf. In his excellent discussion of Cercidium species in the "Vegetation of the Sonoran Desert" Shreve (1951, p. 145; 1964, p. 153) unfortunately confused C. sonome with the much more abundant C. praecox. The description and distribution given under the heading "Cercidium sonorae" and the common name brea apply to C. praecox. The strong, tough wood of Cercidium X sonorae is used to make stirrups, hence the name, palo estribo. 4. Cercidium floridum Bentham ex Gray. [Synonymy and references given under subspecies. 1 Trees or shrubs 2.5-8.0 (-12) m. tall, bark of main branches above main trunk green, horizontally striate; leaves 1 or 2 per node, petiolate; pinnae 1 (occasionally 2) pair, bearing 2-4 pairs of leaflets; flowers yellow, the petals clawed, the claw of the upper petal longer than those of the other four petals; ovary glabrous at anthesis (occasionally a very few hairs present) ; legumes linear-oblong, flat, cuneate at base and apex, coriaceous, white-dotted (stomata) when immature, obscurely so when mature, inconspicuously reticulately-veined ; seeds oblong to ovate, somewhat flattened, brown, the lighter margin sometimes slightly mottled. 4a. Cercidium floridum Bentham ex Gray subsp. floridum. Cercidium floridum Bentham ex Gray, 1852, Plantae Wrightianae, vol. 1, p. 58, subsp. floridum. Parkinsonia florida Watson, 1876, Proc. Amer. Acad., vol. 2, p. 135. Parkinsonia torreyana Watson, 1876, Proc. Amer. Acad., vol. 2, p. 135. Cercidium torreyanum Sargent, 1889, Garden & Forest, vol. 2, p. 388. Tree 4-8 (-12) m. tall, the crown spreading, branchlets of mature trees usually slender, flexuous or drooping, the trunk gray -green; bark with fine horizontal striations, often dark in the indentations; branches glabrous or 34 CALIFORNIA ACADEMY OF SCIENCES LProc. 4th Ser. sparingly pubescent; thorns absent or 1.8-5.4 mm. long (mean 3.6 mm.), slender, solitary; bud scales not accumulating at the nodes to form conspicuous dark clusters; foliage blue-green; leaves with petiole (1-) 2-5 (-11) mm. long bearing one pair of pinnae (occasionally more) 0.3-1.0 (-1.6) cm. long; leaflets mostly 3 pairs per pinnae, but often 2 and occasionally 4 pairs, oblong or obovate, sometimes emarginate, 4-8 mm. long, 2-5 mm. wide, slightly pubescent or glabrate; inflorescence open, glabrous or nearly so, borne on the younger branches; rachis of the racemes 1.0-4.5 (-7.0) cm. long, bearing 1-7 (-10) flowers on pedicels 6-12 (-20) mm. long; petals bright yellow, the upper sometimes orange-dotted near base of limb, the upper petal 9-15 mm. long, the limb 5-9 mm. long, 6-13 mm. wide, broadly ovate, cordate at base, the claw 3-5 mm. long, shorter than the limb, the other four petals slightly shorter than the upper, the limb broadly ovate, the claw short, 1.5-2.0 mm. long; legumes linear to elliptic, not or only slightly narrowed between the seeds, 3-11 cm. long, 1.0-1.5 cm. wide; seeds usually 3 per legume, 9-12 mm. long, 5-7 mm. wide. Type. "Sonora Alta," Mexico in 1830, Thomas Coulter (Trinity College, Dublin, Ireland; also 2 duplicate specimens bearing name in Bentham's hand, K! ) The nomenclatural problem concerning Cercidium jloridum and C. torrey- anum is ably discussed by Johnston (1924a; and Benson (1940). Representative specimens. MEXICO. Sinaloa: La Constancia, Munic. El Fuerte, December 1924 [sic! on original label at DS; US copied label is 1926], Jesus Gonzalez Ortega 6199 (DS, GH, MEXU, US); vicinity of San Bias, 22 March 1910, Rose, Standley & Russell 13203 (GH, US), 28 January 1927, M. E. Jones 23086 (RSA) ; Sonora: vicinity of Guaymas, 23 April 1910, Rose, Standley & Russell 15038 (US); plain N. of Empalme, 6 April 1962, Carter 4359; 20 mi. [32 km.] from Guaymas on Hermosillo road, 20 March 1934, Ferris 8752 (DS, US) ; 3.5 mi. [5.6 km.] (by road) N. of Desemboque, 29 April 1964, Turner & Hastings 64-49 (ARIZ, DS, SD); El Alamo, near Magdalena, 25 May 1925, P. B. Kennedy 7010a (UC, US) ; W. side of Isla Tiburon just N. of Punta Willard, 19 March 1962, Wiggins 17150 (DS) ; Isla Tiburon, 19 March 1962, Moran 8723 (RSA, SD), desertico espinosa, altitude 60 m., 3 May 1971, C. L. Diaz Luna 2266 (ENCB) ; Molina Crater, crater region NW. of Pinacate Crater, 29 April 1951, Kamb 2014 (ARIZ). [The sheets of this collection number at UC and DS are mixed collections, the branch bearing only flowers being a putative hybrid between Cercidium microphyllum and C. jloridum, ,] Papago Tanks, Pinacate region, 15 March 1959, R. M. Turner 59-30 & C. H. Lowe 2979 (ARIZ, CAS); Pozo de Luis, 4 January 1894, Mearns 2696 [Int. Bound. Commission] (US). UNITED STATES. Arizona: Sabino Canyon, Santa Catalina Mountains, Pima Co., 25 May 1917, Shreve 5202 (ARIZ); vicinity of Coolidge Dam, 13 May 1935, Bassett Maguire 11321 (UC); Black Canon Road, 23 mi. J 3 7 km.] N. of Phoenix, Maricopa Co., 21 October 1931, Vol. XL] CARTER: CERCIDIUM 35 John W. Gillespie 8657 [2 sheets, one fl., one fr.] (DS, GH, UC, US) ; washes W. of Castle Dome Mts., Yuma Co., 22 April 1949, /. H. Thomas 396 (DS) ; Topock, Mohave Co., 24 May 1919, A. Eastwood 8886, 8887 (CAS, GH), 29 May 1950, /. T. Howell 26618 (ARIZ, CAS). California: Whipple's Expl. R. R. route, 35th parallel, 1853-54, Bigelow (K); 20 mi. [32 km.] NE. of Ogilby, Imperial Co., 6 April 1932, Munz & Hitchcock 12166 (F, GH, MO, RSA, UC); Box Canyon at Shafer's Well, Chocolate Mts., Riverside Co., 14 June 1918, Ferris 977 (CAS, DS); 10 mi. [16 km.] W. of Coachella, Riverside Co., April 1905, H. M. Hall 5784 (ARIZ, UC, US) ; Palm Springs (Agua Caliente), desert base of San Jacinto Mt., Riverside Co., 4-13 April 1896, 5. B. Parish 4115 (BM, GH, K, MO, US); 4 mi. [6 km.] NW. of Desert Center on Aqueduct Road, Riverside Co., 10 April 1947, P. A. Munz 11722 (ARIZ, RSA, SD) ; ca. 5 mi. [8 km.] E. of Clark Lake in Borrego unit of Anza State Park, San Diego Co., 21 April 1955, Guy Fleming 45817 (DS, SD) ; near Needles, San Bernardino Co., 3 June 1929, H. L. Mason 5362 (DS, UC) ; base of Whipple Mts. adjacent to Colorado R., 2 mi. [3.2 km.] above Earp on road to Parker Dam, 15 July 1940, Wolf 9723 (RSA). Flowering from March to June with the peak in April and with occasional off-season blooming from August to November. The blue palo verde, as it is called in Arizona, occurs principally in fine soil along washes and on flood plains, for the most part at altitudes below 1100 m. Its water requirements appear to be greater than those of C. microphyllum. It occurs in the Colorado Desert area of southeastern California, in southern Arizona from 35°N. near the Colorado River, southeast to ca. 33 °N. on the eastern border of Arizona and southward through the coastal and middle portion of Sonora to 27°30'N. Outlying collections have been made in coastal Sinaloa as far southward as 26°N. Cercidium jloridum and C. microphyllum are the most widely distributed members of the genus within the Sonoran Desert, but C. jloridum subsp. jloridum is the only one restricted to the mainland. Although current floras indicate C. jloridum subsp. jloridum as occurring in Baja California, and Goldman (1916, p. 335) cites it as being there, no specimens have been seen which support such a distribution. In most areas, the peak of the blooming season of Cercidium jloridum is about two weeks in advance of that of C. microphyllum. 4b. Cercidium floridum Bentham ex Gray subsp. peninsulare (Rose) stat. nov. and comb. nov. Cercidium peninsulare Rose, 1905, Contrib. U.S. Nat. Herb., vol. 8, p. 301. Shrub or tree 2.5-8 m. tall, openly branched, but with a dense, symmetrically- rounded crown; bark gray and scaly-fissured at base, bright green above, inconspicuously horizontally striate; branchlets not flexuous, having pilose or 36 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. sparse, appressed-villous hairs when young, glabrate to glabrous in age; a single axillary thorn usually present at a node, 2-3 (-7) mm. long (mean 3.87 mm.), slender to stout; axillary bud scales, when present, not dark and conspicuous; leaves with petiole 1-4 (-10) mm. long bearing 1 (rarely 2) pair of pinnae 0.3-1.7 (-2.1) cm. long, the leaflets mostly 2 pairs per pinna but many (25 percent) with 3 pairs, obovate and sometimes emarginate, 6-11 (-18) mm. long, 3-7 mm. broad, sparingly pilose to glabrate; inflorescences borne on old wood as well as on younger branchlets, more or less compact, the racemes with rachis sometimes almost absent, but usually 3-10 (-20) mm. long, pubescent, bearing 3-5 flowers on pubescent pedicels 4-9 (-12) mm. long; calyx sparingly pubescent to glabrous; petals deep or bright yellow, the upper 7.5-9.0 mm. long, the limb 5-7 mm. long, 5.0-6.5 mm. wide, ovate or rhombic ovate, the claw 2 mm. long, the other four petals slightly shorter to subequal to the upper, limb ovate, the claw short, 1.5-2.0 mm. long; legumes 1.5-5.5 cm. long, 8-15 mm. wide, flat, the margins slightly constricted between the seeds, seeds 1-2 (3) per legume, 9-10 mm. long, 6-7 mm. wide. Type. La Paz, Lower California, Mexico, in 1890, Palmer 112 (US! ). Representative specimens. MEXICO. Islands, Baja California Sur: arroyo leading eastward from old Ruffo Ranch, Isla Ceralvo, 16 April 1962, Wiggins 17758 (DS, ENCB); Arroyo de Agueda, NE. side of San Jose Is., 11 April 1962, Moran 9402 (CAS, RSA, SD); arroyo above Ensenada Ballena, Espiritu Santo Is., 21 April 1962, Moran 9634 (SD, US) ; S. end of Santa Cruz Is., 18 April 1962, Moran 9577 (SD, UC) ; arroyo above spring, Santa Catalina Is., 14 April 1952, Moran 3866 (DS), 10 April 1962, Moran 9369 (SD) ; Puerto Ballandra [Balandra], Carmen Is., 21 May 1921, Johnston 3802 (CAS, DS, GH, K, UC, US) ; 4 June 1963, Carter & Reese 4546. Peninsula, Baja California Sur: San Jose del Cabo, March-June 1897, A. W. Anthony 363 (F, GH, K, UC, US) ; Arroyo San Lazaro, ca. 10 mi. [16 km.] NW. of San Jose del Cabo, 2 May 1959, Thomas 7775 (ARIZ, CAS, DS, SD, UC, US); Arroyo del Salto, E. of La Paz, 30 March 1949, Carter 2595 (DS, UC) ; La Paz, 16 April 1899, E. A. Goldman 388 (GH, US! Type) ; Arroyo San Ramon W. of Rancho San Ramon (25°14.5'N., 111°17'W.), 25 October 1964, Carter h866; vicinity of Puerto Agua Verde, 4 June 1965, Carter & Sharsmith 4943; arroyo S. of Bahia Agua Verde, 5 June 1965, Carter & Sharsmith 4946; Arroyo Peloteado, W. of Rancho Peloteado (25°45'N., 111°30'W.), 21 April 1962, Carter 4410; San Javier, 21 April 1962, Carter 4414; Rancho Viejo, ca. 28 km. from Loreto on road to San Javier, 6 May 1972, Carter $680; Rancho Quihi, on old mission trail SE. of Comondu, 20 April 1955, Carter & Ferris 3423 (DS, MEXU, SD, UC, US); alluvial arroyo margin, Las Cuevitas below Comondu, 12 April 1939, Gentry 4458 (ARIZ, GH, K, MO, UC, US); Valle de Los Encinos, S. side of Cerro Giganta, altitude 750 m., 8 November 1960, Carter 4134, 8 June 1963, Carter & Vol. XL] CARTER: CERCID1UM 37 Figure 6. Bark patterns in Cercidium: a, b, C. praecox showing the quadrate-pustulate surface typical of this species, a, ca. X W2, b, ca. X 10; c, C. floridum subsp. floridum showing the horizontally striate pattern characteristic of this species and of C. micro phyllum. Reese 4573; coastal strand, San Bruno (26°12.5'N., 111°23'W.), 1 June 1963, Carter & Reese 4537 \ Arroyo Mulege, ca. 10 mi. [16 km.] W. of town on road to El Potrero, 11 April 1963, Wiggins 18231 (CAS, DS, K, MEXU, US) ; Picachos de Santa Clara, Desierto Viscaino, 5-10 November 1947, Gentry 7754 (DS); wash 16 mi. [26 km.] from San Ignacio on Calmalli road, 10 March 1934, Ferris 9037 (DS, US). Flowering is from March to June with the peak in April, the branches usually bearing leaves at the same time. The flowering period precedes that of its congeners in the same area. It grows on coastal plains and along washes and canyons into mountain valleys from sea level up to 750 m. This tree, known locally as palo verde is the only taxon in Cercidium restricted to Baja California, where it is moderately common from near the tip of the peninsula northward to San Ignacio (27°25'N., 112°52'W.). As noted by Gentry on one of his collections (4458), this tree has a denser, leafier, more symmetrical crown than any other members of the genus. Also, it retains its leaves for a greater portion of the year than any of the others. The upper petal of C. floridum subsp. peninsular e is not as conspicuously elevated above the other four petals as it is in the other taxa. Discussion In an attempt to understand the relationships between the Sonoran Desert taxa of Cercidium as set forth above, the following characters were of use and merit some discussion. Bark. Inasmuch as most cercidiums retain only a few leaves during much of the year, or have minute leaves, photosynthesis also occurs in the green bark which is relatively smooth except for the lower trunks of large trees. In C. 38 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. praecox, the bark is bright green clear to the ground, whereas in the others the main trunk becomes grayish and often fissured. The bark of C. praecox differs also in having a minute quadrate-pustulate surface pattern (fig. 6a,b), easily visible with a hand lens or even to the naked eye, and in being conspicuously coated with a waxy substance. Shreve (1951, p. 145; 1964, p. 153) mentioned these characteristics, but he attributed them to C. X sonorae which he at that time mistook for C. praecox. The bark of C. microphyllum and C. jloridum, on the other hand, is finely horizontally striate (fig. 6c) and, especially in C. jloridum subsp. jloridum, often dark in the indentations, thus giving a 'dirty-neck' appear- ance. These striations are less conspicuous in C. jloridum subsp. peninsulare. The bark of C. X sonorae is usually smooth, but sometimes horizontal striations are visible with a hand lens. On some specimens of C. X sonorae there is a thin, inconspicuous layer of waxy material which suggests C. praecox. Pubescence. Young branchlets of Cercidium show varying degrees of pubescence; older branchlets are glabrous to glabrate. In C. jloridum subsp. jloridum, branchlets are essentially glabrous, but sometimes with a few hairs, especially southward in its range. In C. jloridum subsp. peninsulare, on the other hand, the young branchlets are conspicuously pubescent with pilose or sparsely appressed villous hairs. In C. microphyllum there is some correlation between geographic distribution and type of pubescence on the stems: all of the Baja California material is strongly strigose; the Sonoran collections studied are puberulent or pilose (especially in the interior) ; in Arizona and California some collections have glabrate branchlets. Branchlets of C. praecox vary from glabrate to pubescent. In C. X sonorae the branchlets may be strigose, short villous, or pilose, or occasionally the branchlets are glabrous. The pubescence of the ovary at anthesis is of diagnostic value. In both subspecies of C. jloridum and in C. praecox the ovary is glabrous; in C. micro- phyllum and C. X sonorae it is lightly pubescent to strigose at anthesis. Armature and branching pattern. Most treatments of Cercidium refer to the thorns as "nodal" or "stipular," but Kearney and Peebles (1960, p. 407) describe them as "rudimentary branches transformed into spines." This concept (except for the use of the term "spine" instead of "thorn") is in accordance with Blaser's (1956) definition, which is followed herein, of a thorn being a reduced determinate shoot. Leaves of seedling plants and of young seasonal growth of Cercidium bear a pair of inconspicuous, slender, foliaceous stipules; these do not form spines, but are fugaceous. Thorns, when present, are the first shoots to develop in a primary (first to develop at a node) leaf axil. Their shoot nature is clearly shown in the early stages by the presence of a cluster of rudimentary leaves at the distal end of the developing thorn (fig. 7c). These soon disappear as the thorn increases in size and hardens (fig. 7d). Subsequent growth at a node results in either inflorescence, leaf, or shoot development from buds on an Vol. XL] CARTER: CERCIDIUM 39 Figure 7. Developmental stages in leaves and thorns of Cercidium. a, three week's old seedling of C. praecox showing simply pinnate basal leaves and bipinnate upper leaves {Turner) ; b, five week's old seedling of C. micro phyUnm with petiolate primary leaves {Carter & Reese 4531), X 2 ; c, young branch of C. praecox showing immature axillary thorns bearing rudimentary' leaves at apex {Carter, Hastings, & Turner 5597) ; d, more mature thorn of C. praecox after rudimentary leaves have fallen {Carter, Hastings, & Turner 5597). extremely shortened stem in the axil of the primary leaf. However, the positional relationship of the thorn primordia to the shoot, leaf, and flower primordia cannot be determined without more extensive morphological investigation. In Cercidium micro phyllum, none of the buds in a primary leaf axil develop 40 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. C microphyllum C. X sonorae C praecox C. floridum subsp. floridum subsp. peninsulare Figure 8. Branching patterns in Cercidium, based on number and length of branchlets in the terminal 25 cm. of a stem. into a short leafless thorn; rather, at the majority of nodes one bud develops into the stout, determinate, leafy branch, spinescent at the tip, which is so characteristic of this species. Cercidium praecox is consistently the thorniest of the Sonoran Desert members of the genus inasmuch as at each node one, and occasionally two, axillary buds develop into thorns. When two thorns occur at a node, one is always larger. In C. floridum subsp. floridum, comparatively few axillary buds develop, — the resulting mature tree having long, slender branchlets, a less divaricately branched appearance, and fewer thorns (especially in the northern part of its range) than do most other species. However, Shreve (1951, p. 142; 1964, p. ISO) states that young trees of C. floridum less than 2 m. in height have very thorny twigs. Cercidium floridum subsp. peninsulare and C. X sonorae are variable in the number of axillary buds which develop into thorns. In many instances, C. X sonorae is thornless. Cercidium microphyllum, which, as noted above, has no axillary thorns but bears only spine-tipped, determinate branchlets, has the greatest number of branchlets (mean, 8) in the terminal 25 cm. of a branch, whereas C. praecox, the thorniest of the species, has the fewest branchlets (fig. 8). In C. micro- phyllum, there is usually a gradual reduction in length of the branchlets toward the distal end of the branch in contrast to the other taxa where reduction in branchlet length often shows no correlation with position on a stem segment. Cercidium X sonorae has an intermediate number of branchlets in comparison with its putative parents and the branchlets are longer and more flexuous than those of either parent. The average number of branchlets developing in the terminal 25 cm. of the branches of the two subspecies of C. floridum is rather Vol. XL] CARTER: CERCIDIUM 41 close: 5.4 branchlets in C. jloridum subsp. jloridum and 6.1 branchlets in C. jloridum subsp. peninsulare, but the mean length of the branchlets diverges significantly. The flexuous branchlets of C. jloridum subsp. jloridum have a mean length of 13.3 cm. and the stouter and stiffer branches of C. jloridum subsp. peninsulare have a mean length of 9.3 cm. (fig. 8). These means for number and length of branchlets help to indicate their relationship within the group, but in all cases, the range of these measurements within each species and the overlap with other species, is too great for use of the characters in keys. Leaves. The seedlings of all Sonoran Desert taxa of Cercidium have con- spicuous cotyledons, 23 to 25 mm. long in C. jloridum and 6 to 18 mm. long in the others. In the seedlings, each of the first few nodes (usually up to five) bears a pinnate leaf with several pairs of large leaflets; following these, the bipinnate leaves more typical of each of the various taxa develop (figs. 3, 7a). The inconspicuous, foliaceous stipules are caducous. The pinnae (i.e., the secondary rachises with their leaflets) are short (0.3 to 6.5 cm. long) and the leaflets are opposite. A small, foliaceous or semi-indurate bract is borne at the apex of the leaf rachis and at the apex of each of the pinnae. Because of the need to discuss them separately, the first leaf developing at a node is herein designated as a 'primary' leaf and those developing subsequently in the axil as 'axillary' leaves. Buds produced in the axil of the primary leaf may develop into a thorn (a short determinate shoot), or into foliaceous or floral shoots. The foliaceous shoots are usually so reduced that the leaves borne on them appear to arise directly in the axil of the primary leaf or leaf scar; one to eight such leaves may be present at a time. Leaves are deciduous in most taxa, usually falling before the blooming period or with periods of drought. Leaves of Cercidium microphyllum differ from those of the other taxa in being non-petiolate, except in the seedling stages wherein the primary leaf may have a slightly winged petiole from 1 to 4 mm. long which merges inconspicuously with the broadened rachis area where the pair of pinnae arises. In later stages, the primary leaf pinnae are borne on an indurate, scale-like structure, the basal portion of which usually is not distinguishable from the rachis as a petiole. The indurate terminal bract is also indistinguishable from the rachis. The two pinnae fall separately from this scale-like rachis and it remains on the branch for some time before falling and exposing the primary leaf scar. (Sometimes in C. microphyllum, the leaflets fall before the pinnae do.) In the other species of Cercidium the entire leaf abscisses at the base of the petiole. (Occasionally in C. jloridum subsp. jloridum the pinnae fall from the apex of the petiole, or from the rachis if the leaf bears two pairs of pinnae, and the petiole, or petiole and rachis remain on the stem for a short time.) In specimens of C. X sonorae with subsessile leaves, the shortened petiole and the rachis become somewhat indurate, 42 CALIFORNIA ACADEMY OF SCIENCES LProc. 4th Ser. C.floridum subsp. florid urn subsp. peninsulare 5 10 15 Length of petiole (mm.) 5 10 15 Length of pinnae (mm.) Figure 9. Comparison of length of petioles and pinnae in Cercidium floridum subsp. floridum and C. floridum subsp. peninsulare. but they fall with the rest of the leaf instead of being tardily dehiscent as in C. microphyllum. In C. praecox and C. X sonorae, the scar of the primary leaf is conspicuous on younger branches; in C. floridum subsp. floridum it is inconspicuous. In C. floridum subsp. peninsulare, the scar of the primary leaf is often obscured by a cluster of reddish-brown glandular hairs in the axil. In Cercidium praecox and C. floridum the bract at the apex of the rachis tends to be linear and foliaceous rather than indurate. In specimens of C. X sonorae with petiolate leaves, the elongate apex of the rachis is usually slightly indurate, while in those having subsessile leaves the rachis is an indurate, long-tipped scale similar to that of C. microphyllum. Cercidium floridum leaves have pinnae usually less than 1 cm. long and Vol. XL] CARTER: CERCIDIUM 43 mostly bearing 2 to 4 pairs of leaflets, whereas the other three taxa (C. micro- phyllum, C. praecox, and C. X sonorae) have pinnae up to 6.5 cm. long bearing 4 to 9 pairs of leaflets. At a significance level of 1 percent the means for both characters differ between members of the two groups. As indicated by the graphs (figs. 9, 10), the two subspecies of C. jloridum are similar in characters of petiole length, pinna length, and leaflet number. Each subspecies exhibits its greatest variability in length of pinnae. Mean length, however, is about the same for the two (7.8 mm. for C. jloridum subsp. jloridum, 7.9 mm. for C. jloridum subsp. peninsular x) and the difference between the means is not statistically significant. At a 1 percent significance level the only leaf difference is the mean number of leaflet pairs. In the primary as well as axillary leaves, C. jloridum subsp. jloridum usually has 3 pairs (mode 3, mean 2.7) and C. jloridum subsp. peninsulare, usually either 2 or 3 pairs (mode 2, mean 2.5). In all of these structures, the measurements show too much overlap to make them principal key characters. That these characters differ in degree rather than in kind supports the subspecific disposition of C. jloridum peninsulare. Among the other three taxa, petiole length is useful in recognizing entities: the leaves of C. microphyllum are non-petiolate, i.e., the pinnae are sessile (except in seedling stages); the leaves of C. praecox always have conspicuous petioles; those of C. X sonorae vary from no petioles up to petioles 12 mm. long (fig. 11). The mean petiole length for each taxon differs significantly (P = 0.99) from that of the other two. As with C. jloridum the pinnae in each species exhibit great variability in length, but the means, although clustered, are not so close (table 1). At a 1 percent significance level the only difference occurs between those for C. X sonorae and C. praecox. As to the leaflet number, the members of this group overlap considerably and show no significant difference in mean number of pairs. Both in this character and in petiole length the mean for C. X sonorae is inter- mediate between that for C. microphyllum and C. praecox. In pinna length, however, the mean for C. X sonorae exceeds that of the other two taxa; but its value is close to the mean for C. microphyllum, and the two values do not differ significantly. As suggested by the common name for Cercidium jloridum, blue palo verde, leaf color has been utilized in identification, but because interpretations of color differ so markedly, no attempt has been made to include precise shades of green in the detailed descriptions. The following field observations, however, point up the foliage color differences in the taxa of Cercidium as they occur in Sonora (correspondence: Hastings to Carter, August 10, 1971). "I made a trip to Sonora again week before last and the summer rains have wrought their usual alchemy there. The parched country we saw in April is green and incredibly lush. Also incredibly muggy, buggy and muday. 44 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. o .2 § !<-, <~r> 00 ■* <*3 lO 1 d l o rvj ro "* £ C ^5 O ^ a 2 -* X 00 w >0 Vol. XL] CARTER: CERCIDIUM 45 Cfloridum subsp floridum subsp. peninsulare 4 5 2 pairs Pairs of leaflets 3 pairs 4 pairs Figure 10. Comparison of number of pairs of leaflets in Cercidium floridum subsp. floridum and C. floridum subsp. peninsulare. "All four palo verdes were in full leaf and seeing them that way did a good bit to restore my ego. Not only is it possible easily to tell Cercidium floridum from C. micro phyllum, [but also] C. praecox and C. X sonorae can be distinguished at a glance. Cercidium praecox shows up as blue — even bluer than C. floridum. C. X sonorae, on the other hand, is relatively yellow — not the nearly chlorotic yellow of C. microphyllum, but about the color of the stems on both C. X sonorae and C. praecox." Inflorescence. In all of the Sonoran Desert taxa of Cercidium, except C. floridum subsp. peninsulare , when both flowers and leaves are present con- currently, the inflorescences exceed the leaves in length. In C. microphyllum, C. praecox, and C. X sonorae, the inflorescences normally develop before the leaves; if leaves are present, they are usually on separate branches. In C. floridum subsp. floridum, flowering precedes leafing on about 30 percent of the collections studied. In C. floridum subsp. peninsulare, where both flowers and leaves are usually present on a given branch, the inflorescences are equal or subequal to the leaves; occasionally, flowering precedes leafing. The flowers are borne in compact or open racemes, which may be either single or fascicled at the nodes. For the most part, the inflorescences are borne in profusion on the terminal branchlets, but C. praecox departs strongly from this pattern in having its usually compact inflorescences strongly fascicled along the stems, none of which are as branched as in the other taxa. Shreve (1951, p. 145; 1964, p. 153) pointed out this striking characteristic of C. praecox, but here again, he mistakenly attributed it to C. X sonorae. In C. floridum subsp. 46 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. C. microphyllum C.X sonorae C. praecox 5 10 15 Length of petiole (mm) Figure 11. Comparison of number of pairs of leaflets and length of petiole in Cercidium microphyllum, C. X sonorae and C. praecox. peninsular x, some racemes are borne on older branches; therefore the trees do not have the strong terminal flowering aspect of C. jloridum subsp. jloridum. Inflorescence rachis length varies both within the taxa and between them. In C. jloridum subsp. jloridum the rachises vary from 1.0 to 4.5 cm., while in C. jloridum subsp. peninsulare they are usually 0.3 to 1.0 cm., but occasionally up to 2 cm. In C. microphyllum they are 0.2 to 4.5 cm. long, while in C. praecox they are 0.2 to 1.0 cm. long. Those of C. X sonorae (0.5-4.0 cm.) fall between those of C. microphyllum and C. praecox. Vol. XL] CARTER: CERCIDIUM 47 Figure 12. Flowers of Cercidium: a, front view showing larger, elevated upper (posterior) petal (C. X sonorae, Carter 4411) ; b, back view (C. floridum subsp. peninsulare, Carter 4410) ; c, side view showing elevated upper petal bearing cinnamon-colored spots (C. praecox, Carter 4412). Numbers refer to Carter and/or Carter et al. collections; data for these are given under "Representative collections cited" for the respective taxa. The pedicel is jointed, with the joint in the distal half. The distance between joint and calyx varies less than that between joint and rachis. In total length of the pedicel, as in rachis length, C. floridum subsp. floridum and C. floridum subsp. peninsulare fall at opposite ends of the spectrum for the five taxa. Their respective means, 10.0 mm. and 5.9 mm., differ significantly at P = 0.99. So do Figure 13. Comparison of petal shape and size in Cercidium: a, C. microphyllum {Carter 5683) ; b, C. X sonorae {Carter 5679) ; c, C. praecox {Carter 5678) ; d, C. floridum subsp. floridum {Turner, Arizona, pickled material); e, C. floridum subsp. peninsulare {Carter 5680). Numbers refer to Carter and/or Carter et al. collections; data for these are given under "Representative collections cited" for the respective taxa. 48 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. the means for C. microphyllum (6.0 mm.) and C. praecox (8.1 mm.). The average pedicel length for C. X sonorae (8.0 mm.) falls between those of its putative parents, but does not differ significantly from that of C. praecox (table 1). As for the caesalpinoid flowers (figs. 12, 13), there is size variation within each taxon, but the flowers of C. jloridum subsp. jloridum are the largest in the group and those of C. microphyllum and C. jloridum subsp. peninsulare are the smallest. Flowers of C. microphyllum differ also in having the limb of the long- clawed upper petal white, or occasionally cream or pale yellow and the other four petals rhomboidal or lanceolate rather than broadly ovate (figs. 12, 13). The upper, long-clawed petal of C. praecox nearly always bears a cluster of small orange dots near the base of the limb; these are also often present in C. jloridum subsp. jloridum, but they are lacking in C. jloridum subsp. peninsulare. The upper petal of C. X sonorae is extremely variable. It may be white, creamy, or light yellow, and it may or may not bear orange dots. The tricolpate pollen grains, which vary in diameter from 18 to 34 microns, are prolate spheroidal and supra-reticulate (Carter and Rem, 1974). The lumina are reduced in diameter and depth near the margin of the furrows. Legumes and seeds. The venation of the valves of Cercidium legumes is of diagnostic value: the valves of C. praecox are somewhat papery with conspicuous reticulate venation; those of C. microphyllum and C. X sonorae are striate — especially noticeable in the latter; and those of C. jloridum are smooth with veins scarcely visible to the naked eye. The surface of the valves of immature legumes of C. jloridum bear conspicuous white-dotted areas. These are stomatal pores; apparently the young fruits as well as the bark carry on photo- synthesis at a time when the trees are nearly leafless. Stomata are present also in the legumes of the other taxa, but they are somewhat obscured by the prominent venation. Although Brenan (1963, p. 207, table) used the upper suture of the pod as a distinguishing character, this is more variable within our Sonoran Desert taxa than he indicated. The legumes are tardily dehiscent along the sutures or indehiscent; in Cercidium microphyllum and C. X sonorae the thin-walled valves also sometimes break irregularly. Differences in shape and size of legumes and seeds may be compared in figure 3. In characters of legume venation and shape, C. X sonorae approaches C. microphyllum more closely than it does C. praecox. As to surface pattern, seeds of C. praecox and C. X sonorae are gray-brown with various degrees of brown mottling on the flat surfaces; seeds of C. micro- phyllum and C. jloridum are brown with the marginal area somewhat lighter and sometimes faintly mottled. In all of the taxa the hilum and micropyle are subterminal; in some species they are in a slightly recessed or notched area Vol. XL] CARTER: CERCIDIUM 49 CEFL a CESO CEPR CEMI CEFL CEPE PAAC PAAF C Seed coat '//, Endosperm Cotyledons Figure 14. Comparison of seeds (cross section and position of hilum and micropyle) in Cercidium and Parkinsonia: CEFL, C. jloridiim subsp. floridum, a, side view to indicate position of cross section (b) to entire seed (E. E. Schellenger 6, Chucawalla [Chuckawalla] Bench, Colorado Desert, California, 25 July-14 August, 1903, UC) ; CESO, C. X sonorae, note variation in form of cotyledons {Carter & Sousa 5212) ; CEPR, C. praecox (Carter 4146) ; CEMI, C. micro phyllum, note thick cotyledons {Ferris 9902) ; CEPE, C. floridum subsp. peninsulare (Carter 4821) ; PAAC, Parkinsonia aculeata (Harbison & Higgins, San Telmo, Baja California, 17 December 1953, UC) ; PAAF, Parkinsonia ajricana (Rodin 2156, Namib Desert near Swakop River, Southwest Africa, 31 October 1947, UC). Diagrams drawn after seeds had soaked in water overnight. Seeds, X 2M>, side view of apex, X lVu- (fig. 14). As seen in cross section, the thick, oblong or subglobose seeds of C. microphyllum are distinct in having large cotyledons in comparison to the amount of endosperm (fig. 14). The other taxa, with more flattened and oblong or oblong-ovate seeds, follow much the same pattern in cross-section (fig. 14) as that illustrated by Boelcke (1946) for C. australe Johnston, the Argentinian species closely related to C. praecox. Chromosome numbers. Chromosome counts of 2n = 28 have been reported for Cercidium floridum, C. microphyllum and C. X sonorae (Turner and Fearing, 1960). Although contraindicated in the publication, no voucher was preserved SO CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. for the C. X sonorae count (correspondence, B. L. Turner to Carter, 5 July 1972). It is probable, however, that the seeds provided for this count were those of C. praecox rather than of C. X sonorae (correspondence, R. M. Turner to Carter, 23 June and 18 August, 1972). Hybridization. Few putative hybrids between Cercidium micro phyllum and C. jloridum subsp. jloridum have been noted, whereas C. X sonorae, which occurs in the area of overlap of C. microphyllum and C. praecox, is relatively abundant. A six-year record kept by W. G. McGinnies for trees in the foothills north of Tucson, Arizona, shows that C. jloridum subsp. jloridum blooms about two weeks earlier than C. microphyllum and C. praecox (the latter being a cultivated tree in that area) . So, although there is some overlapping of the blooming period of C. jloridum subsp. jloridum and C. microphyllum in the Tucson area at least, the peak is reached at different times (correspondence, Hastings to Carter, 24 January 1972) ; no putative hybrids have been reported from this area. On the other hand, in the lower Colorado River basin between Arizona and Colorado, these two species do reach peak of bloom at approximately the same time and a small percentage of putative hybrids has been noted (pers. comm., Jones to Carter, September 1973). Ultraviolet light studies show that the flowers of these two species appear different to pollinating insects: C. microphyllum petals absorb UV light and all are "bee purple," while in C. jloridum subsp. jloridum only the upper petal absorbs UV light and the other four reflect it. The pollinators appear to be highly selective in accordance with these ultraviolet patterns (pers. comm., C. E. Jones to Carter, September 1973). In the case of C. X sonorae, on the other hand, whose putative parents, C. microphyllum and C. praecox, bloom concurrently and have the same ultraviolet absorption patterns as the above mentioned taxa (Carter, 1974), it appears that their pollinators are not selective in their visits inasmuch as there is apparently ample cross-pollination. No putative hybrids have been noted between C. jloridum subsp. peninsulare and the other three taxa occurring with it in southern Baja California. Normally this species blooms before its congeners; its ultraviolet absorption pattern is the same as that of C. jloridum subsp. jloridum and of C. praecox, i.e., the upper petal absorbs and the other four reflect ultraviolet light. Relationship between Cercidium and Parkinsonia. As indicated in the key at the beginning of this treatment, a number of characters serve to distinguish members of the New World genus Cercidium from Parkinsonia aculeata Linnaeus, the only member of the latter genus considered to be native to the western hemisphere. Treatments of these two genera fail to clarify the structure of the primary leaf and its relationship to thorns or spines; many floras merely state "stems armed" or "stems unarmed." As detailed above, armature in Cercidium, when present, consists of thorns which develop in the axil of a primary leaf. Vol. XL] CARTER: CERCIDIUM 51 In Parkinsonia aculeata, on the other hand, there are no axillary thorns. The conspicuous armature consists of the short petiole and rachis of the primary leaf itself, which has become enlarged, thickened and indurate, plus the tip of the rachis which has developed into a long, stout, sharp spine. Hutchinson (1969, pp. 70-71, fig. 64E), misinterpreted this structure, calling it a stipule. Also, he failed to differentiate between primary and axillary leaves. The stipules are either early deciduous or develop into short, indurate lateral spines, which, however, may fall long before the indurate petiole and rachis. One to three pairs of long (10 to 60 cm.) pinnae are produced on the short leaf rachis. In both primary and axillary leaves the secondary rachises are conspicuously flattened (phyllodial). In her anatomical comparison of Parkinsonia aculeata and Cercidium torreyanum (C. jloridum), Scott (1935) found that only in the rachises of the pinnae and in the pulvinae is there a marked difference between these two taxa. She did not include armature in her studies. The small, caducous leaflets borne along the margins of the pinna rachis may be either opposite or alternate on the same pinna. The long phyllodial pinnae usually remain until the onset of winter and lend the trees a graceful, drooping aspect. When pinnae of the primary leaves fall, the indurate, spinose-tipped petiole and rachis remain on the branch. On older branches, this structure (spine) also eventually falls, leaving a conspicuous scar at the node. Burkart (1952, fig. 36) illustrates the spinose nature of the leaf rachis. He does not point out, however, that these stout spines are restricted to the first leaf at a node of a main branch and are only slightly, if at all, developed on leaves of the usually short axillary branches which are from 2 to 12 (-23) mm. long in contrast to the completely reduced axillary shoots in Cercidium. Until the recent work of J. P. M. Brenan (1963), only one other taxon has currently been considered as belonging to the genus Parkinsonia, the African species P. ajricana Sonder. The primary leaf of P. ajricana is similar in some respects to that of Cercidium micro phyllum. In Parkinsonia ajricana a single pair (occasionally one of the pair aborts) of pinnae 4 to 17 cm. long arises from a small, bractlike, nearly sessile leaf rachis which has a short, slender, foliaceous tip; stipules are inconspicuous, caducous, bractlike structures; the pinnae are not flattened, but are elliptical in cross section and bear tiny opposite, caducous leaflets, as is the case in the axillary leaves. Troll (1939; p. 1611, fig. 1372; p. 1612, fig. 1373), considers these leaves to be a reduced form of the typical caesalpinoid leaf which is petiolate and bears several pairs of pinnae. Armature, when present, is a short, stout, determinate stem (thorn) in the axil of the primary leaf. Johnston (1924a, p. 63) erred by indicating in his key to Cercidium and Parkinsonia that both P. aculeata and P. ajricana have similar armature. Presence or absence of these axillary thorns is extremely variable in P. ajricana, as it is also in some taxa of Cercidium. In all of these characters except the 52 CALIFORNIA ACADEMY OF SCIENCES LProc. 4th Ser. somewhat longer pinnae with leaflets caducous, Parkinsonia ajricana resembles Cercidium micro phyllum more than it does Parkinsonia aculeata. Torrey (1859, p. 60) reached a similar conclusion. The inconspicuous, bractlike rachis of the primary leaf in both Cercidium microphyllum and Parkinsonia ajricana might well be considered a rudimentary expression of the highly developed spine of the primary leaf of Parkinsonia aculeata. So, in this character, Cercidium microphyllum tends to resemble Parkinsonia more closely than it does its congeners with petiolate leaves (both primary and axillary) which fall entire. However, in the short pinnae which are not flattened and in the opposite leaflets which usually are not caducous, it falls within Cercidium. It has not been possible to make careful examination of the other African trees placed in Parkinsonia by Brenan: P. anacantha Brenan and P. scioana (Chiovenda) Brenan \Peltophoropsis scioana Chiovenda] ; so comparisons cannot be made regarding the structure of the primary leaves (matters which are not adequately treated in the descriptions of these two taxa). There is a significant difference in pedicel length (P = .99) between Parkinsonia aculeata and P. ajricana, on the one hand, and the species of Cercidium studied, on the other. The mean pedicel length for the taxa of Cercidium ranges from 5.9 to 10.0 mm., with the ratio of total pedicel length to length from pedicel joint to calyx ranging from 3.5: 1 to 4.3: 1. In Parkinsonia aculeata the mean length is 14.5 mm.; the ratio, 6.4:1. In P. ajricana the mean pedicel length is 13.0 mm., not significantly different from that of P. aculeata. But the ratio is only 2.4:1, i.e., the joint occurs almost at the midpoint of the pedicel, whereas in P. aculeata it is relatively close to the calyx. In Cercidium, the joint is in the distal half of the pedicel, but never proportionately so close to the calyx as in Parkinsonia aculeata. Material was not available to carry the comparison to the other African taxa placed in Parkinsonia. Pollen studies in relation to hybridization in Cercidium and Parkinsonia have shown that putative hybrids between taxa in Cercidium have a much higher percent of presumably viable pollen (as indicated by staining with aniline-blue lactophenol) than do the few known hybrids between Parkinsonia aculeata and species of Cercidium (Carter and Rem, 1974). In addition, pollen grains of putative hybrids between species of Cercidium are not malformed whereas those of putative hybrids between Parkinsonia aculeata and species of Cercidium are irregular in shape (Carter and Rem, 1974, fig. 1). No putative hybrids between C. microphyllum and Parkinsonia aculeata have been noted even though these two taxa sometimes occur in the same area. As seen under the scanning electron microscope, pollen grains of the several species of Cercidium studied and those of Parkinsonia aculeata have similar Vol. XL] CARTER: CERCIDIUM S3 Figure IS. Pollen as seen under scanning electron microscope: a, Cercidiutn jloridum subsp. peninsulare (Carter 2595) ; b, Parkinsonia africana (Seydel 1230, Namibrand: Karibib, Okomitundu, Southwest Africa, UC) ; c, Parkinsonia [Peltophoropsis] scioana (Burger 2731, Erer Rarea, 60 km. west of Dire Dawa, Ethiopia, 9°31'N., 41°25'E., 21 April 1963, K). All ca. X 1000. Photographs taken at Electronics Research Laboratory, University of California, Berkeley. supra-reticulate sculpturing (Martin and Drew, 1969, 1970; Carter and Rem, 1974). Pollen of Parkinsonia ajricana (fig. 15), on the other hand, although being reticulately sculptured, has many shallow lumina as well as the deep lumina characteristic of Cercidiutn and Parkinsonia aculeata. Furthermore, in P. ajricana the lumina are much smaller and the muri thicker than in the above- mentioned taxa. Pollen of Parkinsonia [Peltophoropsis] scioana (fig. 15) is triporate with three large pores; its sculpturing differs markedly in that the reticulate surface appears rugose and the lumina are small and the muri wide. Unfortunately, the material of P. anacantha available contained insufficient pollen for a scanning electron microscope preparation or for satisfactory light microscope examination. The legumes of Cercidiutn microphyllum resemble those of Parkinsonia aculeata more than they do those of other species of Cercidiutn. They are thin- walled and strongly torulose, i.e., much constricted on either side of the 1 to 4 seeds; dehiscence is irregular, as well as occurring along the sutures. Legumes of C. X sonorae resemble those of C. microphyllum in dehiscence. Those of other species of Cercidiutn appear to be indehiscent, or only slightly dehiscent. Legumes of Parkinsonia aculeata are sinuate or somewhat torulose and 1 to 5 seeded; otherwise they are similar to those of Cercidiutn microphyllum. Legumes in Parkinsonia ajricana material at hand are sinuate, flattened, and strongly dehiscent, while those of P. scioana and P. anacantha are flattened, but not constricted; those of P. scioana, at least, are indehiscent. Also, P. scioana is unique in the group in having long-funicled seeds. Although Cercidium micro- phyllum strongly resembles Parkinsonia aculeata in the character of its legumes, it stands apart from that species, and from other species of Cercidium, in a seed 54 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. character: the cotyledons are thick in relation to the endosperm (fig. 14). Also, in dehiscence of the pinnae from the scale-like leaf rachis, C. micro phyllum ap- proaches Parkinsonia aculeata, but in the majority of characters it is more at home in Cercidium than in Parkinsonia. As stated in the introduction, resolution of the generic relationship of Cercidium and Parkinsonia .should be based upon extensive comparative biological, morphological, cytological, and genetic studies. On the basis of present information, however, one can say that Cercidium forms a discrete, easily recognizable unit confined to the Americas, with C. micro phyllum most closely related to Parkinsonia. Study of Brenan's comparative chart (1963, pp. 206, 207) shows that among the taxa included, Parkinsonia aculeata is the most discordant element. Furthermore, the American species of Cercidium included agree in more gross morphological characters than do the African taxa placed in Parkinsonia. Pollen of two African taxa, as seen under the scanning electron microscope is of two distinct types whereas in Cercidium it is all similar. Pollen studies (Carter & Rem, 1974) indicate that species of Cercidium are more closely related to each other than they are to Parkinsonia aculeata, a species thought to be native to America, but which has been introduced widely throughout the warmer parts of the world. I would like to suggest that Parkinsonia be considered a monotypic genus comprised of P. aculeata Linnaeus, and that the relationships of the three African species (P. africana, P. anacantha, and P. scioana) be considered further before accepting them as congeneric with Parkinsonia and Cercidium as proposed by Brenan. Acknowledgments In conclusion, I wish to express my sincere appreciation to the various governmental departments in Mexico which have been instrumental in granting me permission to carry on field work in their country, especially those of the Director General de Aprovechamientos Forestales and the Consejo Nacional de Ciencias y Tecnologia as well as to Dr. Arturo Gomez-Pompa, Jefe del Departamento de Botanica, Instituto de Biologia, Universidad Nacional Autonoma de Mexico. In La Paz, Sr. Guillermo Arrambidez A., formerly Jefe de la Division Hidrometrica de Baja California, Secretaria de Recursos Hidraulicos, has through the years enabled me to time my field trips to take the best advantage of rains. Many residents of Loreto and the Sierra de la Giganta have facilitated my work by their friendship and cooperation. Also, I am deeply indebted to the Belvedere Scientific Fund of the California Academy of Sciences and to the Director of the Academy, Dr. George Lindsay, for their generous support of my field work in Baja California during recent years. To Dr. Lincoln Constance and other members of the Staff of the Vol. XL] CARTER: CERCIDIUM 55 Herbarium of the University of California, I owe special thanks for facilitating my work on Baja California materials. To Miss Nel Rem I am indebted for her assistance with studies on pollen and to Miss Charlotte Mentges for preparation of drawings and graphs. Drs. R. M. Turner and the late J. R. Hastings have shared with me the results of their extensive field work in the Sonoran Desert and generously aided me in many ways during the course of the study. And, last but not least, to the several fellow-botanists who have set forth with me from time to time on field trips in the Sierra de la Giganta, I give special thanks. They never knew just what lay ahead of them on these trips! Literature Cited Benson, L. 1940. Taxonomic contributions. I. The native palo verdes of Arizona. American Journal of Botany, vol. 2 7, pp. 186-187. Blaser, H. W. 1956. Morphology of the determinate thorn-shoots of Gleditsia. American Journal of Botany, vol. 43, pp. 22-28. Boelcke, O. 1946. Estudio morfologico de las semillas de Leguminosas, Mimosoideas y Caesalpinioideas de interes agronomico en la Argentina. Darwiniana, vol. 7, pp. 240-321 plus 11 plates. Brenan, J. P. M. 1963. Notes on African Caesalpinioideae: the genus Peltophoropsis Chiov. and its relationship. Kew Bulletin of Miscellaneous Information, vol. 17, pp. 203-209. Burkart, A. 1952. Las Leguminosas Argentinas, silvestres y cultivadas. Buenos Aires, xv + 568 pp. Carter, A. M. 1974. Evidence for the hybrid origin of Cercidium sonorae (Leguminosae: Caesalpin- ioideae) of northwestern Mexico. Madrono, vol. 22, no. 5, pp. 266-2 72. Carter, A. M., and N. C. Rem 1974. Pollen studies in relation to hybridization in Cercidium (Leguminosae: Caesalpin- ioideae). Madrono, vol. 22, no. 6, pp. 303-311. Gentry, H. S. 1942. Rio Mayo Plants. Carnegie Institution of Washington Publication 52 7, pp. vii + 328, 29 pis. Goldman, E. A. 1916. Plant records of an expedition to Lower California. Contributions from the United States National Herbarium, no. 16, pp. 1-371. Hastings, J. R., and R. R. Humphrey (editors) 1969a. Climatological data and statistics for Baja California. Technical reports on the meteorology and climatology of arid regions, no. IS, iv + 96 pp. University of Arizona, Institute of Atmospheric Physics. 1969b. Climatological data and statistics for Sonora and northern Sinaloa. Technical reports on the meteorology and climatology of arid regions, no. 19, iv + 96 pp. University of Arizona, Institute of Atmospheric Physics. Hastings, J. R., R. M. Turner, and D. K. Warren 1972. An atlas of some plant distributions in the Sonoran Desert. Technical report 56 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. no. 21, xiii + 255 pp. University of Arizona, Institute of Atmospheric Physics. Hutchinson, J. 1969. Evolution and phylogeny of flowering plants. Academic Press, London, xxv + 717 pp., 557 figs. Johnston, I. M. 1924a. Taxonomic records concerning American spermatophytes. 1. Parkinsonia and Cercidium. Contributions from the Gray Herbarium of Harvard University, n.s., no. 70, pp. 61-68. (April). 1924b. Expedition of the California Academy of Sciences to the Gulf of California in 1921. The botany (the vascular plants). Proceedings of the California Academy of Sciences, ser. 4, vol. 12, pp. 951-1218. (May). Kearney, T. H., and R. B. Peebles 1960. Arizona flora. University of California Press, Berkeley and Los Angeles. 2nd ed. 1085 pp. Lanjouw, J., and F. A. Stafleu 1964. Index Herbariorum, Part 1, The herbaria of the World. Fifth edition. 251 pp. International Association for Plant Taxonomy. Utrecht, Netherlands. Martin, P. S., and C. M. Drew 1969. Scanning electron photomicrographs of southwestern pollen grains. Journal of the Arizona Academy of Science, vol. 5, pp. 147-176. 1970. Additional scanning electron photomicrographs of southwestern pollen grains. Journal of the Arizona Academy of Science, vol. 6, pp. 140-161. Munz, P. A. 1959. A California flora. University of California Press, Berkeley and Los Angeles. 1681 pp. Sargent, C. S. 1889. Notes upon some North American trees. V. Parkinsonia. Garden and Forest, vol. 2, p. 388. Scott, F. M. 1935. The anatomy of Cercidium Torreyanum and Parkinsonia microphylla. Madrono, vol. 3, pp. 33-41. [In a later paper (Madrono, vol. 3, p. 190, 1936) Scott stated that in the above title and throughout the paper Parkinsonia aculeata should be substituted for P. microphylla.] Shreve, F. 1951. Vegetation of the Sonoran Desert. Carnegie Institution of Washington Publication, no. 591, xii + 192 pp., 37 plates, 27 maps. [Republished in 1964 as part of vol. 1, Shreve and Wiggins, Vegetation and flora of the Sonoran Desert. Stanford University Press, Stanford, California.] Shreve, F., and I. L. Wicgins 1964. Vegetation and flora of the Sonoran Desert. 2 vols. Stanford University Press, Stanford, California. 1740 pp. Torrey, J. 1859. Botany of the Boundary [United States and Mexican Boundary Survey], vol. 2, pp. 30-2 70, 61 pis. Troll, W. 1939. Vergleichende Morphologic die Hoheren Pflanzen. Erster Band: Vegetationsorgane. Zweiter Teil. Gebruder Borntraeger, Berlin. Vol. XL] CARTER: CERCIDIUM 57 Turner, B. L., and O. S. Fearing 1960. Chromosome numbers in the Leguminosae. III. Species of the southwestern United States and Mexico. American Journal of Botany, vol. 47, pp. 603-608. Watson, S. 1876. Descriptions of new species of plants, chiefly Californian, with revisions of certain genera. Proceedings of the American Academy of Arts and Sciences, Boston vol. 11, pp. 121-148. Woods Hole, Mass. | PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FOURTH SERIES Vol. XL, No. 3, pp. 59-86; 4 figs.; 2 tables. October 30, 1974 THE COMPARATIVE MORPHOLOGY OF EXTRINSIC GASBLADDER MUSCULATURE IN THE SCORPIONFISH GENUS SEBASTES (PISCES: SCORPAENIDAE) By Leon E. Hallacher San Francisco State University San Francisco, California 94132 1 Abstract: Phylogenetic relationships within the rockfish genus Sebastes (Scorpae- nidae) are not well understood. Study of variation in the structure of extrinsic gasbladder musculature within this group may help clarify understanding of evolutionary relationships. Eighty-two species of rockfishes were dissected for examination of gasbladder muscles. Possible evolutionary implications are discussed. The function of the gasbladder muscles is sound production, but they may also be used for sound reception. Introduction The rockfish genus Sebastes, of the family Scorpaenidae, contains approxi- mately 100 species, ranging in size from about 200 mm. to 1000 mm. standard length as adults (Phillips, 1957; Eschmeyer and Hureau, 1971). Unlike most species of scorpionfishes which are tropical in distribution, rockfishes inhabit cold temperate seas. They are found principally in the North Pacific from Japan to Mexico, although a few species can be found in the North Atlantic and in the temperate Southern Hemisphere (Eschmeyer and Hureau, 1971). Fifty species of rockfishes may occur at a single latitude with up to 10 species occurring at a single collecting station. This high degree of congeneric sympatry Present address: Department of Zoology, University of California, Berkeley, CA 94720. This study formed the basis of a Master's thesis at California State University, San Francisco. [59] 60 CALIFORNIA ACADEMY OF SCIENCES (Proc. 4th Ser. is of interest to community ecologists studying division of resources and isolating mechanisms. Although rockfishes are important to commercial and sports fisheries, little is known about their biology. All rockfishes are believed to be generalized large- mouthed carnivores. They vary from benthic hole-dwelling solitary species to schooling species that live off the bottom in the kelp canopy or in open water. They are ovoviparous, giving birth to large numbers of yolk-sac larvae (Moser, 1967). Field observations made in 1970 by the author indicated that several species can be stimulated to produce sounds under stress. Recent work by Mclnerney and Yearsley (in press) shows that rockfishes produce sounds during agonistic encounters. The sound producing mechanism consists of a pair of muscles, one on each side of the midline, originating on the occipital portion of the cranium and inserting on the gasbladder. Gasbladder muscles of this sort were first called "extrinsic" by Dufosse (1874) because they inserted on the gasbladder at one end and elsewhere at the other end. Muscles with both points of insertion on the gasbladder wall, such as those found in triglids (Tower, 1908), were labeled by Dufosse as "intrinsic." In this paper, muscles originating on the cranium and inserting on or near the gasbladder will be called extrinsic gasbladder muscles or, simply, gasbladder muscles. The first major work which described these gasbladder muscles for a variety of scorpaenid fishes was that of Matsubara (1943). All the scorpaenids dissected by Matsubara have some type of muscle associated with the gas- bladder, although a variety of morphological patterns exist. For Oriental rockfishes, he showed that intrageneric variations in the structure of these muscles do exist. Dissections made on 82 species of rockfishes in the present study also reveal great variations in the patterns of the gasbladder muscles within Sebastes. The species in the genus Sebastes are well known from several regional works (Matsubara, 1943; Phillips, 1957; Barsukov, 1964), but a worldwide review has not yet been attempted. One character that may be useful in an attempt to clarify phylogenetic relationships within this genus is the variation in the structure of the extrinsic gasbladder musculature. As a result of the attempt to clarify the possible origin and evolution of the various rockfish muscle patterns, dissections were made on representatives from 23 other genera in the family Scorpaenidae and on fishes from 9 other families in the order Scorpaeniformes. The classification of groups within this order is uncertain (Greenwood et al., 1966), and it is possible that an eventual overall survey of gasbladder musculature will aid in the understanding of phylogenetic relationships in the so-called mail-cheeked fishes. Another character not surveyed in this study that may also prove useful Vol. XL] HALLACHER: GASBLADDER MUSCLES 61 in clarifying phylogenetic relationships within Sebastes is variability in gas- bladder structure itself. The morphology of rockfish gasbladders appears to be diverse. Variability in the thickness of gasbladder walls, in attachment of the bladder, and in division of the bladder into two chambers is apparent (see also Matsubara, 1943, p. 126-147). Acknowledgments Specimens used in this study were obtained from a variety of sources. Eastern Pacific material came principally from the California Academy of Sciences fish collection, although some material was obtained from the collection at California State L ; niversity at San Diego and the Scripps Institution of Oceanography. Western Pacific material came from the California Academy of Sciences fish collection and from the Museum of Zoology of the University of Michigan. Field collections were made during the course of this study to provide fresh material. I wish to thank the following people for their aid: Dr. George W. Barlow of the University of California at Berkeley who first advised the author of sound production by rockfishes; Dr. John S. Stevens of Occidental College and the crew of the research vessel Van Tuna, who provided specimens from Catalina Island off southern California; Dr. Lo-chai Chen of California State University at San Diego and the Scripps Institution of Oceanography for supplying a boat for collections made off La Jolla, California; Dr. Tomio Iwamoto of the California Academy of Sciences, who collected specimens off British Columbia from the research vessel G. B. Reed; Dr. William N. Eschmeyer of the California Academy of Sciences, who collected specimens off the Oregon coast from the Oregon State University vessel Yaquina; Dr. Eschmeyer, Dr. Chen, Christopher Tarp, Frederick Jones, and Ernest W. Iverson, who assisted in making collections from sportfishing boats; David W. Behrens, Kenneth R. McKaye, and divers from the University of California at Berkeley Diving Program, who aided the author in netting and spearing shallow water species; Daniel J. Miller and Robert N. Lea of the California Department of Fish and Game, who donated specimens; Dr. Reeve M. Bailey, Dr. Robert R. Miller, and staff of the Museum of Zoology of the University of Michigan, who provided the assistance to Dr. William N. Eschmeyer which resulted in the loan of Oriental specimens from their University collection; Dr. John E. Mclnerney and Dr. John H. Yearsley for allowing me to read their manuscript on sound production by rockfishes. Additional help came from many persons. Dr. Richard Winterbottom of the National Museum of Canada provided information regarding nomenclature on the subject of gasbladder musculature. Daniel J. Miller, Robert N. Lea, Dr. Chen, Dr. Eschmeyer, and W. I. Follett assisted in identification of specimens. Pearl M. Sonoda, James E. Gordon, Beverly J. Wesemann, and David W. Behrens 62 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. provided curatorial assistance. Cherryl P. Pape assisted in the typing of the manuscript. Katherine Keeney Smith did all drawings. This study was supported by National Science Foundation (NSF GB 34213), Dr. Chen and Dr. Eschmeyer co-principal investigators. The California Academy of Sciences provided working space and access to collections. Dr. Eschmeyer provided background information on scorpionfishes, and he, Dr. Chen, and Dr. Margaret G. Bradbury gave advice during the course of the study. I wish to thank the members of my thesis committee, Drs. Margaret G. Bradbury, William N. Eschmeyer, and Jerry W. Gerald, for their suggestions and comments. I would further like to acknowledge the following persons for their assistance in revising the manuscript for publication: Dr. Eschmeyer, Dr. Walter R. Courtenay, Lillian J. Dempster, and W. I. Follett. Methods In this paper the term gasbladder has been applied to the structure more commonly referred to as the airbladder or swimbladder. This follows the terminology suggested by Lagler, Bardach, and Miller (1962). Since the gasbladder in rockfishes is physoclistous (no duct connects the gasbladder to the alimentary canal) and inflated by means of a gas gland, the term gasbladder seems more appropriate than either alternative expression. In accordance with this terminology, any specialized muscles associated with the gasbladder have been referred to by this author as gasbladder muscles rather than the more commonly used 'swimbladder muscles'. Extrinsic gasbladder muscles were exposed for examination by dissection. These muscles appear to have arisen from the ventral fibers of the epaxialis (the dorsal half of the lateral body musclature). All species have a pair of muscles, one on each side of the midline. They extend posteriorly from their origin on the occipital region of the cranium to the gasbladder or vertebral parapophyses where they insert by means of tendens. Since their position relative to skeletal structures is constant throughout the genus, one standard dissection technique was sufficient for all rockfish species and could be done with a scalpal and forceps. In preliminary dissections, the muscles from both sides of specimens were examined with no structural differences being observed. All subsequent dissections were made on the right side of specimens, with only the muscle on the right side being examined. A large patch of skin was removed, its perimeter extending along the rear of the cranium and pectoral girdle, posteriorly along the base of the spinous dorsal fin to its end, downward to the ventral half of the fish's side and forward to the pectoral fin axil. All exposed muscle tissue lying dorsal to the right dorsal ribs was removed. The right gasbladder muscle was exposed to view by gently breaking away the dorsal ribs. Care was taken while removing the first two dorsal ribs since the Vol. XL] HALLACHER: GASBLADDER MUSCLES 63 gasbladder muscle lies just ventral to them. In most species of rockfishes the muscle passes between the second and third ventral ribs so that it is found on the visceral cavity side of the third through eighth ventral ribs. Once the gasbladder muscle was exposed, all skeletal muscle tissue surrounding it was plucked away cautiously to expose the muscle from its point of origin on the cranium to its points of insertion on the gasbladder or vertebral parapophyses. The insertion points were determined by clearing away all intercostal muscle between the ventral ribs. Once the first eight or nine ribs were exposed, the dense tendon tissue of the gasbladder muscle was seen extending laterally just under the ribs. These whitish tendons were followed to their points of insertion on the vertebral parapophyses. In species where these tendons inserted first upon the gasbladder, an incision into the body cavity was made to examine posterior insertion sites. Another method was sometimes used in species where the insertion points on the vertebral parapophyses were not readily determined from a dissection of lateral body musculature. The body cavity was entered from the ventral side and the viscera and the gasbladder were removed. Once the gasbladder was removed, the whitish tendons of the gasbladder muscle could be seen inserting on the vertebral parapophyses. Dissections of other members of the order Scorpaeni formes were conducted in much the same manner since their gasbladder muscles were located in the same general area of the lateral skeletal musculature. The notable exceptions were members of the suborder Cottoidei. In species of this group, muscles termed "cranioclavical muscles" by Barber and Mowbray (1956) extend from the posterior cranium to the pectoral girdle. To expose these cranioclavical muscles, only musculature situated directly behind the cranium was removed. At the onset of the study, radiographs were taken of dissected specimens to clarify the topographic relationships of the origin and insertion points of the extrinsic gasbladder musculature. Once the anatomy of these fishes became familiar, radiography was no longer required for the majority of species. Specimens used in the study are listed in the appendix. Results Dissection of 82 species of rockfishes reveals considerable variation among species in gross morphology of the extrinsic gasbladder musculature. In species where both sexes were examined no sexual dimorphism was observed. Several insertion points may be listed for one muscle because of the fact that in most cases, the striated muscle tissue of the gasbladder muscle ends in one to several tendons after the muscle passes under the third rib. In several species, only one or two tendons develop from the muscle, but later branch to insert on several different vertebral parapophyses. Table 1 summarises origin and insertion points for the gasbladder muscles of all rockfishes examined. 64 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. Table 1. Muscle type and points of insertion of the gasbladder muscle for all species of Sebastes examined. Where more than one specimen of a species was examined, the information on each specimen examined is listed. Abbreviations for the points of insertion are as follows: gb represents the gasbladder, r-1 thru r-4 represent the first through the fourth ventral ribs, and v-7 through v-11 represent the seventh through the eleventh vertebrae. Muscle category includes major type {either Type I or Type II) and the subdivision within a type: a-z = aleutianus-zacentrus, a-v = atrovirens-vexillaris, i-v = ijimae-vulpes, p = paucispinis, s = serriceps, t = taczanowskii, m = marinus. Muscle Category Points of Insertion Species gb r-1 r-2 r-3 r-4 v-7 v-8 v-9 v-10 v-11 aleutianus I a-z X X X alutus I a-z X X X X ii I a-z X X X X X auriculatus I a-z I a-z X X X X X X ii I a-z X X X aurora I a-z X X X atrovirens II a-v X X X X M II a-v X X X X babcocki I a-z X X X baramenuke I a-z X X X X borealis I a-z X X X brevispinis I a-z X X X capensis I a-z X X cam at us II a-v X X X X caurinus II a-v X X X ii II a-v X X X X chlorostictus I a-z X X X chrysomelas ciliatus II a-v I a-z X X X X X X X constellatus I a-z X X X crameri dallii I a-z I a-z X X X X X X diploproa I a-z I a-z X X X X X X elongatus I a-z X X X X emphaeus I a-z I a-z X X X X ensifer entomelas I a-z I a-z X X X X eos I a-z X X exsul flame us I a-z I a-z X X X X X X X flavidus gilli glaucus * I a-z I a-z I a-z X X X X X X X X X goodei I a-z X X X * Not examined. Information from Matsubara (194.?). Vol. XL] HALLACHER: GASBLADDER MUSCLES 65 Table 1. Continued Muscle Category Points of Insertion Species gb r-l r-2 r-3 r-4 v-7 v-8 v-9 v-10 v-11 helvomaculatus I a-z I a-z X X X X X X hopkinsi I a-z X X hubbsi I a-z X X ijimae II i-v X X inermis II i-v II i-v X X X X X X X X iracundus * I a-z X X X itinus * I a-z X X X jordani I a-z X X joyneri I a-z X X X kawaradae * I a-z X X X X lentiginosis I a-z X X X levis I a-z X X X X X longispinis I a-z X X X macdonaldi I a-z X X X X X X X maliger II a-v II a-v X X X X X X marinas I m X X X X ti I m I m X X X X X X X matsubarae I a-z X X X X X X m elan ops I a-z X X X melanostictus * I a-z X X X X melanostomtts I a-z I a-z X X X X X X X X miniatus I a-z X X X mystinus I a-z See Table 2 nebulosus II a-v X X X nigrocinctus I a-z X X X nivosus II i-v X X oblongus II i-v X X ovalis I a-z X X X X owstoni I a-z X X X pachycephalia II i-v X X X pancispinis II p II p X X X X phillip si I a-z X X pinniger I a-z X X X X proriger I a-z X X X X rastrelliger I a-z X X reedi I a-z X X X X rosaceus I a-z X X X rosenblatti I a-z X X X * Not examined. Information from Matsubara (1943). 66 CALIFORNIA ACADEMY OF SCIESCES [Proc. 4th Ser. Table 1. Continued Species Points of Insertion Muscle - Category gb r-2 r-4 v-7 v-8 v-9 v-10 v-11 ruberrimus I a-z rubrivinctus I a-z X rufus I a-z saxicola I a-z X schlegelii II II i-v i-v scythropus I a-z serranoides I a-z serriceps I s X simulator I a-z sinensis I a-z steindachneri I a-z taezanoivskii II t X thompsoni II i-v trivittatus II i-v umbrosus I a-z variegatus I a-z vexillaris II a-v X vulpes II i-v wakiyai * I a-z wilsoni I a-z zacentnis I I a-z a-z * Not examined. Information from Matsubara (1943) X X Points of insertion decayed X /, X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X For the purpose of determining the extent of variation within a species, dissections were done on 17 specimens of the blue rockfish, Sebastes mystinus. These dissections revealed only slight variations in the size and shape of the extrinsic gasbladder muscles, but they did indicate great intraspecific variation in posterior insertion points (table 2). No apparent sexual dimorphism was found. The size range of this series, consisting of eight males and nine females, was 70 mm. to 376 mm. On the basis of morphological variations, two basic divisions in gasbladder muscle structure are recognized for the genus Sebastes. These divisions have arbitrarily been designated Type I and Type II. Type I consists of all muscle variations that attach firmly to the pectoral girdle as they pass posteriorly from their origin on the occipital cranium to their points of insertion on the vertebrae (fig. 1). Type II is comprised of muscle variations that bypass the pectoral girdle as they pass posteriorly from their origin on the occipital cranium to their points of insertion on the gasbladder or vertebrae ( fig. 2 ) . Each division is further subdivided into subtypes on the basis of additional structural variations. Vol. XL] HALLACHER: GASBLADDER MUSCLES 67 Table 2. Muscle type and points of insertion of the gasbladder muscle for 17 specimens of the blue rock fish, Sebastes mystinus. Sex and standard length (S.L.) are listed for each specimen. Abbreviations for the points of insertion are the same as in table 1. L. M uscle v Points of Insertion Se> ; S.L. Muscle Points of Insertion Sex s.: -7 v-8 v-9 v-10 v-7 v-8 v-9 v-10 $ 222 mm. I a-z X X 9 112 mm. I a-z X X S 231 mm. I a-z X X X 9 157 mm. I a-z X X 6 251 mm. I a-z X X X 9 191 mm. I a-z X X c? 258 mm. I a-z X X 9 264 mm. I a-z X X X 6 260 mm. I a-z X 9 265 mm. I a-z X & 267 mm. I a-z X X X 9 294 mm. I a-z X X $ 275 mm. I a-z X X 9 342 mm. I a-z X X $ 289 mm. I a-z X X 9 376 mm. I a-z X 9 70 mm. I a-z X X X Type I contains three subdivisions that are nominally denoted: aleutianus- zacentrus, serriceps, and marinus. Type II contains four subdivisions: paucispinis, ijimae-vulpes, atrovirens-vexillaris, and taczanowskii . Subdivision nomenclature was determined by simply listing all species with a particular type of musculature in alphabetical order. The first and last species names on the list were taken to denote the subdivision name for that particular muscle type. Hence, the aleutianus-zacentrus subdivision is the pattern in all species in alphabetical order from Sebastes alcutianus through Sebastes zacentrus that possess dorsoventrally flattened muscles, originating on the cranium, attaching to the cleithrum, and inserting on the vertebral parapophyses. Four of the subdivisions are found in only one species each, so only one species name is listed for each of these four systems. Type I All species with musculature of this major category have extrinsic gasbladder muscles that attach firmly to the pectoral girdle as the muscles pass posteriorly from their origin on the occipital cranium to their insertion points on the vertebral parapophyses. Subdivisions of this major category have been formed on the basis of additional structural variations. aleutianus-zacentrus subdivision. This was the basic muscle structure found in the genus Sebastes with 62 of the 82 species dissected having this morphology. (See fig. 1A). L T nlike some of the other muscle types which are found only in one geographic region, this type occurred in species taken throughout the geographic range of the genus. The gasbladder muscles of this group are dorsoventrally flattened with the point of origin being the cranium. The specific site of attachment to the cranium is only slightly variable, and is usually around the exoccipital-opisthotic suture, C A LI FOR XI A ACADEMY OF SCIEXCES IProc. 4th Ser. o t3 c J* ,■; ~o ■& .2 ; .£? c I — ' c/) So C M ^J ^ "O D o -S s £ £ o 0, 3 CO c Vol. XL] HALLACHER: GASBLADDER MUSCLES 6'J although in some species the attachment site seems wholly on one or the other of these two bones, or on the opisthotic. As the muscle passes posteriorly, some of its fibers attach to the supracleithral bone near the distal attachment site of Baudelot's ligament. Insertion by tendons occurs on the ribs or vertebral parapophyses. The points of insertion, which tend to vary interspecifically, are listed in table 1 for all 62 species included in this muscle category. Unless otherwise stated in table 1, only one specimen of each species was dissected. The species of Sebastes examined, listed alphabetically, are as follows: aleutianus, alutus, auriculatus, aurora, babcocki, baramemike, borealis, brevispinis, capensis, chlorostictus, ciliatus, constellatus , cramcri, dallii, diploproa, elongatus, cmphaeus, ensijer, entomclas, eos, exsul, flameus, jlavidus, gilli, goodei, helvomaculatus, kopkinsi, hubbsi, jordani, joyneri, lentiginoses, levis, longispinis, macdonaldi, mat sub arm, melanops, melanostomus, miniatus, mystinus, nigrocinctus, ovalis, owstoni, phillipsi, pinniger, proriger, rastrelliger, reedi, rosaceus, rosenblatti, ruberrimus, rubrivinctus, rujus, saxicola, scythropus, serranoides, simulator, sinensis, steindachneri, umhrosus, variegatus, wilsoni, and zacentrus. Species of Sebastes dissected by Matsubara (1943) possessing this type of musculature are as follows: baramenuke, flameus, hubbsi, longispinis, matsubarae, owstoni, scythropus, and steindachneri; plus those examined by Matsubara but not by the author (Matsubara's results are included in table 1): glaucus, iracnndus, itinis, kawaradae, mclanostictus, and wakijai. serriceps subdivision. This muscle arrangement is found only in the treefish, Sebastes serriceps, an eastern Pacific species. The gasbladder muscles of this species are very similar to those of the aleutianus-zacentrus subdivision, being dorsoventrally flattened muscles with the point of origin on the cranium. As the muscles in this species pass posteriorly, a portion of the muscle fibers attach to the supracleithral bone. The mode of insertion of the gasbladder muscles in 5. serriceps is however quite different from that found in the species listed in the aleutianus-zacentrus group. In the treefish, the posterior margin of the gasbladder muscle inserts primarily on the second rib, without first becoming a tendon. Careful examination revealed one small tendon coming off the muscle that passed between the second and third ventral ribs and inserted to the parapophysis of the seventh vertebra (see fig. IB ) . marinus subdivision. 1 The extrinsic gasbladder muscles of Sebastes marinus are similar in overall structure to those found in the aleutianus-zacentrus group (fig. 1C). Each muscle originates on the occipital region of the cranium and attaches to the pectoral girdle as it passes posteriorly to insert by means of tendons to the vertebral parapophyses. The major difference between this species and all other rockfishes is that the muscle crosses between the third and fourth 1 See addendum. 70 C ALIFORM A ACADEMY OF SCIENCES IProc. 4th Ser. u ^5 C o oj 43 H ,r O c 43 E £ c CI T) 1) >o C (33 u s bl) '5 ^ g, 11 — 1-1 o o ■a c 4-. 4= uj .H ^ Vol. XL] HALLACHER: GASBLADDER MUSCLES 71 ventral ribs as it passes posteriorly towards the vertebrae, instead of passing between the second and third ventral ribs as in all other rockfish species examined. See table 1 for points of insertion in all Sebastes marinus dissected. Type II All species having muscular of this major category have extrinsic gasbladder muscles that bypass the pectoral girdle as the muscles pass posteriorly from their origin on the occipital cranium to their insertion points on the gasbladder or vertebral parapophyses. The muscles of this category tend to be more massive and cylindrical than those of the Type I category. Subdivisions of this Type II category are delineated by the author on the basis of additional structural variations. paucispinis subdivision. This muscle arrangement is found only in the bocaccio, Sebastes paucispinis, an eastern Pacific species. In this species the gasbladder muscles have the point of origin on the cranium. As the muscles pass posteriorly, they do not make a connection with the pectoral girdle. This species, unlike any others in the genus, has the striated muscle tissue of the gasbladder muscles divided into two parts by a thin fascia (fig. 2A). The gasbladder muscles insert, each by means of a single tendon, on the posterior portion of the gasbladder. This tendon, while firmly fastened to the gasbladder wall, curves dorsally to anchor the gasbladder to the parapophysis of the tenth vertebra. ijimae-vulpes subdivision. The gasbladder muscle pattern included in this group was found in nine species, all from the Orient. In these fishes, as in the rest of the genus, the extrinsic gasbladder muscles have the area of origin on the cranium, near the suture of the opisthotic and exoccipital. Like the muscles found in Sebastes paucispinis, these do not make a connection to the pectoral girdle as they pass posteriorly. The muscles have a striated portion that is slightly more massive than those described in previous groups. They are not quite as broad as the dorsoventrally flattened muscles of the aleutianus-zacentrus or serriceps subdivisions, but are of greater girth, being more cylindrical in shape. They appear to insert by tendons to the vertebral parapophyses without first attaching to the gasbladder (fig. 2B). Points of insertion for all species in this group are listed in table 1. The species of Sebastes falling into this group are as follows: ijimae, inermis, nivosus, oblongus, pachycephalia, schlegeli, thompsoni, trivittatus, and vulpes. atrovirens-vexillaris subdivision. The muscles of this grouping and the next (taczanowskii subdivision) are the largest in the genus Sebastes. These gasbladder muscles originate on the cranium. As they pass posteriorly they do not make a connection to the pectoral girdle. They insert as tendons that attach firmly to the gasbladder before swinging up and attaching to the vertebral parapophyses (figs. 2C and 3). In several of the species, one of the two tendons 72 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. >, CO (J si) .£ J2 -* D. o 3 ^ s 3 ~ O TT -4-> _£2 e 3 en be rC -^ -s SO ,Q