Dee : Fi < ene pel ele Sra rat ace SUA SLES i satanitarte as talticgee eG. RE ror rena 1 neato aoe MNase tT a F Roacle thay : Renew eset tT Retreat sath Ree Rap enone ae ea SGU TRCN SS ae Sols torent rete a a “hg a et ; epee RET Letrarsen sgn eH at OES arretaeaeiennonettn thelr nai eel hea ae fra Ls corner + Ura erate ty ‘ euch aieh ee ats Dnecies incr Sastry air ‘ Sake ean sy ie a0) : te Nee at ore tut east pvureD Cerra iether temp es eater ee apes pugs cr Iptirsees reanexerer ea eee ae punta ie fechas oan apa ain mE ea Beal uscriraen reenact note ria St e Bars gees tars lb Fe Y per wists Cin eye a waht tian ir RA rs ray a i ‘ ute ‘ ; Lae ret re Lae MUON eal Pe Wa ela elkestosica fi Por scr ie year ew Sw y ie YOO WS Sor iy ees . ea Wate waite 4 ee ar ¥ f ‘ a Rare Ey Ur Wk Py HP ur fs é rt} ihe i i a oka miiacchentig east ; ae PAR ES Gea ie scape Bruin ote aw sine ¥ 3 a RoC = : es Laren iap SAR saute east cpacratar ia cea aca ag RR aac ug real Rp " eels MURR ei eines i le ssi tetas ‘ ee ne i Lviors rh ci Warder ast sth His seltiea ‘ina dase fad os eta On ai Geta iy | eeu Se NA cretae boa G aig potion toed ra NR ape axe aE asa ma ihedael ae Uhr aia ite Peashisl tack a Ra arith he ed Nn Me MI AiLe ft eT ae sGaeRERERRLS ct i oe San ENR Ria ater ss ; , aout ; My th i Poa te Ber PO ee ES CMLL ary nt aR fates ae PENRO Ni oid aero et i aU sae) Wie i tre Seana wt Sane Uae aie VeerE aL Nb Secs ANE cRrRce art F ovate eke erepey AUT eth Citron eer ) isn Ae saa E aM a aut t tes ts ¥ stra rd tee TRPURL RTs a a ith 4 o aS hitetie pee 4 i Loe erart re eee ae hea rset ar Ev ee erie f eit aceseet gt A toh A ee ea oe vata Deeg ae : Ltr grist £ Apt bees Oh : in Gace me SPST We linters Dey ear cha in Cool an hy bed diet ( r Pant fess bay ch mt oe sae Pi teNr ae ia E ay aa hah ‘ LOLA oc ' t sare ekertrenit ete He Tonhan vegan PRA Meron A Mair heMatt irk BCH es ash Mtn ie balsas Di i sf motets ‘=| f i eres > is Hide He ear Tok j is " a. fe DEE ms ahd eaii es penne re het ey sli icra riplemee tint cot A Hirt ey Ee oceans meats is wt 7 Ln fuga Pikelrie st Attell es aie me ae facgenes fa Pern, Lee hy n eteea: fi oF OO Rare i A Peli SVR R CHM AH biel iy Lease CR is eas v i aie SR ar Pa Be em, oy +. ine 7 i 4 ee mae Ror 3h a Ht ea Ks ‘i : 4 + " rotte “ : Derehiet seen ots haere MDE Hens eA A gt) re i y otros Cia a . LP enit ¥ seaieyee a S Soni greehaae tag ETE eter rene peer elah re HN MOC Arad CA Rc a CL Cit i ' eT rf “sha s ~ Abe Meee ene be i Sh) rh ti AN Glad tf OE ee botat Res t ¥ if ae Rear a ie a r y eRe! nee PUF APL mis WH co HORM Re i Bid eet peda t Peart / ry es , erro ee rn net fats RANE OINRN Rati Fiera tn Bria nena nine dire Geeeatces enamine Soe eee Lats veaper i : ee ‘ - tri 44s. etree SH rma wad si} Peet tor : EMI tw reer ih PU Ho MOAN te Lie ear te ar i eae hark rae SAP Rb Ra as CCC OCI ACM eter 0 ” ek > veg py [+i ais ST | why 8 tel hay Wed 0 c nell Uni 3 The Salton Sea; a study of the geo iii ersi Qk graphy, | 02 07 iii THE SALTON OFA A STUDY OF THE GEOGRAPHY, THE GEOLOGY, THE FLORISTICS, AND THE ECOLOGY OF A DESERT BASIN BY D. T. MACDOUGAL and Collaborators WASHINGTON, D. C. PUBLISHED BY THE CARNEGIE INSTITUTION OF WASHINGTON 4 ¢ \ N289064% CARNEGIE INSTITUTION OF WASHINGTON PUBLICATION No. 193 PRESS OF J. B. LIPPINCOTT COMPANY PHILADELPHIA PREFACE. By D. T. MacDovaat. The geological and biological processes of the surface layers of the earth’s crust are profoundly different under subaerial exposures to those which prevail under subaqueous conditions. Any region, therefore, which may be subjected to submergence and to weather- ing, alternately, will offer a changing complex of environic conditions, with accompanying disturbances in the balance and distributional movements of the organisms of the region. The solid ground, in one case, is subject to the erosion of precipitation-water and stream-flow, coupled with a leaching action of the water, while surface material will be variously transported and deposited. Wind effects will vary widely with aridity and other climatic features, while precipitation, evaporation, and temperature will act as determi- nants as to the character of the living forms supported. The submersion of any area may be expected to be followed by the complete or nearly complete destruction of the land flora and fauna. The temperature of the substratum is equalized, variations in moisture disappear, wind comes in only as an agency for transport- ing propagative bodies and as a cause of wave-action, sorting, wearing, and depositing material along shore lines. The deposition of sedimentary material in submerged areas under lakes takes place in such manner as to be easily distinguishable from the effects of stream-action. The superposition of the two groups of effects in any region such as the lowermost part of a desert basin, especially when submergence and desiccation alternate at intervals of sufficient length to give full force to the two extremes, might be expected to offer highly unusual physical conditions of the surface layers and of the soil, to which organisms would be expected to display reactions of interest and importance useful in the interpre- tation of phytogeographical phenomena in general. The great Cahuilla Basin, which lies to the westward of the lower or southern part of the main delta of the Colorado River, has been the scene of alternations of the kind in question. The lower part of this basin has been submerged and desiccated many times in the last few hundred years, as attested by the numerous beach or strand formations and layers of travertine on the shores. The making of the lake in 1904, 1905, and 1906, at the time of the organization of the work of the Desert Laboratory, offered opportuni- ties for some studies the results of which are presented in the present paper. The facts to be taken into account were so diverse in character, and the necessary methods of cali- bration and estimation so unlike, that the cooperation of a number of workers in various branches of science was enlisted. First of all the senior contributor was so fortunate as to secure a general sketch of the geology and topography of the basin by the late Professor William Phipps Blake, whose barometric measurements as a member of the Williamson expedition in 1853 first established the fact that the region was an inclosed basin, the lowermost part of which was below sea-level. The history of earlier travel and the general geography of the basin is described by Mr. G. Sykes, geographer of the Desert Laboratory. The analyses of water samples have been carried out under the direction of Professor R. H. Forbes, at the Agri- cultural Experiment Station of Arizona, by the aid of Dr. W. H. Ross and Professor A. E. Vinson. The surface geology, with especial reference to the soil formations, has been the subject of much painstaking examination by Mr. E. E. Free, formerly of the United States Bureau of Soils. Professor G. J. Peirce, of Stanford University, has contributed the results of studies of organisms living in brackish and saline waters, which are of especial concern Il Iv PREFACE. with the subject, especially in the stages of the desiccation of the lake yet to come. Professor M. A. Brannon, of the University of North Dakota, has carried out a series of cultures by which the specific action of the micro-organisms upon the sulphates, iron, and calcium com- pounds has been determined and some important facts bearing upon the changes in woody tissue probably initial to the formation of coal have been obtained. Professor J. C. Jones, of the University of Nevada, has made brief studies of the travertine coatings found on the solid granite near the level of the highest beach line, establishing the participation of biological agencies in their formation. Mr. 8. B. Parish, of San Bernardino, has written a description of the plants which form the vegetative setting of the drama of the appear- ance and disappearance of the lake; the senior contributor has devoted attention chiefly to the revegetation of the areas laid bare by the desiccation of the lake. The field parties have been accompanied by scientific visitors at various times, from whom valuable sug- gestions as to interpretation have been received. TABLE OF CONTENTS. Page PRB AGB Se soi 5 cecaice Bec a9 oS IIS EIS ES ASF WSS ES OES TOE RI RRS EE ey EAA Bales bealee eee aes iil List OF TLLUSTRATIONS 333 cic.cts os: acshhs Stpcks Congas sesndiots bs dada data edd ws band Cok Mie ee wes aas Ra use aE viii Tur CanvILLA Basin AND Desert oF THE CoLorapo. By WILLIAM Puiprs BLAKE ............0..-000 00s 1-12 BD}. 0) (0 21 (0) ¢ Oa ee ee eee eee 1 DAD AGOTEONIO : PAs cesar esse ca oes ky see dea tes asa ans gee ate Tol Pte gd bg DGPS Ng Gu Ped a cea ota oS 1 Geologic history. cca 3 dine acnerewe eetenkohe os Ru ae Poked aes ted sores Be ie Shawl a eae are Gade euiuiecee 2 Evidence‘of uplifts «si esas: tee nacvad swans Ga Sa welts Bate Os ee Se Fa VE a ea ad ERTS OR es ORDER ORES 2 Lake: Cahuillais, 2 sccj5529:4 ge kd Ger gl nso Ge CCR Pw BERS ERS OG i es Oe RS La Se te ee 3 Desiccation:of: Lake: Cah will aii :ccih case ie a Patke ds Seek wh ONE ERs OA Re Be RW Re ANE Ree elgaees 5 Dalton Sea: ces cesdsre cictaececare laa dex 16 saasue fo soaseeels Syveua tac gece oes aeat eh ceo Rada be astedes acd Mule ecee nGEd SIM URNS DLR 5 Rate Gf: evaporation sa.2.24.c508 choles aut satin SPA ead wkina OMire asataleieG wlanleS Wale Mewcnilaa ROMA G Wade LS 5 Colorado: desert. icc care es 3 Be ea eee hah e ok es ea Sok A eae. BRN ens Da wee 6 Fertility of desert: soils; Abd Guia E hie takes extn 115-172 Desert basins and revegetation.......... 0... c cece ce eee eee nee eee teen eee tent ne eet neenenetnne 115 HE SACOM SU es asi scm he easy. ha a aeette Sepa eR Ee a MEI Od ARGH PROS ge BON ee I RA he 116 Laguna: Maquata scsinccts diss wea on tis angers aud aa aeae d pae a deaene a os ama nei oleae sce 118 Plan: of field workin ‘the Salton:Sinks¢:2.:.2.ac:0.00 G03 death es oka hoe PE EN RAS EES RECO CRE RS eee 119 Facts to becascertained c3.0 ea.cae. scenes cates Heese Ri ee ESS ae BSE EEA ORS Be DEEE MEE ee Sapa oak dul 120 Reoccupation of the strands of 1907......... ccc ccc cc tenet teen eee e tebe beet nnn 121 Reoccupation of the strands of 1908...... 0... ccc ccc ccc ce cere eee e ene b nee teen beet tenes 131 Reoccupation of the strands of 1909.0... .. eee eee ene eee ete e eee etenne eee 134 Reoceupation of the: strands Of 1910 sacs dsaacs, ca negara eins «deus + coe ds waa ea eg Ba ea ee bh pwearas Gow Bos Meee Peg eee da nae de aeeeed 144 Occurrence and behavior of various species on the beaches......... 00. s eee e cee eee tee eee e teen eens 153 The movements of plants into sterilized areas......... 0. e eee n ee eee eens 159 TSG OTA CA cic ic Ses, cage vast ee esos echo au oeseee SANE eg EET RUE EL BIE Sou BG city o Sea uAAD Ba Hotes Ge wd ta OE ease a 160 The physical conditions of the Salton Sink... 10.0... Leen eee 160 Naturesof the @vid nGe ssc cg.ave ic -cscesepndee Sets ain GANG A eche eR dma hand De bee dG Deedee aa Rea 161 Germination; flotation, and survivals. cj04.6- A. conn cee Seek Ga een ais Se Ne isi Sa Awe wath ane ars 161 Species appearing earliest on the beaches............. 006. c ccc eee eee eee beeen eee 165 Suecessions and eliminations, »\ic.tiwes sess eek ae ba ohne eee boa eb ew aa ea Ped Ae ay eau Poe awees eae 166 Ancient:strands s 2 s/s asians tee aged Meee es MARS RO TE SSO Y a Ge bem Hea Pau S ERM ee de Ads 168 The reoccupation of sterilized islands......... 0.0... cece eee eee ee teen eee tenes 168 GeneraL Discussion. By D. T. MacDOouGAal....... 0... ccc ccc ce teen eee tee e nee eneenans 173-182 LIST OF ILLUSTRATIONS. PLATES. pice 1. San Gorgonio Pass. The late Professor W. P. Blake. Saline plain of Salton Sink................. Facing 2 2. Map of the Desert of the Colorado..... 0.0.0... ccc ccc cent tent tenet nett eee e ne eeneenns 6 3. Map of the Gulf of California by Castillo... 00.0.0. ccc ccc ccc tenet eee ence ten cent eeeenens 14 4. Map of the regions about the Gulf by Wytflict...........0.0 0.00 c ccc cece cence tee ee eee e nn eeee 16 5. Rocque’s map of North America....... 0.06. eee een nee eee bet e eee n ett e enna 18 6. Accretion dune. Crescentic dunes. Old Beach line of Blake Sea. Big Island.................-....05% 22 7. San Jacinto Mountains. Foothills of Santa Rosa Range............0. cece eee e eee eect e tenes 26 8. Northwest part of Cahuilla Basin. Clay lumps on beach........... 0... eee eee eect eee eee nee 30 Q, OPPAMISHAS 11 OPN 5.5505 sigs atsicais esc ets og cue scar aw ea wow ARES Sees RGR can RULES waded ois eS 62 10. Cultures of chromogenic bacilli. iis... 00005 ce ceea esses sea cece ec dsenee eee teeth eseoepenuegeee’s 64 11, Cultures‘of ‘chromogenic bacilli: csscesies snes 4 69a bine 8G Se On ERS CRUE ote Aen new ee ene 66 12. Prosopis glandulosa. Dune surrounded by water. Tongues of sea extending up washes................. 72 13. Prosopis glandulosa and P. pubescens in various stages of decortication..............0 02 ccc ceneeeeceaee 74 14. Decortication of Larrea tridentata; Culture of Beggiatoa........ 0. cc cece cece tee cence e een n reese 76 15 Map of SaltoneSinke sisie ie ceves-Sco cece crocs x Savin poeseatiin eee aw Shave ea OS Gas Daas SiG ate Rava aden weuteae CER IO 116 16. Emersion of 1907, Imperial Junction Beach. Big Island and Rock Island, from Obsidian Island.......... 120 17. Travertine Wash, February 1907. Travertine Wash, February 1909........... 0.2... c cece cece eee eens 124 18. Upper limit of Travertine Terraces, February 1908. Strand of 1907, Travertine Terrace, November 1908.... 126 19. Travertine Wash, November 1909. Channels of desert washes after submergence ................0eeeee 130 20. Emersion of 1907, Imperial Junction Beach, October 1911. Emersion of 1907, Travertine Terraces, GEOR E OT siesta st Aan ace wc tata tde ick shards sxe dopas'n, oe ad os coaudasul ave weltva baatnedssd cso leuaas ais Autuee A Sato SUITS anh eee 132 21. Southeast point of Obsidian Island, October 1912. Strand of East Bay, Obsidian Island, October 1912.... 134 22. Strand of 1909, Travertine Terraces, October 1912. Ranch on emersion of 1910, Mecca, October 1912.... 136 23. Emersion of 1910, Mecca, October 1912. Emersion of 1911, Mecca, October 1912............-......005 138 24. Uppermost portion of strand of 1911, Travertine Terraces, October 1912. Strand of 1911, Imperial Junction Beach; October 1912) icaiss cateas caeraiaien main a arte vse Gan ibaa amines MBN Setehec mere oa Sas arose 138 25. Travertine Wash, October 1912. Cut bank of mid-winter 1911-12, Travertine Terraces................. 140 26. Cut bank at upper limit of strand of 1913. Cut bank of 1913, three months later ...................05- 142 27. Flotation of seedlings of Atriplex lentiformis............. cc cece ccc cece cence eee eee e nee eeere 146 28. Strands of 1909 and 1910, Imperial Junction Beach, October 1911. Dead stems bearing fruits and seeds on submerged areas. Rank of Suda on strand near Salton station............ 0... cece ene ee eee eeee 152 29. Prosopis and Cucurbita on strand of 1907, Obsidian Island. Rank of vegetation on ancient strands; high beachsline.of BlakerSeais eo icecscadissecntiea oa eve Pao aa asta sane ase arma sae lmauh eR Mos Mek ea ee 154 30. Ranks of ancient vegetation of Blake Sea. Heliotropium as pioneer on emersed island.................. 168 31. Pluchea and Spirostachys as pioneers on islands. Baccharis as pioneer on Cormorant Island............. 170 32. Islands and bars, southwest shore of Salton Sea. Spirostachys near cormorant’s nest on new island...... 170 FIGURES IN TEXT. 1. Detail of Rocque’s map showing body of water separated from head of Gulf of California................. 15 2. Diagram showing fluctuations of water and organisms in brine...........-.. cece eee eee cette teen teen een 61 3. Curves showing alterations in level of Salton Sea, 1905-12, with estimated inflow..................00 00005 138 4, Diagram showing recession of sea at Travertine Terraces, 1907-12...... 0... cece eee e cece eee e cent eee eenes 138 Ix THE SALTON SEA A STUDY OF THE GEOGRAPHY, THE GEOLOGY, THE FLORISTICS, AND THE ECOLOGY OF A DESERT BASIN By D. T. MACDOUGAL and Collaborators THE CAHUILLA BASIN AND DESERT OF THE COLORADO. By Witi1am Parees BuaKe.! EXPLORATION. The year 1853 was notable in the history of explorations of United States territory west of the Mississippi River. In that year, under the administration of President Pierce, with Jefferson Davis as Secretary of War, four fully equipped expeditions, authorized by Congress, were sent out to explore the almost unknown country lying between the Mississippi River and the Pacific Ocean, to look for and determine a practicable route for a railway. Our knowledge of the country at that date may be summarized as follows: Aside from the early exploration in the northwest, Frémont, in his daring overland explorations, had made us acquainted with the obstacles and perils of a route across the Sierra Nevada; Stansbury had told us of the great Salt Lake; Sitgreaves had crossed Arizona south of the Grand Canyon, entering what is now California near the mouth of Bill Williams Fork. Emory had made a rapid military reconnaissance of the route from Fort Leavenworth, Missouri, to San Diego, in California. To one of the expeditions of 1853 was assigned the duty of following the Sierra Nevada of California southward, to seek for any suitable pass through which a railway might be built. This survey was placed in charge of Lieut. R. S. Williamson, of the United States Topographical Engineers, with Lieut. J. G. Parke second in command, and the writer as geologist. Walker’s Pass, much vaunted at the time as the best and only practicable pass in the Sierra Nevada, was the first objective point. It was most favored by Senator Gwin of California, who had personally taken the field and journeyed as far south as the Tejon, from whose summit he could see a favorable route across the Great Basin eastward. The Williamson expedition made surveys of Walker’s Pass, the Taheechapah (the orthog- raphy of which has been corrupted to Tehachipi), Tejon, Cafiada de Las Uvas, the passes north of Los Angeles, and the Cajon from the Mojave to San Bernardino, without finding any pass that offered an especially favorable and easy route or inviting grades. SAN GORGONIO PASS. Imagine, then, the enthusiasm with which the unknown great break in the mountain range between San Bernardino and San Jacinto was approached by the members of the party as we made our way eastward from the region, then practically unoccupied but now including the towns of Colton and Redlands, and found an easy grade and open country for our train of wagons to the summit, only 2,580 feet above the sea. Here, at last, was discovered the greatest break through the western Cordillera, leading from the slopes of Los Angeles and the Pacific into the interior wilderness. It had no place upon the maps and had not been traversed by surveying parties or wagons. From the summit, we could look eastward and southward into a deep and apparently interminable valley stretching off in the direction of the Gulf of California. This pass was evidently the true gateway from the interior to the Pacific Ocean. (Plate 1 a.) 1 This paper was prepared by Professor Blake two years prior to his death in 1910 with a view to its publication in conjunction with a series of other essays upon the control of the Colorado River by the American Society of Civil Engineers. The manuscript, however, was found to be more suitable to the present volume and was secured for publication here through the kind offices of Mr. H. T. Cory. But few alterations have been made, and these consist chiefly in omissions of sections bearing upon the vegetation and other topics which are more fully treated elsewhere. The paper possesses a peculiar interest, since it is based upon observations begun by the author with his discovery of the basin-like character of the region in 1853 and extended to his last visit in May 1906. 1 2 THE SALTON SEA. The discovery of this practicable and easy railway route determined the construc- tion of a southern railroad and made it necessary to acquire from Mexico the strip of country in southern Arizona since known as the ‘“‘Gadsden Purchase.” We descended with eagerness into this great, unknown valley, carefully reading the barometer at regular distances to ascertain the grade. Proceeding without obstacles, but with no trace of a road, and following the dry bed of a stream, now known as the White- water, we reached the bed of a former lake and found it to be below the level of the sea. Plate 1 c. The es of this pass, as a great natural gateway through the mountains from the ocean to the interior, is emphasized by the magnificence of the mountain masses rising like sentinels, on the north and on the south. These are San Bernardino, over 11,000 feet and snow-capped for the greater part of the year; and San Jacinto, the sharp peak on the south with rugged sides, and similar elevation, which forms the northern end of the Penin- sula Range. This San Gorgonio Pass is the only great break directly through the mountains from Cape St. Lucas to the Golden Gate; and like the Golden Gate, it is a great draft channel for the inrush of ocean winds to supply the uprising heated air of the interior deserts. Topographically this valley into which we descended is the northwestern extension, pro- longation, or head of the Gulf of California. GEOLOGIC HISTORY. That this valley was formerly occupied by sea-water is shown by the reefs of fossil oysters and other marine shells. As these fossils are now above tide-level, it is evident that there has been a considerable uplift of the whole region, and a change from marine to fresh-water conditions. Further north on the west coast of the United States we have another great longitudinal valley separating the Coast Range of California from the Sierra Nevadas. The two great valleys are similar in many respects; they both receive the drainage of the larger rivers of the interior and protect the deltas of their rivers from the direct destructive or modifying action of the sea. In the Gulf of California we find the Delta of the Colorado, and in the California valleys the deltas of the San Joaquin and of the Sacramento Rivers. The California Valley is nearly at the sea-level. The sea has been displaced by alluvium. A depression of the western coast of less than 1,000 feet would flood the valley with sea-water through the Golden Gate from the Tejon to Shasta. A similar or even less depression of the Lower California region would carry the waters of the Gulf far north of Yuma and flood the valleys for 200 miles northwest of the present head of the Gulf. A depression of 2,580 feet would connect the water of the Pacific and the Gulf at the pass of San Gorgonio, where the trains of the Southern Pacific Railway cross the divide, and would make an island of the peninsula of Lower California. EVIDENCE OF UPLIFT. Such, no doubt, were the conditions in Middle Tertiary times. The waves of the Gulf then washed the slope of the San Jacinto and San Bernardino, where we now find arid mountains and desert plains. The silt of the Colorado was distributed far and wide in the interior sea, only partially cut off from the broad Pacific by a chain of islands which now form the crest of the Penin- sula Mountains from San Jacinto to Cape St. Lucas. As the land gradually rose from the waves, the beds of oyster-shells and of other forms of marine life came into view and may be seen to-day, 1,000 feet above the valley on the sides of the San Jacinto Mountains. Such evidences of the former marine occupation of the valley are particularly strong and convincing along the eastern base of the Peninsula Mountains, where marine fossils of the Tertiary period are numerous, especially in the stratified formation along Carrizo Creek. SALTON SEA es \. View in San Gorgonio Pass eastward into the Cahuilla Basin The late Prof. William Phipps Blake standing on Summit of Travertine Rock, May 1906, fifty-three years after his this formation original discovery o iew of Saline Plain or Saltern at bottom of Salton Sink en from near the railroad station of Salton, Februar 903 THE CAHUILLA BASIN AND DESERT OF THE COLORADO. 3 Many of these fossil shells were observed in 1853, but have since been described more in detail by other explorers, notably by Dr. E. E. Stearns of California. Dr. Stephen Bowers, who describes many of the localities, writes of the region generally as follows: “The water of the old Tertiary Sea, which once prevailed here, must have been extremely favorable to the propagation and growth of mollusks, especially oysters. After the vast erosion that has taken place, there are many square miles of fossil beds, especially of oyster-shells, which, in places, are 200 feet thick and may extend downward to a much greater depth. The oysters existed not only in vast numbers but in many varieties, from the small shell which is in evidence over so much of the territory, to varieties nearly a foot long and to others weighing several pounds each. One variety is nearly as round and as large as a dinner plate.” Fossil oyster-shells are perhaps most abundant in the Coyote Wells District, about 7 miles north of the international boundary line and about 375 feet above the level of the sea. Other deposits of marine shells, including shark’s teeth, pectens, and univalves, are reported from one of the branches of Carrizo Creek. But the occupation of the valley by sea-water, while comparatively recent geologically, has extreme antiquity and long ante- dates human history, dating back to the Middle Tertiary. The continental elevation which followed culminated in the Pleistocene, or Glacial period, when the precipitations of rain and snow are believed to have attained their maxi- mum. At that time, the Colorado of the West had its greatest volume and transporting power. Its silt was distributed far and wide in the interior sea, then only partially cut off from the broad Pacific by a chain of islands which now form the crest of the Peninsula Mountains from San Jacinto to Cape St. Lucas. Entering the Gulf just below where the mouth of the Gila River now is, it began dropping its load of débris and silt, forming the raised delta which gradually extended westward and southerly across the upper end of the Gulf toward the Cocopah Mountains and finally to the higher ridges beyond the Pattie Basin, even to the eastern base of the Peninsula Mountains. Aided by the gradual elevation of the land and by the tides of the Gulf, the building up of the delta proceeded rapidly. It assumed the nature of a great dam or levee stretching across the Gulf and diverting the river-water through shifting channels to one side or the other, first to the lower part leading to the Gulf and then to the upper end of the depressed area shut off from the tides. At certain seasons the tides rise to a great height at the head of the Gulf and are accompanied by dangerous bores. Such tides rushing up the mouth of the Colorado have ever been important factors in the formation of the delta. Ordinary tides are said to rise 15 feet, and in extraordinary cases to 37 feet. According to the United States Geological Survey the tidal range is from 14 to 32 feet. (See p. 17.) LAKE CAHUILLA. The head of the Gulf, being cut off by the Delta from the free access of the sea, became an inland lake of salt water, or at least of brackish water, with the great Colorado River at certain seasons and stages of flood flowing into it. This stream then, as now, was laden with the rich alluvial earths of its upper course, torn from the ravines and cajions of the Rocky Mountains and the Grand Cafion of Arizona. This influx of river-water, though variable in duration and quantity, must have exceeded the loss by evaporation. Consequently the level of the lake was raised until the excess overflowed to the Gulf by a lower outlet. That such conditions continued for centuries appears certain, for the enormous accu- mulation of sediment within the old beach-lines tells the story of long-continued lacustrine conditions, of the displacement of the sea-water, and of the final occupation of the valley by fresh water. This is shown to us by the fresh-water shells, not only on the surface but in the blue-clay sediments, in the banks of ravines and arroyos, and in the deep borings for water, showing that the shells dropped to the bottom and were thus entombed. These 4 THE SALTON SEA. fresh-water shells are so abundant in the lacustrine clay of the desert, especially at the northern end, that they accumulate in windrows before the wind. The thin pearly shells of anodonta are common in the clay about Indio. Four or five species of univalves, new to science, were collected in 1853. The long-continued existence of such a lake is shown, not only by the fossil shells, but by the ancient shore-lines and beaches, as fresh as if recently left by retiring waters, and especially vivid and convincing north of the Delta, where they are visible for miles. At an outlying mass of rocks at the base of the main ridges of the Peninsula or San Jacinto Mountains, a deposit of travertine marks the former height of the water by a thick incrustation, covering the granite boulders from view. The foundation rock must have been a small islet of granite projecting above the waves of Lake Cahuilla. It is now known as Travertine Point, and its base was nearly reached by the rising waters of the Salton Sea in 1907. (Plate 1 B.) By the courteous invitation of Dr. MacDougal, I had the pleasure of revisiting this place in the month of May 1906. Crossing the valley from Mecca on the Southern Pacific Railway, we visited the then rising Salton Sea, skirting it to Travertine Point, which I again ascended half a century after its discovery and description in 1853. The old water- lines and beaches were comparatively unchanged in appearance. Concentric lines of sparse vegetation marked where the waters had stood centuries before. Looking out from the summit across the Salton Sea, it was difficult to realize that the old-traveled trail across the desert lay 15 fathoms deep under water, where before not a drop could be found. The former lake, the shores of which are recorded on the rocks and slopes of the Cahuilla Valley north of the Delta, had an area of about 2,100 square miles. It was 100 miles long and about 35 at its widest. It was first identified and described by me in 1853, in a commu- nication to the San Francisco Commercial Advertiser, edited by J. D. Whelpley, in the winter of 1853-54, and later in the reports of Exploration and Surveys for a Railroad Route from the Mississippi River to the Pacific Ocean, volume v. Its boundaries were then approxi- mately shown and its origin explained. I have named it ‘‘Lake Cahuilla,” from the name of the valley and of the Indian tribe. (See pp. 24 and 25.) The name ‘‘Salton Sea”’ is appropriately applied to the recent inflow and partial inundation of the valley covering the salt-beds at Salton, but the ancient lake in its entirety requires a distinctive name. If any precedent is needed for naming an ancient lake which has disappeared, it is found in the naming of the old lake in Utah by Clarence King, as Lake Bonneville. Lahontan is another example. The Great Salt Lake of Utah is the residual lake of Lake Bonneville much as the Salton Sea is the residual lake of Lake Cahuilla. Lake Cahuilla occupied the northwestern end of the basin of the California Gulf— that portion cut off from the sea by the delta deposits. The northwestern part of the valley is also known as the Cabezon or Cahuilla Valley, so named from the Cahuilla Indians, who have inhabited the oases and tillable fringes of the Desert from time immemorial. There is a difference of opinion regarding the proper orthography of this name. It is ably dis- cussed by Dr. David Prescott Barrows in the Ethno-Botany of the Cahuilla Indians of Southern California. He writes: “A word should be said as to the pronunciation and spelling of the tribal name, Coahuilla. The word is Indian, and the tribesmen’s own designation for themselves, and means ‘master’ or ‘ruling people.’ There is some slight variation in its pronunciation, but the most usual is probably kow-wee-yah, accent on the second syllable. The spelling has been various. That used by the early writers and correct, according to the value accorded to Il in Spanish-American, is that adopted here—Coa-hui-lla.”’ The writer, in the year 1853, when passing through the ‘‘Ka-wee-yah” or Four Creek Country in California, with Lieutenant Williamson, in the endeavor to conform phonetic- THE CAHUILLA BASIN AND DESERT OF THE COLORADO. 5 ally to the Indian name, wrote it ‘‘Cohuilla,” and sometimes ‘“‘Cahuilla.” This last form seems to have been more generally accepted and is preferred to Cohuilla, Coahuilla, or any other. DESICCATION OF LAKE CAHUILLA. With our present knowledge of the delta deposits of the Colorado, the varying phases of the stream, the lightness and depth of its deposits of silt, its quicksands, its shifting channels, and uncontrollable ways, it is easy to realize that the inflow to Lake Cahuilla must have been extremely variable and uncertain. We can realize that under favorable conditions the whole volume of the Colorado may have been diverted alternately to the Lake and to the Gulf, and that long intervals of drought accompanied by drying up were often experienced. Writing upon the subject in 1853, attention was directed by the writer to the tradi- tions of the Cahuilla Indians, as follows: “The explanation of the formation of the lake and its disappearance by evaporation which has been presented, agrees with the traditions of the Indians. Their statement that the waters retired poco-a-poco (little by little) is connected with the gradual subsidence due to evaporation, and the sudden floods of which they speak undoubtedly took place. It is probable that the lake was long subject to great floods produced either by overflows of the river at seasons of freshets, or by a change in its channel, or by a great freshet combined with a very high tide, so that the river became, as it were, dammed up and raised to an unusual height. The present overflows, though comparatively slight, are probably similar; and yet it is possible that the interior of the Desert might be deluged at the present day, provided no elevation of the land has taken place, and the river should remain at a great height for a long time—long enough to cause the excavation of a deep channel for New River.’ SALTON SEA. This is precisely what has recently happened by the cutting of irrigating canals and by the uncontrolled flow of the Colorado water: deep and destructive channels were cut, a partial flooding of the desert followed, and the ‘“‘Salton Sea’’ was formed. The body of water which so recently threatened the restoration of the former lake conditions, by the month of February 1907 had attained a length of 45 miles, a maximum breadth of 17 miles, and a total area of 410 square miles, with a maximum depth of 83 feet. It extended from Imperial Junction nearly to Mecca Station. It submerged railway stations and neces- sitated the removal of the track of the Southern Pacific for 67 miles to a higher and more northern bed. By the great and masterful exertions of the engineers in charge, seconded and supported by the Southern Pacific Railroad, the destroying deluge was stopped in the month of February 1907 and the gradual disappearance of the Salton Sea by evapora- tion commenced and is now in progress. In this we have immediately before us a prac- tical exhibition of what must have happened many times before. Evidently in the case of the ancient Lake Cahuilla, with the loss of the supply of water from the Colorado the lake disappeared by evaporation. The conditions for this were extremely favorable. Of the rate of evaporation and the time required for the com- plete desiccation of the valley, we have no direct evidence, but there is every reason to accept the statement of the Indians that the water retired little by little, or very slowly, and no doubt years passed before the lake dried up. RATE OF EVAPORATION. Experiments by me upon the rate of evaporation in the Tulare Valley, California, in 1853, indicated 0.25 inch per day, or between 7 and 8 feet yearly.?, Dr. Buist found 1 Report Geological Reconnoissance in California, p. 238. ; * Report Geological Reconnoissance in California, p. 195, and Trans. Geog. Society, vol. rx, p. 39, 1849-50. See also, Trans. National Institute, Washington. 6 THE SALTON SEA. that the amount of evaporation from the surface of the water at Aden, on the Indian Ocean, was about 8 feet per annum. At the rate of 8 feet yearly, the 83 feet of water now covering the Desert, and known as the Salton Sea, will require ten and a half years for its complete evaporation. Mr. H. T. Cory, the engineer who had charge of rediverting the Colorado River to the Gulf of California in 1906, states! that if there is no further inflow of the Colorado River to the Salton Basin, the sea will practically dry up by evaporation in about eighteen years, and that the actual evaporation from the Salton Sea from February 1907 to July 1912 has been almost exactly at the rate of 5 feet per annum. From measurements of the evaporation from a tank at Calexico by Mr. Peck, of the California Development Company, the annual evaporation was shown to be about 6.73 feet, as will be seen by the following tabular report: Taste 1.—Evaporation from a water surface at Calexico. Month. 1904. 1905. 1906. Month. 1904. 1905. 1906. Inches. Inches. Inches. Inches. Inches. Inches. January...... 4.39 2.72 2.57 August...... 10.98 8.52 8.47 February... .. 6.32 1.47 2.43 September... 8.61 7.83 6.73 March....... 8.86 4.44 5.06 October..... 8.78 6.77 5.45 April........ 9.55 4.74 5.99 November... 5.40 3.23 3.61 May scncs cass 10.91 8.38 6.84 December. . . 3.48 3.43 2.40 June......... 13.89 12.86 7.41 a =} | JULY.s Sass oe 12.47 10.43 6.76 Total..... 103.64 75.00 63.66 COLORADO DESERT. The drying up of Lake Cahuilla left a broad region at the head of the Gulf; a depressed area below the sea-level, a trackless waste of nearly level land extending, including the Delta, for some 200 miles northwesterly beyond the present limits of tide-water in latitude 31° 30’ N., approximately 80 miles south of the mouth of the Gila River at Yuma on the Colorado. The limits of this desiccated area are approximately marked indelibly on the ground by the shore-lines and beaches of Lake Cahuilla, extending on both sides of the valley from near Yuma to Indio and beyond. The name ‘‘Colorado Desert” was given to this region by the writer in 1853. This was before the State of Colorado received its name. It was deemed most appropriate to connect the name of the Colorado River with the region, inasmuch as the desert owes its origin to the river by the deposition of alluvions and the displacement of the sea-water. A tendency is shown by some writers to extend the area known as the Colorado Desert so as to include the arid regions north of it, especially the mountainous region along the Colorado and the Mohave, partly known to-day as the ‘‘ Mohave Desert.” This was not the intention or wish of the author of the name. It was intended to apply it strictly to the typical desert area of the lacustrine clays and alluvial deposits of the Colorado where extreme characteristic desert conditions prevail, such as arid, treeless plains, old lake-beds and sand-hills—such conditions as are found in the Sahara of Africa and in the delta regions of the Nile. The appellation may properly be confined to the regions rcached by the deposition of the silt of the Colorado, whether in the form of deltas or at the bottom of ancient lakes. I should also include the bordering detrital slopes from the contiguous mountains. So restricted, the area is practically coterminous with the ancient beach- lines and terraces of the lakes which occupied the valley. Its area is estimated at not less than 2,100 square miles; its breadth east and west opposite Carrizo Creek about 33 miles. Its height above tide ranges from 135 feet above 1 Proceedings of American Society of Civil Engineers, November 1912. PLATE 2 \16° pS +34 ° Oo St Hy, # v Dn 0, ). GEOGRAPHICAL FEATURES OF THE CAHUILLA BASIN. 19 range, some 7 miles north of Julian. It drains the country to the south and west of the Santa Rosa Mountains and forms a junction with Carrizo Creek in about 33° 6’ N. and 115° 57’ W. Carrizo Creek itself has two forks, one coming from near the international boundary-line, and the other from the south side of the Fish Creek Mountains. The combined creek thus formed runs north to its junction with San Felipe Creek, as above mentioned, and carries on its own name down into the Salton Sink. Here, too, the water- flow is very irregular, long periods of total dryness alternating with occasional heavy floods. THE CHUCKAWALLA WASH DRAINAGE. The only drainage system of any importance which reaches the Salton Sink from the northeast is that known as the Chuckawalla Wash. This is a large but generally dry desert wash, which drains an area of some 250 square miles, beginning with the slopes of the Chuckawalla and Eagle Mountains. It spreads out and is virtually lost in the extensive playas which lie northeast of Durmid, but at times, during some of the heavy and violent rains to which this region is occasionally subjected, a portion of its water may pass on down into the lowest part of the Sink by way of the Salt Slough. RECENT HISTORY. During the summer of 1890 the water from the Colorado River filled many of the small channels and lagoons toward the southwest, and in 1891 flowed through into the Salton Sink and formed a lake several miles in length. The intervening region was com- paratively little known and its drainage system hardly comprehended at that time, and the appearance of such a large body of water in close proximity to the Southern Pacific Railroad attracted much attention and gave rise to some of the wildest of rumors and hypotheses as to its origin. William Convers, followed by one or two others, succeeded in making the journey by boat from the Colorado to the lake, and so the mystery was solved. Mr. H. T. Cory, who has a comprehensive knowledge of the conditions in the Delta of the Colorado, concludes that some flood water has found its way down the channel of New River toward the Salton every year since the inundation of 1891, and cites opinions of old settlers who allege that water came into the Salton in 1840, 1842, 1852, 1859, 1862, and 1867. The mail stage service between Yuma and San Diego was interrupted by the flood of 1862 and a flatboat was used for crossing New River for several weeks in the summer of that year. (Transactions Amer. Soc. Civil Engineers, vol. Lxxv1, pp. 1204, 1571.) In 1900, a company having been formed for the purpose, work was begun upon the task of connecting and clearing the various channels which formed the natural waterway between the river and the basin; and by the middle of 1901 water was flowing upon the irrigable lands of what has since become known as the Imperial Valley. It had been deemed advisable by the promoters of the scheme to take the water from the river in United States territory, and so the upper section of the canal was cut almost parallel to the river for several miles and with a very low gradient. This circumstance, together with the general unsuitability of the site selected for the head works, caused considerable trouble for two or three years, as more and more water was required to fulfil the demands of the growing communities in the desert; and so various openings were made between the river and the canal in order to furnish a more adequate supply. Then in the winter of 1904-05, one of the infrequent winter floods in the Colorado, coincident with a tremendous rush of storm waters from the Gila, found before itself the unprotected head and comparatively steep downward grade of the canal, and at once began to cut and enlarge the channel. The ordinary summer flood of 1905 also poured its water through the opening, and it was soon realized that the outpour had got beyond control. Practically the whole of the Colorado was now flowing into the Salton Basin and an- other flood in the following November (1905) made the task of closing the breach seem 20 THE SALTON SEA. almost hopeless, although the most strenuous efforts were being made by the engineers; and it was not until February 1907 that the Colorado was finally returned into its former channel. Here again, however, the vigorous vegetation of the Delta had played its part, for the river bed had in the meantime become so choked by plant growth and the deposi- tion of silt that the water since made repeated attempts to escape, first towards the southeast through the Santa Clara Slough, directly towards the head of the Gulf, and since—in spite of some rather hastily planned and inadequate efforts to control it in that direction—into the head of Bee River and the Pescadero and so by various ways into the Hardy, which is now, in its lower reaches, carrying virtually the whole volume of the Colorado. This surcharging of the Hardy Channel had the further effect of allowing a large quantity of water to flow over its western bank and towards the Pattie Basin, and a large lake now fills the lower part of this Sink. During the summer of 1906, and at the time when the maximum inpour was reaching the Salton Sink, the channels of the Alamo and the New River began to cut backward from the lower end, and the soluble, loess-like soil along their courses was carried bodily into the lake, leaving the deep, precipitous-sided valleys through which the present streams flow. Much apprehension was felt at the time by the engineers lest their cutting action should reach the Colorado and so preclude all possibility of repairing the breach, but the closure was effected in time and the danger averted. Recent observations at the eastern end of the Salton Lake (October 1912), now that the water is receding, show that there has been an enormous deposition of this or other sedimentary material subaqueously; and with further recession it will doubtless be found that this deposition has taken place far out into the lake. Some idea of the magnitude of this displacement and redistribution of material may be obtained from a statement made by Mr. Cory,! that the total yardage thus moved is over 450,000,000 cubic yards, or almost twice that of the Panama Canal. A good deal of water still passes through both the Alamo and the New River, but this is merely the overflow from the irrigation canals, and it is quite improbable that with the interests now at stake in the Imperial Valley and the close watch kept upon the river by the engineers, any further uncontrolled incursion of water will be allowed to take place. In the region of the lower Delta, however, conditions are very different; here we have a large, wayward, and silt-laden river, thrown out of balance by a temporary diversion, and always hampered at its mouth by great and violent tides, wandering at present virtu- ally unchecked over a large area of friable alluvium with downward grades in several directions. All these conditions tend, as may be readily imagined, toward a condition of instability and possible geographical change; in fact, it is probable that even if the Colorado and the general drainage conditions through the Alamo and its associated channels had not been interfered with in any way by the operations of the irrigation engineers, another diversion of the river water towards the west was about due from the natural causes outlined above, and would in any case have ensued within a few years. It is furthermore evident that as so much of the flow of the river during the growing season of the early spring is now diverted and utilized for agricultural purposes, and as the bed of the river in its lower reaches is left practically dry during the period of most rapid growth of the Delta vegetation, its obstruction and elevation will be more rapid and the stability of the irrigation and protective works menaced more and more unless adequate measures are taken for controlling and storing the flood-waters of the early summer upon the upper Colorado. 1 Transactions of American Society of Civil Engineers, vol. uxxv1, pp. 1204, 1571. / SKETCH OF THE GEOLOGY AND SOILS OF THE CAHUILLA BASIN. By E. E. Frees. INTRODUCTION. The first geological examination of the region covered by this volume was made by the late Professor Blake as a part of his work for the Pacific Railway Survey and is de- scribed in the reports of that survey,! and more fully in the present volume. Following Professor Blake’s examination, the region received only incidental notice? until 1909, when Mendenhall* published a geologic sketch of the desert in connection with a discussion of water resources. More recently Harder‘ has published some observations incidental to a study of the region just to the north. The conclusions of the present paper are based mainly upon personal study of the region, though much use has been made of the data of Blake and Mendenhall. On various trips to the region the writer has had the advantage of accompanying the late Dr. W. J. McGee, Dr. J. M. Bell, and Professor J. C. Jones, as well as others associated with prepa- ration of this volume, though not concerned with the geology or soils. Acknowledgments are due to all of these gentlemen for many valuable suggestions. The soil studies were made while the writer was connected with the Bureau of Soils of the United States Depart- ment of Agriculture, and were carried out with the permission and advice of Professor Milton Whitney and Dr. F. K. Cameron, of that Bureau. The writer is indebted to these gentlemen both for facilitating the course of the work and for permitting the present use of the data obtained. During these soil studies the writer was accompanied by Mr. L. D. Elliott, to whom thanks are due for most efficient assistance. DESCRIPTIVE GEOLOGY. The general topography of the Cahuilla Basin and its bordering mountain ranges is outlined in Mr. Sykes’s paper in this volume and need not be reviewed. In its major features it differs little from the topography characteristic of nearly all desert basins of North America—rugged bordering mountains half-buried in long smooth-sloped aprons of mountain waste, which merge finally in a central salt flat, in this case covered by the Salton Sea. The main exception in the Cahuilla Basin is the openness of the southern end toward the Gulf of California. Here the mountain rim is lacking and the basin is separated from the Gulf only by the alluvial ridge of the Colorado Delta described by Mr. Sykes. Perhaps the depression as a whole is better described as a ‘‘trough” than a “basin.” The geological structure of this trough is not well known. The bordering mountain ranges have not been studied in detail and little is known of their structure, except that it isnot simple. The tilted block structure, so frequent in the ‘‘basin ranges”’ to the north, is not discernible here. There is some reason to believe that the axis of the trough is the locus of a major fault-line continuous with the great Cajon fault to the west, and running substantially northwest by southeast. Whether or not this be the case, there has un- doubtedly been much faulting in all the bordering mountains and the trough in general is certainly structural, whatever may have been its origin in detail. 1 Pacific Railway Reports, Sen, Ex. Doc. No. 78, 33d Congress, 2d Session, vol. 5, part 2, 1856. 2 See especially, Bailey, Saline Deposits of California, Cal. State Mining Bureau, Bulletin 24, 1902. 3 U.S. Geological Survey, Water Supply Paper 225, 1909. 4U. 8. Geological Survey, Bulletin 503, 1912. 21 22 THE SALTON SEA. Petrologically, the available information is similarly meager of detail. On the north and northeast the Chocolate, Cottonwood, San Bernardino, and San Gabriel Mountains form one range, nearly continuous from the Colorado River, until they merge with the Sierra in the highlands of Tehachapi. Where this range has been examined it is composed almost entirely of granites, gneisses, and ancient schists. From his studies in this range and to the north, Harder distinguishes two general divisions of the granitic rocks; the first is probably Pre-Cambrian and associated with gneisses and schists; the second is a later intrusive, probably Mesozoic. Overlying the early granitics, but antedating the intru- sives, are fragments of slates, quartzites, and limestones of unknown age, and usually much metamorphosed and disturbed. The studies of Mendenhall’ indicate a similar character for the Santa Rosa Mountains, Fish Creek Mountain, and other outliers of the Peninsula Range which border the basin on the southwest. Along the basin-ward foot of the San Bernardino Range, throughout nearly its whole extent, are irregular and much eroded hills of poorly consolidated conglomerates, sand- stones, and shales. In most cases these strata have been greatly disturbed and their original relations are not clearly decipherable. It seems probable, however, that they consist of three fairly well-defined members: (1) a basal conglomerate resting normally upon an eroded surface of schists and granites, such as make up the core of the range; (2) lying conformably upon this, a thick member of coarse, arkose sandstone, usually reddish in color and of quite uniform texture; (8) an upper member of very variable sandstones and clays, mostly thinly bedded and showing many probable minor unconformities. Aside from the vertical variation as thus noted it seems also that the fineness of the material in each of the main divisions increases as one recedes horizontally from the mountains. Thus the lower conglomerate decreases in thickness and its constituent pebbles become smaller, the middle arkose member shades off into a sandstone of much finer and more uniform grain, and the top member, while almost entirely of sandstones at the foot of the mountains, becomes almost entirely of clays as one approaches the basin floor. In many places these clays contain thin crusts of gypsum and in one place, about 5 miles north of the railway station of Pope, they contain an interbedded stratum of mirabilite or hydrous sodium sulphate, several feet in thickness. This series of strata will be designated the Mud Hill Series, from the Mecca Mud Hills, where it is perhaps best developed. At that locality the section corresponds to perhaps 3,000 vertical feet,? but it is impossible to be sure of this or to draw a general section because of the very broken character of the strata as exposed and the impossibility of determining whether observed differences in the visible fragments of the strata indicate strata of different vertical position in the section or simply variations in the same stratum at different distances from the mountains. Beyond the southeastern end of the Santa Rosa Mountains similar sandstone and clay strata surround the granite core of Superstition Mountain and form several series of low hills which flank the outliers of the Peninsula Range much as the San Bernardino Range is flanked by the strata just described. It is scarcely possible definitely to correlate these strata with the similar series on the northeastern side of the basin, but the lithologic similarity is apparently complete and it seems very probable that both exposures belong to the same series and to the same genetic condition, whether or not they be exactly syn- chronous. Provisionally, all will be regarded as belonging to the Mud Hill Series. No fossils have been found in the Mud Hill Series at the type locality at Mecca or, indeed, anywhere on the northeast side of the basin. In the valley of Carrizo Creek, however, Blake discovered strata of partially consolidated clay and sand containing oysters, pectens, and other marine shells of Miocene age. According to Mendenhall these fossiliferous beds are here the basal member of the sedimentary series. This, in connection with its general 1 Loc. cit., pp. 10-14. ? Mendenhall estimates 4,000 to 5,200 feet, loc. cit., p. 12. SALTON SEA in Carrizo Region with Prosopis Shrubs he foreground Accretion Dune Barchan or Crescentic Dunes in Carrizo Region with wind ripples in Salton Li > ake at level of 1912 on left Lighter @ | cuties sus many z., some v. Oia co Betaesn Osten 215 east COP diiscssed do 7/25 |28 | 1.240 Osc ars as MALY waiei cs eacas O's ncvecece 2 athe Oe st eankiss shyt 0% agiccih's all ponds very red, water temp. taken under and through crust 14/25 |32 | 1.2425 (1: pee ene many z. and v. |more.,......... GOs: cadiave sty ni amen artemias many 21/19 |20.5)1.225 | hidden......... Os 2 scpesuiis FOV sts Soars 2 cae GOs soc esiese a eer ets artemias none 28 | 24 25 1.235 | bright ......... GO!S Sadeudevdave Many .......... Osc encarta det teh oO Oct. 5/23 |29 | 1.240 GOn es dsees oie WO’: s essicave saa few. snes ters Owe seieitiy ea some few and small 13} 19 | 21.5 | 1.245 OO ceva Oe Aiidean is ate Oe citi sas DOs colorless’. cc, s.iscizsys.||) Sines crystallized out; 8|16.5]/16 [1.115 GO Yeritersr dn many (div.), few v. Oe p-cchices -eeninen AOR, eet) || eaeiee: || ease het rain for 2 days before 14)10.5]14 | 1.145 GOs) sss Rteee GE Obs cscciese: || atten ee Sa Asa HME | PSE ERROR Eas Dl chee || PHGEAE BREET? no artemias 22/17 11.5 | 1.1675 DOYS rie ne oe many z., few v...| 80Me€......-++-- greenish.......- aiegs | beatabene Sousa : do 29 | 115] 11.5] 1.1575 DOr Sietvoh done Many z., Many v. AO}s oh pds os CO vice watered |] episcere I) aeeenind Ss BORA rained 3 days, no 1912. artemias Jan. 4 | 14.5) 11.5 ]1.110 AGS earl ites Oi... Souenret: do 11 |14.5|14 |1.1075| dark clouds..... Oss sibisre tates do 18 /12.5}15 |1.125 | raining......... GO; cfd gas ae do 25/11 13.5 |1.135 |rain...........- MADY sive 2e 225 brownish........ é ; Feb. 7/19 |21.5|1.120 | clear.........-. many z., few v... greenish ......-. | ceeee | eee rere eee E. ditch many uni- cell. greens 8116 |20 | 1.120 ri (cee ES dig. egress do........5- 15/14 |18 {1.135 DOag ee ceeed & Ojos sapere. brownish........ 22 | 22 20.5 | 1.1425| bright.......... | oe ee ee ee ee GOs eee coarse 29/19 |2251/1.155 j cloud .......... DOvidedes stab 3 greenish........ Mar. 7/11.5]20 {1.130 |clear........-.- dO sscxe sah 3 brownish ....... 14]/14 [21 | 1.115 Onan MAES DO scusess santa es near colorless 21 /17.5/18 1.125 Ot eceee voter very few z...... brownish........ 28 )15.5}18 | 1.135 do.......... | same as in No. 1 colorless........ | -+2++ [oeeeeeeeeeee a few artemias Apr. 3] 16.0} 24 1.1625 Oi iinats ence ss small few, no large asin No. 1...... 10)12 |14 | 1.145 CO}. ssid eke MANY 2: 2s eesees colorless........ 17 | 14.5 | 24.5 | 1.170 GOs vasa ened PD itarieeed wig sr's asin No. 1...... 24/16 24 1.1975 Oe Saeatonc unk do, some dead | many, some dead | pinkish......... May 2 | 16.5 | 25.5 | 1.215 Ow s wee acs as in No. 1 and Goncewaess No. 2 Tank No. 4. 1911, Aug. 16 | 20.5;}28 |1.253 | bright.......... many (fewlargev.)| many .......+.- PINK sess eceyenavene, 3 . leryatallized out; artemias dead Aug. 23 | 22.5) 33 | 1.235 OO ie akie MARY. 4s dsveae 2 LOWiieck se bsialaidase do LA ce dO. eee kcdavs artemias many Sept. 1 | 16.5 | 25.5 ) 1.237 DO veges gcerees ett conjugation?.... | more........... some OE: sansa do 725 (28 | 1.225 Wilind eaneeee 8 (es et cx] nae e awe woe dw E6534 33 do 14 | 26 28 1.235 dOee see aas many z., some v. |many........-- Os Sco Ste 21)19 |20 1.220 | hidden, showers.. BG, udu tude ee POW a ceyagen o4 3 do, vsass artemias none 28/23 | 24.5 | 1.2455} bright.......... DO wisteed wissen MANY ....------ dG: s5 eae do Oct. 5/22 |27 | 1.245 owiicrensnens dO sciedigeie 3 Oieceynssie | Ox agetis yee sei DO vk tne artemias few and some dead 13.|17.5) 21 | 1.251 CO causes latin DO tne At Ok eee es GOs eer orig) tone dO: eehne's do 19|26 |27 |1.2325 do... seve dO isin ea es OO aces tee OGRE eee crore | SARS dO: waciets artemias some 2617 | 16.5) 1.200 | clouded ........ GO sats toisebiien Dc achsid abies colorless........ Geaxiaey do Nov. 2|22 |25 |1.285 |clear........... dois iaytaneie.: doin Gis vats PinkkiShhss/secses 2:4] ease Oar kensrs do 9/18 |21 | 1.2425 | clear but clouds.. es riche GBs 6 hat gibeas red or brown.... | some | being removed| mother-liquor 16/17 |19 |1.230 |clear..........- DO's 2 nig iace 5 O's crased senate pinkish (drained) | ..... | ..........-5 practically mother- liquor 23/11 |20 | 1.200 GOs cea eae asin No. 3...... Dee. 1/25 |17 | 1.215 Oia 2ecoscunn' nvera agin No. 2...... 8/16 |11 | 1.0725 GG ace gape some z., some v.. | few..........-. colorless........ MABRY | esi ww cde rained for 2 days before 14 | 10.5 | 13.5 | 1.115 dO Sais ieies GO vinci < 22% very few ....... dO snnssetars BOME: | woakavene yes probably overflowed by tide, no arte- mias 22 | 16 11 1.11 GO ence ae many z., some Vv. | some....... ... | greenish........ GOS Wise peek siete cite and 2 no artemias 29)11 [41 | 1.215 OO se retaiwaise Oe ceaviwccvas WGiia: cides colorless........ GO; lace ce saroteees rained 3 days, no aie artemias Jan. 4 ve 11 1.110 F GO vnc asde sss many z., many v. GOidiwe a saeye-t do. do 11 14 | 1.1075 | dark clouds..... DO ietr” aayaneetie {: (sere eee do. i 18]12.5/15 [1.120 |raining......... dos: cane: Pedi) udomaaeashirs dow hd ras aa 25 ) 11.5 | 13.5 | 1.1275 dO vice ices ee MADY? 5 sesoies oy oe fewer (somereat.) |brownish.......) | Feb. 1/19 21 LATS clear. ccna 8 | 15 20 1.120 GO sie): 2a sasax 15 | 14 18 1.1375 | clear........... 22 | 21.5 | 20 1.140 | bright.......... 29/18 {22 |1.160 |clouded........ se Mar. 7/11.5|20 /|1.125 DO jsyesesnie tearing Obese cet sneve's BONG... cece wrens faintish . 14]}14 |21 [1.1075] clear........... Ose hee Os es suuar snc colorless 21)17.5118 |1.115 DOteveswaveex OY. asad rreisare OS aig oval ae DOs iissesecs some 28|}15 |18 | 1.130 COS nid 8 a8 DOs saad Segotue very few........ dog moa Apr. 3/15 | 24 |1.160 Oe sanss sank (eee reer CGincs a. onex ve asin No. 1...... 10} 11.5]14 |1.1375] do......... do Messen Ow ann anh colorless... ... 17|14 |24 | 1.180 Oe secisite mesa DO 50 saesd os a CO ese aratscod as in No. 1 24/16 124 |1.2175 Ore wet ate sedate Osx. sah eacay MANY, x waies a2 pink. May 2/16 | 25.5 | 1.225 Ban. wiki snce BG cake does doviaki cs THE BEHAVIOR OF CERTAIN MICRO-ORGANISMS IN BRINE. TaBLE 22.—Record of living contents of samples of brine taken from six salterns—Continued. 59 Tank No. 5. Tem.| Tem. ie Bre No. Pyra- Date. ie elas ae Sun. o Cae gti Color of water, ee Salt. Remarks. 1911. Aug. 16/20 |28 | 1.235 same as Noel. 25. | ase || ie sicaods aes artemias few 23 | 23 32 D286: |), [dO kde sone inn VS MNO ep cos, Sghce ay arakadeieavssend acks'l| Moseuspiuasieus ed: audcard Webi] e-omuade Ide wend Us artemias many Rept. Lae (2058 2400) | dle uenees Wt dO: onan ania Il eu ealenmase name ad line oe wader e aitn goal atedats, dave sey auawiels Seg do %\20528) 1.235) dO. wens coes [very many “240 |p MANY dicen. oo: alee cauteen ecauee of] Moean | Seow vere ee artemias few, many dead 14 | 26 29 Pee. Seeks taiee | Ob es cece Rei uslnddes anu Pande uamwaree lll waiein | Abe gedaan artemias none 20:20) 11005: | 1.2376 |‘out-atter shower: |:ag in: Nos Tocris, |) aseaoe agieieid « wate ol] cede dieals wa, Seasa 4) || secesaanyes | rete eemrac odode do 28 | 23 24 V2425)) DrIgb by assess eset fe ADisnastes ani af: ll geo ceaehein werneess oe wll ean wearer | <2 eae I hee Wee aes artemias few and small Oct. 5 | 22 26 Q2g5 | EOkeoss ete Many ea few Vevs | MANY < iccee eee as ce deena vsres | caacee | dsencux esas do LB TE (20) 22S dO] dO eee ws.) dOsere secu [resend oom || ater || aa ee eemens artemias some 19 | 26 27 1.2452 pgnaty I epleal! Sat ealsnaates es do 26 | 17 17 1.170 very greenintube| ..... | ..........6- do Noy. 2 | 22 25 12375 | cleariicons occ egae many z., many v. Oj tte ae pinkish......... some 9/18 21 1.2375 | clear but cloudg.. GO exwieas 4 end some........... (mother-liquor) do reddish and re- filled 16 | 17.5 | 19 1.220 OS 2. Saas, Suet many z., few v... Osa a atts vi pink-brownish. . . 23 )11 |20 | 1.1975 O's. cawedownd asin No.3...... Dec. 1 | 25 17 1.2122 Owe osiaee ease some (div.),some v| some (div.)..... colorless........ 8 | 16 16 1.1225 Oy chistes some z., Many v. DO aed eyes greenish........ some |........005 . | rained for 2 days 14) 10 138.5,| 1.1325 Oks median an? some z., some v.. GOs Sie es OGisAe psaticas ¢ c Go! + \lagea eee yae probably overflowed by tide, no arte- mias 22 | 16 1 1.145 CO}. ssotixiyd Sake many Z., some v. DO rxate's siecle Golorléass % siculse-|| eeniua-|| wes x iseues 4s no artemias 29} 11 11 1.120 COs indies She do OO suiicce seesys S5 Oi sseticcerienn some |............ rained 3 days, no artemias 1912. Jan. 4/14 |11 | 1.1075 Oia er teen many z., many v.| few....... ied Se LG weiSsaie szncyisata ers do 11 | 14 14 | 1.0975 | dark clouds..... Mc: wey ng oe SOME: 6 ssc: 44° GO; « Wksasicarecanieoe os do 18 | 12 14.5 | 1.1225 | raining......... Gn 20-2 ceniscer de AG sei cco es ducecine faint greenish... |/few | ......---0.5 do 25/40 718 11.135 i brownish....... Feb. 1/18 21 1.115 greenish........ 8 | 14.5 | 20 1.120 15 | 13 18 L130: fh. GOvectnanes gee ll dO ~~ MOiacediaess | Colorlessoxccces as) Seeger || Wa eames protozoa well fed! 22 | 21 20 1.1475 29 | 17 22 1.150 Mar. 7/11 20 1.1325 14 | 14 21 1.120 21/517 18 1.1275 some 28 | 14 18 1.1325 Apr. 3/14 24 1.155 10 | 11 14 1.1625 17|14 |24 |1.175 colorless........ some 24/16 23 1.215 Pinko es eee May 2/16 24.4 | 1.220 Once on nuess x 1911, Aug. 16 | 20 29 1.220 | bright.......... many Z......... very few........ colorless........ .. lerystallized out) filled from pickle- pond No. 3, many artemias 23 | 25 33 1.230 Os 52h ech ee conjugating?.... WO wecnscnaes pinkish... 2e:.44 | esos (: (cer pickle-pond No. 3 refilled: greenish, large artemias, eggs Sept. 1/165) 25 | 1.232 Os: ssceeegees many z.,few adults| faint pinkish.... | ..... Ordo ds 7 | 26 29 1.233 Ds 5 dete seers many Z......... DOs casa Sek sees do...... 14 | 24 wees | 1.2250 Oe agansgi.ossnscens many z., some v. 21)20 |19 | 1.2325] bright after show- GOs. eseear eas DO textecin have CO igi sz artemias none ers 28/24 |25 |1.2375)|bright.......... Os sees MANY v9 sees e < Omens and x AO! es 54 do Oct. 5/22 |26 | 1.245 CO! cies wie many z., few v.. GO sicces eee 5 Ose is arg se's |! es eevee GO! cecets artemias many but small 13)19 | 21.5 | 1.245 COs vay 2 wees GOis4.2 eget ety very few........ Goieksus one « ||| Kaa de AO: veal ipaee few and smal 19 | 27 | 27.5 | 1.250 Gs icoc's Kees BG uci bare DOleeeee eaters $ WO peyis artemias some 26 \378:) 17 | 1,190 | chowded .....0u, ile wa aleieiesy ¢, | COLOLISSS 0 were iol] Axes vie GO? cessive do Nov. 2 | 23.5 | 24.5 | 1.2475| clear........... OF scch e's do 9/19 |21 {1.250 | clear but clouds. some dividing?.. | reddish.........] ..... GO wsigices.s 16/17 18.5 | 1.245 Os jag tas conjugating?.... | pinkigh.........] ..... GO! sissiuccs 4 dividing...... 23/13 |20 |1.225 Oi sae estes some z., few v... | few............ eolorlega. s 5 eigen | ay xx UO odeas 60 THE SALTON SEA. TaBLE 22.—Record of living contents of samples of brine taken from six salterns—Continued. Tank No. 6—Continued. Pyra- .|Tem.| Sp. gr. No. of No. of y Date. tem bee | eet Sun. D. viridis. D. salina. Color of water. mim- Salt. Remarks. 1911. Dec. 1]|25 | 16.5 | 1.2475} clear but clouds.. | some z., few v...|few..........+5 brownish....... | ..... 8/18 | 15.5 | 1.185 Os 3 sear. cody many z., few v... Oi 6 5sie 9 S3 greenish........ many divid- ing 14 | 11.5 | 14.5 | 1.205 WO): sive anaes Oz aise + tues dO: eveceeees colorless........ some: fi vised ss auked sess | fb Les emriees, lll elas eta | Aakeoned |] hae woken Man CH ana score sie Se acleaes's aiaerayscechansts o: trace 0.013 0.017 .021 0.025 Chlorine: (CD esi a esvecieg oe tees é 169.75 204.05 240.90 280.93 339.42 Sulphuric acid (SQu).. 2.0... 6. eee ee 47.60 56.74 65.87 76.36 91.67 Carbonic acid (COs)..........--.005 6.58 7.66 7.34 6.38 5.78 Bilicio: (Si@a)vcsoe.c stain teens eee 8 Y 1.41 1.43 1.59 1.55 1.83 Phosphoric (PO) ... 0.009 0.011 0.01 013 trace Nitric (NO3)........ oe 0.18 0.20 none none none Nitrous (NO2)....... -..| Done trace 0.0006 none none Oxygen consumed.........-....--.-- 0.093 0.059 0.068 .045 0.063 Borieacld wea ects einin snacks os daketsa| cea trace!) nasa seats trace trace A critical examination of the substances present in the Salton Sea waters does not indicate that they are active in disintegrating the cell walls of woody tissues or necessarily cause even a breaking down of the tissues in herbaceous plants. However, the sulphates present in the water have a very definite relationship to certain bacteriological processes supposed to accomplish decomposition of cortical portions of woody stems. No analyses were made with reference to the gaseous contents of the Salton Sea water, but it contains manifestly an abundance of oyxgen, inasmuch as large numbers of fish were present during the summer of 1912. The water has a specific gravity of 1.001. It is clear and quite free from sediment and may show an osmotic pressure of 5.5 atmos- pheres. The currents of the water are largely confined to those produced by the wind, since there is no outlet to the basin. A review of the physical facts indicates that the decomposition of vegetable tissue immersed in the water, even though some of the woods were covered for five years, could not be attributed to the action of non-biological agents. This conclusion was supported by the laboratory tests conducted with fresh woody tissues of Prosopis glandulosa, Prosopis pubescens, and Larrea tridentata, which remained absolutely intact during the eight months of submergence in sterilized Salton Sea water. Biologically considered, the submerged condition of all of the herbaceous and woody plants would be, of course, fatal to those organisms within a short time and would re- move them, therefore, entirely from the question of resistance to the physico-chemical fac- tors of the Salton Sea. On the other hand, the temperature, the light, the gas-content, and the food supply were extremely favorable for the development of a rich bacterial flora. This was demonstrated by the fact that there were several thousand bacteria per cubic centimeter of water when first collected from the Salton Sea. These organisms were represented by several species, some of which were found to have a definite relation to the change in the chemical composition of the water and to the decomposing processes under- gone by dead, submerged organisms. Inasmuch as the plankton did not seem to enter into this study all reference to that biological phase of the Salton Sea was omitted. The specific investigations of this problem were directed (1) to an anatomical study of the species submerged from one to five years, and (2) to a bacteriological study of con- siderable quantities of the water itself and of processes undergone by fresh woods sub- merged in containers filled with Salton Sea water and maintained at room temperature in the Botanical Laboratory. The investigations concluded with an anatomical investi- gation of the fresh woods which had been kept under control in the laboratory during the months that the bacteriological experiments were carried forward. SALTON SEA PLATE 12 A. Prosopis glandulosa, the base of which was submerged in 1906 and laid bare late in 1907. Photographed late in 1908 by D. T. MacDougal. B. Segment of Crescentic Dune surrounded by Rising Waters of Lake in vicinity of Carrizo Creek, February 1907. C. Tongues of Water from the Rising Sea in the Channels of Desert Washes, Imperial Junction Beach, February 1907 THE ACTION OF SALTON SEA WATER ON VEGETABLE TISSUES. 73 ANATOMICAL STUDIES OF THE SPECIMENS SUBMERGED ONE TO FIVE YEARS. Obviously it was important to learn definitely what changes, if any, had taken place in the tissues of these woody plants during their term of submergence, as indicated in Plates 13 and 141. It was found that almost the entire cortex had been lost from all specimens except those which emerged in 1907, after a single year’s immersion. A detailed study of the cell walls was necessary in order to answer a question which had arisen relative to the possible procedure of petrifaction. In addition, to give opportunity for study of anatomical structures, a series of sections were made in the transverse, tangential, and radial planes. These were placed in various liquids in order to soften them for sectioning on a special type of swinging microtome. Sterile water, unsterile water, and concentrated hydrofluoric acid were used. It was impossible to secure sections thinner than 25 microns in any of the specimens not freed from mineral contents by hydrofluoric acid. In the specimens that were treated from one to three weeks in concentrated hydrofluoric acid and thoroughly washed it was possible to section 5 to 10 microns. In all cases the sections prepared from the water or hydrofluoric acid softened specimens were stained with a water solution of safranin for 24 to 36 hours and counterstained with aniline blue for a few seconds. These sections showed entire absence of epidermis, cortex, and phloem. The woody cylinder was unchanged in the xylem, ray, and pith regions. These conditions were the same in both Prosopis and Larrea. Microscopical measurements of the walls in the cells of the ray regions and pith and the xylem were made in order to determine whether there were variations in the tissues of the woods that had been submerged one, two, three, four, and five years. In no case was there any difference in the same region of the same plant. The walls of the ray cells were 1.9 to 2.5 microns thick; those of the wood fiber cells were 5.7 to 6.6 microns, and those of the trachez were 5.5 to 9.5 microns for Prosopis glandulosa; for the same regions of Larrea they were 1.4 to 2.3 microns, 0.9 to 4.8 microns, and 5.5 to 6.5 microns thick, respectively. Sections of the woods softened in water preparatory to cutting were carefully examined for evidence of mineral deposition. In no case could crystals other than calcium oxalate be found. These crystals were present in the same relative places in the sections of the fresh woods and in similar quantities as they were in the dried woods that had been sub- merged. In view of this situation, one was forced to conclude that petrifaction had not been initiated in any of the woods submitted for study. Since the samples of wood that were sent from the Salton Sea represented the final stages of the decomposition processes which had decorticated them it was thought best to secure water from the Salton Sea and reproduce, so far as possible, what had taken place in nature when the samples forwarded underwent decortication. Hight 5-gallon carboys of Salton Sea water were sent to Chicago, four of which were labeled (A), (B), (C), and (D). Pieces of fresh Prosopis glandulosa were placed in (A), of fresh Prosopis pubescens in (B), and of fresh Larrea tridentata in (C); and fresh specimens of all three woods were placed in (D). These cultures were started on December 9, 1911. The tightly stoppered carboys were maintained at a temperature of 22°C. After five days, there was a pro- nounced odor of hydrogen sulphide when the stoppers were removed from the carboys. A milky condition developed in all of the containers supplied with fresh woods. Carboy (C), containing Larrea tridentata, soon showed a remarkable growth of white threads fes- tooned just below the neck of the carboy. These proved to be growths of Beggiatoa, which is semi-anaérobic. The threads were neither on the surface nor at the bottom of the carboy, but in an intermediate position with reference to the oxygen supply (Plate 14m). Later, white films formed on the surface of carboys (A), (B), and (D). A careful examina- tion of these showed that they were composed of filaments of Beggiatoa and free sulphur liberated by the Beggiatoa when they used the hydrogen sulphide for energy releasal in 74 THE SALTON SEA. place of free oxygen. According to Omelianski,! Beggiatoa is dependent upon the action of Spirillum desulphuricans (which reduces the sulphates of mineral waters and liberates the hydrogen sulphide that serves the Beggiatoa for food supply), and the hydrogen sulphide is acted upon by an organism that oxidizes it and forms sulphuric acid. It was found that this acid acted upon iron and formed ferrous sulphide which dropped to the bottom of all of the carboys in considerable quantities as the foregoing processes proceeded in the car- boys of Salton Sea water. Chemical analyses of water from the various carboys were secured through the kindness of Dr. Julius Stieglitz, of the Department of Chemistry of the Uni- versity of Chicago. (Table 25.) Carboys (A) and (B) contained Salton Sea water, while carboys (C), (D), and (E) contained Salton Sea water into which specimens of Prosopis glandulosa, Prosopis pubescens, and Larrea tridentata respectively had been placed six months previously. TaBLE 25.—Analyses of water of Salton Sea, in grams per 100,000 c.c. (C) Prosopis glandulosa. (D) Prosopis| (E) Larrea (A) (B) pub tridentata. No.1 No. 2 Chlorine (Cl)............ 394.735 397.512 400.146 401.100 397.040 393.760 Sulpliuri¢: (BOs «6 v cecas 4 105.233 107.650 75.96 75.39 101.056 105.800 Hydrogen sulphide (H2S)..) ....... | ....eee 1.9979 1.0718 - 6088 . 08616 In the waters which contained an appreciable quantity of hydrogen sulphide there was a decided decrease in the amount of sulphate ion, showing that the SO, had been reduced. From the chemical analyses it is shown that every 100,000 of the parts of the Salton Sea water contain from 105.233 to 107.650 parts of sulphate ion before receiving the different woods. After retaining the wood specimens for six months the amount of sulphuric (SO,) varied from 105.800 in Larrea tridentata to 75.39 parts in the culture of Prosopis glandulosa. It was necessary to determine whether or not these changes might have resulted from the direct action of substances in the water. According to the work of Beijerinck ? this reduction of sulphates may have been accom- plished by micro-organisms. One of the prominent anaérobic forms is Spirillum desul- phuricans. He later*® discovered forms that were less definitely anaérobic, which were able to aid in sulphate reduction. The final products of this bacterial action include sulphureted hydrogen as one of the main constituents. Thus we have an explanation of the large quantities of hydrogen sulphide that appeared in the cultures of Prosopis and Larrea in Salton Sea water. There were similar but less pronounced results in the cul- tures of these same woods in Lake Michigan water, and similarly the same when Robinia pseudacacia was placed in Salton Sea water and Lake Michigan water cultures. The next modification in the chemical change of the output of these first bacterial forms is caused by Beggiatoa, according to Winogradsky.‘ This organism uses the hydrogen sulphide as a source of energy releasal. The hydrogen sulphide is oxidized to sulphuric acid, and free sulphur is first stored in the cell and then liberated in a free condition. The sulphuric acid acts upon the carbonates or some of the bases, as iron, and forms sulphates and ferrous sulphide. This accounts for the large amounts of free sulphur which collected in the bottom of all of the culture carboys of Salton Sea water. It also explains why con- siderable ferrous sulphide collected in the bottom of the vessel in which the greatest amount of sulphuric was decomposed, #.¢., Prosopis glandulosa, where the 105.233 parts per 100,000 of Salton Sea water were reduced to 75.39 parts after the H.S was largely oxidized. 1 Callens Ws a Centralbl. f. Bakt. 2 Abt., vol. vimr, p. 193, 1902; (2) vol. x1, p. 369, 1904; (3) vol. xm p. ; ; , # ~ ’ * Beijerinck: Centrbl. Bakt., vol. 1, 1, p. 1, 1895. 3 Beijerinck: Cent: ‘Winogradsky: Ueber Schwefelbacterien, Bot. Ztg., vol. Tie, é “og Acer pany oreo Be ete Deraits oF Puate 13. Figs. A. B C. D Prosopis glandulosa submerged in the Salton Sea 1906 to 1908. Only the wood cylinder was left in this specimen, the cortex and phloem having been entirely decorticated. . Prosopis glandulosa after continuous submergence for two months in Lake Michigan water. Very little decortication had taken place at the completion of this period. Specimens shown in B after two months more of submergence in Lake Michigan water. Slight decortication is shown in the two middle transverse sections. . Sections of Prosopis glandulosa after two months of submergence in Salton Sea water. The longitudinal sections show far more decortication than was indicated in B, the specimens that had been submerged an equal time in Lake Michigan water. . Four of the specimens shown in D have now advanced greatly in decortication during four months more of submergence in Salton Sea water. All of the phloem and cortex is raised from the ends of the two long sections and entire decortication has occurred in the two small transverse sections. Hydrolysis has advanced far in the cambial regions of the sections which have been submerged constantly for six months in Salton Sea water. . Prosopis glandulosa in transverse section. This shows the cortex (c), phloem (p), and xylem (x). In the phloem (p) the zones of alternating lignified (lig.) and unlignified (unlig.) tissue are shown. Hydrolysis of the unlignified (unlig.) tissue here coincides with the broken-down layers in D and E where the lignified (lig.) tissue is lifted in well-defined layers, after the unlignified (unlig.) tissue was hydrolyzed. . Prosopis pubescens photographed after two months of submergence in Lake Michigan water. Decortication is slight as in B. . Same culture as G after two months of further submergence in Lake Michigan water. The upper transverse section shows considerable decortication after four months of sub- mergence in the Lake Michigan culture. . Prosopis pubescens after two months of submergence in Salton Sea water. Some decortication is shown in one transverse section. . Same culture as shown in I after two months more of submergence in Salton Sea water. Marked decortication is indicated by the four transverse sections. . Prosopis pubescens showing the cortex (c), phloem (p), and xylem (x) in an interesting relation when compared with F. Here closely associated lignified (lig.) zones are separated by very thin unlignified (wnlig.) regions, consequently Prosopis pubescens does not decorti- cate after the manner of Prosopis glandulosa, A comparison of K and F, and compari- sons of B and C with G and H, also a comparison of D and E with I and J are instructive. SALTON SEA PLATE 13 Prosopis glandulosa and Prosopis pubescens in various Stages of Decortication. The Water Cultures in this work were maintained at Room Temperature of 22° C. THE ACTION OF SALTON SEA WATER ON VEGETABLE TISSUES. 75 It is not understood that these analyses represent the chemical conditions of cultures, in which the various woods were placed, during the entire time of the investigation, inasmuch as the chemical changes were taking place constantly due to the bacterial organisms which were reducing the sulphates and the organisms which oxidized the sulphureted hydrogen. It is quite certain, therefore, that chemical analyses taken at other periods would have shown a different relationship existing in the quantities of sulphates contained in the cultures of these various woods. For instance, in the culture of Larrea tridentata the evidence of the evolution of hydrogen sulphide preceded that in the other culture. An analysis gotten at the time the supply of hydrogen sulphide was especially abundant in the Larrea culture would certainly have given somewhat different results from those indi- cated in the above table for Larrea tridentata. Chemical evidence shows that whenever an appreciable quantity of hydrogen sul- phide is present in water whose sulphates have been acted upon by bacterial organisms, the reduction of sulphates has taken place and the decrease in the amount of sulphates is approximately proportional to the increase of the hydrogen sulphide. Since the Beggiatoa lays hold of the hydrogen sulphide for energy releasal and oxi- dizes the sulphide, giving rise to acids, the question arises, might not any break-down of woody tissue present in a culture where this process was going forward be due to the pres- ence of newly formed acids? This question was disposed of by the fact that the quantity of acids is very small and is immediately cared for by bases present in the water. Therefore, interesting as are these chemical changes, due to the reduction action of Spirillum desul- phuricans and the oxidizing action of Beggiatoa, they can not be held to explain the de- corticating changes observed in the wood sections placed in cultures of Salton Sea water. DECORTICATING PROCESSES IN SPECIMENS OF FRESH WOOD IMMERSED IN SALTON SEA WATER. If the substances in the water of the Salton Sea, either before or subsequent to the processes carried on by the organisms above mentioned, were not active in producing decortication, there is only one possible avenue of investigation open in search for the causal factors of cortex removal. Certain enzyme-producing bacteria would in all probability have access to the tissues which were decomposed. Investigations with reference to these forms were instituted in two directions; first, a series of controlled cultures were prepared consisting of three sets: A, a series of sections from Larrea tridentata, Prosopis glandulosa, and Prosopis pubescens were sterilized and placed in cultures of unsterilized Salton Sea water; B, sec- tions of the same woods unsterilized were placed in cultures of sterilized water; C, sections of the same woods were sterilized and placed in cultures of sterilized Salton Sea water. These cultures were kept at room temperature and series A and B soon gave evidences of bacterial activity. The changes in B series were considerably greater than those in A, while series C remained inactive. Throughout the whole test it was noted in the sections from series A and B, after three weeks of culture, that the cortex and phloem loosened and slipped from the woody cylinder with ease. In addition to the controlled cultures, sections of fresh Prosopis glandulosa were made in transverse, longitudinal, and tangential sections and placed in a 5-gallon carboy. Also specimens of branches 15 to 25 c.c. long were immersed in the same gross culture. Some of these branches were paraffined at the end, so that no bacterial organisms could secure a portal of entry to the cortex except through a superficial break in the epidermis. Other specimens were broken in such a way as to leave a ragged and exposed surface at the ends, giving the most ready access to the cortex, phloem, and cambium. Similarly 5-gallon car- boy cultures were made of Prosopis pubescens and Larrea tridentata. These gross cultures kept under observation for six months gave distinct results with reference to decortication 76 THE SALTON SEA. wherever a portal of entry to the tissues of cortex, phloem, and cambium had been pro- vided. On the contrary, the specimens whose ends had been protected by paraffine caps possessed a firm, close cortex and phloem which adhered as tightly to the woody cylinder five months after the cultures were made as when the specimens were first immersed. The various stages of progressing decortication are indicated in Plates 13 and 14. In the latter stages the effects of decortication are pronounced, notably in Larrea tridentata (Plate 14, p, @, 0.) Moreover, it is evident that wherever the branches had sharp and even sections (Plate 13, B, c, G, H, 1, and Plate 14, transverse sections of N, 0, P, Q), the progress of decortication was far less rapid than in those branches which were broken at an angle producing ragged, jagged fractures, which furnished an admirable portal of entry to any bacterial organisms present in the culture. (Plates 13, p and z, and 14, N, 0, P, and Q.) In order to make microscopical investigations of the changes in the immersed speci- mens, sections of the various woods were prepared. Different methods of technique were employed, but the most successful results were secured when specimens of wood were softened for three weeks in hydrofluoric acid, thoroughly washed and imbedded in gelatine, hardened in formaline, and sectioned in a swinging microtome, stained 36 hours with safranine followed by a very brief treatment with aniline blue. Studies of these woods showed that the tissues of the woody cylinder were unchanged, likewise the lignified re- gions in the phloem were unaltered. (Plate 13, rF and kK, and Plate 14,R.) A distinctly different condition was present in unlignified parenchyma zones which alternated with the lignified zone in the phloem; and a notably disintegrated state of affairs was found in the region of the cambium. The walls of the meristem were totally dissolved in places and disintegration following the hydrolysis of these delicate walls had progressed so far that the slightest disturbance caused a breaking away of the whole cambial region from the woody cylinder, marking, of course, a completion of the decorticating process (Plate 13, D and &, and Plate 14, p, a, and 0). The two factors entering into this lack of uniform decortication are the number of layers of cambial cells intervening between the phloem and xylem and the degree of accessibility afforded to the agents which might act upon the walls of the cambial cells. This, in a measure, would explain why some trees, submerged for five years in the Salton Sea, were found to have patches of cortex remaining after they were exposed by the lowering of the water level. The breaking down of the walls in the zone of the cambium and in the meristematic regions of the unlignified zones in the phloem (see Plate 18, p and 8, and Plate 14, p and Q), can not be accounted for merely by reason of their having more delicate structure than that which characterizes the cells in the woody cylinder and in the lignified zones of the phloem. It is true that it would be possible to rend delicate tissues of this character if considerable physical force were applied to external parts of the uninjured specimens. Such an explanation, however, does not fit this case of decortication. There is a difference not only in the strength of the cell walls in these different regions, but there is a difference in the chemical composition of the walls of the cells of the cambium, the woody cylinder, and the lignified regions of the phloem. This was set forth by Mangin' who held that the cellular membrane in young tissues differed from that in adult tissues. He definitely denied that cell walls were composed of pure cellulose. He affirmed that they were always associated with groups of pectins which are essentially distinguishable by color reactions and by certain optical properties, and by great mutability under the action of acids and bases. After conducting extended researches, he concluded that tissues might have their cell walls disassociated by four different processes: (1) prolonged boiling in pure water; (2) prolonged boiling in a 2 to 5 per cent solution of caustic soda or caustic potash; (8) continued action in a cold, weak acid and solvents of pectic acid, alkalines, alkaline salts, ammonia water, and organic acids; and (4) it was possible that certain organisms, which 1Mangin: Jour. de Bot., vol. vir, p. 336, 1893. Detaits oF Puate 14. Figs. L. M. Larrea tridentata submerged 1906 to 1908. Only the woody cylinder was left in the specimen submitted for investigation. Carboy neck showing filaments of Beggiatoa in great abundance. This culture was two wecks old, having developed in the carboy shortly after specimens of fresh Larrea wood had been immersed in the carboy of Salton Sea water. N. Larrea tridentata after two months’ submergence in Lake Michigan water. Considerable decortication is shown in the long sections. . Culture shown in N two months later, when decortication has advanced far in the long sec- tions whose ends had been twisted prior to submergence. . Larrea tridentata two months after being submerged in Salton Sea water. The twisted por- tions of the long specimens show great decortication, while the transverse sections are entire, and there is no breaking away of the cortical zone. . Culture P two months later, when the decorticating process has advanced further. The transverse section has begun to break away in the cambial region after four months of Salton Sea submergence. . Larrea tridentata showing cortex (c), phloem (p), and xylem (x). The unlignified zone (unlig.) here is very wide, and the lignified zones are less numerous than in Prosopis glandulosa or Prosopis pubescens. This is closely associated with the different types of decortica- tion indicated in D, E, I, and J of plate 13, and N, P, O, and Q of plate 14. SALTON SEA PLATE 14 Se ETE Larrea tridentata in different Stages of Decortication, Transverse Section of Culture of Beggiatoa. Fresh Non-immersed Stem, and also a The Water Cultures were maintained at Room Temperature, 22° C THE ACTION OF SALTON SEA WATER ON VEGETABLE TISSUES. 77 nourish themselves upon pectic compounds, could effect a disassociation of tissues. In this group of organisms he included a group of bacteria known as the Amylobacter group. Mangin in this last conclusion was supported by previous workers, van Tieghem! and Trecul.? Louis Gaucher’ held that the cell membranes were degenerated through the action of bacteria bringing about what he termed an abnormal pathological condition. Speaking of the relation of gums and pectin,‘ he states that a member of the Leguminose family, Acacia vereck, breaks down the walls of cells, disengaging all the vascular parts of the plant. In that case the pectin would be hydrolyzed by enzymes produced within the tissues of the growing plants. In other cases he quotes Mangin, who claimed that gum might be made to appear in the cortical parts of plants by injuring their branches with repeated blows upon the bark. In discussing these phenomena, however, these investigators note that the cell walls are composed of a series of complex carbohydrate compounds including pectic acid, pectose, pectin, and cellulose. They note that these substances may be modified in various ways through abnormal and normal processes. But in every case the modification is brought about either through action of a hydrolyzing agent produced entirely by the cells of the living plant itself, sometimes in a normal and sometimes in an abnormal condition, or else by the presence of invading organisms such as the Amylobacter which had gained access to the tissue whose cell walls had not become wholly lignified but were in the early stages of being laid down. Subsequent to the completion of the experimental work with the fresh woods received from the Salton Sea region, it was suggested by Professor Joseph S. Caldwell that the decortication processes were similar to those involved in the retting of flax hemp. This suggestion was supported by an examination of the work of M. 8. Winogradsky,® who carried on numerous experiments and investigations with waters concerned with the mass erosion of vegetable tissues. In these researches he came to the conclusion that he had to deal with a specific bacterial organism which he isolated in his fermentation experiments. This organism was apparently widespread and, according to Winogradsky, might be looked for in any waters where the successful retting of flax was carried forward. In a later work, Professor Dr. J. Behrens investigated the retting of flax and hemp.® Professor Behrens, by means of extensive laboratory experiments, opposed the view of Hauman, namely, that the retting of flax and other vegetable tissues could be carried on successfully by many different species of bacteria. In a table containing a report upon a series of controlled cultures, he itemizes the organisms which would be able to produce so-called retting of flax and hemp. His general conclusions were that this work should be referred to definite and specific organisms. Some of the best work in investigating cellulose fermentation was carried on by W. Omelianski.’ His investigations were concerned with several phases of cellulose fermenta- tion, such as the hydrogen fermentation of cellulose, the methane fermentation of cellulose, and the fermentation of cellulose through the dentrifying bacteria, aérobic bacteria, and mold fungi. Following his suggestions, it was possible to isolate an organism which grew upon sterile filter paper, which was apparently hydrolyzed by its action. In harmony with Omelianski’s findings, this organism, which was isolated from Salton Sea water, required a long time for its incubation upon cellulose. The solution employed was composed of phos- phate of potassium 1 gram, magnesium sulphate 0.5 gram, sulphate of ammonia 1 gram; a very small quantity of sodium chloride in a liter of distilled water used as a solvent. Steril- 4 van Tieghem, Ph., (1) Comptes rend. del’Acad., 1879, vol. 88, p. 205. (2) Also vol. Lxxxrx, pp. 25 and 1102, 1879. (3) Bull. dela Soc. Bot. de France, vol. xxtv, p. 128, 1877; vol. xxv1, p. 25, 1879; also vol. xxvii, p. 243, 1881. 2 Trecul, A., (1) Comptes rend. de l’Acad., vol. Lx1, pp. 156 and 436, 1865; also vol. txv, p. 513, 1867. 3 Gaucher, Louis: Etude générale de la membrane cellulaire, 1904. 4 Gaucher, Louis: Ibid., p. 207, 1904. 5 Winogradsky, M. S., Sur le rouissage du lin et son agent microbien. Comptes Rendu, vol. cxx1, p. 742, 1895. 6 Behrens, J., Ueber die Taurotte von Flachs und Hanf. Parasitenkunde und Infektion krankheiten, vol. x, pp. 524-530, 1903. 7 Omelianski, W.: See note 3. 78 THE SALTON SEA. ized strips of filter paper were then placed in this solution, and flasks containing these prepa- rations were inoculated with 10 to 50 c.c. of water from the different cultures of the Salton Sea woods. These were incubated at room temperature and, of course, accompanied by controls. At the expiration of a week’s time, a slight cloudiness was observed in the inocu- lated cultures. Within a few days small colonies were observed growing on the filter paper. These conformed in essential respects to the group of organisms described by Omelianski as capable of carrying on hydrolysis of cellulose. They were isolated most satisfactorily from the culture of Salton Sea water containing Prosopis glandulosa. There was every evi- dence that successful cultures depended upon using large quantities of the inoculating fluid, though it is probable that small pieces of the infected wood might have been far more effective as an inoculating agent than was the water in which the wood was immersed. In view of the findings of the authorities quoted there seems to be no question that the disintegration of cambial cell walls and consequent removal of the cortical and phloem portions of woody plants submerged in brackish waters is to be attributed to the action of bacterial organisms belonging to the Amylobacter group. From this it would seem that the present problem and the mass erosion carried on by substances present in the water where flax and hemp are retted are related to such economic problems as the breaking down of cellulose or its related compounds when they pass through the digestive tract of animals, or through the septic tank of sewage-disposal plants, and also to the ultimate breaking down of the mantle of humus overspreading the earth.! SUMMARY. 1. Woody plants submerged by the flooding of the Salton Sea were found to be decor- ticated after a period of one year. 2. Microscopical preparations of samples of fresh woods placed in Salton water and kept under control were found breaking down in the zone of meristematic cells, notably in the region of the cambium and in the zones between lignified regions of the phloem. 3. The chemical composition of Salton water could not account for the decortication of woody plants submerged in the sea for a period of one to five years. 4. Sterilized specimens of fresh woods placed in sterilized Salton water did not decorticate during the ten months they were kept under inspection. 5. Bacteriological cultures made it possible to isolate the bacterial organisms, belong- ing to the Amylobacter group, which produce an enzyme capable of hydrolyzing pectins which are considered by some chemists to be the principal substances in young cell walls. 6. A microscopical study of woods which emerged annually from the autumn of 1907 to 1911 did not show breaking down of cell walls in any portion of the wood cylinder. Cell walls of the same tissue in the woody specimens of all the species examined were characterized by the same general thickness. Consequently, it is believed that the action of the Salton Sea water on tissues of woody plants is wholly related to hydrolyzing agents having a bacterial origin ; and, furthermore, evidence is lacking that petrifaction had begun in the tissues of the woody plants submerged for four years in the Salton Sea. It is a pleasure to make grateful acknowledgment of the helpful suggestions received from Prof. John M. Coulter and Dr. W. J. G. Land during the time that these investi- gations were continued. I am especially indebted to Dr. Land for valuable directions in developing technique suitable for making microscopical preparations of the bone-hard specimens of dead woods which were received from the Salton Sea for study. U.8, Dept-of Agneultare wasrocelved, This bulletin ropenee rescle aie oe bn the Bureau of Plant Industry of the ant Soil Mycologist F. M. Scales, while they were investigating ‘The Destruction of Cellulose by Bacteria and Fila- mentous Fungi.” In determining the work of cellulose-dissolving bacteria they made “examinations of sewer slime, of manures, and of the soils of the United States.’ Apparently their studi i i s i i 4 ti Galan Ben PYolleas-ot celluloas hy avslect, pp y their studies are allied to, but in no sense identical with, THE TUFA DEPOSITS OF THE SALTON SINK. By J. CiaupEe Jones. One of the most characteristic deposits left by the Quaternary lakes of the Great Basin are the masses of calcareous tufa found in their former basins. Those of ancient Lakes Bonneville and Lahontan especially have been noted by the geologists who examined them and they have been thoroughly studied by Gilbert! and Russell.2 In accounting for these deposits it was supposed that they were due to chemical precipitation resulting from the concentration of the lake waters by evaporation or through the mechanical expulsion of loosely combined carbon dioxide through wave-action. An examination of specimens of the tufa at present forming in the shrinking Salton Sea disclosed the intimate association of vegetation with the deposit and suggested an alternate hypothesis as to its origin. The present study, made possible through the kind- ness of Dr. D. T. MacDougal and Mr. E. E. Free, has indicated that this hypothesis applies to the older tufa deposited in the same basin by the former Blake Sea, and it is not impossible that much of the tufas of the other Quaternary lakes have a similar origin. THE PRESENT TUFA DEPOSITS. During 1906 and 1907 the rising water of the present Salton Sea killed the shrubs and bushes growing on its former shores. As the water receded the dead vegetation remained standing and in 1910 it was noted that a calcareous incrustation or tufa was forming on the submerged twigs and branches. When this tufa is first taken from the water it is soft, gelatinous, and dark greenish-gray in color. On drying it hardens, becomes much lighter in color, and takes on the characteristic appearance of a tufa. The deposit is a millimeter or so in thickness, with a roughened surface covered with minute warty excrescences. It is not a simple uniform coating, but is made up of a multi- tude of protuberant individual deposits. These vary in size from a pin-head up to 0.5 centimeter in diameter and, broadening in their outward growth, coalesce, forming a fairly uniform coating with interspaces scattered here and there where little or no deposit has formed. A hand-lens disclosed many thread-like green alge originally projecting beyond the mass of the tufa but matted down over its surface with the drying out of the material. By dissolving the tufa with weak acetic acid the alge were isolated. This material was referred to Mr. C. L. Brown, of the University of Nevada, who identified the alge as a species of Calothrix, probably either C. thermalis or parietina. In thin sections of the tufa the carbonates were found to be very finely crystalline and with a characteristic cloudy appearance. While it was possible to recognize individual crystals under the higher powers of a compound microscope, yet the cloudier material could not be resolved into separate grains. Under an oil immersion objective the cloudy appearance disappeared almost entirely and it is possible that it was caused by the reflec- tion of light from the surfaces of submicroscopic crystals. 1G. K. Gilbert, U. 8. Geol. Surv. Monographs, vol. 1, 1890. ? I. C. Russell, U.S. Geol. Surv. Monographs, vol. 11, 1885. 79 80 THE SALTON SBA. The mineral matter was grouped about the algal threads and partially filled the interspaces. As a result, the tufa is very porous and many vacant interspaces remained. Just beyond the main mass of the deposit minute isolated crystals of the carbonates could be noted attached to the free ends of the alge. An analysis of the tufa, by Mr. C. N. Catlin, of the Arizona Experiment Station, shows it to have the composition shown in table 26. Tapue 26. The undetermined substances were stated to con- : s . - . 7 ; SilleBe a cuee ay oa dees pales Bee 3.36 sist largely of organic material which it was impossible | Torna citron ee to separate entirely from the deposit. The analysis | Calcium oxide.................... 40.65 e 7 a Magnesium oxide...............-. 2.23 shows the tufa to consist essentially of caletum carbon- | Water at 100°............... 000. 2.02 Carbon dioxzide................045 33.42 ate with minor amounts of calcium sulphate and magnesium carbonate. The organic matter was evi- dently the included alge. Omitting the undetermined substances and recalculating the analysis to 100 per cent, it is interesting to compare its composition with that of other tufas and a marl deposited in Michigan that has been shown to originate through the activity of alge (table 27). Sulphuric anhydride.............. 2.04 Undetermined by difference........ 14.40 TaBLe 27. No. 1 No. 2. No. 3 No. 4. Remarks. BOR oe scceascacseiete s 5 4.16 5.06 8.22 2.40 No. 1 = Salton tufa, omitting organic matter AU OG 8 vss cosyse 8 cence tet 1 06 | 1.29 1.20 0.89 and recalculated to 100 per cent. B@2O tievtioa aa: vides : tr i No. 2 = Lahontan dendritic tufa, U.S.G.S8. CROE ii soteais Staeinda 47.49 49.14 46.50 52.52 Monographs, vol. 11, p. 203. Me Obs: 2 sts eieice ad 2.61 1.99 3.52 1.41 No. 3= Bonneville tufa, 40th Parallel Sur- Nave ives: chennia. aatelp (Bfouttase di tebdinle On54 I asenaties vey, vol. 1, p. 502. KeOirececacwetin won! seeoets | vavsens O22) |) wseece No. 4 = Mineral incrustation of ‘‘ Chara” FSO wasteciusise de tdi 2 aes 2.36 2.01 1562) off xasowes forming chief constituent of marls, C. A. COis ena 2 ae nee 39.04 40.31 38.33 42.78 Davis, Jour. Geol., vol. 8, p. 492, 1900. BOR ed dneetes wise ea ae 2.38 tr earner | eedepshery Totaliceci ses 100.00 99.80 100.14 100.00 The general similarity in composition of the present Salton tufa to the older tufas of the Quaternary lakes is very striking. Especially noteworthy is the constant small amount of magnesia found in all tufas. The significance of the presence of sulphates in the recent tufa in contrast with its absence in the older deposits can not be stated, but it is not impossible that if the former were exposed to the weather for any length of time the more soluble sulphate would be leached to a large extent. The similarity of analyses No. 1 and No. 4 is still more suggestive. Davis! has demon- strated that the formation of the Michigan marls is in a large measure due to the activi- ties of both Chara and the blue-green alge. His analysis was made on the material leached by hydrochloric acid from the stems of Chara; and its close similarity to the Salton tufa, both in composition and the close association of the alge with the deposits, implies a like origin. Aside from the submerged stems and branches of the dead shrubs about the shores of the lake the only other locality where calcareous deposits are forming at the present time is on the submerged volcanic hills at the southern end of the lake. The writer was unable to visit them, but was informed by Mr. Free that a tufa was there forming on the rocks similar to that found on the shrubs. As far as is known at present, there is no general deposition of calcareous material other than the tufa and no cemented sands or oolites have been discovered on the present beaches. 1C. A. Davis, Jour. Geol., vol. vit, pp. 485-497, 1900 ; vol. rx, pp. 497-506, 1901. THE TUFA DEPOSITS OF THE SALTON SINK. 81 THE TUFAS OF BLAKE SEA. The ancient lake, Blake Sea, that formerly filled the Salton Basin, left among the other records of its existence an extensive tufa deposit. As is the case with the present tufa, it is limited in its distribution to the localities where the waters bathed resistant material. The older tufa is found only along the northwestern shore, where the granites of the Santa Rosa Mountains were washed by the sea. The principal deposit is displayed at Travertine Point, although tufa is found on the cliffs and on boulders of granite scat- tered over the desert sands for several miles to the south. At the time of the Blake Sea, Travertine Point was a low-lying serrate spur or group of granitic hills extending in a semicircle a half mile to the northeast from the mountains. These were submerged by the rising waters until only a few feet remained above the surface of the sea. During the time that the water stood at its highest level, bars of granitic sand were built out from the shore until as the waters began to subside the hills had been connected with each other and the shore as typical land-tied islands. With the retreat of the sea the slopes of the bars were slightly notched by the waves, leaving many distinct recessional terraces. Although these were cut but a short distance in the yielding sand, there has not been sufficient wind erosion since to efface them and they are practically as the sea left them. At the outermost of these land-tied islands, the original granite hill rose abruptly some 175 feet from the desert sands at its base. Its sheer cliffs were concealed nearly to the top by a steep talus composed of huge boulders up to 10 feet in diameter that had broken and fallen from the cliff. In the cave-like recesses left between the boulders one may penetrate at times 25 or more feet from the surface of the cliff. The tufa covers the boulders on all surfaces where they are not in contact. It is dendritic in character and at first sight seems to be a vegetative growth fixed through petrifaction on the surfaces of the rocks. The illusion is heightened by the tendency of the projecting columns and spurs to turn upward and outward towards the light. On the outer surfaces of the boulders the tufa is uniformly about 30 inches thick over a vertical range of 120 feet. In places the growing tufa had evidently been broken from the rocks by the waves and the vacant space healed over by a fresh growth, leaving abrupt local changes in the thickness of the deposit. These were entirely local and meas- ured by a few feet, with no relation either to each other or to vertical distribution. At the base of the cliff the tufa rather abruptly diminished in thickness, probably due to the recession of the water. At the top a mere film near the high-water mark indi- cated the first deposition of tufa. Within 4 feet of the high-water mark the tufa was well developed and several inches in thickness. Around the sides and back into the recesses it diminishes in thickness, until in the darker places it is but a mere film and has lost its dendritic character. Here it is a solid, thinly banded coating similar to the lithoid tufa of the Lahontan Basin. The gradation between the lithoid and dendritic types is gradual, the projecting columns and plates of the dendritic tufa decreasing in size and becoming less prominent with the decrease in the thickness of the deposit until the uniform lithoid coating results. The dendritic tufa is occasionally banded, due to a greater coalescence of the indi- vidual columns forming the tufa, but there seemed to be no definite relation either in the distance between the bands or in their number at different parts of the deposit. Con- siderable sand is intermingled with the tufa, in part blown in by the winds since the retreat of the water, and in part included in the tufa and cemented by it. Many snail shells are inclosed by the dendritic tufa, showing that they lived at the time of the formation of the tufa. The lithoid tufa was much purer, containing little if any sand and none of the shells. On breaking the tufa it is seen to be built up of rudely circular columns or thin plates approximately 3 mm. in thickness. The interior of these is composed of a finely crystalline 6 82 THE SALTON SEA. and porous aggregate. Coating this are usually several concentric layers somewhat similar to the lithoid type. While the columns and plates retain their individuality for the most part, yet they coalesce and anastomose frequently, thus binding the mass into a firm and resistant whole. Where the tufa is attached to the sides or beneath the boulders the plates and columns are usually more distinctly separated. The projecting shafts and blades bend outward and upward, as if seeking the stronger light. This tendency is the most striking feature of the deposit and the abundant recesses in the cliff gave an exceptional opportunity for thorough observation of this relation. It was found without exception that the tufa turned towards the entrance of the caves in the direction from which the stronger light was com- ing. As the caves penetrated the cliff in all directions this relation of the direction of the growth of the tufa towards the stronger light was tested under all conditions. As the thickness of the tufa developed also depends on the amount of illumination received the influence of light is one of the factors that must be considered in accounting for the origin of the deposit. Thin sections of the tufa show it to be composed of minute crystals of calcium car- bonate and masses of cloudy material similar to that composing the recent tufa. The thin banding noted in the lithoid type and around the individual plates and columns of the tufa is due to the alternation of denser material with the more porous. Occasional spots of yellow iron stain are found wherever sand grains are included and are a result of the oxidation of the ferromagnesian minerals of the sand. Frequent thread-like lines of slightly darker color, the same size and form of the alge, found in the present tufa and radiating at right angles to the banding, suggest algal filaments. Rarely, in favorable locations, the original algee may be seen preserved in their original position and faint green color. ORIGIN OF THE TUFAS. The source of the calcium carbonate which forms the greater part of the tufas is found in the waters of the lakes. This is self-evident, since the tufa forms wherever a firm point of attachment is offered, regardless of the lime content of its support. The second column of table 25 shows that the waters of the present Salton Sea contained the stated amounts of calcium as shown by the yearly analyses. From the perennial increase in the amount of TaBLe 28, Parts per Per cent Per cent CaCO; |Ca( Year. 100,000 | calcium in a one COs in ee ; ae calcium. | total solids. 2 total solids. | 100,000 100,000 1907.... 9.95 2.4 6.58 1.8 10.97 8.88 1908 ....| 11.87 2.5 7.66 Led, 12.77 10.35 F089 cil! E276 2.4 7.34 1,4 12,23 9.92 1910....| 13.67 2.2 6.38 1.05 10.63 8.62 Tiles.) 15,82 2.2 5.78 0.8 9.63 7.81 calcium, it is evident that the lake waters have not yet reached the point of saturation for the lime. In the third column the percentages of the total solids contained in the water are figured and the partial deposit of the lime is shown. In the fourth and fifth columns, the more significant figures of the carbonate radicle show that it has been decreasing since 1908. Assuming for the sake of argument that all of the carbonate radicle was combined with the calcium either as the carbonate or bicarbonate, we find the amounts given in columns 6 and 7 would have been present at the times that the samples were taken. Of the two compounds the bicarbonate is by far the more soluble and its solubility is largely increased by the presence of carbon dioxide in the water. If either compound THE TUFA DEPOSITS OF THE SALTON SINK. 83 is present in the water it is probable that it is the bicarbonate. The solubility of the calcium carbonates is shown by table 26, taken from Seidell.? It is evident that all of the carbonate radicle could not be combined with the lime in the form of the carbonate, for the water would be supersaturated and over 90 per cent of the carbonate would be deposited. It is possible, however, for the bicarbonate to be present, as the water would then contain only a third of the amount possible, even if no carbon dioxide were present. If by any means the bicarbonate should be decomposed to the carbonate a deposit would necessarily result. Owing to the comparative dilute solution of salts in the lake water it is not possible for mass action to cause {Calcium carbonate 20°C. 1.2 parts per 100,000. a deposit and other than simple chemical methods must be in operation. Calcium bicarbonate 15° C. Of the various methods by which the decomposition | °° ©? Per 100 c.c. |Ca(HCO:)2 per 100,000 of the bicarbonate may be accomplished, only two (me- TaBLe 29. chanical agitation and vegetation) need be considered. ee aes Since the tufa forms beneath the surface of the water it et 82.1 - .25 59.5 can not be a result from the evaporation of spray. 0.08 40.2 As was pointed out by Gilbert, in his study of Lake en sed Bonneville,? the thickest deposits of tufa are found where the waves were most active. This is the case with the older tufa of this basin. The deposits are best developed where the waters dashed against the cliffs of Travertine Point and the Santa Rosa Mountains. Nevertheless, the tufa was well developed on the small boulders lying on the gently sloping shores to the south. The essential factor in both the present and older tufas seems to be a firm support. Wherever the Blake Sea washed the unconsolidated sands or easily disintegrated Tertiary sediments the tufa apparently did not develop. This is strikingly brought out in the deposits at Travertine Point, where the sands of the bars connecting the heavily coated cliffs show no evidence of cementation or tufa deposits. The present beach sands similarly show no evidence of the deposition of calcium carbonate, although the tufa is forming on the submerged branches and twigs of the dead shrubs but a few feet away. If the formation of the tufa is due to the agitation of the waves alone it would seem strange that it should be so restricted in distribution. If, on the other hand, the tufa is formed primarily through the activities of the alge its better development where the waves are active would naturally follow, for it is well known that the sedentary aquatic life thrives best where the water is in active circulation. The alge associated with the tufas are known to produce calcareous deposits else- where.? The evident relation of the thickness of the deposit of the older tufa and exposure to light, its growth towards the light, its local development, its uniform thickness over a large vertical range, the abundance of snails implying a food supply, the presence of the alge, the porous, dendritic character of the deposit—all point to its vegetative origin. The details of the process by which the alge decompose the bicarbonates and force the deposition of the carbonates are not well understood and remain to be worked out. When this is done, we can say why and under what conditions tufa deposits are formed. 1 A, Seidell, Solubilities of inorganic and organic substances, pp. 86-88. 2G. K. Gilbert, op. cit., p. 168. 3 W. H. Weed, Formation of travertine and siliceous sinter by the vegetation of hot springs. U.S. G.S., 9th Ann. Rept., pp. 619-682, 1889. See also papers cited. J. Tilden, Minnesota Algze, vol. 11, pp. 268-271, 1910. C. A. Davis, Jour. Geol., vol. vitr, pp. 495-497, 1900. PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. By S. B. Parisu. TERRITORIAL LIMITATIONS. The area the vegetation of which is here discussed lies between the margin of the pre- historic Lake Cahuilla and that of the present Salton Sea. The boundaries of this ancient body of water are still indicated in many places by easily recognized ‘‘old beaches,’’ the gravelly ridges formed by the action of its waves; or, where rocky promontories abutted upon the water, by the heavy coating of travertine deposited upon the subaqueous por- tions. Everywhere the soil is full of bleached shells of the small mollusks which inhab- ited this long-vanished lake. Time and again the lake bed has been partially refilled by the irruption of the Colorado River, to again disappear by the slow evaporation of its waters. The most recent replenishment occurred in 1905, through human negligence, and again this present lake is contracting and dwindling to extinction. To the bed of Lake Cahuilla the name Salton Sink has been given. Its upper margin is about 20 feet above present mean tide level, and its lowest point, now beneath the water of the present lake, is 286 feet below. In the central depression lies Salton Sea. The whole catchment area whose drainage is towards the sea is known as the Cahuilla Basin. The Sink is oblong in shape, its longer axis lying southeast and northwest, a distance of about 80 miles, in an air-line from Indio to Calexico. Its greatest width is measured by a line running slightly south of west through Imperial Junction, and is about 30 miles. On either hand rise arid and sun-scorched mountains of low altitude. Desolate and barren as these appear, and as they in reality are, their slopes and waterless carfions sustain a scanty but interesting flora, the consideration of which, however inviting, is beyond the present purpose. The southern boundary is formed by the outward slope of the great Delta which has been deposited by the turbid waters of the Colorado River, whose uncer- tain channel follows the ridge it is continually building. BOTANICAL HISTORY OF THE SINK. Within the last dozen years an irrigation system has brought the water of the Colorado River to a large area of the southeastern end of the Sink, now known as Imperial Valley, and as a result about 275,000 acres of previously arid desert have been brought into culti- vation, railways have been built, and busy towns have sprung up. At the northwestern end a considerable but less extensive reclamation has been made by the development of artesian water. These agricultural operations have entirely changed the character of a large part of the Sink and have complicated the study of its flora. Especially is this the case in Imperial Valley, where the peculiar conditions present problems at once interesting and difficult. It is to be regretted, therefore, that the references in botanical literature to the conditions formerly existing in the Sink are both scanty and unsatisfactory. The papers mentioned below comprise all, so far as I have been able to ascertain, which relate to the subject. The survey of the boundary between the United States and Mexico, made in 1856, passed across the southwestern border of the Sink, and the first volume of Emory’s Report of the Survey contains a botanical paper by Dr. C. C. Parry, the botanist of the party, 85 86 THE SALTON SEA. and a longer article, from the same pen, forms the introduction to the fifth volume of the report. In both papers a few paragraphs are devoted to the vegetation about New River. In 1853 Dr. W. P. Blake made a survey for the proposed Pacific Railroad, which passed through the western border of the Sink. Over much of this route he was the first explorer, but his account, while interesting for the geologist, contains little for the botanist. There are some references to plants of the basin in Torrey’s Appendix. Blake’s report forms a part of volume 5 of the series of Pacific Railroad Reports. Dr. E. L. Greene made a pedestrian journey in February 1877 from San Diego to Fort Yuma, along the old stage route, passing through the New River country. He gave an account of his journey in the American Naturalist, vol. x1v, pp. 787-793, 1880, and vol. xv, pp. 24-32, 1881, under the title “‘Botanizing in the Colorado Desert.”’ A report on the ‘‘Lands of the Colorado Desert in the Salton Basin” was published in 1902, as Bulletin 140 of the Agricultural Experiment Station of the University of Cali- fornia. It is devoted to an investigation of the agricultural capabilities of the soil, but contains an annotated list, contributed by J. Burtt Davy, of 22 species of plants collected in Imperial Valley by F. J. Snow, who had made the examination of the soils. A brief but valuable account of the botanical aspect of Salton Sink, shortly before its recent inundation, is given by Coville and MacDougal in Carnegie Institution Publi- cation No. 6, pp. 20-22, pl. 23-26, 1903. The following recent papers by MacDougal contain much information bearing upon present conditions in the Salton Sink and the Colorado Delta: The Delta of the Rio Colorado. Bull. Amer. Geograph. Society, vol. xxxviu1, pp. 1-16; map and figures. 1906. Vegetation of the Salton Basin. Fifth Year Book Carn. Instit. of Wash., pp. 119, 170, pl. 10,11. 1907. Movements of Vegetation in the Salton Basin. Eighth Year Book Carn. Instit. of Wash., pp. 57, 58. 1908. Surface Geology and Vegetation of Salton Basin. Tenth Year Book Carn. Instit. of Wash., pp. 49, 50. 1910. The Desert Basins of the Colorado Delta. Bull. Amer. Geograph. Soc., vol. xxxrx, pp. 1-25, figs. 1-11, map. 1907. ZONAL POSITION OF THE FLORA OF THE SINK. Before entering upon a discussion of the flora of the Sink, it is desirable to indicate the position which it occupies in relation to the general scheme of plant distribution in the Colorado Desert. This will be best understood by a short account of the several floral zones traversed in passing from the northwestern rim of the Colorado Desert at Banning to the dunes which border the sea-level line near Indio, and then considering the zones which intervene between the southeastern margin of the Sink and the Colorado River. At Banning, altitude 2,300 feet above sea-level, begins a tension zone in which the plants of the desert contend for the possession of the soil with those which have crossed over from the coastal slope of the San Gorgonio Pass. Rapidly the species of the latter contingent are forced to drop out, and are replaced by successive reinforcements from the former. The ground is fully occupied, and while woody and suffruticose plants are most prominent, there are many perennial and annual herbs, including grasses. This tension zone is a narrow one, and already at Cabezon, only 5 miles from Ban- ning, but 540 feet lower in altitude, it begins to be succeeded by an Opuntia-Yucca zone. The dominant plants are Yucca mohavensis, an inornate tree, 10 to 20 feet high, clothed with dagger-like leaves, and Opuntia echinocarpa, here a robust shrub of upright growth, its numerous branches forming a compact head. Subordinate to these are many low, xerophytic shrubs, in the shelter of which, and in the intervals, herbaceous vegetation finds a place. The soil is less closely occupied than in the tension zone, and only here and there is to be found an intruder from the west. The aspect is distinctly desert-like. First the Opuntia thins out, and then the Yucca, and the creosote bush (Larrea, tri- dentata) becomes increasingly abundant, until at Whitewater, a further descent of 650 feet in a distance of 10 miles, it is the dominant plant of a well-defined Larrea zone. In- dividuals of this species are to be found scattered over most parts of the desert beyond, PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 87 but are seldom prominent, and then only over limited areas. Here they are both numerous and well grown. There is an increasing amount of bare sand and gravel, but as the moun- tains are near enough to draw from the clouds a certain amount of annual rainfall, these intervals are often clothed in the spring with bright-flowered annuals. By another 450 feet of descent in a distance of 5 miles, Palm Springs railway station is reached, and the Larrea zone has already been left behind. Here is a scene of desolation indeed. A vast, apparently level, plain of gray, wind-swept sand stretches before the eye. Picked up by the wind from the detrital fan at the pass, the sand is carried along with a force that eats into wood and etches stone. Larrea remains the most abundant shrub, but the individuals are separated by great distances of bare sand, and of many the form has undergone achange. They lie prostrate on the sand, a long, narrow line of green, point- ing away from the direction of the wind. By its varying action a hillock of sand may be heaped about them, or much of their roots may be exposed. Here and there a sparse fringe of Panicum urvilleanum serves to partially fix the sand, and where there is a little shelter the surface may be thinly dotted with small tufts of Chamesyce polycarpa, an inch or two high. This sand desert continues for nearly 20 miles, until it becomes, along the margin of the Sink, a series of dunes, each formed by an overwhelmed mesquite tree, whose projecting ultimate twigs crown it with verdure.! Between the southeastern margin of the desert and the Colorado River the zonaliza- tion is less extensive, since the dividing ridge which is here passed over has an altitude of only about 400 feet. The zones are, however, well marked. The old sea-beach is bordered by mounds heaped about living mesquites, and smaller ones in which are buried larreas, which do not long survive. Beyond is a stretch of drift- ing hillocks, totally bare of vegetation. These are not ‘‘sand hills,” as they are usually called, but are dunes in form and in every respect, except that they are formed of fine particles transported by the wind from the alluvial soil of Imperial Valley. Where a little basin has collected the water of a thunderstorm, almost always are to be found mats of Pectis pappose, perhaps Chamesyce setiloba, Lepidium lasiocarpum, Baileya pauciradiata, or a few other annuals. Beyond these mounds the imperceptible dividing summit is occupied by an Opuntia zone. To this succeeds a pebble-strewn plain of so slight an inclination as to appear level, scattered over with the tall, slender stems of the ocatilla (Fouguteria splendens), and con- tinuing quite to the verge of the broad bottom-lands of the Colorado River, which are grown up to mesophytic thickets of cottonwoods and willows. The general zonal disposition of the flora of the Colorado Desert, on a line drawn from the summit of San Gorgonio Pass through Salton Sea, to the Colorado River, is shown diagramatically below: Diagram of the floral zones of the Colorado Desert. Tension Zone, Cottonwood-willow Zone, Yucca-Opuntia Zone, Fouquieria Zone, Larrea Zone, Opuntia Zone, Sand-drift Zone, Alluvial-drift Zone, Mesquite-dune Zone, Mesquite-mound Zone, Atriplex Zone of Salton Sink. FACTORS DETERMINING THE DISTRIBUTION OF THE SINK FLORA. The causes which determine the character and the distribution of vegetation are general in their nature, and apply alike to the parched desert and the humid valley, but the resultant problems which they present are simple or complex in proportion to the modifying influences which are involved. 1 The distances here given are along the Southern Pacific Railway; by an air-line they would be shorter. 88 THE SALTON SEA. The Salton Sink is practically a level plane, everywhere exposed to an equal insola- tion, no irregularities of surface serving to vary the effects of sun and wind; with a unl- form temperature and a uniform deficiency of atmospheric humidity and of rainfall. With these variations of environment eliminated or negligible, the question of distribution is left to depend directly upon the chemical and mechanical character of the soil and its water-content. WATER SOURCES OF THE SINK. There are four sources from which the Sink derives its supply of water. (1) The precipitation upon its immediate area; this is scanty and irregular, and its effect is almost negligible. (2) The run-off from the surrounding mountains; these are subject to torrential downpours, which running at once down their steep and naked acclivities are gathered by the branching cafions and poured out into the so-called ‘‘washes.” These are shallow channels, broadening as they proceed, and often permitting the water to “flood out” over wide areas of mesa. In a few hours the torrents cease and the washes are again empty. (3) A third water supply is derived from the San Gorgonio and San Jacinto Moun- tains, some 50 miles northwest of the Sink. These are respectively 11,600 feet and 10,800 feet high, and enjoy a precipitation considerable for the region. Their eastern drainage is carried into the great cone of stones, gravel, and sand which fills the San Gorgonio Pass, into which it promptly sinks. Passing down by percolation, it reaches the northern rim of the Sink, where it is brought so near the surface that, raised by capillary action, it produces the extensive area of alkaline flats between Indio and Mecca. The percolating water itself is remarkably pure, as is demonstrated by the artesian wells which have been sunk into it, but in passing upward it becomes heavily impregnated with mineral matter. The same percolation waters are the sources of the springs which occur along the upper end of the Sink, and they supported the saline marsh which formerly occupied the lowest part of the depression, now drowned beneath the Salton Sea. (4) The fourth water source of the Sink is supplied by the Alamo! and the New Rivers, two diffluents of the Colorado, forming a part of the intricate distributory system of chan- nels, sloughs, and lagoons, which, in its Delta, relieves the floods of that stream. By the irruption of the Colorado, in 1906, the character of these diffluents has been greatly changed. They now carry considerable volumes of water throughout the year, and run for most of the way through channels cut 20 to 50 feet deep in the alluvial soil, whose perpendicular banks inclose a strip of moist bottom-lands which becomes wider as the rivers approach their point of discharge in Salton Sea. Previous to the settlement of Imperial Valley, and the consequent changes which have taken place in the character of the diffluents, these appear to have consisted of a series of lagoons, or shallow lakes, connected by much less deeply eroded streams. The names of many of these lakes still appear upon the maps, but they no longer exist; they have been drained and their beds are in cultivation. Annually, or at least in most years, these streams and lakes were refilled by the spring floods of the Colorado River, and when these ceased they contracted as their waters evaporated. Some became dry, but others were permanent. At all events, the New River country was regarded by desert travelers as affording reliable watering-places at all seasons of the year. About these natural reservoirs, and in the land moistened by them, doubtless grew most of the paludose and halophytic vegetation which now lines the rivers and irrigation canals of Imperial Valley. Parry speaks of camping by one of these lagoons in September, where the party found sufficient grama grass for the mules, and where he noticed an abundant vegetation. The Alamo appears on most old maps as “Salton River,” and on at least one as “East River.” PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 89 SOILS OF THE SALTON SINK. The soils of the Sink may be classified under three general types. At the upper or northern end the soil is a very fine sandy loam, deposited under water at the time that the Sink was still an arm of the sea, as well as during the lacustrine period. It has a high capillarity, so that the artesian water is able to seep up towards the surface, carrying with it alkaline salts, which are concentrated on or near the surface. The water-table is high, and the alkaline content is so great as to confine the vegetation largely to plants of specialized type. Along the northeastern slope of the Sink, extending from Mecca to Frinks, the sur- face is covered with wash from the adjacent mountains, consisting of sand, gravel, and stones, assorted in increasing coarseness as the upper contour is approached, where it is cut up by the broad shallow channels through which the flood waters are carried. This soil is readily permeable by water, so that floods and rainfall sink deep, and are thus protected from evaporation. The same porosity renders it easy for the roots of the plants to penetrate and avail themselves of the stored moisture. It is in this wash-soil that the greatest variety and abundance of xerophytic vegetation is found. The Imperial soils, which occupy the lower or southern part of the Sink, were deposited by the Colorado River. They are loams of very fine compact grain, varying in the amount of clay which they contain, but with very small percentages of sand. They are permeable by water only to a slight degree. Consequently the water which they may receive readily runs off, and only an inch or two of the surface is wetted, and this is rapidly dried again. The mounds also are composed of similar loam, but of lighter particles which have been selected by the winds and are consequently of a more permeable texture. All the Im- perial loams are rich in materials for plant growth, needing only water to be abundantly productive. DERIVATION OF THE SINK FLORA. A student of the flora of the Colorado Desert speedily discovers evidences of a migra- tory movement from the south and east. Some plants from Arizona reach the Colorado River, but fail to cross it, or, like the saguaro, barely gain a foothold on the California side. Fouquteria splendens abounds on the desolate plains from Fort Yuma to Acolyta on the Southern Pacific Railway, and then falls back to the mountains, along which it is able to persist as far north as to Red Cafion, near Mecca. The palo verde (Cercidium torreyanum) pushes on farther and reaches Palm Springs. A few, like the desert willow (Chilopsis linearis) and the mesquite (Prosopis glandulosa), even straggle over San Gorgonio Pass into the cis- montane region of southern California. Instances of like character might easily be multi- plied; indeed the one prominent fact in the distribution of the xerophytic flora of the desert is the successive dropping out of species as one proceeds from the southeast to the north- west. The Colorado Desert is the western fringe of the arid flora of Sonora, Arizona, New Mexico, and regions still to the east and south, the so-called Lower Sonoran flora. Of this general flora of the Colorado Desert the xerophytic vegetation of the Sink is a part, differentiated mainly by the great preponderance of Atriplices in its composition, so that it may be fittingly denominated the Atriplex zone. As it is only in very recent times, as geological periods are reckoned, that it has ceased to be submerged, it must have directly received its population from the more elevated land bordering it. It has, indeed, again and again been freed of vegetation over areas of greater or less extent, by the repeated partial refilling of the depression, to be again repopulated as the water receded. The plants which border the diffluent rivers of Imperial Valley were brought in from the delta of the Colorado. Most of them may be found in the river bottoms at Fort Yuma. One most interesting member of this association has descended the Colorado from its distant headwaters in Wyoming and Utah. 90 THE SALTON SEA. These two sources account for the greater part of the plant population of the Sink. Some individuals belonging to the cafion-flora of the bordering mountains are sometimes carried down by the flood waters, mostly as seeds, into the Sink, but they do not increase or probably long endure. Travertine Rock, of which some account will be given farther on, introduces a few montane plants. Two small “Cismontane islands’ : are formed by small groups of Baccharis viminea at Dos Palmas and of Hriodictyon californicum at Indio. The endemic flora of the Colorado Desert is not extensive, and but few of its species are found in the Sink. Of these Washingtonia filifera is the most important." DISTRIBUTION OF THE SINK FLORA. The distribution of the flora of the Sink is determined in dependence upon the edaphic conditions outlined on a preceding page, and may be conveniently considered in its relations to them. The term “formation” is applied to the wider groupings, and the term “‘associ- ation” to the minor societies of which the formations are composed. HYDROPHYTIC FORMATION. The hydrophytes are very feebly represented in the flora of the Sink. The water of the rivers and canals is so heavily loaded with silt as to prevent the growth of these plants and the amount of clear water is very limited. The springs, and the pools about them, are nearly free from alge, such as commonly mantle the surface or float below it in similar situ- ations elsewhere. Nor does one find the beds of dried pools coated with the felted filaments of alge which had grown in the water. The few species which do occur are mostly to be found in artificial pools and streams, such as are formed by the drip or waste of artesian water, and even there they are not abundant. Occasionally a few filaments may be seen on mud banks. No collections were made to determine the presence of desmids or diatoms. In only a single instance was a submerged spermatophyte observed. This was in Salt Creek, between Seely and Dixieland, in the extreme southern part of the Sink. Salt Creek is a shallow stream, some 50 feet wide, flowing between high and perpendicular bluffs of alluvial soil, from the bases of which the salty water oozes by seepage. Its bed is completely filled with a thick growth of Ruppia maritima. HELIOPHYTIC FORMATION. The heliophytic formation of the Sink is far richer in species and greater in extent than the hydrophytic. It may be considered under two associations, that of the springs at the upper or northern end of the Sink and that of the rivers and canals at the lower or southern end. SPRINGS ASSOCIATIONS. The vegetation growing in the springs, of which there are several on both the east- ern and the western sides of the upper end of the Sink, is composed of two species of very extended distribution, Typha latifolia and Scirpus olneyi. In none of them was there apparent evidence of the presence of plankton or of any subsidiary associations. They consist of pools, having little or no run-off, very thickly grown up with the two plants mentioned. Short descriptions of three of these springs, each of a different type, will suffice to give an idea of all. MORTMERE. Mortmere is near the northeastern end of the Salton Sea, at an altitude of about 200 feet below mean sea-level. Its position is marked by its clump of green vegetation, which 1 The following species are endemic in the Sink, so far as present knowledge of their distribution indicates, but eed een Pace wre el Be epee to ae the range of at least some of them: Astragalus aridus . ymatus Sheldon, Atriplex saltonensis Parish, Cham lt is Millsp. . Spheralcea orcuttii Vasey & Rose, Phellorina macrosperma Ubyd, ck pn bared PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 91 contrasts vividly with the gray and parched aspect of the gravelly desert plain in which it is situated, and above which it is elevated on a low mound of black earth, formed by the decay of its vegetation and the accumulation of the dust carried by the wind and detained by the moisture of the spring. The water stands in a pool in the summit of this elevation, but there is no run-off. It is strongly impregnated with sulphureted hydro- gen and is also slightly saline. The entire pool is filled with a dense growth of robust Scirpus olneyt. Surrounding this, the damp soil of the mound is covered with a sod of Distichlis spicata, and bordering this is a sparse fringe of the shrubs Spirostachys occi- dentalis and Pluchea sericea. DOS PALMAS. This spring is situated also in the northeastern part of the Sink and only 2 feet below its upper contour line. It is at the mouth of a wide wash which drains a large part of the Chuckawalla Mountains, from which source its waters are derived. Below it is a broad playa, flooded occasionally by the torrential storm waters which come down the wash. It consists of a swampy pool of considerable area, retained by a slightly elevated border of black earth, which is strongly impregnated with alkali. The water is tepid and very slightly saline, and only sufficient in volume to maintain the pool without needing an outlet. It is filled with robust Scirpus olneyi, for the most part so dense as to exclude other plants, but in places intermixed with Typha latifolia and a little Juncus coopert. In the middle of the scirpetum grows a fine specimen of Salix nigra and the two lofty Washingtonias which give the spring its name. On its margin Pluchea camphorata is mingled with the sedge, and in the moist saline soil which borders the swampy pool Anem- opsis californica forms a narrow intermittent belt, in which grow scattered shrubs of Baccharis viminea. Beyond the immediate border, but still in soil somewhat damp, are considerable areas of Distichlis, with a scattered growth of Pluchea sericea, Prosopis pubes- cens, and large domes of Atriplex lentiformis. AGUA DULCE AND FIGTREE JOHN SPRINGS. There are several of these springs, alike in character, scattered within a distance of 2 or 3 miles on the southwest side of the Salton Sea, which at the height of its recent re- plenishment reached most of them. Its shore is now fully half a mile distant. Rising, as these springs do, from an artesian source, the water, while tepid, is pure and sweet. Their character is similar to that of Dos Palmas—swampy pools grown up with Scirpus olneyt and Typha latifolia. Some afford sufficient water to irrigate small gardens, and there are large tracts of damp soil covered with a sod of Dvzstichlis, in which grow clumps of Juncus coopert. About them are well-grown trees of black willow, delta cottonwood, and Washington palm, all of which have probably been planted by the Indians, who have made their homes here for generations. SCIRPUS-TYPHA ASSOCIATION OF ALAMO AND NEW RIVERS. The margins of the two diffluents of the Colorado which flow through Imperial Valley are bordered by a close growth of Typha latifolia and Scirpus paludosus, which pushes into the water until its depth exceeds 6 or 8 inches, and creeps up on the mire of the low muddy banks. The width of this zone, therefore, varies in accordance with the character of the stream. Where, in its meanderings, it crowds closely to the bluffs there is room for but a narrow belt, but on the opposite side it may extend, in the shallow water and over the mud flats, as much as 200 to 300 feet, or even more. The two species do not inter- mingle, but are mutually exclusive, nor is their respective occupancy determined by hydrophytic conditions. Hither may occupy the entire border of the stream, or either may have possession of the water front, leaving the landward margin to the other; but always the gregarious colonies are of considerable extent and exclude all other vegetation. 92 THE SALTON SEA. It is probable that the species which first obtains possession of an unoccupied mud bank so fills it with its rhizomes as to keep out the other. The character of Typha is too well understood to require any comment, but Scirpus paludosus is a little-known species and until now its presence was not even suspected beyond the marshes of Wyoming (where it was first discovered) and of adjacent Utah. It must have descended the Colorado from its distant tributaries, and its colonies will doubtless be discovered along the line of its descent when the flora of these rivers becomes better known. It is plentiful along the Colorado River in the neighborhood of Fort Yuma and, like its companion Typha, has entered the Sink from the delta of that river. Scirpus paludosus is a vigorous perennial sedge, 1 to 214 feet high, and produces an abundance of seed. Each plant throws out one or more rhizomes about 3 mm. in diam- eter, at the ends of which ovoid tubers are formed which develop new plants. When young these tubers are often an inch in diameter, free from fiber, and edible. As the new plant matures the starch is absorbed and there remains a woody enlargement with a cluster of fibrous roots. The young plants soon become independent by the death of the connect- ing rootstocks and in turn begin to throw out innovating rhizomes of their own. It will be readily understood how well such a plant is suited to fluviatile transporta- tion. Not only are the abundant seeds discharged upon the water, but as shifting currents disintegrate a bordering marsh they set free a quantity of young plants, each provided with a store of food-stuffs at its base and ready at once to strike vigorous root as the falling floods deposit it upon some newly formed mud-flat. The same tubers would also serve to carry plants over the seasons of drought to which aquatic vegetation is subject in this region of uncertain water supply. In the irrigation canals of Imperial Valley Scirpus paludosus is universally present and a source of great annoyance. Its rhizomes fill the mud banks (from which it is prac- tically impossible to remove them), and spread out into the water, checking its current. Typha also is frequent along these artificial streams. Indeed, throughout the Sink, wherever an artificial pool or reservoir is made, Typha promptly appears and takes pos- session. Cyperus erythrorhizos, an annual sedge, attaining in its best development a height of over 2 feet, and bearing an umbel more than a foot in diameter, but often much reduced in height and inflorescence, is a semi-halophyte, or amphyphyte. It is frequent along the two rivers, but by reason of its reproduction (being entirely by seed), and of its annual duration, its growth is not gregarious. It occupies the banks at the rear of the Scirpus- Typha association, in soil not quite wet enough for them, but when opportunity offers it flourishes in water 2 or 3 inches deep. This is not permitted by the vigorous preemptors of the river borders, but it may be seen occasionally along the canals. HALOPHYTIC FORMATION. Some halophytic associations, limited in extent, have already been indicated as bor- dering the springs described on previous pages. Here are also to be included the salt-grass meadows at the southwestern end of Salton Sea, consisting of a sod of Distichlis spicata which holds entire possession of the ground by virtue of its stout, matted rhizomes. MECCA-INDIO CHENOPODACEOUS ASSOCIATIONS. Far more important are the associations occupying the extensive alkaline flats, stretching from Indio to the borders of Salton Sea near Mecca, a distance of fully 15 miles. Owing to causes already explained, the soil is strongly charged with soluble salts, consisting of sodium chloride, sodium carbonate, and sodium sulphate in differing proportions, the total saline con- tent varying from 0.5 to over 3 per cent. Over much of the area the water-table is sufficiently high to permit the capillarity of the soil to raise the water to, or nearly to, the surface. PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 93 A careful study of the vegetation of this area, correlated with analyses of the soils in which the different plants grow, would doubtless yield definite and accurate results explanatory of its distribution. In the absence of such studies only a cruder and more general account can be presented. The dominant plants of this area are all of Chenopodaceous genera. They are Atri- plex lentiformis, A. polycarpa, A. canescens, Sueda torreyana, and Spirostachys occi- dentalis. A composite, a variety of Isocoma veneta, occupies an important subordinate position. These several species do not grow intermixed, or at most are intermingled only on the borders of the portions respectively occupied by each. Whether the soil shall be monopolized by Atriplex, or by the other two chenopods, appears to depend upon the alka- line content, the latter enduring larger percentages. ATRIPLICETA OF THE INDIO-MECCA FLATS. The three species of Atriplex above named hold possession of the larger part of this area, their distribution being determined by the amount of water in the soil. Those parts in which the water-content is greatest is occupied by Aériplex lentiformis. This is a vigor- ous species with leaves broader and greener than those of the others. It grows either in thickets close as a hedge, or isolated in great domes, 6 to 8 feet, and exceptionally even 12 or 15 feet in height and base diameter. The same requirement of a very damp soil is manifested elsewhere in its position about springs and on river banks. Where the soil is somewhat drier Atriplex polycarpa forms a nearly pure stand. Its habit of growth is similar to that of A. lentiformis, but mostly in thickets, and less fre- quently in dense entangled individuals, in neither case much over 3 feet in height. Its leaves are narrowly linear, short, and in color gray. The third species, Atriplex canescens, has great ecological adaptability, so that while reaching its best development on the drier margins of these flats, it also is frequent in the most arid soils. In such situations it grows in individual isolation, one low shrub more or less distant from its neighbors, but on the flats it is gregarious and forms close thickets, about 3 feet in height. It has the fewest, the smallest, and the grayest leaves of any of the three species, the foliage development of each being in accordance with the amount of moisture in the soil in which they grow. In October and November, the season of their fruitage, the Atriplices of the flats are so laden with heavy panicles of fruit that they seem as if thatched with them. These are of a bright green color, which is especially manifest in the broad fruit-wings of A. canescens, and the fruit of A. polycarpa is frequently tinged with red. By this adaptation the fruit obtains the benefit of a large chlorophyllous surface at the time of ripening, while the sparse foliage of small gray leaves minify the excessive insolation of summer. When thus loaded with fruit these homely shrubs have a cheerful and elegant appearance. SU/EDETA AND SPIROSTACHETA OF THE INDIO-MECCA AREA. Those parts of the flats where the alkaline content exceeds the amount tolerated by any of the Atriplices, and perhaps differs in saline composition, are given over to two other chenopodaceous plants, Sueda torreyana and Spirostachys occidentalis. As it grows in these flats, Sueda is an erect suffrutescent plant, 2 to 4 feet high, with numerous slender, much intertangled branches, which are distantly set with small, terete, succulent leaves. It is often gregarious in habit, and is notably high and thick-set on soil which has been denuded of its original vegetation. In such places it is the first plant to establish itself. Sptrostachys occidentalis also has terete leaves, which are larger, more numerous, and more succulent than those of Sueda, the only other true succulent of the flats. Both appear, where an open interval permits, as subordinates in the Atriplex associations, but 94 THE SALTON SEA. as dominants they are able to hold possession only of soils too alkaline to permit the growth of the latter. In tracts so heavily charged as to inhibit all the other chenopods of the flats, Spirostachys holds undisputed possession. Here it forms a low, sparse growth, the individuals separated one from the other by intervals of bare, alkaline-incrusted soil. The stress of the conditions is indicated by the dwarfing of the plants, hardly a foot high, and by the yellowish color of the foliage, in striking contrast to the dark green of the vigor- ous bushes two or three times their size which flourish under more favorable conditions. Forms of Isocoma veneta var. acradenia, a many-flowered composite, appear as subordi- nate members of all of the associations which have been passed in review. Occasionally, in drier and sandier spots, they form a pure growth, marked, among their somber neighbors, by the bright yellow of their abundant autumnal blossoms. In a variety of unstable eco- logical forms, some of which have received names as species, Isocoma veneta is the most widely distributed plant of the desert. In other forms it is found in all except the moun- tainous parts of southern California. Its value as a phyto-geographical index is, therefore, inconsiderable. In the drier parts of these alkaline flats it attains its greatest size and its most abundant inflorescence. It is here several-stemmed, and the clusters of a few heads, borne on pedicles 1 or 2 inches long, are produced in large open panicles. In all these associations there is a singular dearth of herbaceous plants. A not infre- quent Heliotropium curassavicum spreading over the ground and an occasional Conyza couliert are the only species of any importance. Not less than five trees grow in this formation, but a consideration of them is reserved for a subsequent page. HALOPHYTIC ASSOCIATIONS OF IMPERIAL VALLEY. The southern extremity of the Sink also has its halophytic associations, whose char- acter varies in accordance with the nature of the soil. The plains of compact Imperial clays receive but little water from any source, and this little they absorb with difficulty and lose with rapidity. The water-table is far below the limited capillary capacity of the soil. In many places the percentage of alkali is considerable. The factors inimical to vegetable life are at the maximum, and there are wide expanses absolutely devoid of a single plant save in the infrequent furrows and channels which constitute the drainage system. In these the passage of the water has somewhat leached the soil and rendered it more open and retentive. Here scattered and stunted suffrutescent plants are able to exist. Mainly they are Sueda, reduced to a low and compact form, with an occasional equally stunted Atriplex canescens. In the minor drain furrows the plants are fewest and smallest, and their number and development, never great, increases in proportion to the size of the channel and the consequent greater leaching effect of the water discharged through it. A second association, which the railway traveler may observe near Estelle, is com- posed of Spirostachys, growing at intervals of 4 to 10 feet, each the center of a small mound built up, to all appearance, by wind action. The intervals are flat and perfectly free from vegetation. Other parts of these flats are occupied by an open growth of Sueda, usually inter- mixed with more or less Atriplex canescens. This soil has a less alkaline appearance and probably a less alkaline content, for when reclaimed and irrigated it is considered suitable for agriculture. SESUVIUM-SPIROSTACHYS ASSOCIATION. At points on the southeastern borders of Salton Sea, where the ground is moist and moderately alkaline, the zone nearest the water is covered with Sesuvium sessile, growing in large, prostrate mats, circular in shape. Behind this is a mixed zone of Sesuvium’ and Spirostachys, both growing up through mounds which have been heaped about them by PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 95 the wind. These increase in size with the distance from the shore line. It is a condition which soon proves fatal to Sesuviwm, so that the outer and drier soil is left to the more resistant Spirostachys. Among these drier mounds Astragalus limatus grows in consider- able abundance, and a little Oligomeris glaucescens appears as an interval plant. HALOPHYTIC ASSOCIATIONS OF THE RIVER BOTTOMS. When the interval between the bluff banks of the rivers is sufficiently wide, it is occu- pied by low bottoms, little above the water-level of the channel which winds through them, the whole wet or damp in proportion to the distance from the stream. The paludose associations which occupy the immediate margins of the streams have already been de- scribed. The remaining vegetation consists of halophytic and mesophytic plants, fairly well segregated in small societies in the wider bottoms, but much confused in the narrower ones. The mesophytic element, consisting of willows and cottonwoods, will be considered later. The halophytic societies occupy open intervals which are probably due to soil conditions. Sida hederacea holds possession of some of these. The individuals are small, the declined stems seldom over 6 inches long, and usually the individuals are separated by an inch or two of bare soil. Lippia nodiflora covers small tracts, but covers them compactly, since the stems root at the nodes and continually spread. Atriplex fasciculata, an herbaceous species with prostrate stems, a foot or less in length, is a component of these associations. Other less prominent members are Dicorea canescens, a bushy annual, Sesuvium, and all the chenopods of the Indio-Mecea flats. SESUVIUM-PHRAGMITES ASSOCIATION OF SALT SLOUGH. The so-called Salt Slough is a dry wash, heading in the playa below Dos Palmas and reaching Salton Sea near the Salton railway station, a mere section house. It is at this point quite deep, with bluff banks of clay, and at the present stage of the sea it forms an inlet of it. From one of the banks a seepage of salty water serves to moisten a strip along its base, so that there is a narrow, more or less muddy bottom, to the maintenance of which the water of the inlet doubtless contributes. The vegetation is disposed in the following manner: The wettest part of the seepage soil, close to the bluff, is occupied by Scirpus paludosus, but this does not enter the water of the inlet, which is bordered by a shore entirely bare, a feature doubtless due to the rapid recession of the water. The greater part of the flat is spread with broad circular mats of Sesuvium, some of them as much as 6 feet in diameter; with this grows Heliotropium curassavicum, a semi-succulent herb. This, in turn, is bordered by a fringe of Phragmites communis, which throws out an abundance of stout stolons, 6 or 8 feet long, wandering over the surface of the sand without taking root. MESOPHYTIC FORMATION. As will be readily inferred from the climatic conditions, mesophytes are but poorly represented in the flora of the Sink. Even those plants which must be included in this class possess certain xerophytic characteristics. The leaves of the delta cottonwood are but half the size of those of the cottonwood of cismontane California. The leaves of the willows are very narrow and have a cuticularized epidermis. Baccharis glutinosus, also a narrow-leaved species, has the surface of its leaves protected by the varnish which sug- gested its specific name. Yet these few are all the plants which, by any construing of the term, can be reckoned as mesophytes. MESOPHYTIC ISLETS IN THE MECCA FLATS. In the lower end of the Indio-Mecca flats, and mostly in the parts occupied by Atriplex lentiformis, there*occur small tracts of mire, covered with a surface layer of firmer soil, which yields under the foot, and through which animals are likely to break. They are, 96 THE SALTON SEA. in fact, blind springs, and perhaps a little water may at times appear on the surface. The center is mostly occupied by T'ypha, with a considerable admixture of Pluchea camphorata. In this, and surrounding it, is an open thicket of slender black willows and delta cotton- woods, with more or less Baccharis glutinosus. Usually there is an outer margin of Dis- tichlis sod. It will be seen at once that this is a society very like that previously described as growing about springs, and it might with some propriety be included with them. It is differentiated, however, by the predominance of the mesophytic trees. SALICETA OF IMPERIAL VALLEY. It has been already mentioned that the flora of the river bottoms of Imperial Valley is, in part, mesophytic. This consists of small clumps of delta cottonwood, but more especially of thickets of Salix. In places the same willow also appears densely bordering the banks of irrigation canals. XEROPHYTIC FORMATION. It is not altogether easy to draw the line between the halophytic associations of alka- line soils and the distinctly xerophytic associations, for the latter often include such typical halophytes as Sueda and Spirostachys. But these are here very subordinate members in the associations, most of whose characteristic plants never enter into the composition of true halophytic societies. The xerophytic formation occupies an area exceeding the combined areas of all the others, comprising, indeed, the greater part of the whole Sink. It can be divided into certain associations which, while including in their composition some plants common to all, yet are in other ways sufficiently differentiated. SOME GENERAL CHARACTERISTICS OF THE XEROPHYTIC FORMATION. The most marked general characteristic of the flora of southern California is the prevalence of shrubs and suffrutescent plants. These are less accompanied by other forms of vegetation in direct proportion to the aridity of the environment. In the Salton Sink the climax is reached. The species of shrubs are few, but they constitute the greater part of the whole plant population, plants of other forms being but an insignificant part of the whole. Individual shrubs are separated by intervals often great and almost wholly bare. They are dwarfed by lack of moisture; as there is no competition for sunlight, but rather a need of protection from excessive insolation, they are low, compact, and more or less rounded in outline, this being the shape best adapted to shelter them from the light and the heat of the sun’s rays, and from the evaporating effect of the strong, dry winds. This is well exemplified by Larrea, which in some other parts of the desert, where the condi- tions favor a social growth, attains a height of 6 feet and has a loose and open form; while in the Sink it seldom exceeds 2 feet in height and is almost as compact as the garden box. A noticeable feature of the xerophytic vegetation of the Sink is the absence of certain types of plants and certain modifications of others usually present in similar environments and recognized as characteristic of arid regions. Thus, in the mountains about the Sink, and almost to its very margin, in the San Gorgonio Pass down to as low as 400 feet altitude, and on the summit of the broad open plain which divides it from the slope towards the Colorado River, the Cactacee surround the Sink on every side, but do not enter it.' So closely do certain opuntias approach the margin of the Sink, along the flanks of the Chuckawalla Mountains, that there seem to be no differences of temperature, aridity, or soil 1 Near Figtree John Spring a single specimen of Echinocactus cyli i ich i 4 | n J E is cylindraceus was seen in a wash, down which it had evidently been carried from the neighboring mountains, where the species is frequent, and a solitary cylindraceous Hae was seen near Mecca under similar conditions; but neither can be considered members of the true flora of > PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 97 to which to attribute the presence or absence of these plants; but in an equilibrium of deli- cately balanced conditions a slight cause may have a determining effect. Be the cause what it may, the fact remains that the Sink is below the altitudinal limit of succulent plants. Induments of various kinds, and aromatic exhalations, are characteristic of xerophytic plants and are often extensively developed; but here, where conditions of aridity are extreme, protections of this nature have not been evolved to a marked degree. Larrea has a highly varnished surface and a strong odor, but it properly belongs to a higher zone. The Atriplices have a scurfy coating, and other shrubs, such as Franseria dumosa, are moderately pubescent or hirsute. Among herbs, annual and perennial, this character is more pronounced, as exem- plified in Psathyrotes ramosissima, Coldenia plicata, and Encelia eriocephala. None of the trees have developed characters of this kind beyond a very slight degree. Nor are there in the flora of the Sink, with two exceptions, plants which possess any form of storage root, a development so often found in arid regions. The exceptions are Cucurbita palmata, which has moderately large tubers, and Hesperocallis undulata, which has a deep-seated bulb. The first is not uncommon in most parts of the Sink, but is more abundant in the cajions of the adjacent mountains; while the latter, although common in the southeastern borders of the Californian deserts, has as yet been detected in the Sink only at Mecca. The results of the arid conditions under which the xerophytes of the Sink maintain their existence is doubtless manifested in important anatomical and physio- logical modifications, but these are not here considered; only the evident exterior adapta- tions of the plant organs can be noted. The agents by which these are produced are the excessive insolation in a region of ever-cloudless skies and the great evaporating effect of a dry and windy atmosphere on plants inadequately supplied with water. It is to these causes that the distinctive aspect of the desert flora is due, so different from shrubs of more favored regions. The condensed growth of the plants, affording the maximum of shade and protection, has been already adverted to. The same causes have produced the small, narrow leaves, sparse and often early deciduous, so that some are leafless for much of the year, which is so common a feature of the desert flora. Atriplex hymenelytra is an exception in this respect, but its broader leaves are thickened and the epidermis permits so little transpiration that branches broken off are weeks in becoming dry. The prevailing grayness of the desert vegetation is a response of the isolated and exposed plants to the intense sunlight which pours upon them. It affords to the chloro- plasts a needed shade. Considerations such as these are commonplaces in treating of desert floras. A study less hackneyed, more difficult of prosecution, but promising important results, relates to the character and the extent of the ligneous and fibrous root systems of the desert xero- phytes. An example of the method and the value of such investigations is offered by Cannon’s “‘Root Habits of Desert Plants.’# To such a study the flora of the Sink offers an inviting and an untouched field. The soils differ widely in their composition and texture, and consequently in their capacity for admitting and retaining water, in the depth to which it may descend, or from which it may be raised by capillarity, and in the facility by which they may be penetrated by the roots of plants. While experimental proof is wanting, it is certain that these factors are largely those which determine the distribution of the xerophytic flora. The revelations made by the washing away of bluff banks show that even in compact soils roots penetrate to great depths. There are certain anomalies to which such a study would probably furnish the key. The two species of Prosopis have an abundant, although finely divided, foliage, and they flourish best and attain their greatest size in damp soil, yet they are to be seen in very arid soil, and P. glandulosa seems unaffected even when buried to the tips of its branches in 1 Carnegie Institution Publication No. 131, pp. 59-61, 96, pl. 23, 1911. 7 98 THE SALTON SEA. dunes of drifted sand; but in the very compact soil they are not found. The inference is easy that they largely owe their extended range to the ability of their roots to go deep for water in soils that permit penetration. A certain endemic perennial herb, Astragalus limatus, which is scattered all over the Sink, at once impresses one as often out of place in its surroundings. It belongs to a genus well represented in the arid west, where numerous species exhibit the reduced foliage, the protective indument, or the limited growing period, which adapts them to their environ- ment. This species has numerous and fairly broad leaflets, which are but slightly thick- ened; only the growing apex is protected by a sparse pubescence, and while the lower leaves die those above continue green and growing throughout the year. It is, in short, a plant of more mesophytic than xerophytic aspect. Probably an examination of its root- system would explain the apparently incongruous conditions under which it often grows. AGE OF THE XEROPHYTIC TREES AND SHRUBS. The student of the arid mesas of the Sink is at once impressed by the apparent great age of most of the ligneous vegetation. It is true that this appearance may be aided by the unsymmetrical forms of many of the trees, by the sparseness or absence of foliage, and by the grayness of the shrubs. No investigations directed to this subject have been made in the Sink, but this aspect of age is characteristic of the sylva of the deserts, only accentuated here in its most arid part. Shreve! has recently made determinations of the age of 146 specimens of Parkinsonia microphylla, a leguminous desert tree growing near Tucson. He found the age of 50 of them to be 200 years and upwards, the oldest exceed- ing 400 years. Only 10 were of the comparatively youthful age of 75 years. It is probable, therefore, that the aged appearance of the desert trees and shrubs is not altogether decep- tive. The proportion of very old trees may be expected to be greater in the Sink than at Tucson, where the aridity is much less. It is greatly to be desired that extended investi- gations of this nature should be made on the Salton trees. JABSENCE OF SEEDLINGS. The absence of seedlings and of juvenile and adolescent individuals is a topic of interest, in connection with the question of age. During the course of more than a month devoted, at several periods of the year, to a study of the flora of the Sink, the writer did not see a single seedling or the withered remains of one, save of annuals, in any of its arid portions. This deficiency is not due to a lack of seed production; on the contrary the vegetation is exceptionally fertile. Atriplex canescens is literally weighed down with its heavy fruitage. The mesquite produces abundant crops, and the screwbean is little less prolific. Beneath Parosela schottit the ripened pods lie thick. In short, all the perennials, from the trees to the radiate Chamesyces that lie prostrate on the soil, are most prolific in the production of seeds, and the annuals are equally fertile. Instructive experiments carried on by Dr. MacDougal, to ascertain the readiness of germination and the duration of the viability of the seeds of the desert plants, are detailed in another section of this report. Incidental facts shed some light upon these points. Cercidium torreyanum has been extensively planted as a street tree in the town of Brawley, in Imperial Valley. The trees have now reached the fruiting age, and prove objectionable because of the great number of seedlings which spring up in the lawns and streets. In the same valley, while annuals are far from common, there seems to be an accumulation of their seeds in the dry earth, for no sooner is a place well wetted than there springs up an abundance of certain species. It is altogether probable that the absence of seedlings is due, not to any cause inherent in the seeds themselves, but to the want of proper conditions for their germination. These 1Shreve, F. Establishment behavior of the palo verde. Plant World, xiv, p. 293, 1911. PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 99 not only need moisture sufficient to induce the seeds to germinate, but that it be con- tinued long enough to enable the young plant to establish itself. Such conditions can occur on the desert but exceptionally, perhaps only at intervals of 100 years or more; or, it may be that an investigation of the ages of the desert trees would afford indications of secular changes in the desert climate. Only one of the desert trees forms an exception to the absence of adolescent individuals. These may be seen of various ages among the groups of Parosela spinosa, a tree which grows almost exclusively along the dry water-courses of the desert; it is, indeed, as certain an indication of a “‘dry wash’’ as is a line of willows of a living stream in a well-watered district. Soaked by the occasional torrents which pour down these washes, their soil affords an exceptional seed bed to the bordering Paroselas. ASSOCIATIONS OF THE XEROPHYTIC FORMATION. While the xerophytic flora has many points in common throughout the whole Sink, it is differentiated into three associations, dependent upon the nature of the several soils. They may be distinguished as the associations of the detrital slopes, of the Imperial clays, and of the drifted mounds. THE DETRITAL ASSOCIATION. The plants of the first of these associations occupy the detrital soils along the north- eastern margin of the Sink, extending from the high contour line either to the alkaline flats at the upper end, or, at the lower, as at Durmid, to the present shore of Salton Sea. Eventually, as the Sea evaporates and disappears, the whole lower border will be along the lacustrine flats and marshes of the central depression. A less well-marked area of this association is situated on the opposite side of the Sink, at its upper end. As has been already explained, these areas of washes and detrital fans have soils more permeable to water and to the roots of plants than the other arid soils of the Sink. They are also more subject to torrential floods and have, perhaps, a slightly greater annual precipitation. Consequently the plant population is comparatively denser, although still open, and there is a greater diversity of species. Some of the leading species, as Atriplex canescens, are equally prominent elsewhere, but others, which are here frequent, are seldom or never present in the other associations. This is especially true of those which grow along the upper margin, where a few of the cafion plants come down into the Sink flora. In passing upwards to this margin there is a certain succession of plants encountered, indicating a zonalization, which depends partially upon soil and, probably, partially upon deep-seated water tables. For, arid as these mountains are, to a certain extent they are reservoirs of water, as is demonstrated by the occasional springs found in the cafions. It is evident that as the drainage slowly percolates through the detrital cones the parts nearest the base of the mountain will contain the greatest and most reliable amount of water, and this can not but have its effect upon the plant covering. This zonal distribution is readily observable in passing upward from Mecca to the mouth of Red Cafion, in the Chuckawalla Mountains. Leaving the alkaline flats, which have been elsewhere described, a narrow zone of dry lacustrine soil is encountered. It is covered with an open growth of low bushes of Atriplex canescens, the open intervals quite abundantly grown up with dwarfed Plantago fastigiata, Cryptanthe sp., and other annual herbs. Beyond this begins the detrital soil, where the Atriplex is still abundant, but is mixed with much Petalonyx thurberi and Franseria dumosa, and a few Larreas. A little higher leguminous trees appear, Prosopis pubescens, P. glandulosa, Cercidium torreyanum, Olneya tesota, and Parosela spinosa, all but the first two being confined to this association. These grow along and near the washes, where the soil is most open and where there is probably more deep-seated water. The principal herbs here are Coldenia plicata, Psathy- rotes ramosissima, and Chamesyce parishii. Towards the upper limits of the Sink Parosela 100 THE SALTON SEA. schottit and P. emoryi are very abundant, and there is considerable Croton californicus and Bebbia juncea. Among plants seen only in this association are Chorizanthie rigida, species of Hriogonum, Chamesyce serpens, Helianthus tephrodes, and Encelea farinosa. Atriplex hymenelyira and Cladothrix oblongifolia are abundant here but are seldom seen elsewhere. IMPERIAL ASSOCIATION. The association of the Imperial clays is distinguished by its much sparser plant popu- lation, made up of fewer species. Adtriplex canescens is by far the most abundant plant, and a low form of Isocoma veneta, with condensed inflorescence, is also common. Sphe- ralcea orcuttit, a suffrutescent endemic species, 2 or 3 feet high and flowering the most of the year, is confined to this association. The only trees, and those mostly in the extreme south, are the two species of Prosopis. Where the soil is wetted, under present conditions mostly by leakage from irrigation canals, Bouteloua arenosa and Lepidium lasiocarpum appear in abundance. In some parts there are considerable societies of Larrea. MOUND ASSOCIATION. The principal plants of the mounds at the southern border of the Sink are Prosopis glandulosa and Larrea tridentata. Both are deeply buried in the loose drifting material of the mounds, the former without apparent detriment to its vitality, but the latter to its ultimate death. Baccharis sarothroides, a compact evergreen shrub, 3 or 4 feet high, is confined to this association. The annuals Pectis papposa, Chamesyce setiloba, Baileya pauciradiata, and Palafoxia linearis are more frequently seen here than elsewhere. The intervals between the ligneous plants are mostly bare. TRAVERTINE ROCK. Travertine Rock is situated at the northwestern end of the Sink, not far from Fig- tree John, and a mile or more from the present shore of the Sea. Seen from a distance it appears to be the culmination of a projecting spur of the western mountains, but it is, in reality, separated from them by a broad expanse of sandy wash, which continues past it towards the Sea, bordered by the usual groups of Parosela spinosa. The rock is a precipi- tous mass of light-colored quartzose, nearly 300 feet high. Its base is heaped about with great angular blocks which have fallen from it. In the far-off time when Lake Cahuilla filled the Sink to its brim, the sharp summit of Travertine Rock projected some 25 feet above its waters. All the once submerged part is coated with eray travertine, 1 to 20 inches thick. A distinct horizontal line divides the travertine from the whitish quartzose above and records the former high-water level. A few small palo verdes are rooted in the sand at the base, and all about grow the gray shrubs of the desert. Travertine Rock itself is almost destitute of any vegetation, but the little there is has the interest of introducing within the Sink limits a few migrant plants common in the neighboring desert mountains. The interstices of the tuberculated surface of the travertine contain, in a few places, a minute black lichen, infertile, and which could not be identified, but resembling one seen in Red Cafion. In sand pockets among the fallen blocks about the base grow sparse tufts of Stipa, and in the crevices a few specimens of Nicotiana trigonophylla, Hofmeisteria pluriseta, and Peucephalum schottii, the first two chasmophytes, and all common in the cafions of the adjacent mountains. PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 101 THE TREES OF SALTON SINK. In the preceding pages the trees have been referred to only incidentally, the purpose being to reserve further remarks concerning them for a few separate paragraphs. They fall into three classes: those which grow only in wet soil, those which tolerate both wet and superficially dry soils, and those confined to arid soils. The first class comprises the palm (Washingtonia filifera), the black willow (Salix nigra), and the delta cottonwood (Populus macdougalii). There are but few indigenous palms in the Sink—the two at Dos Palmas and a group of half a dozen in the alkaline flats near Mecca. There are said to be some others in the flats, but they were not found by the writer. They are abundant along the bases of the mountains northeast of Indio, and in various places in the cafions, but beyond the borders of the Sink. The desert palm has a columnar trunk 50 feet or more in height and 6 to 8 feet in circum- ference, surmounted by a crown of large leaves, whose fan-shaped blades are borne on long and stout petioles. Their functional life is about a year, and as they die they continually add to the mass of deflexed dead leaves hanging beneath the living ones. The palm flowers in May and June, each tree producing five or six large branched panicles of chartaceous blossoms. The fruit, a thin edible flesh surrounding a large bony seed, ripens in early fall. As it grows in the Sink, Salix nigra has a rough-barked trunk seldom exceeding 6 feet in height. An unusually large specimen measured 82 inches in circumference at 3 feet from the ground. The black willow fruits in May or June, and so abundantly that often the branches are white with the silky coma of the seeds as it bursts from the opening capsules. Large mature trees are usually found as solitary specimens, or in small groups, but there are often small thickets of pole-saplings along the rivers. The delta cottonwood has a clear trunk of 15 feet to the branches. One near Mecca was 86 inches in circumference at 3 feet from the ground, but larger ones were seen. None comparable in size to that attained by the common California cottonwood (Populus fre- montit) are to be found in the Sink or have been seen in the bottom lands of the lower Colorado River. Both this and the black willow require abundant and easily reached water, but neither will endure a large excess of alkali. The two species of Prosopis constitute the second class. Both are tolerant of widely varying edaphic conditions. They reach their best development in the drier and less alkaline parts of the flats about Indio and Mecca, but are also found in those which have large alkaline and water content. Towards the Mexican boundary, and still more beyond it, they cover large tracts of alluvium in open orchard forest. The occurrence of P. glandu- losa buried in dunes and mounds, at either margin of the Sink, has been more than once alluded to. P. pubescens and more rarely P. glandulosa are sometimes seen as solitary specimens in dry detrital soils; but they probably never grow where permanent moisture is beyond the reach of their deeply penetrating roots. Indeed, the desert inhabitants hold them to indicate the presence of water at no great depth. It is seldom that a mesquite (P. glandulosa) can be found with a single trunk more than a foot to the point of branching. One at Mecca had a trunk 18 inches to the branches and 42 inches in circumference; another, at Indio, had a trunk 30 inches high and 78 inches in circumference. Usually a number of large stems start from the surface, broadly spread- ing and becoming much deflexed, so that it is difficult to penetrate the outer circle of branchlets. The tree flowers in early spring, and in the fall the clusters of thick pods, 5 to 8 inches in length, ripen and eventually fall to the ground. The crop is usually abun- dant, but there are occasional failures. The screwbean is a smaller, less branched, and more erect and slender tree, rarely exceeding 15 feet in height, but usually with a single trunk. Two specimens at Mecca had trunks respectively 38 and 42 inches in height and 20 and 24 inches in circumference. 102 THE SALTON SEA. Its fruit is a pod tightly coiled in a cylinder 1 or 2 inches long, borne in close spikes of from 2 to 12 pods. Both species have abundant bipinnate leaves with numerous leaflets, which remain verdant throughout the summer, even in the more arid situations. Chilopsis linearis should not be included among the trees of the dry soils of the Sink, for although of this class it is represented only by a few migrants which have been brought down a wash from the neighboring cafions, where it properly belongs and where it is abundant. Those of this class which are authentic members of the Sink flora are Olneya tesota, Cercidium torreyanum, and Parosela spinosa. All are here small trees, although elsewhere Cercidium attains large dimensions. They grow in the detrital soils at the northern extremity of the Sink, usually along or near washes, where there is likely to be attainable moisture within reach of their roots. As their functional activity is largely suspended during the heated months of summer their demand for water must be limited. None was seen in the compact soils at the lower extremity of the Sink, or in the looser soil of the mounds. Olneya is a small tree, rarely over 15 feet in height. Two specimens near Mecca had trunks respectively 35 and 49 inches to the branches and 62 and 42 inches in circumfer- ence. The leaves are pinnate with but few and small leaflets, which had nearly all fallen by the last of June, when the fruit was ripe. The pods are brown and closely pubescent, and one- to three-seeded. I have not seen the flowers. Only the ultimate grayish twigs appear chlorophyllous. The palo verde (Cercidium torreyanum) may attain in the Sink a height of 20 feet. A specimen at Figtree John had a trunk 39 inches in the clear and 43 inches in circumfer- ence. Usually there are two to five nearly equal stems from the same root. The leaves are bipinnate and the few small leaflets are early deciduous, so that for the greater part of the year the tree is leafless, but the photosynthetic function is carried on by the bright green bark of the limbs and twigs. The large yellow flowers are produced in March and April, when the leaves have already fallen, and in such profusion that the whole crown is a mass of golden blossoms. The pods are numerous, 2 to 4 inches long, and constricted between the seeds. They were quite ripe late in June and still pendant from the branches, while in October they littered the ground beneath the trees. No specimen of Parosela spinosa with a single trunk was seen in the Sink, but always there were two or more nearly equal stems springing from the same root, so that it appears more a large shrub than a true tree. Twenty feet would be an extreme height, seldom attained. One near Agua Dulce was divided at the surface of the ground into two branches, respectively 44 and 30 inches in circumference. The wood is weak, so that old trees are often split and splintered below by the bending of the stems beneath their heavy load of branches and twigs. The leaves are very few in number and promptly deciduous, and they have but two or three minute leaflets. The numerous branches divide and subdivide into an intricate tangle of branchlets and twigs, almost gray in color. Some twigs are slender spines on which the violet-blue flowers are produced in great profusion in June, when the tree presents a very handsome appearance. The fruit probably ripens very promptly, as do the fruits of the shrubby species of the genus, which grow in the desert. Of the eight trees of the Sink, excluding Chilopsis, five are leguminous, all of the dry-land species being of this family. All five, notably the mesquite, are infested, often heavily, with Phoradendron californicum, a leafless, slender-stemmed mistletoe, producing abundant small red fruit. _ This commensal is an evident drain on the vitality of its hosts and sometimes causes their death. PLANT ECOLOGY AND FLORISTICS OF SALTON SINK. 103 ANALYSIS OF THE FLORA. Of the plants listed in the following catalogue 23 are sporophytes and 179 are sperma- tophytes, a total of 202 species. The sporophytes are represented by 15 families and 21 genera.! The spermatophytes are represented by 43 families (21 of them having but one species each) and 121 genera, 89 of which have but a single species each. Of the 179 spermatophytes, 131 are indigenous and 48 are introduced. All the introduced plants except a few found on the beaches of Salton Sea, are confined to those parts of the Sink which have been reclaimed through the economic operations of human settlement; in no case have they been able to intrude where natural conditions remain. The largest families are: the Composite, with 25 indigenous and 12 introduced species; Graminez, with 12 indigenous and 10 introduced species; the Leguminose, with 10 indige- nous and 2 introduced species; the Chenopodiacee, with 9 indigenous and 3 introduced species; the Polygonaces, with 6 indigenous and 2 introduced species; and the Euphor- biacez, with 7 species, all indigenous. Genera having over three species each are: Atriplex 8, Chamesyce 6, Eriogonum 4, Parosela 4, and Spheralcea 4. All the species are indigenous except one Aériplez. Disregarding 9 species which merely overpass the rim of the Sink, its indigenous flora consists of 8 trees, 23 shrubs, 10 suffrutescent plants, 30 perennial herbs, and 51 annuals. These indigenous spermatophytes may be divided in accordance with their habital character into two classes: (1) those which grow only in water or in moist soils; (2) those which grow in arid soil. The latter constitutes the flora of the absolute desert. Jsecoma and the two species of Prosopis are included in the second class, although they grow also in damp soils. TaBLE 30.—Habital division of the flora. Trees. Shrubs. Semi-shrubs. Perennial herbs. Annual herbs. Total. Ecological classes. < - a 4 4 3 3 af 3 2 3 2 3 ae 3 a WY 3 < el ajelelel2]elelF ei e/2) e181 8) 21 2) 2 ele lela |jel/eael|l|elelalse |e lal e |e} ale |e] oe Hydrophytes, etc....| 2 1 3 1 7 2 0 2 12 8 20 7 4 11 29 14 43 xXecphvies taskes sagaeine a 5 0 5 13 3 16 2 6 8 4 6 10 27 13 40 61 28 79 otal: ¢nctedin du 7 1 8 19 4 23 4 6 10 16 14 30 34 17 51 80 42 | 122 A floral census evidently conveys an inadequate and incorrect presentation of the phytogeography of a region if an equal value is given to species widely distributed and abundant and those which are rare—a single specimen of the latter perhaps having re- warded extensive explorations. With this in view an attempt has been made in the above table to separate those species which are fairly frequent (at least in some parts of the Sink) from those which are rarely seen. In making such a division the tabulator must needs be guided largely by what he has himself observed, so that his results are subject to the revi- sion of other and more extended investigations. Especially does this apply to the annual herbs; for under the irregular meteorological conditions of the Sink an unusual season may bring up in abundance annuals which in ordinary years make hardly any appearance. But it is believed that, on the whole, those species which are indicated in the table as “‘frequent”’ are those which the botanist will certainly find in more or less abundance in some part of the Sink, though not in every part. It appears by the table that of the 79 xerophytes, 51 are common in some parts of the Sink. The species which are abundant everywhere in the arid soils of the Sink are but two: Atriplex canescens and Isocoma veneta var. acradenia. These two are also found in the physiologically dry alkaline soils. In number of individuals they probably equal the united total of all the other plants above the rank of herbs. 1The undetermined moss and two undetermined lichens are not included in this synopsis. CATALOGUE OF PLANTS COLLECTED IN SALTON SINK. This catalogue is founded on collections made by Dr. D. T. MacDougal and by the author. A few additional references incorporate the short list of plants published by J. B. Davy in Bulletin 140, Agricultural Experiment Station, University of California, page 9. The numbers inclosed in parentheses are of the Parish collections; those num- bered lower than 8200 were made between June 27 and July 4, 1912, those between that number and 8400 between October 8 and 24 of the same year, those from 8400 to 8500 in February 1913, and those above 8600 from September 19 to 21, 1913. Stations without numbers are given from the author’s field notes. The MacDougal specimens are at the Desert Botanical Laboratory at Tucson. A set of the Parish specimens is in the author’s herbarium, and nearly complete sets have been deposited in the herbaria of the Desert Botanical Laboratory of the University of California, and in the Gray Herbarium of Harvard University. Synonyms are printed in italic, introduced plants in small capitals. LYCOPERDACE, Podaxon farlowii Massee. In the alluvial soils of the Sink, in places that have been wet and are drying off. Mounds at the old beach east of Holtville (8128), Rockwood (8211), Meloland, El Centro, Brawley, Mecea, Coachella (L. A. Greata). This and the following species were determined by Mr. Lloyd. Phellorina macrosperma Lloyd. A single specimen in alluvium which had been re- cently flooded, on the mesa at Mecca (8212). An African genus, of which, according to Mr. Lloyd, there have been but three previous collections in the United States. POLYPORACES. Polyporus corruscans Fries. In the fracture of a dead and fallen willow, Figtree John Spring (8605). Determined by Mr. Lloyd. ZECIDACEZ. Negredo scirpi (Cast.) Arth. Uromyces scirpi Burr. On the leaves and stems of Scirpus paludosus, New River near Rockwood (8217). Abundant at this station, but not seen elsewhere. It has been collected on Scirpus pacificus at several stations in California. Determined by Dr. Arthur. CHROOCOCCACEE. Chroococcus sp. On a dripping iron water-pipe, Thermal (8201). OSCILLATORIACEZ. Oscillatoria formosa Bory. On mud, Alamo River at Holtville (8204). Phormidium luridum (Kuetz.) Gomont. With the last species (8204a). Phormidium tenue (Mengh.) Gomont. On a dripping iron water-pipe, Thermal (8201). Schizothrix lardacea (Ces.) Gomont. With the last species (82018). 104 NOSTOCACE. Nostoc ceruleum Lyngbye. : In the small stream running from the railway water- works at Mecca (8460). BACILLARIDACEZ. Amphora sp. In the water of Salton Sea (MacDougal, 1908). Gomphronema sp. With the last species. VOLVOCACE&. Chlamydomonas sp. Developed in water taken from Salton Sea at Traver- tine Terraces, Feb. 8, 1913, by Dr. MacDougal, and used in cultural experiments at Tucson. CCELASTRACEZ. Ankistrodesmus falcatus (Corda.) Ralfs. With the last species. The four last_ preceding species were determined by Dr. T. E. Hazen. ULOTHRICACE, Conferva utriculosa Kuetz. (?) In a small stream from an artesian well at Mecca (8402). CDOGONIACES. Cdogonium sp. In_the water of Salton Sea (MacDougal, 1908). Determined by Dr. Hazen. CLADOPHORACEZ. Cladophora crispata (Roth.) Kuetz. In small streams from artesian wells at Mecca (8402, atk 8406, 8407). Determined by Mr. F. S. ollins. Rhizoclonium hieroglyphicum [Kuetz. var. macromeres Wittrock. In the water of Salton Sea (MacDougal, 1908). Determined by Dr. Hazen. A variety of the same et was collected in a small pool at Thermal CATALOGUE OF PLANTS COLLECTED IN SALTON SINK. ZYGNEMACE. Spirogyra sp. Sterile. In a ditch near Seeley (8203). Sterile Spirogyras were also collected in the small streams running from artesian wells at Mecca (8403a, 8404). Mougeotia sp. Sterile. In a ditch near Secley (8203a). The above Alge, except as noted, were determined by Dr. W. A. Setchell. CHARACE. Chara fragilis Desv. In the pool at Figtree John Spring (8614). Determined by Dr. Hordsted. DERMATOCARPACEZ., Dermatocarpon rufescens (Asch.) A. Zahlb. A very few plants, in the shade of a rock, at base of the mountains southwest of Travertine Rock (8472). Determined by Dr. H. E. Hasse. A few sterile thalli of a second lichen were growing with the above species (8472a), and in the corruga~ tions of Travertine Rock (8410). MUSCI. A small patch of a sterile moss was growing with the Dermatocarpon above mentioned (8411). Another sterile moss, apparently a different species, was found on the wet mud banks of the small stream flowing from the railway waterworks at Mecca (8482). NAJADACE. Ruppia maritima Linn. Filling the shallow current of Salt Creek near Seeley (8223). GRAMINE. Houcus gauePensis Linn. Sorghum halepense Pers. A single clump near El Centro. Introduced 25 years ago into southern California as a valuable forage grass and now become a trouble- some weed, but only adventive in the Sink. Pleuraphis rigida Thurb. Rabbit Bay (MacDougal 219). A common grass at higher altitudes in the desert; here a migrant. PasPALUM DISTICHUM Linn. Margins of canals, ditches, and reservoirs. Valley, Meloland (8094), Rockwood. A common weed in cismontane California, but prob- ably entering here from the delta, EcutinocuHLoa coLona Link. A very common weed in irrigated land throughout the Sink. Holtville (8085, 8241), Mecca (8101), river bottom at Rockwell, Imperial, Brawley, el Centro, Thermal. Very rare in cismontane California, and probably entering here from the delta. Echinochloa zelayensis Schult. Very common along the rivers and the irrigation canals and ditches in Imperial Valley. Holtville (8085), New River near Rockwood (8240), El Centro, Meloland, Brawley, Mecca, a single clump along the railway (8618). . The type was collected at Zelaya, in the State of Queretaro, Mexico. It isthe common Echinochloa of Mexico, extending into western Texas and Arizona, and now first reported from California. It occurs in the Colorado River bottoms near Fort Yuma, and is here an entrant from the delta. Readily recognized in the field by its narrow panicle with erect branches. Identified by Mrs. Agnes Chase. Imperial 105 GRAMINEZ—Continued. CENCHRUS CAROLINIANUS Walt. Depauperate annual, 23 cm. high. Plains northeast of Mecca (8440). An occasional weed in cismontane southern California. Setanta Guauca Beauv. Widely introduced, but nowhere abundant, in the cultivated parts of the Sink. Streets of Brawley (8236), Mecca (8244), Thermal. A widely distributed, but not abundant weed, in cismontane southern California. Aristida bromoides H. B. K. Occasional in dry soil. Indio (8237), near Dixieland (8239), base of the range of mountains west of Travertine Terraces (8434). A wade but not abundant grass of the Colorado esert. Aristida californica Thurb. (?) In sand-pockets, Travertine Rock (8238). The spikelets had fallen, so that the identification is not positive. In California the species is confined to the eastern borders of the Colorado Desert, thence eastward to Arizona and New Mexico. Sporobolus strictus Merr. Near Agua Dulce, at the northwestern end of Salton Sink (8242). An Arizona species, now first detected in California. This and the two preceding species were determined by Mrs. Agnes Chase. Sporobolus airoides Torr. A few large tussocks at Indio, and a single one between Calexico and Signal Mountain. A species of wide distribution on the Pacific Slope, extending into Mexico. AVENA FATUA Linn. Brawley, in fields (8613). Very common and long established in most parts of California, but rare in the Sink. PoLYPOGON MONSPELLIENSIS Desf. Imperial Junction Beach (MacDougal 1014). Not seen in Imperial Valley. A common weed of cultivation in California. Mecca. CyNnopon pDAcTYLoNn Pers. Bermuda grass. Very common in irrigated lands, and about houses throughout the Sink. Indio, Thermal, Mecca, Rock- wood, Brawley, Imperial, 1 Centro, Calexico. Common throughout California except in the moun- tains and deserts. CHLoRIS ELEGANS H. B. K. Common in cultivated lands and about streets throughout the Sink, always apparently introduced. Thermal (8247), Mecca. A reduced form 2 to 3 em. high, plains northwest of Mecca (8441). Abundant at the Government Date Garden, Mecca (8100). Brawley, El Centro, Holtville. Native of Texas and Mexico and an occasional waif in cismontane southern California. Bouteloua arenosa Vasey. Desert between Brawley and Salton Sea (8234), bluffs of New River near Rockwood (8235), Brawley and El Centro, about streets, Rabbit Bay (MacDougal 221). A species of northern Mexico, known in California only from the eastern borders of the Colorado Desert. Bouteloua barbata Lag. B. polystachya Benth. In dry soil near Mecca (8102), Town Park, Holtville (8125). In California known only from eastern part of the Colorado Desert, passing east and south to Mexico. 106 GRAMINE—Continued. Leptochloa imbricata Thurb. Alamo and New River bottoms, and now become an abundant weed in irrigated lands and about towns throughout the Sink. Indio, Thermal, Mecca (8119), Holtville (8248), Brawley, Imperial, New River at Calexico, Imperial Junction beach (Mac- Dougal 118). Known in Imperial Valley as “ water grass ’’ because believed to be carried into the fields by irrigating water. Probably an entrant from the delta, as it occurs along the lower Colorado River. A few col- lections have been made in cismontane southern California, probably of waifs. Eracrostis mMeaastacuys Link, A depauperate form 1 to 2 cm. high, plains northwest of Mecca (8442). Common weed throughout the United States and Mexico, but not seen elsewhere in the Sink. Triodia pulchellus Hitche. In sand along the base of the travertine range, south- west of Travertine Rock (8434). A common species of the lower Colorado Desert. Distichlis spicata Greene. Salt grass. Common in moderately damp saline soil throughout the western part of the Sink. Thermal, Mecca, Mortmere, Dos Palmas, Agua Dulce. Common throughout United States and Mexico. Phragmites communis Linn. In alkaline soil, Salton slough, near the Southern Pacific Railway station of that name (8068). Abun- dant at Travertine Terraces. Frequent about springs in the Colorado Desert, but rare elsewhere in California. CYPERACE. Eleocharis capitata R. Br. Dos Palmas (Hall 5984, May 4, 1905). An infertile Eleocharis was seen at Travertine Terraces. ta species at low altitudes in southern Cali- ornia, Cyperus erythrorhizos Muhl. Wet bottoms of Alamo and New Rivers and along canals and ditches in Imperial Valley. New River near Rockwood (8234), canal at Holtville (8237), New River at Calexico (8379, 8380), ditches at Imperial and El Centro. A widely distributed species eastward and collected at several places in central California, but not other- wise known from the State. It grows along the lower Colorado River and is probably an entrant from the delta. CYPERUS ESCULENTUS Linn. Abundant in the Government Date Garden near Mecca. A common weed in many places in cismontane south- ern California. Cyperus speciosus Vahl. An immature specimen, only 3 cm. high, which prob- ably belongs here, was collected on Obsidian Island (MacDougal 304). Mecca, in the stream from the railway waterworks (8616). The species is common in the Colorado River bottoms at Fort Yuma. Cyperus levigatus Linn. ee oe of the pool at Figtree John Spring A species of wide distribution in warm latitudes, but in the United States known only in southern Cali- fornia, where it is frequent. THE SALTON SEA. CYPERACEA:—Continued. Scirpus americanus Pers. In saline seepage, Travertine Terraces (8427). Very common in saline marshes in the Colorado Desert, and rare in cismontane southern California. Scirpus olneyi Gray. : Common in springs at northwestern end of the Sink. Figtree John (8378). Widely distributed in North America. Dos Palmas, Mortmere, Travertine Point (MacDougal 409, Parish 8426). Scirpus paludosus A. Nelson. . Common in the margins of New and Alamo Rivers and of the irrigation canals and ditches of Imperial Valley. New River at Rockwood (8249), Seeley and Calexico, Alamo River at Holtville, Irrigation ditches at Imperial (8376, 8377), Meloland (8093), and El Centro, Mecca, beaches (MacDougal 404). The type was collected in Wyoming, and the species has been known heretofore only from that State and a few places in Utah. It grows along the Colo- rado River at Fort Yuma and is here an entrant from the delta. Identified by Dr. Nelson. TYPHACE. Typha latifolia Linn. Common in the margins of water and in marshy places throughout theSink. Indio, Thermal, Mecca, Rock- wood, El Centro, Calexico. : . A cosmopolitan species common everywhere in Cali- fornia. JUNCACER. Juncus cooperi Engelm. . 3 Dos Palmas (8244), Figtree John Spring, Travertine Terraces (8428). A rare endemic species of the California deserts. Juncus balticus Willd. Ditch along railroad south of Mecca (8455). A rush of wide distribution, abundant in most parts of California. JUNCUS TORREYI Coville. In drain of railway water-tank at Mecca (8619). An occasional species in cismontane southern Cali- fornia, but here evidently introduced. LILIACE. Hesperocallis undulatus Wats. A few plants on plain northeast of Mecca. Abundant further east in both deserts, extending thence into Arizona. The present station is prob- ably the western limit. PALMACES. Washingtonia filifera Wendl. A group of a few trees in the alkaline flats near Mecca, two trees in Dos Palmas Spring. The species is endemic in Cahuilla Basin, and there are extensive groves, not far from the margin of the Sink, north of Indio. The two trees at Dos Palmas have leaves whose petioles are unarmed throughout, the first mature trees thus unarmed which have been seen, except in cultivation. They belong, therefore, to the variety microsperma Beccari. But the varieties of this species are founded on insufficient characters and are merely orms. SAURURACE., Anemopsis californica Hook. Moist saline soil bordering Dos Palmas Spring. Frequent in similar soil, notably in southern California and as far north as Sacramento River. CATALOGUE OF PLANTS COLLECTED IN SALTON SINK. SALICACES. Salix nigra Marsh. River banks, springs, and damp soil throughout the Sink. Mecca, Dos Palmas, New River near Rock- aa Holtville (8079), Rabbit Bay (MacDougal Possibly the variety vallicola Dudley (S. vallicola Britton), but our material is insufficient for certain determination. S. gooddingii Ball is apparently the same as Dudley’s variety, but I have seen no authen- tic specimens. While the reference of these desert trees to the species S. nigra is not satisfactory it may suffice for the present. Salix exigua Nutt. 8S. longifolia Muhl. in part. (?) Abundant in river bottoms and along irrigation canals in Imperial Valley. New River near Rockwood (8383) and Calexico, Alamo River at Holtville and east of Calexico, irrigation canals at Meloland (8092) and Brawley. Throughout the southern California deserts, and reaching the cismontane borders. Populus macdougalii Rose. In the bottom lands of the Alamo and New Rivers, as an entrant from the delta. Also, frequent in cultivation in all parts of the Sink. Mecca (8130, 8385, 8471, 8607, and MacDougal 128), Indio (8470). According to Indian testimony the delta cottonwood was unknown in the upper end of the Sink before the construction of the Southern Pacific Railroad, when trees were brought from Yuma and planted at several of the stations. From these trees were de- rived those now growing about the Indian setile- ments. The only self-planted cottonwoods in this part of the Sink are the young trees which are spring- ing up in abundance on the moist borders of Salton Sea and along irrigation ditches from wind-sown seeds of the cultivated trees. The delta cottonwood occurs along the Colorado River in the neighborhood of Yuma and is especially abundant bordering the diffluents of the delta. It differs from P. fremontii, which is frequent in central and southern California, in its smaller size, the lighter color of the mature trunks, the whitish gray of the bark of the limbs, and its smaller leaves, which are truncate, instead of cordate, at base, and are twice as broad (10 to 12 cm.) as high, and short-pointed at the apex. LORANTHACE. Phoradendron californicum Nutt. Common on Prosopis glandulosa, P. pubescens, and somewhat less so on Olneya tesota, Cercidium tor- reyanum, and Parosela spinosa, throughout the ranges of these trees in the Sink. Colorado Desert and east into Arizona. POLYGONACE:, Rumex berlandieri Meisner. Along rivers and in moist soil in Imperial Valley. Holtville (8078), Meloland, El Centro. Also on Obsidian Island (MacDougal 30, 1094) and Imperial Junction Beach (MacDougal 32). Eastward through Arizona and New Mexico to Mexico. An entrant from the delta. Not previously reported from California. Rumex crispus Linn. Along the stream from the railway waterworks, Mecea (8617). This European dock is abundantly naturalized in California, but only adventive in the Sink. PoOLYGONUM LAPATHIFOLIUM Linn. In a few places along an irrigation canal, half way between Calexico and Signal Mountain (8078). Similarly at Meloland. 107 POLYGONACEA—Continued. POLYGONUM LaPaTHiroLium Linn.—Continued. Frequent in cultivated grounds in central and southern California. Eriogonum trichopes Torr. In dry detrital soil near Mecca (8291). A common species of the Californian deserts. Eriogonum thomasii Torr. Obsidian Island (MacDougal 27, 114). A common species of the Colorado Desert. Eriogonum plumatella Dur. & Hilg. Caleb (8290), Durmid (8060), Obsidian Island (Mac- Dougal 205). A widely distributed species of the Californian deserts. Eriogonum deserticola Wats. Big Island (MacDougal 416). An endemic species of the Colorado Desert. Chorizanthe rigida T. & G. In dry detrital soil at Durmid (8061). Common in similar places throughout the Colorado esert. CHENOPODIACE. CHENOPODIUM MURALE Linn. An occasional weed about houses. Mecca, Brawley. Common in many parts of California. CHENOPODIUM ALBUM Linn. A few plants in fields at Mecca. This widely distributed weed is abundant in California, but was not seen elsewhere in the Sink. ATRIPLEX SEMIBACCATA R. Br. Abundant along the railway near Imperial (8252), and in the streets at Brawley. Introduced into cultivation in southern California as a forage plant, but not proving of value; now a common weed in many places, and especially abundant about San Diego. Atriplex fasciculata Wats. Streets of El Centro (8090), Imperial Junction Beach, and Obsidian Island (MacDougal). ake type was collected near Daggett in the Mojave esert. Atriplex lentiformis Wats. In wet saline soil throughout the Sink. Indio, Ther- mal (8266), alkaline flats, Mecca (8116, 8264, and MacDougal), Dos Palmas (8265), Caleb along ditches, Imperial, bottom lands of New River at Brawley and Calexico, mouth of Salton Slough, Travertine Wash (MacDougal 9), Obsidian Island MacDougal 411), Imperial Junction Beach (Mac- Dougal 106). A species of the California deserts extending into Arizona; also collected in lower San Joaquin Valley. Atriplex polycarpa Wats. Common in damp alkaline soil throughout the Sink. Indio, Thermal, alkaline flats, Mecea (8262), beach at Mecca (MacDougal 405), Travertine Terraces (MacDougal 821, 222), Obsidian Island (MacDougal 45), Imperial Junction, El Centro, banks of New River at Calexico (8261). A species of the Californian deserts and adjacent Arizona. Atriplex linearis Wats. In dry detrital soil, Durmid (8073), Imperial Junc- tion beach (MacDougal 108), in alluvial soil, Holt- ville (8258), between Brawley and Salton Sea (8253). A Sonoran species, reaching Arizona, and now first reported from California. 108 CHENOPODIACEZ:—Continued. Atriplex hymenelytra Wats. Desert Holly. In dry detrital soil in the northeastern part of the Sink, where it is abundant, and occasionally else- where. Mecca, Caleb, Salton, Durmid, Bertram, Obsidian Island (MacDougal 33, 48, 414). A species of the Californian deserts extending into Arizona. Atriplex canescens James. One of the most abundant and widespread plants of the Sink, in both moderately damp and arid soils. The wings of the fruit vary greatly in size and in the development of the teeth. : ; Wings deeply toothed: Bluffs of New River at Calexico (8074), Holtville (8077), between Brawley and Salton Sea (8254), Caleb (8255, 8256), El Centro, Durmid, Obsidian Island (MacDougal 41, 47, 118), Imperial Junction Beach (MacDougal 52). : Wings entire, or nearly so: Holtville (8257), Imperial (8259), Rockwood. A species of extended range in the arid west, through- out the deserts, and occasionally in other parts of southern California. Atriplex saltonensis Parish. Mesa at Mecca (8452, 8600). Iknown only from the type station. Spinostenhys occidentalis Wats. Allenrolfia occidentalis ‘untze, In alkaline soil throughout the Sink. Alkaline flats at Indio, Thermal and Mecca (8099), mouth of Salton Slough (8071), Caleb, Mortmere, Figtree John, Travertine Terraces in fruit (8437), Rockwood, borders of Salton Sea near Westmoreland, New River at Calexico. A species widely distributed in the western deserts. Suzda torreyana Wats. Dondia moquinii A. Nelson. In alkaline soil throughout the Sink. Alkaline flats at Indio and Thermal, Mecca (MacDougal), Rock- wood, Imperial, El Centro. A species widely distributed in the western deserts and frequent in cismontane southern California. AMARANTHACEA., Amaranthus palmeri Wats. River bottoms in Impcrial Valley and now an abundant weed in irrigated lands. Holtville (8269), Imperial, Brawley, New River at Rockwood and Calexico, East Bay of Obsidian Island (MacDougal 412). A species of the southern Colorado Desert extending into Arizona. Here an entrant from the delta. A. chlorostachys Willd. is reported in Davy’s list. There is no specimen from the Basin in the herba- rium of the University of California, and Davy’s record was doubtless founded on an erroneous de- termination of a specimen of A. palmeri. AMARANTHUS GRECIZANS Linn. A single specimen in a field at Mecca (S609), A common weed throughout California, but not seen elsewhere in the Sink. AMARANTHUS CALIFORNICUS Wats. Common in Imperial Valley in cultivated and waste grounds. Throughout California northward to Oregon; abun- dant in bottom lands of the Colorado River at Fort Yuma, whence probably introduced into Imperial Valley through irrigation canals. Cladothrix oblongifolia Wats. Frequent in dry detrital soil at the northeastern end of the Sink, and occasional elsewhere. Durmid (8067), Caleb (8298). Mountains and mesas of the eastern part of the Colo- rado Desert and adjacent Arizona. THE SALTON SEA. NYCTAGINACEX. Abronia villosa Wats. var. aurita (Abrams) Jepson. In light soil throughout the Sink. Dixieland, West- moreland, Calexico (8294), Mecca (8474), sands near Travertine Rock. Frequent in the Colorado Desert. AIZOACEZ. Sesuvium sessile Pers. . . Frequent in wet alkaline soil throughout the Sink. Mecca, mouth of Salton Slough, borders of Salton Sea northwest of Brawley, New River bottoms near Rockwood and Calexico, Imperial Junction (MacDougal 110), Carrizo Sands (Mac- Dougal). ; In California this species has been collected in San Joaquin Valley and in both the Mojave and the Colorado deserts. CARYOPHYLLACE, AGrosTEMA GiTHaGo Linn. : ‘ : A single plant noted along the railway in Imperial Valley. . : : As yet known in California only as an occasional waif. PORTULACACES, PoRTULACA OLERACEA Linn. A very rare weed about houses. Mecca, Brawley. This cosmopolitan weed is abundant in the older- settled parts of California. Calandrinia ambigua Howell. Claylonia ambigua Wats. Indio (March 1881, April 1884, Parish). An endemic species of the Colorado Desert, seldom collected. CRUCIFERE. Lepidium lasiocarpum Nutt. In alluvial soil, coming up freely in placcs where water has dried off. Abundant on mesa land which had been flooded by leakage from a canal, Rockwood (8296), Imperial Junction Beach (MacDougal 51), in a small basin in the mounds east of Holtville, near Westmoreland. Alamo River, according to Davy’s list. A species of the western deserts, Dithyrea californica Harv. Plains near Mecca (8451, 8457, 8478), sands near Travertine Terraces. Common northwest to San Gorgonio Pass. Streptanthus longirostris Wats. Greene. Mesa northeast of Mecca (8444, 8473). A common species of the Colorado Desert. Brassica NIGRA Koch. Sparingly along a roadside near Imperial. A common weed in the cultivated parts of California. CAPPARADACEZ. Wislizenia refracta Engelm. Borders of Salton Sea (MacDougal). From Texas to the desert borders of California, and in the San Joaquin Valley. RESEDACE&. Oligomeris glaucescens Camb. Borders of Salton Sea (MacDougal 112, 114), Obsidian Island (MacDougal 41), dry desert between Braw- ley and Salton Sea, near Mecca (8443). In California in the deserts along the southern borders of the State, and occasionally on the lower coast, without doubt indigenous. Guillenia longirostris CATALOGUE OF PLANTS COLLECTED IN SALTON SINK. LEGUMINOSZ. Prosopis glandulosa Torr. Mesquite. A tree of wide distribution in the Sink, occurring wherever its roots are able to reach moist soil. Abundant in the alkaline flats from Indio to Mecca. river bottoms in Imperial Valley, dunes at Indio, and mounds east of Holtville and Calexico, Dos Palmas, Westmoreland, Alamo River near the Mex- ican boundary (8270), Obsidian Island (MacDougal 1308), Rabbit Bay (MacDougal 210), Travertine Terraces. A common species of the Californian deserts and occa- sional in cismontane southern California. Prosopis pubescens Benth. Strombocarpus pubescens Gray. Screwbean. Range of the preceding species, except in the dunes. Thermal, Mecca, Caleb, Dos Palmas, Durmid in dry channels of the mesa, Calexico, Rabbit Bay (MacDougal 210), Obsidian Island (MacDougal 30, 117), Figtree John, Travertine Terraces. A common species of the Californian deserts, thence southward into adjacent Mexico. A single group grows on the dry banks of the Santa Ana River be- tween San Bernardino and Redlands in cismontane southern California. Smaller than the preceding species and apparently able to endure greater aridity. Cercidium torreyanum Sarg. Parkinsonia torreyana Wats. Palo verde. In detrital soil at northern end of Sink, most frequent along washes. Between Mecca and Red Cajion, Caleb, Travertine Rock. Colorado Desert and adjacent Arizona and Mexico. Olneya tesota Gray. In detrital soil at northeastern end of the Sink, along or near washes; not seen on the western side. Near Mecca (8123), ‘Mortmere, Dos Palmas. Southern part of the Colorado Desert and adjacent Arizona and Mexico. Parosela mollis Heller. Dalea mollis Benth. Infrequent in arid soil. Mecca (8312, 8450), Dixie- land (8311). A common species of the California deserts, but appar- ently infrequent in the Sink. Parosela emoryi Heller. Dalea emoryi Gray. . Widely distributed in arid soils throughout the Sink and notably abundant in the washes and detrital slopes of its northeastern border. Near Mecca, mounds east of Holtville (8080), Caleb, Figtree ae Westmoreland, Obsidian Island (MacDougal A frequent species of the Colorado Desert. Parosela schottii Heller. Dalea schottii Gray. Abundant along the washes and on the detrital slopes of the northeastern part of the Sink towards its upper border. Between Mecca and Red Cafion, near Dos Palmas. Frequent in the Colorado Desert and eastward into Arizona. Parosela spinosa Heller. Dalea spinosa Gray, Abundant along washes, in the northeastern and northwestern borders of the Sink, and increasingly so towards its upper margin. Mecca, Caleb, Dos Palmas, Figtree John. A common species, bordering dry washes and in cafions, in the Colorado Desert, extending into Arizona, The first leaves of P. spinosa seedlings, 5 to 6 in number, are oblong, 2.5 to 2.75X0.5 to 0.75 cm., glandular-serrate and sparsely guttate-glandular. At the height of about 3 cm. they are succeeded by spines 2 to 2.75 cm. long, sparingly set with 109 LEGUMINOS—Continued. Parosela spinosa Heller. Dalea spinosa Gray—Continued. pointed glands. The whole plantlet is densely ap- pressed-pubescent. SESBANIA MACRocARPA Muhl. Along an irrigation canal near Imperial (8306) and said to occasionally occur along other canals. At the above station it was spreading into the adjacent irrigated alfalfa field. It was here a weed of culti- vation, brought in by the canals from the Colorado River, in whose delta it is abundant and indigenous. Although not seen along the Alamo or New Rivers, it is not improbable that it may also be a natural entrant. It is a weed in the Government Date Garden near Mecca, where it was sown to test its value as a cover crop, its roots being the host of nitrifying bacteria. Me.itotus inpica All. Adventive at Imperial. A common weed in cold, damp lands in many parts of California, Astragalus aridus Gray. Coachella (L. A. Greata, Herb. Parish), sandhills near Mecca (8467). An endemic annual of the Sink. Astragalus limatus Sheldon. Frequent in arid soil throughout the Sink. Near Mecca, Caleb, Durmid, between Brawley and Salton Sea (8271), El Centro, Meloland, Holtville (8081), Seeley, Westmoreland, Calexico. It attains a much more luxuriant development when growing in moderately moist soil than in dry. Tra- vertine Terraces (MacDougal 408, Parish 8429), An endemic perennial of the Sink. ZYGOPHYLLACEZA. Larrea tridentata Coville. sote bush. Frequent in arid soil throughout the Sink, but scat- tered and seldom dominant. Caleb, Figtree John, mounds east of Holtville and Calexico, between Calexico and Signal Mountain. Near Westmore- land it occurs in abundance over small isolated tracts. The most characteristic shrub of the entire Lower Sonoran life-area. TRIBULUS TERRESTRIS Linn, Common along railway tracks throughout their entire length, and extending into the streets of the towns. Indio, Mecca, Durmid, Imperial Junction, Brawley, Imperial, E11 Centro, Holtville, Calexico. This weed has entered California in recent years along the Southern Pacific Railway, and is now common about most railway stations in southern California and spreading in streets of towns. GERANACEA, Erodium texanum Gray. Plains northwest of Mecca (8449). A common species of both deserts, and eastward to Texas. Also rare in cismontane southern California, Eropium cicurariom L’Her. In irrigated fields, Mecca (8469). A common naturalized species throughout California, except the deserts and higher mountains. EUPHORBIACEZ. Croton californica Muell. In detrital soil towards the rim of the Sink near Mecca. Abundant northward in the Basin and in many parts of southern California. Covillea tridentata Vail. Creo- 110 EUPHORBIACEZ—Continued. Chamesyce eeeseene Small. Euphorbia cinerascens Engelm. Bana aah near Figtree John Spring (8306). Not before reported from California. ; Founded by Engelmann on plants from northern Mexi- co (Bot. Mex. Bound. Surv., vol. 2, pt. 1, p. 186), with a variety apendiculata from San Felipe, the latter reduced (Bot. Cal., vol. 2, p. 73) to a form of EF. polycarpa. Chamesyce parishii Millsp. ined. Hwuphorbia parishii Greene. Sand wash between Mecca and Red Cajfion (8113). The type of this little-known species was collected at Daggett, in the Mojave Desert. Chamesyce polycarpa var. hirtella Millsp. ined. Huphor- bia polycarpa Benth. var. hirtella Boiss. Dos Palmas (8304), sand wash near Dixieland (8308), Figtree John (8245), shores of Rabbit Bay (Mac- Dougal), sands between Travertine Rock and the adjacent mountains, abundant. Abundant northward in the Colorado Desert. Chamesyce saltonensis Millsp. ined. n. sp. Old beach-line near Calexico (8302), streets of Brawley (8305). An endemic species of the Sink, known only from these collections. Chamesyce setiloba Millsp. ined. Engelm. In sandy or loose soil, in which it is often partly buried. Mounds at the old beach-line east of Holt- ville (8087), sand wash in the desert southwest of Brawley (8301). A species of the southern parts of the Colorado Desert, the type collected at Fort Yuma. Euphorbia setiloba ee serpens Millsp. ined. Huphorbia serpens H. In hard detrital soil at Durmid (8066). South to Mexico and east to Kansas and Illinois. before reported from California. The above species of Chamesyce were determined by Dr. Millspaugh. Not MALVACE. Malvastrum exile Gray. Plains northwest of Mecca (8447). Common northward in the Colorado Desert and in the Mojave Desert. Matva PARVIFLoRA Linn. Adventive in several of the towns of the Sink, but nowhere abundant. Mecca, Brawley, El Centro. An abundant weed in California. Sida hederacea Torr. Frequent in extensive socicties, in moderately dry subalkaline soil, in the river bottoms of Imperial Valley. New River at Rockwood (8316) and Calex- ico. Also a troublesome weed in irrigated fields. poaenens El Centro, Meloland, Seeley, Westmore- and. An entrant from the delta, where it is common. Spheralcea orcuttii Vasey & Rose. Frequent in the arid alluvial soils of the southern and eastern parts of the Sink. Desert between Brawle and Salton Sea (8317), El Centro, Meloland (8097), HoH Dixieland, Westmoreland, Calexico, Dur- mid, Suffrutescent; flowering at most seasons of the year. An endemic species of the Sink. This and the following Sphzralceas were identified by Dr. Robinson. THE SALTON SEA. MALVACEZ—Continued. Spheralcea emoryi Gray. (?) Along railway tracks at Indio (8321). ; Corollas lilac rather than brick-red, but otherwise agreeing with the ordinary form of this species, which is common in the Californian deserts. Spheralcea fendleri Gray. Streets at Mecca (8318, 8451). 3 : A species of western Texas, New Mexico, and Arizona, Not heretofore detected in the Colorado Desert, but occasional, in a varietal form, in cismontane south- ern California. Spheralcea angustifolia Spach. var. cuspidata Gray. A number of plants near Indio (8319). This also is a plant of the same region as the last; not detected heretofore in California. TAMARICACEA, TAMARIX PALLASII Desv. A single shrub on the moist banks of Salton Sea at Travertine Terraces (8512). Determined by Dr. Niedenzu. In full flower in Sep- tember, and perhaps not identical with the shrub flowering in early spring, cultivated in southern California as T. gallica, an older name, according to the Kew Index, for that here adopted, but recog- nized as a distinct species by Niedenzu in Pflanzen- familien vol. 2, tiel 6, p. 294, 1895. LOASACEZ. Petalonyx thurberi Gray. Common in detrital soil throughout the Sink. Mecea, (8121), Caleb, Durmid, Figtree John. A common species of the Colorado Desert. ONOGRACEZ., Cnothera trichocalyx Nutt. Anogra trichocalyx Small. Desert between Brawley and Salton Sea (8326). Gnothera scapoidea Nutt. var. aurantiaca Wats. Chy- lisma claveformis Heller. Obsidian Island (MacDougal 29), abundant on plains northwest of Mecca (8445), and in sand at Traver- tine Terraces. The above Ginotheras are common plants of the Colo- rado Desert. LYTHRACEZ, LyYTHRUM CALIFORNIcUM T. & G. A single plant, at Figtree John Spring (8611). A common California species, not reaching the deserts, and here a waif. CACTACEA, Echinocactus cylindraceus Engelm. A single large specimen in a wash near Figtree John, evidently carried down the wash when small, from the a mountains, syoere the species is fre- quent. oes not properly belong to the flora of the Sink. ne : Opuntia echinocarpa Engelm. (?) A single plant in a wash near Mecca, without fruit or flower. Like the previous species it was a migrant from the adjacent mountains. ASCLEPIADACES, Asclepias subulata Decsne. A few plants in a wash near Agua Dulce at the south- western end of the Sink. Frequent in the washes of the mountains which border the Sink, here a migrant. CATALOGUE OF PLANTS COLLECTED IN SALTON SINK. ASCLEPIADACEZ—Continued. Philabertia linearis B. & H. var. heterophylla Gray. Philabertella hartwegti var. heterophylla Vail. Occasional in dry soil throughout theSink. Agua Dulce, near Brawley (8327), Holtville (8081), mounds near the Alamo River east of Calexico (8327). A common species of the desert, abundant in the Colorado River bottoms, and passing westward into cismontane southern California. CUSCUTACEZ. CuscuTA CALIFORNICA Choisy. A few plants, parasitic on alfalfa, in the town park at Holtville (8080). Slight infestments on alfalfa, seen elsewhere in Imperial Valley, were probably the same. A common indigenous species in many parts of Cali- fornia. POLEMONIACE. Langloisia schottii Torr. Gulia schottit Gray. A few plants in sandhills northeast of Mecca (8466). A common species of the Colorado Desert. HYDROPHYLLACEA. Heliotropium curassavicum Linn. A subordinate member of halophytic associations throughout the Sink; rarely dominant over limited areas, as on flats near Caleb. Mecca (MacDougal 5, 105), near Salton (8072), Obsidian Island (Mac- Dougal 45), Dos Palmas, Rockwood, borders of Salton Sea near Westmoreland, El Centro, New River at Calexico. Widely distributed in California and in most of the warmer parts of North America, Phacelia crenulata Torr. A single plant, in the sandhills northeast of Mecca (8464). Infrequent in the southeastern Colorado Desert and thence to Arizona and New Mexico. Eriodictyon californicum Greene. A single insulated society, occupying an acre or more near Indio. This common species of cismontane southern Cali- fornia occurs in a few places on the desert slope, as at Whitewater. The present station is its farthest known eastern and southern limits. Nama hispidum Benth. “ Alamo river-bed ”’ according to Davy’s list. The specimen (Davy 7965) is in the herbarium of the University of California and the locality noted on the label is Calexico. It has also been collected at Palo Verde on the Colorado River. BORAGINACE:, Cryptanthe angustifolia Greene. In a desiccated pool at the old beach east of Holtville (8124). Common species of the Colorado Desert, extending into Arizona. Cryptanthe barbigera Greene. (?) Obsidian Island (MacDougal 25, 26). The specimens are too young for positive determina- tion, but the species is common in the Colorado Desert and in Arizona. Cryptanthe costata Brandegee. In sands near Travertine Terraces (8429). Sandhills near Mecca (8465). An endemic species of the Colorado Desert. Deter- mined by Mrs. Brandegee. 111 BORAGINACEZ—Continued. Krynitzkia micrantha Gray. Greene. Sandhills northeast of Mecca (8464). A common plant of arid soils in southern California, extending eastward to Arizona. Eremocarya micrantha Coldenia plicata Coville. C. palmeri Gray. A common xerophyte of the Sink, especially in detrital soil, but occasional elsewhere. Mecca, Durmid, Imperial Junction, Figtree John, E] Centro, West- moreland. A common species of the Colorado Desert. Pectocarya penicillata A. DC. Mesa, at Mecca (8448). Common in arid soils in southern California. LABIAT. Menrtua citraTa Ehrh. In the stream from the railway waterworks, Mecca (8462, 8608). Naturalized sparingly in cismontane southern Cali- fornia. VERBENACES. Lippia nodiflora Michx. In the moist soil of river bottoms in Imperial Valley. New River at Rockwood (8330) and Calexico (8332), Obsidian Island (MacDougal 38). A cosmopolitan tropical species; in California known only along the lower Colorado and as above. An entrant from the delta. SOLANACE&. SoLANUM ELEAGNIFOLIUM Cav. About railway station yards and tracks. Durmid, Imperial Junction (8104), Imperial, Meloland. This weed entered southern California along the Southern Pacific Railway, and is now widely dis- tributed throughout its length in southern Cali- fornia, but is not abundant. Physalis wrightii Gray. Rare in river bottoms, but very common along canals and in irrigated lands, in Imperial Valley. Brawley, el El Centro, Meloland (8091), Rockwood 8338). An entrant from the delta and apparently owing its presence here mainly to the irrigation system. It occurs also in the Colorado River bottoms at Fort Yuma and is now for the first time reported from California. The type was collected in southwestern Texas. Physalis crassifolia Benth. A single plant at the base of the range southwest of Travertine Rock (8435). A migrant from the desert mountains, where it is a common chasmophyte in both deserts. Datura piscoLor Bernh. Streets at Mecca (8335), and along a road in the flats a few miles west, along a canal at Dixieland (8336), in somewhat damp soil not far from the borders of Salton Sea at Caleb and Agua Dulce. This Mexican species grows in the bottom lands of the Colorado River at Fort Yuma, where it appears in- digenous. In the Sink it is an entrant from the delta, apparently recent. The stations at Agua Dulce and the Mecca flats are all on land flooded by the recent Salton Sea; that at Caleb is slightly above the late high-water mark. It thus maintains the ambiguous character assigned it by Gray in the Synoptical Flora, where it is insufficiently de- scribed. 112 SOLANACEZ—Continued. Datura meteloides DC. Indio, Mecca; rare. . Common in southern California, occasionally reaching the western borders of the desert, and reappearing in the bottom lands of the Colorado River at Fort Yuma. Nicotiana trigonophylla Dunal. 3 A few plants in a wash at Caleb (8333), and three in crevices of Travertine Rock. A common chasmophyte in the cafions of the desert mountains, and occasionally in washes from seed carried down by floods. In the Sink only as a rare migrant from higher altitudes. NicoTIANA GLAUCA Graham. A single young plant at Mecca (8622) and two at Calexico. Common in cismontane southern California, where introduced from Mexico; adventive in the Sink. BIGNONIACE. Chilopsis linearis Sweet. C.saligna Don. Desert willow. A few trees at Caleb, where the seed had evidently been brought down a wash; common in cafions and along the courses of washes in the detrital soils of the Basin, but it belongs to higher altitudes than those of the Sink. PLANTAGINACEE. PLANTAGO LANCEOLATA Linn. Abundant in a few lawns at El Centro. A common weed in southern California. Plantago fastigiata Morris. Abundant on the plains northeast of Mecca (8446). Dried_ remains collected near Dixieland are prob- ably the same. A common species of the Colorado Desert, extending into Mexico. CUCURBITACEZ. Cucurbita palmata Wats. Wild gourd. Occasional throughout the Sink, notably in washes. Mecca, Caleb, Westmoreland. A common species of the Colorado Desert, but most abundant in cafions. LOBELIACEZ. Nemacladus adenophorus Parish. On dry mesas. Salton (Hall). An ephemeral winter annual of detrital plains, common in both deserts. COMPOSIT. Hofmeisteria pluriseta Gray. A ioe vlants growing in the crevices of Travertine ck. The species is a common chasmophyte of the moun- tains of the Colorado Desert, whence here a migrant. Brickellia californica Gray var. desertorum Parish. A few plants at the base of the range southwest of Travertine Point. Frequent in the desert mountains, from which it is here a migrant. HETEROTHECA GRANDIFLORA Nutt. Along the boundary-line road near Calexico; evidently a recently introduced weed. Indigenous, but more commonly a weed of cultivation, in cismontane southern California. THE SALTON SEA. COMPOSITZ—Continued. Isocoma veneta var. acradenia Hall. J. acradenia and I. eremophila Greene. . The most widely distributed plant of the Sink, where it is common in both physically and physiologically dry soils. Mecca (8117, and MacDougal 402), Imperial, Westmoreland, Brawley, El Centro, Tra- vertine Terraces (MacDougal 23), Indio (8353), Thermal (8343) in damp alkaline soil, are tall forms, with large branching inflorescence. Mecca (8350), in like soil, has few-toothed leaves and represents the I. eremophila form. Desert between Brawley and Salton Sea, in dry alluvium (8351), has rather broad, entire leaves and agrees with I. acradenia. Dixieland (8353) on artifically moistened alluvium, is 5 em. tall, the numerous involucres solitary at the summit of pedicels 2 to 5 cm. long. About Durmid and Imperial Junction low forms, with the inflorescence almost capitately condensed, are com- mon. A plant of such wide distribution as Isocoma veneta, and capable of accepting such varied edaphic con- ditions, necessarily responds with a complex of ecological forms. The student who has before him even large suites of herbarium material may be con- tent to define what appear to him valid species, but which field study would convince him could be sustained at most only as more or less uncertain forms and varieties. Treated in this light it might be advantageous to give subspecific names to some of the more marked of these forms, but if too rigidly defined they might be identifiable only with the type specimens. Aster spinosus Benth. Moist soil along the Alamo and New Rivers, and a pestiferous weed in irrigated lands, in Imperial Valley. New River at Rockwood and Calexico, Brawley, Imperial, El Centro, Meloland, Holtville, Westmoreland. Salton (MacDougal 120), Obsidian Island (MacDougal 202, 203). An entrant from the delta. Known to farmers of Imperial Valley as ‘‘ wild asparagus,” from a resem- Diente of its matted roots to those of that vege- table. ASTER EXILIs Ell. var. AUSTRALIS Gray. A frequent weed about towns and in fields. Thermal, Mecca, Brawley, Imperial, El Centro, Westmore- land, Obsidian Island (MacDougal 306, 413). A common weed in California; probably introduced in Imperial Valley by irrigation, as it is the form most abundant in the Colorado River bottoms. ERIGERON CANADENSIS Linn. A frequent weed about towns and in fields. Thermal, Mecca (8358), Rockwood, Brawley, El Centro. Common in many parts of California. Conyza coulteri Gray. In the alkaline flats at the northwest end of the Sink. Thermal (8357), Mecca (8356). Occasional at low altitudes in southern California, in soil more or less alkaline. Baccharis sarothroides Gray. Mounds along the old beach at southeastern border of the Sink. Near Alamo River east of Calexico (8349), and east of Holtville. In California confined to the southern border, and thence into Arizona and Mexico. Baccharis sergiloides Gray. A co oe in subalkaline soil, near Agua Dulce A species of the Colorado and Mojave Deserts, ex- tending into Nevada and Utah. CATALOGUE OF PLANTS COLLECTED IN SALTON SINK. COMPOSITA—Continued. Baccharis glutinosa Pers. Common along rivers and about springs in most parts of the Sink. New River bottom neary Brawley, Rockwood (8354), and Calexico, Alamo River near Calexico, Thermal, Mecca (8127, 8347), Obsidian Island (MacDougal 44, 118, 204, 302). A species extending south and east to Arizona, Colo- rado, and Mexico, and extensively distributed at low altitudes in California; but most abundant in the Colorado Desert. Baccharis viminea DC. About the moist borders of Dos Palmas Spring (8348). A common species of cismontane southern California, here quite out of its normal range. Pluchea camphorata DC. About springs and in marshy places in the north- western parts of the Sink. Alkaline flats at Mecca, beach at Mecca (MacDougal 403), Imperial Junc- tion beach (MacDougal 129), Figtree John Spring, Dos Palmas. Frequent in like places, at low altitudes, throughout California. Pluchea sericea Coville. Arrowweed. Very common in river bottoms, about springs, and in damp soils generally, throughout the Sink. Ther- mal, Mecca, Mortmere, Dos Palmas, Rockwood, Imperial, El Centro, Calexico, Penguin Island (Mac- Dougal 201). Common in southern California, most abundant in the desert. Dicorea canescens T, & G. Occasional in moist, subalkaline soil. Indio, as a weed in a young date garden (8366), New River bottom at Calexico (8365). Colorado Desert and thence into Arizona. Hymenoclea salsola T. & G. Detrital mesas near Mecca, Rabbit Bay (MacDougal 209). A common species of the Colorado Desert. AmpBRosia PsILosTacHya DC. Adventive at Imperial. A common weed in California and in many parts of North America. Franseria dumosa Gray. Gartneria dumosa Kuntze. Frequent in detrital soil and occasional in light allu- vium. Mecca, Caleb, Durmid, Westmoreland. A characteristic species of the Californian deserts, extending into Arizona and Mexico. XANTHIUM COMMUNE Britton. Frequent throughout Imperial Valley, Calexico, Holt- ville, Meloland (W. E. Paccard), Brawley. Also at Mecca (8438, 8610). Common in the Colorado bottoms at Fort Yuma; not reported otherwise from California, but it occurs at San Diego. Eciipra auBa Linn. In the river bottoms and very common along canals and in irrigated lands in Imperial Valley, and in waste places about the towns of the Sink. Mecca (8362), Rockwood, Brawley, E1 Centro (8089), Holtville, Mel- oland, Calexico, Obsidian Island (MacDougal 4,303). A widely distributed subtropical weed, common along the Colorado River bottoms at Fort Yuma and an entrant into the Sink from the delta. Not before reported from California. Bebbia juncea var. aspera Greene. . Oceasionul along washes at the upper margin of the Sink at its northeastern end. Near Red Cafion, Dos Palmas. ; : A apecies of the cafions of the desert mountains, which barely enters the borders of the Sink. 8 113 COMPOSITA!—Continued. HELIANTHUS ANNUUS Linn. This common California weed is beginning to appear in the fields and towns of Imperial Valley. Braw- ley, Meloland, El Centro. Also at Mecca, rare. Encilia farinosa Gray. Occasional on detrital soil at the northeastern end of the Sink. Salton, Caleb. A common species of the California deserts, reaching westward into San Bernardino Valley, and east- ward into Arizona. Encelia frutescens Gray. Rabbit Bay (MacDougal 214). A widely distributed species in the cafions of the deserts of California, extending to Arizona and Utah. Here a migrant from the mountains. Encelia eriocephala Gray. Common on the plains at Mecca (8345). Island (MacDougal 33). Abundant in sandy soil in the Mojave and Colorado deserts. BipEns Pitosa Linn. A single plant by the railway at Mecca. Common along ditches and small streams in cismon- tane southern California, here a waif. BIDENS EXPANSA Greene. A few plants in the stream from the railway water- works, Mecca (8462). An abundant indigenous species of cismontane south- ern California, but here evidently introduced. Baileya multiradiata Harv. & Gray var. pleniradiata Coville. A single plant in sandy soil, near Meloland (8095). Common at lower altitudes in the California deserts, thence east to New Mexico. Baileya pauciradiata Harv. & Gray. Meloland (8095), Dixieland (8368). Abundant in places in the mounds near the Alamo River, east of Calexico, and in sand at Travertine Terraces (8430). A species of the eastern parts of the Californian deserts and adjacent Arizona and Mexico. Palafoxia linearis Lag. Indio, as a weed in a date garden, Westmoreland. Abundant in places in the mounds near the Alamo River, where it becomes lignescent at base and more than annual (8364). Sands at Travertine Terraces. A widely distributed species of the Colorado Desert, less frequent in the Mojave Desert, and extending into adjacent Arizona and Mexico. Pectis papposa Harv. & Gray. Frequent in alluvial soil, notably in the mound region of Imperial Valley. Flats near Mecca (8371), bluffs at Rockwood (8370), mounds of the old beach east of Holtville (8083 epappose, 8084 pappose), mounds of the Alamo River east of Calexico, where plentiful in places, in both pappose and epappose forms (8372), streets of Brawley as a weed. Southern border of the Colorado Desert, eastward to New Mexico. ANTHEMIS COTULA Linn, Adventive at a few places in Imperial Valley. Im- perial. This old-world weed is common in southern California, whence it has probably been introduced here. Peucephalum schottii Gray. A few individuals grow in the crevices of Travertine Rock a at the base of the range southwest of it (8431). This is a characteristic species of the cafions of the arid mountains of the Colorado Desert, and enters the Sink only under the exceptional circumstances here presented. Obsidian 114 THE SALTON SEA. COMPOSITZ—Continued. COMPOSITZ—Continued. Psathyrotes ramosissima Gray. Soncuus asPer Vill. Frequent along washes and on detrital mesas at the Occasional in the cultivated parts of Imperial Valley. northeastern end of the Sink. Between Mecca and Brawley, Imperial, Mecca (MacDougal 123), Obsi- Red Cafion (8115), Obsidian Island (MacDougal dian Island (MacDougal 31). 116). g Li Common in similar places in the Colorado Desert, and ae CLEHACEUS ann. : 3 : 2 : i ery common in the cultivated parts of Imperial east and south into Arizona and Lower California. Valley. Brawley, Imperial, El Centro, Mecca Flats Stephanomeria runcinata Nutt. (MacDougal 127). A few plants at Caleb (8367). These two sow thistles are abundant weeds of culti- Occasional in the California deserts, and thence into vation in California, and owe their presence here Lower California, to the recent settlement of the valley. TABLE 31.—Appropriate altitudes of the localities, in feet. Agua Dulce............... — 175 Hetelle. +. <6.e40e stew eas — 180 New River at Calexico...... — 20 BEAWICY 5 3 ¥esg.cisen sae ec ash eae — 115 Figtree John............... —195 New River at Rockwood... — 185 Cale Dy sc eaiiied acashtn va Mene ae 6 — 200 Holtville..............004. — 15 Rockwood................. — 160 alexi COs. 2 siete ancee nes - 2 INdiOv nas orweaey eines — 20 Dal COM sce: ais 2 nthe ecg geadten — 200 Coachella...............-. — 71 TM Per al edie tei rdew mare oin — 70 DOCLEY ices dacs oa tan oantesin tees — 50 Dixielan diss acco scsex awarwigs we — 650 Imperial Junction.......... —125 Thermal. aioseca sacs ra greck aces — 130 Dos Palmas............... - 3 Mecca neti xv haainccic oes ates —190 Travertine Rock (base)..... —150 AD UMA pape xsccena obese soo ene ee — 200 Meloland nic. oso es tees — 50 Travertine Terraces........ — 190 MOVEMENTS OF VEGETATION DUE TO SUBMERSION AND DESIC- CATION OF LAND AREAS IN THE SALTON SINK. By D. T. MacDovaat. DESERT BASINS AND REVEGETATION. Extensive basins occur in various parts of North America, Australia, Africa, and Asia, in which the climatic and hydrographic conditions are such that the lowermost parts of the depressions are at times occupied by bodies of water which may disappear or show wide fluctuations of level or volume. Alternations of this kind are followed, of course, by the annihilation of the terrestrial species, as lakes are formed or as they increase, and by the revegetation of emersed areas laid bare by receding waters. It is fairly evident that occurrences of this kind have taken place in the great basins of Nevada and Utah, in the Oteri and other depressions in New Mexico and Arizona, in the Pattie Basin in Mexico, and in the Cahuilla Basin in southern California.!_ In some the alternations date far back in geologic time and many thousands of years must have elapsed since the last change occurred. In others, such as the Pattie and Cahuilla Basins, the transformations follow each other rapidly, and although they may have begun far back in time, yet they continue up to the present. The evidences of differences in level at which bodies of water have stood in the lower- most parts of inclosed drainage systems during historic times are indisputable, and con- sideration of the facts obtained from the study of these basins and of the records of long- lived trees leads to the conclusion that the variations in the level of some lakes without outlets in North America were connected in a direct way with variations in major climatic factors. Such variations would have great influence upon the movements of plants about the world, but would lie beyond the scope of this discussion.? The chief biological interest of the present paper centers in the fate of organisms over- whelmed by floods, in the physical changes which follow emersion, and in the biological mechanism of reoccupation of sterilized areas as they emerge from the water, episodes which must have been duplicated in their main features innumerable times in the history of the surface of the earth. Many of the important features of distribution of plants and animals at the present time may be due to antecedent facts of this character. Opportunities are not common for studies of this kind and for analysis of the means and manner by which living things move onto a sterilized area or into a sterilized medium. Krakatau, an island in the Indian Ocean, was devastated by a volcanic eruption in 1883 which destroyed practically all of the higher plants growing upon it. The place was visited by botanists in 1886, 1897, and 1906, and brief analytical studies were made of the establishment and constitution of the flora at these widely separated dates. Little attention could be paid, however, to the successions or to the progress of physical condi- tions upon which the dissemination and establishment of seed-plants depended.® ‘Gilbert, G. K., Lake Bonneville. Misc. Doc. House of Representatives, first session, Fifty-first Congress, 1889-90. Russell, I. C., Geological History of Lake Lahontan. Misc. Doc. House of Representatives, first session, Forty-ninth Congress, 1885-86. . Gregory, J. W., The dead heart of Australia, London, 1906. _ . 2 See Huntington, E., The fluctuating climate of North America. The Geographical Journal, vol. xL, pp. 264-280, 392-411, 1912. 3 Ernst, A., The new flora of the volcanic island of Krakatau, translated by A. C. Seward, 1908. Campbell, D. H., The new flora of Krakatau. Amer. Naturalist, vol. xii, August 1909. 115 116 THE SALTON SBA. The effects of voleanism have furnished other opportunities for observation upon the invasion of sterilized areas on the surfaces of cooled lava flows. The physical condi- tions of the substratum are highly specialized under these circumstances, and moreover it has been found difficult to make any definite analysis of the determining factors in the movements of the vegetation appearing in such places, although some extremely interesting records of these phenomena have been made. The most important and most recent work upon this subject was done by C. N. Forbes, in the Hawaiian Islands." (See Note below.) Newly made lands about the mouths of rivers generally furnish a substratum highly favorable to the growth of a number of species of higher plants as soon as the deposition reaches a stage where the surface is brought permanently above the level of the tide, and the conditions are very favorable for observation of the successions.? Such occupations of bare soil by open formations and later by closed ones have also been described by Flahault and Combes.? Newly made alluvial was first occupied by Salicornia macrostachya, fol- lowed by Salicornia sarmentosa, Atriplex portulacoides, and Dactylis sarmentosa. As may be seen from the contents of the present paper, the procedure on the emersed lands of the Salton is widely different from that in any of the above cases. THE SALTON SINK. The Cahuilla Basin lies in the most arid part of North America, and although the making of the lake in the Salton Sink or portion of the basin below sea-level may be finally due to climatic factors, yet it is caused directly by overflow from the channel of the Colorado River. The geological record seems to indicate that the Sink has been filled and dried out at intervals over a long period extending up to the present time. Circumstantial evidence points to the conclusion that hardly twenty years have passed without some inflow from the river into the Sink, but the only available actual record of any other inundation besides the one the effects of which are considered is that of the overflow of 1891, when the passage from the river to the small lake formed in the Sink was made by a single man in a boat on the inflowing current for the purpose of ascertaining the source of the water forming the lake. (See p. 19.) (Plate 15.) The formation of the lake in 1905, 1906, and 1907 occurred under circumstances that gave unexcelled opportunities for a study of the attendant phenomena. A number of naturalists had visited the Sink in the half century following the Williamson expedition through it (1853) and something of the nature of the surface formations as well as of the plants and animals had become known. Furthermore, a short visit to the bottom of the basin near the railway station of Salton had been made by Messrs F. V. Coville and D. T. MacDougal in 1908, and a brief description was published as to the aspect of this part of the Colorado Desert, including an area to the northwestward. Photographs of parts of the Sink which were afterwards submerged were also obtained. 4 In 1904, an expedition from the New York Botanical Garden, in which Dr. D. T. MacDougal, Mr. G. Sykes, and Prof. R. H. Forbes participated, started from Yuma in February and went down the river through the Delta and to the western shore of the Gulf of California.’ One camp was made a short distance to the southward of the boundary between the United States and Mexico, near an intake of an irrigation canal leading to an area in the basin being brought under cultivation, and brush dams were seen extending out into the stream for the purpose of forcing the current into these artificial channels. a ss : : 2 Oliver. W, The Bouche d'isrquy, New Phytolecee cok ooo aaa ead peu: ote Vol. v, p24, 1912, 5 Sur la flore de la Camargue et des alluvions du Rhone, Bull. d. 1. Soc. Botan. d. France, vol. xr, 1894. 4 See Coville and MacDougal, The Desert Botanical Laborat i ituti 2! icati Wee Gan ee. Wat, Nomi toe aboratory of the Carnegie Institution, pp. 20-22. Publication 5 eile ee D.T., Botanical explorations in the Southeast. Journal of N. Y. Bot. Garden, vol. v, p. 89, 1904. Nots.—A copy of a paper by Mr. W. N. Sands giving “An account of the retur i : agriculture in the area devastated by the Soufriere of St. Vincent in 1902-3 Ke ecuador 33, 1912) has been received since this paper was put into t i i r i oe, Late) eae a pap p o type. Some interesting facts as to revegetation of de- PLATE IS fo SY, yr Hype IN INS pe ‘ne SS ee = 2 Rosman -—%, ae ee AANA SO Sask] —— * oe Mri Za iN ‘S Oo had * Pre sean Tee - s. Pa “2 ~— ee | M et * ‘S190 —> I) wit ieee oe Sst 1a Se & See eae A ee a ¢ o- FYOuHS do 5 Y We gli” * BS \F NX N., SY n = oN MTNS 0 BS al ys Oo u zy? eu zane Pe Nc K PY, ey Meeeea REP” Xe Ue “ee K Wy SD q WZ Y SVE OTISSSS | : willy 3 SS = NW Wid ee wy = W8 ‘Muggs w aul Aw? i ee o- < Ye Ces oe 7a ) or ee = = o¥ {lft § - 5. °9 om BO = 9, &? ly av oSy, 15 Miles MAP OF THE SALTON SINK ComPILED BY G. Sykes 10 °| MOVEMENTS OF VEGETATION IN THE SALTON SINK. 117 A year later a second expedition, starting 300 miles up the Colorado, was made down ae the Delta, returning to the northward along the western margin of the alluvial ands. The southernmost of the canal intakes had been eroded deeper and wider by the inflow- ing current and a stream large enough to float a steamer flowed through it to the westward, spreading out over the alluvial lands, but finally collecting in two main channels known as New River and Alamo River leading down into the Sink. New River was crossed at Calexico in small boats, and it was at this time simply a muddy stream a few feet in depth, but the water was nearly level with the banks. Later, as is noted elsewhere in this volume, a great amount of erosion occurred which excavated a deep and wide channel. Returning to the main line of the Southern Pacific Railway the Alamo was crossed, but no special observations were made on its flow. A party from the Desert Laboratory proceeded southward from Mecca to Travertine Rock in May 1906, and made an examination of the shore of the lake, which at this time had attained a maximum depth of 34 feet, with a total area of nearly 300 square miles. The party included Professor W. P. Blake, who made the trip to verify his observations of 1853 upon the travertine formations which had been observed by him fifty-three years earlier. (See Plate 18.) Samples of the water were taken (see page 35) and the shore phe- nomena were observed. The scope of the opportunities for research upon several geographi- cal problems was recognized and comprehensive plans for dealing with the entire matter were formulated during the year.? The inflow of the water proceeded, but early in 1907 the engineering operations of the Southern Pacific Railway promised to close the breaks and place the inflow under control, and a small sailboat was constructed at Mecca early in January for purposes of exploration. A party with full field equipment embarked in this boat for a trip around the shores of the lake and to the islands on February 7, and made daily camps ashore during the following ten days. The highest level of the lake with a maximum depth of 84 feet was reached on February 10, at which time the party was camped near the southern end among the great crescentic dunes of the Carrizo sands. During all of this voyage landings were made with the greatest difficulty, as the steady rise of the lake had simply pushed a sheet of water out over the loose alkaline soil of the desert and made a soft mud into which one would sink to a depth of 2 or 3 feet in a few seconds, and the saturated soil would be moist and soft for some yards away from the water. Long tongues of water extended far up the channels of the dry washes and no true beaches or strands had yet been formed by the sorting or solidification of the soil material. (See Plate 12 c.) So rapidly had the level risen that several of the great crescentic dunes had been surrounded by water and their long ridges still showed above the water like curved whale-backs. (See Plate 128.) Thou- sands of rounded masses of pumice, from 1 to 6 inches in diameter, floated on the surface, as well as many of the hard-shelled globular fruits of Cucurbita palmata. The water at this time contained about 0.25 per cent of dissolved salts, which is near the limit of pota- bility. This was denoted by the fact that it could be used by some members of the party, but not by others. Four hills in the southeastern part of the basin stood above the water as islands. The southernmost was known as Big Island and rose in conical form to a height of over 100 feet and showed the terraces of previous lake formations far above the recent level. (Plate 6 p.) Its base was fringed by lower flattish hills on the southward, separated from it by narrow chan- nels, and the recession of the waters rapidly increased the exposed area, while the top of a hill to the northward came above the water in 1908. (Plate6p.) Rock Island was the crest of a 1 MacDougal, D. T., Botanical explorations in Arizona, Sonora, and Baja California. Journal of N. Y. Bot. Garden, vol. vi, p. 91, 1905. ? Report of Department of Botanical Research for 1906, Year Book Carn. Inst. Wash., No. 5, 1906. 118 THE SALTON SBA. hill which bore no seed-plants by reason of lack of soil and rose 40 or 50 feet above the water. No landings were made on it. Obsidian island was of dumb-bell shape nearly a mile in its main axis, which lay nearly north and south. The shores of the bays on the eastern and western sides were of sand and alkaline clays and this formation extended southward on the eastern side. These conditions, together with the fact that the island lay in the drift from the Alamo inflow, cooperated to yield some very interesting features. Cormorant Island, 2 miles to the northward, was awash during the period of maximum depth of the lake and was completely sterilized. It is to be noted that the names given to these islands were applied solely for the convenience of the workers of the Desert Laboratory and as they were not formally published have not been generally recognized. Other names for the same formations have probably been used by visitors to the lake, but so far as the author is aware none have been formally presented, and as a matter of course the sub- sidence of the lake will restore the elevations to their status as hills, which may or may not have received definite appellations (Plate 16 B). The islands bore a sparse vegetation, including four species of Atriplex. Lizards and small rodents were found on Big Island and Obsidian Island. A coyote had survived for a short time on Obsidian Island. The amount of food on Big Island was such that in Feb- ruary 1907, a year or more after this elevation had been surrounded by water, rabbits and a raccoon were still alive, as indicated by their tracks and burrows. The isolation of the tops of these arid hills as islands by the waters of the lake made them attractive to aquatic birds, and Obsidian Island became the site of an extensive nesting-place for pelicans as early as 1907, while some cormorants made nests here and also on the top of the smaller island to the northward, which was barely awash at high water and was named Cormorant Island. Representatives of the fishes of the Colorado River had come into the Sea, but did not seem to multiply, with the exception of the introduced carp, which afforded a food supply of some importance to these birds. Nothing definite may be said about the move- ments of other flying species, but it is well known that those mentioned above make long flights southward through the Delta to the Gulf of California and along the course of the Colorado. The possibility of seed-introduction by these birds awakened the liveliest inter- est, although it was probable that the plants which might thus be carried to the islands were likewise the ones which might be most readily borne there by flotation. In any analysis of possible introductions, therefore, the positions of the established plants with regard to the water-level would need careful consideration. LAGUNA MAQUATA. After a brief survey of the lake an excursion two weeks in length was made to another sunken basin to the southward, which lies to the westward of the Cocopah Mountains in Baja California, a province of the Republic of Mexico. The Sink of this basin had been seen to be receiving an inflow from the Colorado River around the southern end of the mountains in April 1905, but nothing was known as to how long this continued. It may be assumed, however, that it ceased with the passing of the summer flood of the Colorado in June of that year, since the water in March 1907 was 12 to 15 feet below the maximum level, which loss or recession could be attributed chiefly to evaporation, which for adjacent regions has been calculated as over 100 inches per year. The body of water formed here has been variously known as the Laguna Salada and Laguna Maquata. The lake was known to be in existence in 1884, was represented by a chain of saline pools in 1890, was seen in much the same condition in 1892, and filled again in 1893. Exact records of the cycle of the lake are not available, but it is known to have become entirely dry at various times up to 1905, when it was seen to be filling as noted above. The rate of evaporation would probably result in the desiccation of the lake in 1909, but early in April 1910 it was again reported to be filling in connection with serious disturbances in the lower course of MOVEMENTS OF VEGETATION IN THE SALTON SINK. 119 the Colorado River. The main current of the Colorado no longer followed the channel which had carried the main flow for a long period and now made a more southwesterly course down the Bee River, which brought its waters directly against the southern end of the Cocopah Mountains. It appears, therefore, that water was poured into this basin in such manner that the lake was maintained continuously from 1910 until late in 1912, at which time this manuscript was prepared. The maximum-level strand was marked by a heavy band of Prosopis, where it cut across bajadas or detrital slopes, and by a distinct water-line along the rocky mountain wall of the Cocopahs. The high beach of 1905 was marked by a zone of Sesuviwm sessile and by the remains of other halophytes, which had perished with the desiccation of the salty soil. This preliminary visit showed the importance of carrying on observations in this basin concomitantly with those to be made in the Salton Sink, but the internal disturbances of Mexico were followed by military operations at Calexico and along the northern margin of the Pattie Basin, which made field work impossible south of the international boundary. PLAN OF FIELD WORK IN THE SALTON SINK. An ideally perfect method of investigation would have been to make frequent exami- nations of the entire shore-line of the lake, but as this was over 150 miles in length and as parts of it lay below deserts accessible with difficulty and was fringed with mud flats, this was practically impossible. Attention was therefore devoted to areas, a mile or less in width, back from the shore, which extended down the slopes with the recession of the water. One such area was located at the extreme northwestern end of the lake near Mecca. The soil consisted of a clayey alluvium, uneven as to saline content and detail of surface, but in general showing a slope in which the fall was about 1 in 400, giving many effects of a flat surface. The moister places maintained Typha, Populus, Salix, and Phrag- mites; the saline spots bore such halophytes as Atriplex, Sueda, and Spirostachys, while mats of Distichlis were found scattered over the area. A second area was selected on the shore to the southwestward of Imperial Junction, ‘ on a clayey slope in which the adobe soil was high in salt content and bore scattering plants of Atriplex, Sueda, and Spirostachys. The slope here was slightly steeper than the one at Mecca, the fall being about 1 in 300. The steep detrital slope running down from the mountains to the southward of Traver- tine Rock on the southwestern side of the lake bore a xerophytic vegetation, among which were a few halophytes, such as Afriplex; the remainder were chiefly spinose and sclero- phyllous forms. Bands or rows of these plants indicated ancient strands, some as high as the maximum level of the ancient lake. (See Plate 18.) The slope here was about 1 in 20. On account of these old terracing effects, which were duplicated by the present lake, this place was designated ‘‘The Travertine Terraces.” The shore-line of Obsidian Island, which lies to the westward from the Imperial Junc- tion beach, about 8 miles from the recent high-water level, was examined over its entire length of sand and gravel strands and rocky shores. The upper parts of the island bore four species of Atriplex, some Spirostachys, and one or two other halophytes. In places the beaches showed a fall of 1 in 7 or 8. A few examinations were made at various places on the western shores of the lake, which lay across the lower parts of long bajadas or detrital slopes coming down from the distant mountains. Some sand dunes were present on these slopes and the vegetation included a variety of halophytes and sclerophylls. Sterilized islands, in the form of hill-tops emerging from the receding water, came in for very detailed attention. The surface of Cormorant Island, north of Obsidian Island, was not submerged, but was sterilized by the salty water at the high level. Various smaller 120 THE SALTON SEA. islands in the vicinity of Big Island in the southern part of the lake were taken into account as they came into view and the character of the surfaces was noted in every case. During expeditions to these places once or twice yearly, collections of plants were made, and notes and photographs were taken of the formations encountered. A large sample of water was taken annually in June from the surface of the deepest part of the lake in order to follow the changing composition of the dissolved salts, and also from other locations and depths for comparison. A chemical and physical examination of the soils on the bared parts of contiguous areas was made, together with every effort to determine the factors cooperative in influencing the behavior of vegetation as it followed the receding lake. In addition to the sail-boat built in 1907, a steel-hulled power launch, built in the shop of the Department at Tucson, and a small sectional steel rowboat were used in travers- ing the lake. Land trips back from the lake and around its southern end were made by camp wagons and field equipment. Some use was made of motors, despite the deep arroyos and great stretches of soft sand encountered. A traverse from the extreme northwestern part of the basin by motor to the northeastern part was made in September 1910, when Messrs. Sykes and MacDougal carried a line of observation on the drainage and surface conditions from the San Gorgonio Pass down to the lake at, Mecca, thence by way of Dos Palmas up the Chuckawalla Wash to the Chuckawalla Divide, and down eastwardly to the Colorado River, a distance of 150 miles. Messrs. MacDougal and Sykes came through the San Gorgonio Pass and went down the drainage of the Whitewater to the northwestern end of the lake in June 1912. Here the party was joined by Professor Brannon and a route was followed along the western and southwestern sides of the lake and across the inflowing channel of the New River. From here a line was run to the western margin of the basin and across the range by Mountain Springs to San Diego. Other trips were made from time to time by the collaborators in this volume and the results of their field work were made available for all those engaged in the work. FACTS TO BE ASCERTAINED. The energy of the entire staff of collaborators was concentrated in an effort to obtain facts which might bear upon the more important problems presented by the receding lake, which for convenience may be arranged in the following topical form: 1, voppoon and nature of the flora of the basin, with chief attention to the species inhabiting e Sink. . Influence of the lake upon vegetation above the flood-level, either in increased humidity or hemmed underflow. . Endurance and survival of the vegetation in the shallower marginal portions of the flooded area. . Geographical relations of the Sink, with especial consideration of the contributing drainage. - Physical and chemical analyses of soils of the Sink, with comparisons between those unaffected by the recent submergence and those taken from the bared strands. . Composition of the water of the lake as varying with its concentration. : ae ene bacterial flora of the lake, and influence of these plants upon the composition of the water. . Alterations in plant tissues induced by submergence. - Reoccupation of the bared strands left by the receding lake by plants. ; Alterations or successions in the plant inhabitants of the strands with increasing aridity. - Environic response of plants gaining a foothold on the strand and later becoming subject to desiccation. . Pioneer occupants of sterilized islands emerging from the water as a result of lowered level. : ee stetvs In carrying seeds, spores, and propagula to bared strands and isolated areas slands. . Introductions, or invasions of the Sink by species not hitherto native to the region. et m © 6 CO NO Ot & no aed — lao WwW bo It need hardly be said that the evidence obtained on some parts of this comprehensive program was very meager, and that some features, notably those of the animal inhabitants SALTON SEA PLATE 16 A. View northward parallel to the shore from middle of Emersion of 1907, Imperial Junction Beach, taken February 1908. Atriplex, Sueda, and Spirostachys zone in middle. Denser Vegetation on right is above Flood Level, and Boat is moored near Shore on left. B. View from Obsidian Island toward the southwest, showing West Bay of that Island; Rock Island in middle distance and Big Island to the right MOVEMENTS OF VEGETATION IN THE SALTON SINK. 121 and plankton, were not seriously touched upon, except in so far as the work of Dr. Brannon upon the bacteria contributes to the matter. A number of the above subjects are considered in full in this section, while others are treated in separate chapters, as indicated in the table of contents. The chief features of the phenomena of revegetation of the beaches may be best pre- sented by a reduction of the field notes to a history of the strands or zones emersed during each year of the recession of the lake. Any intelligent consideration of such descriptions must be made in the light of the following conditions affecting the strands: 1. The observational areas selected represent widely different habitats as to soil composition and other environic factors or components. 2. The lake rose quickly to its maximum level and receded rapidly. 3. The infiltration and leaching of the soil varied year by year as affected by the concentration of the water on one hand and the time of submergence on the other. 4, The salt content of the water was least during 1907 and increased about 18 to 20 per cent in each succeeding year. 5. Every emersed strand would therefore be saturated with a soil-solution resulting from the infiltration of the lake water of the concentration and composition prevalent in the period preceding emergence. . The desiccation of the emersed strands would proceed at a rate determined by the character of the soil and by the composition of the infiltrated water. 7. The rising water of the lake picked up seeds lying on the surface, and their survival constituted a means of revegetation, chiefly of the strand bared in 1907. 8. The rates of evaporation and of recession of the lake varied with the season, being most rapid in June to August and slowest in December and January. The possible total may be estimated at 116 inches per year. 9. The rainfall data of the Sink, obtained from the records of the U.S. Weather Bureau made at Indio, which is located near the extreme northwestern end of the Sink, only a few feet below sea-level, show an average of 2.74 inches per year. 10. Rapid recession of the water would result in separating stranded seeds quickly from the margin of the water, with consequent rapid desiccation of the surface layer of soil, which would be unfavorable to germination and survival. 11. The shallow water lying on wide mud flats fringing the shores was raised to a much higher tem- perature (15° to 20° F.) than the body of the lake during even the winter season, thereby greatly increasing its toxicity for seeds, plantlets, and propagating bodies. The greater number of the seeds falling into the lake would be subjected to this action. The muddy flats fringing the shores at all stages of the lake must, therefore, be considered as a barrier of some magnitude which would be crossed by a plant carried out into the lake and again when de- posited on a beach. (See also p. 140.) o>) REOCCUPATION OF THE STRANDS OF 1907. The level of the lake reached its highest point about February 10, 1907. A record quoted, by the writer (Desert Basins of the Colorado Delta, Bull. Amer. Geog. Soc., December 1907) is to the effect that the rise in the water continued until in March, but the numerous observations on the beaches late in February showed that a fall had taken place beyond doubt and the record for March may be assumed as an instrumental error. The immediate and rapid rate of fall of the water-level consequent upon the diminution of the inflow which was brought about on February 10 may be ascribed to the enormous evaporation and infiltration. This was sufficient to lower the level by a foot at a date early in June 1907, and in one year from the time of cessation of the rise the fall was between 40 and 42 inches. The record of the change in level of the lake and the varying area covered are shown by figure 3, page 138, reproduced from Mr. Cory’s records (see fig. 31, Trans. Amer. Soc. Civil Engineers, vol. xxxviu1, 1912). The slope lying above the observational area at Imperial Junction beach showed a surface of adobe containing such an amount of dissolved salts that it bore a distinctly 122 THE SALTON SEA. halophytic vegetation, inclusive of Spirostachys, Sueda, Atriplex, and Franseria, while Olneya and Larrea were represented by a few individuals back some distance from the shore. A few drainage channels or dry washes cut to a depth of a yard or less ran directly down the slope and offered special conditions when submerged. The water rose rapidly to the maximum level, and the extreme upper strip covered was submerged for a few days, less than a week, even when wave action is taken into account, and then the recession began as illustrated by the curve in figure 3. The bared zone had a width of about 1,000 feet in February 1908, and the average rate of horizontal recession was therefore about a yard daily. Much less than this would be laid bare during the winter months, while during the season of maximum evaporation in the summer as much as 1.5 yards must have been uncovered daily. The first critical examination of the Imperial Junction area was made in February 1908, or almost exactly a year after the recession began. The extreme high-water level was marked by a ragged band of timbers, drifted wood, and rounded masses of pumice for many miles along this shore, which was exposed to almost direct action of storms coming from the northwestward through the San Gorgonio Pass. The abruptly walled channels of the shallow washes which had been submerged were filled with mud by the action of the waves, and when these deposits were uncovered offered slightly different soil condi- tions from the contiguous slopes upon which there had been very little sorting of material (see Plate 12 c). No strand or beach ridges were visible and perhaps the only actual changes were those in the washes just noted and also in the salt and moisture content. The water of the lake when at the high level did not contain over 0.3 per cent of dissolved salts, but by January or February 1908 it had increased to about 0.4 per cent, although no estimation was made until June, at which time the total dissolved salts amounted to 0.46 per cent. This amount was probably much less than that present in the soil of the observational area, in consequence of which a leaching action must have ensued which resulted in fresh- ening the soil to some slight depth. (See Free, pp. 30-33.) The uppermost portion of the strand, or that part of it which must have been flooded by the advancing water in December 1906 and January 1907, and laid bare before May 1907, was occupied by a dense formation of Sueda torreyana, Amaranthus palmeri, Atriplex fasciculata, A. canescens, A. linearis, and Spirostachys occidentalis. The last-named extended below the limits of the others to a point probably representing the stage of the water in June 1907. Immediately below the Spirostachys was a dense band of Pluchea sericea, the individuals of which were from a few inches to a yard in height. Seeds of Sueda, Atriplex, and Spirostachys probably were present in the surface layers of the soil at the time of the inundation, and as the zonation was repeated on the corre- sponding seasonal recession of succeeding years the influence of germination conditions is suggested as the determining factor in all of them. The compositaceous Pluchea did not inhabit this area originally; its crops of seeds are ripened in early summer and are blown about by the wind in enormous numbers. The probabilities are great that some seeds were carried to lodgments on the muddy shore directly by air-currents. Some may have fallen in the water and were then cast ashore by wave-action. Any remaining afloat would be subject to the toxic action of the water, toward which the various species showed specific endurance (see Plate 16 a). A wide zone of hard bare dry soil below the zone of Pluchea was found in February 1908, which was taken to be the strip laid bare in midsummer, during which time the horizontal recession would be 4 or 5 feet daily. This action, coupled with the high summer temperatures, would result in such rapid desiccation of the surface of the soil that deposited seeds would not be under suitable conditions for germination very many days. The actual rate of evaporation and concomitantly that of recession is suggested by the figures given in tables 32 and 33, page 134, which have been plotted f the d iven i i ; plotted from the data given in the section MOVEMENTS OF VEGETATION IN THE SALTON SINK. 123 of this volume by Professor Blake. The relative rates of evaporation by months are: Jan- uary 10, February 10, March 18, April 20, May 26, June 34, July 30, August 28, September 23, October 21, November 12, December 9 (see also p. 134 for evaporation of 1909-10). That the bareness of the soil laid dry in midsummer was due to desiccation and not to dearth of available seeds was made evident by the fact that below this arid belt was a second zone of Atriplex which appeared to occupy a strip uncovered during the autumnal months. It is a notable fact, however, that the formation of two definite bands or zones did not occur in any succeeding year, although there were a few late-summer germinations even as late as 1912, at which time the original invasions of 1906 and 1907 were still recog- nizable. The pelicans and cormorants were frequently observed standing on the shore near the margin of the water, during the middle of the summer especially, and there seems to be greater probability of their agency in bringing seeds to the summer zones than to any other formations examined around the lake. Furthermore, the species represented were those which might most readily have adhered to the legs and feathers of these birds. The greater number of plants invading the strands were carried there in the form of seeds by the wind, by flotation, or by birds, as will be discussed somewhat fully in a later section of this paper, but the emersion of 1907 was characterized by some reoccupations of a different character. The waters of the lake rose rapidly to the maximum level in Feb- ruary 1907 and then fell quickly. In many places not particularly subjected to violent wave-action shrubs and small herbaceous plants stood erect in place, although the water about their bases was perhaps a foot in depth. If, as in the case of Atriplex and Sucda, the branches bore mature fruits, many of these might be retained until after the soil about the base of the plant had been bared, when their dispersal by the simple mechanics of the plant would result in some patterns of reseeding independent of other agencies. The rapid recession of the water down the gentler slopes was not accompanied by any sorting of material in lines parallel to the periphery of the lake at Mecca or on the Imperial Junction Beach. The only specialization occurring was that by which suspended silt was dropped in the channels of the washes formed previous to the inundation. The recession of the water left these as moist strips of mud a few feet or a few inches in depth, which furnished conditions for Heliotropium, Lepidium, Typha, Rumex, and Leptochloa in a loose formation extending radially to the lake. The salt-grass or Distichlis had formed a few mats on the strand laid early in 1907, but these plants were already showing the strain of desiccation, and Leptochloa had also come in at a level corresponding to the autumnal band of Sueda. A number of hard- shelled fruits of Cucurbita palmata had been cast ashore at various places on the slope, but no germinations had been accomplished. The observational area near Mecca was upon a slope in which the fall was not more than 1 in 400, and as the surface was uneven, the vegetational reactions were not so easily interpreted. In fact, the limit at which the pitch of a slope may be sufficient to produce banding or zonation of the plants upon the given rate of evaporation may be taken to lie between that of the Mecea and that of Imperial Junction beach. The circumnavigation of the lake in February 1907 was begun at a point on this area, a boat being brought up the shallow channel made by the advance of the water up a sunken road. The actual margin of the water at this time was an exceedingly crooked line and the high level of the water might not be accurately laid down except at a few points. The second visit to the place in February 1908 showed that in addition to the survival of a few individuals of Atriplex and Sueda within the inundated area, a few trees of Salix and Populus also endured submergence of their roots for several months. A great number of spreading trees of Prosopis pubescens had been killed and the wood of the trunks and branches had become exceedingly brittle. The total width of the emersed strip here was about 1,400 feet. In addition to the moisture effects from saturation, an uneven underflow 124 THE SALTON SEA. was present, since this place lies partly on the detritus of the drainage of the Whitewater River coming in from the northwestward. The crop of seeds from plants of Sueda and Atriplex, living and dead, which had maintained an upright position, were sown in irregular patches by the wind (see Plate 28 B). Oligomeris glaucescens, Chenopodium murale, and a few young plants of some unrecog- nizable species were found on the upper part of the area in the region probably laid bare during the season of February to May. Heliotropium was scattered over the entire area, the older ones, near the high-water line, being in bloom. It was notable that Pluchea sericea and Distichlis, both of which were abundant above the high-water line, had not made their way down the gentle slope, nor had the receding waters deposited any seeds of these plants in such manner as to make possible their successful germination. The results at this place in the following years were vitiated somewhat by operations in locating homesteads and reclamation work, but observations were made from time to time, as noted below. The extension of the water up the bed of a wide dry wash on the western side of the lake in 1907 offered special conditions and it was hoped that some observations of especial value might be made here. The underflow, however, from the extensive and steep moun- tain slopes only a few miles away was heavy and the recession of the water down the slope of the bed was accompanied by the breaking out of numerous seeps and springs above the water-level, which gave confused effects not capable of analysis. It is to be noted, however, that Typha found lodgment here, and a single Rumex grew from seed, while Heliotropium had become established and was in bloom in February 1908. (Plate 17.) The banks of the wash rose perpendicularly to a height of 4 to 7 feet and the ground near-by bore a heavy formation of Sueda, Atriplex, Spirostachys, and Pluchea sericea. The damming of the underflow had also resulted in an increase of the available soil-moisture on either side of the wash, with the result that the species named flourished and showed a marked acceleration in development. These might easily have been mistaken for humid- ity effects (see pp. 139, 140). Southward from the wash an arched slope of detrital material ran steeply down below the high-water level and the waves were cutting the crest of the arch at the high level in 1907. A year later this cutting was seen to have resulted in a bank as steep as the nature of the material would allow and about 12 feet at its maximum height. Masses of dirt had already begun to slide down from the crumbling upper margin, carrying whole plants and seeds from the crest in February 1908. Among these were Coldenia, Pluchea sericea, and Atri- plex canescens. All of the mature plants perished, as was found on the next visit, but in February 1908 the Atriplex had brought down seeds which had germinated, forming a row of plantlets 6 to 8 inches in height (see Plate 18). The recession of the water had resulted in laying bare a terrace about 60 feet in width. A number of small ‘tide-pools” had been formed near the base of the slope or cut bank and on the margin of one of these was an active rhizome of Phragmites, which had been floated to the place. A second one was seen, but in an inactive condition. This and the plantlets of Atriplex mentioned above constituted the only invaders of the strand at this place. The steep slope of the original surface here offered the best opportunities for the analysis of the effects of rapid and slow recession of the water. The low rate of recession in December, January, and February kept the water at nearly the same level, so that the action of the waves resulted in a thorough sorting of the material, and this was heightened by the winter storms. In consequence the midwinter level of the lake was marked by a small beach ridge or by a cut bank which might have a vertical height of over a foot, as illustrated in Plate 25 8. The interval between this and the ridge or cut bank of the fol- lowing winter was taken as the emersed zone of the intervening year. The bases of these ridges or banks offered favorable conditions for the deposition of floating seeds, and the SALTON SEA PLATE 17 A. Sailboat on flood-water of Lake in Travertine Wash, February 1907. Tree surrounded by water is Parosela spinosa. B. Same view, taken February 1908, showing Seepage and Dead Trees MOVEMENTS OF VEGETATION IN THE SALTON SINK. 125 moisture was sufficient to produce germination, with the result that narrow and well- defined ranks of plants were formed in these angles in the succeeding years of the observa- tions. The ancient beach-lines higher up the slopes were taken to have been produced by prolonged maintenance of the water at about the same general level, as the amount of sorting which had been done would not occur within a single season. (Plates 29 a, 30 B.) The recession of the water during the summer months was checked in some seasons by rains, and the consequent maintenance of level, together with storm action, produced another series of beach ridges and cut banks which varied greatly not only in their height and general character, but also in the composition of the rank of vegetation, as the seeds which were being disseminated at this season were of different species in part from those being carried about by winds and waves during the cooler season. It is to be noted that the terraces lay across the crest of a detrital ridge or bajada and that it was thus barred from the liability of invasions by run-off streams from the slopes above, the only introductions from landward possible being those which might fall by gravity from the extreme edge of the crumbling bank. It was therefore determined to make an examination of the slope to the southward, which comes down from a greater distance and at a much gentler gradient from the same range of mountains. Here the recession of 1907 had laid bare a gravelly shore nearly 900 feet in width, with a fall of about 1 in 250. The first glance at this area showed numerous invaders that had come down the run- off streams, which had been strong during the midsummer of 1907. Plantlets of Franseria dumosa, Encelia eriocephala, and an unrecognizable species were found, as well as mature individuals of Wislizenia refracta and Oenothera scapoidea aurantiaca, while Cryptanthe barbigera and Heliotropium curassavicum were abundant and in bloom. Young shoots of Pluchea sericea were also found and this plant and Heliotropium were the only ones which had not arrived by flotation on the run-off streams of the slopes. The results from this place and the terraces previously examined gave in marked relief the relative importance of flotation from the body of the lake and down the run-off streams on the slopes. Nowhere else were the conditions so favorable to invasion by the latter method. Obsidian Island, which lay westward from the Imperial Junction beach about 5 miles, had been selected as offering shore phenomena in good condition for analysis, and this place was reached on the leg of the voyage from the western shore in February 1908. The hill-tops making this island consist of two crests joined by a low neck which was covered, with the exception of a portion about 100 yards in width, by the high water of 1907. The act- ual beach line measured 2 miles as paced with a pedometer in February 1908. The shore was rocky and precipitous in places, but about half of the total was sandy, with the admixture of alluvium in East Bay and of coarse gravel at the southern end and in West Bay. (Plate 21.) A large number of species was found on the beaches at this time and the collections were as follows: Cryptanthe barbigera on upper portions of emersed beach and also above flood-level. Hriogonum thomassii, distributed above and below flood-level, was possibly brought in by the wind or by birds, the soil offering suitable conditions near the water. Psathyrotes ramosissima was found at extreme high-water level, as if the seeds had come in on the waters of the lake. Oenothera scapoidea aurantiaca occurred on the upper level of emersed band; Rumex midway of the beach. Sonchus asper, represented by three indi- viduals in separate places, may have come from seeds floated ashore or may have been carried there directly by the wind. Encelia eriocephala was represented by a single speci- men, which probably reached the beach by the flotation of a seed; it was abundant at Imperial Junction, 8 miles distant and 3 miles back from the shore to the northeastward, but may have come from some other source. Atriplex canescens was found in a tide pool; A. polycarpa was represented by a few plantlets on the beaches. A. hymenelytra and A. fasciculata, with the two preceding species, are indigenous to the island and had come down 126 THE SALTON SEA. on the beaches. Heliotropium curassavicum was abundant in many places on the strand. Spheralcea orcuttii was represented by a single individual, part of which was preserved. Lippia nudiflora occupied a position immediately below the high flood level. Prosopis pubescens, two individuals, and one P. glandulosa arose from a living branch which had been carried to the southeastern point of the island and cast ashore, after which it had been half buried in the sand. The entire southeastern shore of the island was exposed to the direct action of the water which was being emptied into the lake by the Alamo and New Rivers. The feeble current of these streams would be an agency of slight efficiency in carrying plants across the intervening 7 or 8 miles of open water, but once branches or seeds were afloat on the waters of the lake, wave-action might bring many of them ashore on this island. Oligomerts glaucescens was the most abundant of the invaders. Parosela emoryt was represented by a few plantlets on the sorted beach ridge on the shore of West Bay. As will be shown later, this is a species which appears on gravelly shores immediately following the recession of the water and apparently maintains its place in the rank during hundreds of years of aridity. Typha was found at the high flood-level on the eastern shore, evidently arising from a rhizome. Baccharis glutinosa had found lodgment in the mud of West Bay, apparently from floating seeds, and Leptochloa imbricata was established near the high level of the flood on West Bay. The examination of this place in February 1908 completed the initial examination of the observation areas and furnished a census of the pioneer introductions. The fate of the original introductions, together with the accession of additional species, now constituted a subject which aroused keen interest on the part of the workers engaged. The next visit to Obsidian Island was made on May 8, or only three months later. The multiplication of the fish which had come into the lake with the water from the Colo- rado River had made a supply of food for aquatic birds, while the elevated arid summits of this and other islands uninhabited by any carnivorous animals had attracted large numbers of pelicans and cormorants which had occupied areas of several acres in extent with nests constructed among the rocks and desert vegetation. Many of the plants were destroyed, while the continuous trampling on the dry soil offered but slight opportunities for any introduced seed to give rise to a plant that might reach maturity. Nothing had as yet appeared on the part of the beach laid bare by the recession of 1908, but the opportunity was good for observation on the condition of the invaders which had been seen three months earlier. Oligomeris glaucescens, Eriogonum thomassii, Sonchus asper, Cryptanthe barbigera, Psathyrotes ramosissima, Cucurbita palmata, Lippia nudiflora, Rumec berlandieri, Aster spinosus, and Pluchea sericea were maturing seeds and thus furnishing material for reproduction of the pioneers. Distichlis spicata and Leptochloa imbricata were also found. Atriplex canescens and A. fasciculata had maintained themselves on the beach, while A. linearis and A. polycarpa were not to be found. The two young trees of Prosopis pubescens and one of P. glandulosa were thriving. Oenothera, Encelia, and Spheralcea had apparently disappeared, while Heliotropium was represented only by dead stems and prob- ably a crop of seeds on the ground. It might therefore be said that the change in the population included one addition, that of Aster spinosus, and three losses. The shore of the entire northern end of the island was rocky and afforded no lodgment for the higher plants. These were confined to a strip running from East Bay to a point midway of the southern end and to the shore of West Bay. All of these places were sub- ject to heavy storms, so that strands which had been long above the level of the quiet water were subject to the wave-action of the water, which had by this time taken on a con- centration of 460 parts in 100,000. The shores of sand, clayey material, and heavy gravel were steep and the terracing could not be so definitely assigned to any season with cer- tainty, since storms might entirely obscure the effects of a water-level which had remained stationary for some time earlier in the year. PLATE 18 SALTON SEA Tide-pools back of A. Caving Bank marking extreme high-water level at Travertine Terraces as seen February 1908. small barrier at margin of Emersion of 1907. One Rhizome of Phragmites present, also row of Seedlings of Atriplex canescens at foot of wall. B. Strand of 1907 and cut bank at Travertine Terraces as seen November 1908. The Rhizome of Phragmites had elon- gated and branched, Distichlis was forming mats, while Pluchea, Atriplex, polycarpa and A. canescens, Heliotro- pium, Sesuvium, Spirostachys, Prosopis pubescens, Astragalus, Juncus cooperi, and Bouteloua were also present MOVEMENTS OF VEGETATION IN THE SALTON SINK. 127 The gently sloping alkaline beach of Imperial Junction was examined on the same day as Obsidian Island, and as the two places face each other across a 5-mile stretch of open water the difference in results may be attributed chiefly to slope and soil characters. The original belt of Sueda, Atriplex fasciculata, and A. canescens had now become very dense, while some Spirostachys was also present. In addition there had come from additional germinations, Distichlis, Heliotropium, and Sesuvium sessile, and the entire formation was showing indications of the increasing desiccation except Atriplex. Two species of this genus were fruiting abundantly, and as they were already represented down the shore by numer- ous seedlings they might be expected to constitute a strong element in the invasion. The plants in the zone of Pluchea sericea had made but a scanty growth, but some were maturing and casting seeds which were being carried about by the wind. Sesuvium, Spirostachys, Atriplex of the two species mentioned, and Heliotropium also grew among the Pluchea, while these species were diffused over the area which lay above the Pluchea and between it and the Sueda-Atriplex formation. Rumex berlandieri was also repre- sented in the wash transecting the emersion of 1907, together with Typha, Oligomeris glaucescens, Baccharis glutinosa, and Amaranthus palmerit. A comparison with the census of February shows that Oligomeris, Baccharis, Sesuvium, and Amaranthus represent addi- tions to the population, and their appearance is probably to be ascribed to the germina- tion of seeds with the rising temperature on the advance of the season rather than to new introductions. It is of course possible that the seeds of Baccharis may have been brought in by the wind. Cucurbita seen in February could not be found and might be temporarily reported as a loss. The difference between these two observational areas being so great, a visit was made at once to the Mecca area on the following day. Pluchea camphorata, which was native above the flood-level, had invaded the emersion of 1907 and was maturing fruits. Pluchea sericea was also maturing seeds in the 1907 emersion. Distichlis was present in abundance. Sonchus oleraceus was likewise casting a crop of seeds and Rumex was maturing seeds. One plant of Conyza coultert was present. Heliotropium had become very abundant. One young plant of Aster exilis was found. Large numbers of individuals of Oenothera were seen in moister areas, and Atriplex lentiformis and A. polycarpa were represented by large rounded plants, which throve exceedingly in this locality. Sueda and Spirostachys, which had originally occupied the area, were represented by numbers of young individuals. Oligo- meris glaucescens, Chenopodium murale, and an unrecognized species had disappeared in the three months since February, while it is to be seen that Pluchea camphorata and P. sericea, as well as Distichlis, had advanced down the gentle slope into the sterilized area. Sonchus oleraceus and Rumex had also been added to the population, although it could not be sug- gested from what source. A complete round of the observational areas was made in November 1908. The Mecca area appeared to furnish suitable conditions for Sueda, which had made an enormous growth, while the two species of Atripler mentioned above were still active. Heliotropium and Sptrostachys were seen as before. Pluchea sericea was represented by scattering indi- viduals, but P. camphorata was not seen, although, as was demonstrated later, it was present in abundance in seed form. The Imperial Junction beach was examined a day later than Mecca. No marked change from the condition described in May was discernible, except that the stress set up by continued desiccation had limited the growth of some of the species. Typha had appar- ently succumbed, while Oligomeris, Distichlis, Heliotropium, Sesunium, Rumex, Baccharis, and Amaranthus could not be found in the hasty examination made in the twilight of a winter day. The results of later examination, however, make it probable that the non- appearance of these plants indicated a real failure to survive the summer temperature and high evaporating power of the air. 128 THE SALTON SEA. The strand of 1907 on Obsidian Island bore a heavy growth of Pluchea sericea on East Bay in November 1908, with which were growing a few Heliotropium, Sesuvium, Distichlis, and Cucurbita. The exposed gravelly and sandy strip on the southeastern part of the island showed some Lippia nudiflora, Pluchea sericea, one Eclipta alba, Aster spinosus, and a particularly robust form of A. exilis. Baccharis glutinosa and Eriogonum thomasit were still present. One Prosopis was seen on East Bay, one on the southeastern point, and one on West Bay as earlier, and all had made a notable growth. Parosela was established on the shore back from West Bay, but Oligomeris, Sonchus, Rumex, Cryptanthe, Psathyrotes, Leptochloa, and Atriplex were not found. Some of these had doubtless fruited and might appear later as a continuation of the original introduction. (Plate 29 a.) The strand of 1907 at Travertine Terrace bore Sesuvium, Heliotropium, Pluchea cam- phorata, Atriplex polycarpa, and A. canescens, although the rank of seedlings at the foot of the cut bank had perished. Distichlis, Spirostachys, Sueda, Prosopis pubescens, Astragalus, Juncus cooperi, Bouteloua arenosa, and Phragmites were also present. All of these except Phragmites and Atriplex canescens had appeared since the examination in February. The increase in the population was of a character widely different from that of the other areas in which the grains had been balanced or overbalanced by the losses. The high cut bank at the upper part of the area had continued to crumble and also showed marked effects of wind erosion. The location was especially favorable to invasion by wind-borne seeds. (Plate 18 B.) The course of the voyage for circumnavigating the lake in November 1908 gave fav- orable opportunities for examination of the strand of 1907 elsewhere, and a landing was made on the sandy shore of the southwestern part of the lake, which was of such slope as to show a recession of about 200 feet. Spirostachys, Coldenia, Pluchea sericea, Parosela spinosa, Heliotropium, and Eriogonum were the principal members of the invading popula- tion. The last-named species had undergone a remarkable endurance of untoward condi- tions in one place. First a group of these plants, with woody stems less than a yard in height, had been overtaken by a sand dune, with the result that the tips of the stems were bent over until they touched the ground. Upright branches arose from the apices, but now the making of the lake submerged these plants for a period of a few months in water containing from 0.25 to 0.33 per cent of saline material. These plants were in bloom at the time of our visit in November 1908. A second landing on the western shore to the northward of this locality led to the discovery of a plantlet of Salix nigra and representatives of Hilaria rigida, Encelia fru- tescens, Olneya tesota, and Isocoma veneta var. acradenia, in addition to the commoner species named above on the strand of 1907. The beaches received but scant attention in 1909. The midsummer season was char- acterized by precipitation much above the average and a distinct run-off occurred over a large share of the basin. This exerted an influence in two ways. In one the salts were washed from the surface layer, while in another seeds were carried down in drainage lines cutting across the emersed zones. The first visit to the observational areas in 1909 was made during the closing week of October. The strand of 1907 on the Imperial Junction beach showed only Sueda Pluchea sericea, Atriplex, and Spirostachys, and the first-named was showing the stress of aridity distinctly, great numbers of dead plants being present. The washes cutting across this zone, as well as those of 1908 and 1909, had received a large supply of rain-water, which had been held with great tenacity by the adobe soil. In such places Sueda Sesmun) Spirostachys, Pluchea sericea, Cucurbita, Chenopodium, and Baccharis were growing well. All of, the Typha originally noted in this locality had died, but a plant which had been cast ashore this year in the filled wash had survived, making a reintroduction of this plant The invasions on the shores of Obsidian Island were now seen to occupy several soil complexes. Thus, one group of plants, comprising Pluchea sericea, Prosopis pubescens, P. MOVEMENTS OF VEGETATION IN THE SALTON SINK. 129 glandulosa, Baccharis glutinosa, and Distichlis had made a luxuriant growth in the clayey deposits on the shore of East Bay, emersion of 1907. The larger Prosopis glandulosa, on the exposed gravelly and sandy shore of the southeastern part of the island, had attained a height of 10 feet and was in bloom. Four smaller P. pubescens were coming up as seed- lings, although it could not be determined whether they were the offspring of the larger plant or not. Another P. pubescens in bloom was seen on the strand of 1907 West Bay, thus making a total population of six individuals of one species. Cucurbita was present in the same strand. This census indicates no farther introductions into the highest part of the beach, from which (since the previous census of 1908) Heliotropium, Sesuvium, Inppia, Aster spinosus, A. exilis, and Eriogonum thomasti had been lost. Here, as else- where, some of the missing constituents might have been represented by a crop of seeds. A visit to the Imperial Junction beach only was made in May 1910. The two species of Atriplex, Sueda, and Sptrostachys were arranged about as previously described, while Pluchea sericea continued to endure the desiccation. A second inspection of the shore stations was made in September and October 1910. The emersion of 1907 at the Travertine Terraces bore Distichlis, Prosopis, Phragmites, and Astragalus. In the two years since the last visit was made, Heliotropium, Pluchea cam- phorata, Juncus, Bouteloua, Sesurium, Spirostachys, Sueda, Phragmites, and the two species of Atriplex had been lost from the population. The return from this locality was made along the shore, so that many miles of the beach were seen. The surviving pioneers on this gentle slope were Distichlis, Spirostachys, Atriplex, Prosopis, and Sueda. Visits were again made to the shore stations in October 1911. The emersion of 1907 Travertine Terraces bore Distichlis, Pluchea sericea, Salix nigra, Prosopis, and Phragmites, but Spirostachys had disappeared. The shores of Obsidian Island were visited by a boat put off with difficulty from the Imperial Junction beach. The only survivors of the population of the strand of 1907 were Prosopis, Heliotropium, Baccharis, and Cucurbita. A registering thermometer, placed on this island two years before, recorded 41° and 109° F. The losses recorded in 1909 for this place seem to be permanent and the desiccation of the soil may be taken to have reached a point where it would be difficult for any of the lost species to regain a foothold. The Imperial Junction beach was much in the same condition in 1911 that was re- ported in 1910. Two species of Aériplex, Sueda, and Spirostachys comprised the plants that seemed to be holding their own, while the stand of Pluchea was still undergoing great losses from increasing aridity. It being desirable to bring together the results of the various investigations upon the problems of the Salton, at the close of 1912 arrangements were made for a careful examina- tion of the various beaches at some time during the year. Those at Travertine Terraces and Mecca were visited twice. In June Atriplex (two species), Prosopis pubescens, and Sueda were seen in the upper- most portion of the emersion of 1907 at Mecca, and the lower limits of this strand could not be made out on the gentle slope. The emersion of 1907 at Travertine Terraces bore Prosopis pubescens, Distichlis, Astragalus, Salix, and Pluchea sericea, from which it may be seen that there had been no change in the population since 1910. The final examinations were begun at Mecca on October 14, 1912. The gentle slope at this place had operated to prevent any clear delimitation of the limits of the seasonal emersion, but some estimate was made from the known vertical recession of the lake. Sueda, Prosopis, Pluchea sericea, Atriplex canescens, A. lentiformis, with Distichlis probably represented the total population, and a photograph was made with all of these species in view. Reclamation operations had been begun in this region, so that it is probable that 9 130 THE SALTON SBA. serious changes may have come from the agricultural operations. The species named, however, are represented abundantly in the area above the water-level and would be the readiest invaders and occupants of any bare surface in their neighborhood. The expedition proceeded directly from this place to the beach at Travertine Terrace. The strand originally laid bare in 1907 had been open to occupation for six years. Phrag- mites, one of the first two pioneers, was still present. Prosopis pubescens was present in a vigorous young tree, which was originally noted in 1908. Mats of Distichlis were abun- dant, and Pluchea sericea was represented by a small number of individuals. Astragalus was in the form of mature shoots with emptied seed pods, while a plant new to the area, Isocoma veneta var. acradenia, was present, being the only innovation noted here for an extended period. The soil and moisture conditions from a superficial examination appear to be similar to those in which these plants ordinarily grow, with the exception of Phrag- mites, and they may be expected to survive for an indefinite period. After some water travel for the inspection of islands, the results of which will be de- scribed later, a landing was made on Obsidian Island on October 16, 1912. The seven trees of Prosopis (2 species) previously noted were alive and growing rapidly. Heliotropium, Sesuvium, Baccharis, Spirostachys, Atriplex canescens, A. fasciculata, A. hymenelytra, and Parosela emoryi were represented in the flora of the strand of 1907. It is probable that this last-named species had been overlooked on some visits, as it appeared to have had a place since its original introduction. Cucurbita, however, had dropped out, although, as will be shown later, it came on the beaches of succeeding years. Sesuvium likewise was missed on some visits, though probably continuously present. The conditions in this area are much different from those at the Travertine Terraces. In the last-named locality the ancient beaches offer an example of the final fate of strands, and those recently formed will doubtless progress by slow stages to the same condition (see Plate 29 8). Thecontinued flow of the Alamo and New Rivers into the southwestern part of the lake, however, resulted in the deposition of a layer of silt which forms fresh soil with low salt content on the beaches of Obsidian Island and Imperial Junction beach. The return to alkaline desert conditions here would be far slower, and hundreds of years might be necessary for the physical changes which would ultimately bring the lower slopes to the same state as the upper slopes untouched by the last inundation. A résumé of the history of the strand of 1907 during its first six years of desiccation will lead to a more intelligent consideration of the strands of lower level which will be taken up in the succeeding sections. The pioneers on the gently sloping Imperial Junction beach were Atriplex canescens, A. linearis, A. polycarpa, A. fasciculata, Sueda torreyana, Spirostachys occidentalis, Pluchea sericea, and Distichlis spicata on the plane slopes, while the filled channels of the washes with a larger moisture content and lower proportion of salts gave lodgment to Heliotropium curassavicum, Typha angustifolia, Rumex berlandieri, Leptochloa imbricata, and Lepidium lastocarpum. Several hard-shelled fruits of Cucurbita palmata had been cast ashore, but no activity had been shown by the seeds. To this census, made in February 1908, may be added Oligomeris glaucescens and Baccharis glutinosa, on the plane slopes, while Amaranthus palmer: and Sesuvium sessile were added to the population of the filled washes. It may be said at once that no secondary invasions occurred on the plane slopes and that no new species were added to the population of this emersion, although Baccharis, Typha, and Amaranthus appeared in new individuals, which apparently were brought in independently, not originating from those already present. All of the invaders had perished in October 1911, except Sueda, Atriplex, and Spirostachys, while some Pluchea still survived in scattered spots. The slope at Mecca was even gentler than the one just described and the succession of events may not be so succinctly related. The occupants of the strand of 1907, as examined SALTON SEA PLATE 19 A. View of Travertine Wash in November 1909. Mats of Distichlis and scattered plants of Pluchea camphorata, Atriplex, and Spirostachys. Seepage now confined to defined channel. (See plate 17.) B. Heavy growth of Baccharis, Pluchea sericea, and Sueda in the channels of the washes into which the water of the lake had risen at Imperial Junction Beach, October 1909. The denser formations in other washes are to be seen in the distance. MOVEMENTS OF VEGETATION IN THE SALTON SINK. 131 in February 1908, included Oligomeris glaucescens, Chenopodium murale, Atriplex fasciculata, A. lentiformis, Heliotropium curassavicum, and an unrecognizable species. In May the addi- tional species of Baccharis glutinosa, Oenothera, Sonchus asper, Conyza coulteri, probably Pluchea camphorata, and Distichlis spicata were noted. The survivors in 1912 were the two species of Atriplex, Pluchea sericea (which had come in meanwhile), Distichlis, and Prosopis pubescens, which had probably come up from stems or seeds left in place. Phragmites was the sole water-borne introduction to the Travertine Terraces of 1907, seen early in 1908, and Atriplex canescens had fallen down the slope and germinated. In November 1908 the population had increased to include 16 species (see page 128), but a census of the area in 1910 showed but 6 species, 5 of which had maintained themselves until June 1912, the soil moisture being influenced to some extent by a hemmed underflow. A single innovation, Isocoma, was of recent date. The emersed zone of 1907 on Obsidian Island was found to bear 25 species when examined in February 1908 and an additional species was added to the population at the next census in November of the same year, which meanwhile had suffered a loss of 14 species. A year later the population of the zone had fallen to 7 species, and in October 1910 but 6 were present; 10 species were found here in 1912, 4 of which had simply come down the slopes from the drier areas, while only a single one of the other 6 had been present from the beginning. (Plate 21.) The total census of the strand of 1907 in the four main observational areas at the close of the season of 1912 was thus seen to include the following: Atriplex canescens. Spirostachys occidentalis. Pluchea sericea. fasciculata. Parosela emoryi. Astragalus limatus. hymenelytra. Prosopis pubescens. Distichlis spicata. Heliotropium curassavicum. glandulosa. Suseeda torreyana. Sesuvium sessile. Phragmites communis. Salix nigra. Baccharis glutinosa. Isocoma veneta var. acradenia. These plants fall into three groups: xerophytes capable of existence on soil with low moisture-content in an atmosphere of high evaporating power; halophytes which might endure a substratum the moisture of which was highly charged with saline matter; and the single tree Salix, which would grow only with an abundant supply of water with not much dissolved material. All of these plants are native of the Salton Sink in great number and the action of the lake has simply resulted in their establishment on the beaches in zones or bands conditioned by the sorting of the soil-material and the proportion of the water supply, and salts. Not any emersed area at this time had reached a condition equivalent to that in which it had been before the making of the lake. The plant population at Imperial Junction beach, for example, included the species from the slopes above with the addition of Pluchea sericea. The Travertine Terrace had been restored so far as to give a place for Isocoma, but its shelves and beach ridges were as yet widely different from those near the ancient beaches near sea-level. Atriplex canescens had come in, and it appears to stay in place on strands during all of the changes of which these beaches are possible. Parosela emoryi was also included among the species appearing on the beaches of Obsidian Island in 1908 and it is a constituent of the ranks marking the old strands near the ancient sea-level in which the conditions tend to extreme xerophytism. REOCCUPATION OF THE STRANDS OF 1908. The level of the Salton Lake was lowered a net total of about 58 or 59 inches in 1908, and the proportion of dissolved salts rose to over 0.5 per cent, from over 0.4 per cent, during the year. The actual width of the bared beaches would therefore be greater than in the year 1907, in which recession went on only during about ten months and amounted to about 42 inches. Calculated from the slope of 1 in 300, the emersed zone on the Imperial 132 THE SALTON SEA. Junction strand was about 1,400 feet; about 600 feet of this had been laid bare on May 1908, when a visit was made to this place and to the beaches of Obsidian Island. _This visit was opportune, as it confirmed the observation of the previous year that a series of germinations ensued customarily on this beach on the mud laid bare in April. Scirpus * paludosus was in moist depressions near the margin of the water; Leptochloa imbricata was a few yards back from the shore; numbers of plantlets of Atriplex fasciculata were established in the moist soil near the water, a small plant of Typha was found coming up through the mud, probably from a rhizome floated to the place; a few seedlings of Sueda and Spirostachys were also present, and some mature plants of Polypogon monspeliensis were shedding seeds. Numbers of shallow trough-shaped channels, a few inches in width, extended into the soil from the actual margin of the shallow water, and these were filled when the wind was on shore, with the result that loads of débris (including seeds and other parts of plants) were carried up into the land in narrow bands, in which seedlings of an Atriplex polycarpa were found. The steep beach of Obsidian Island, which had been uncovered since the midwinter, was totally bare, and nothing had germinated there when visited again in November 1908. The strand of 1908 at Imperial Junction beach bore only scattering plantlets of Spiro- stachys and Sueda with one Sesuvium in the stretch visited. The lowermost portion nearest the water showed an efflorescence of salts. No change was reported in the constitution of the population in November of the same year. Returning again to the area on the Imperial Junction beach in November, the whole emersion for the year presented the aspect of a stretch of dried and fractured mud, at the upper margin of which could be seen only a few plants of Sueda, Spirostachys, and Atriplex. Typha, Heliotropium, Distichlis, Pluchea sericea, Phragmites, Bouteloua arenosa, Salix nigra, and Populus were present on the Travertine Terrace strand of 1908, the upper margin of this emersion being marked by the beach ridge of midwinter formation. These introductions being evidently due either to the action of the wind or waves, a long stretch of beach to the southeast was visited for supplementary observations. The total census of several miles of this strand gave the following : Heliotropium, Distichlis, Prosopis, Atriplex, Coldenia, Chamesyce, Polycarpa hirtella, Scirpus paludosus, Parosela emoryt, P. spinosa, and Bouteloua arenosa. The effect of the run-off streams coming down the bajadas from the Santa Rosa Mountains to the westward was plainly evident. These streams might also have an additional effect suggested by some observations on the south- western part of the lake, where seedlings of Pluchea sericea were seen pushing up through sand which had been deposited over the seeds by such streams in times of precipitation. Without such a protecting layer the moisture would be insufficient for germination. Only two of the four observational areas were visited in 1909, and this inspection was made in October. The summer had been characterized by a heavy rainfall over most of the Sink, with the result that numerous channels had been formed and that some salts had been leached from the soil, at least from the surface layers. The total precipitation measured at the U. S. Weather Bureau station near Salton, from July to October, was 3.5 inches, of which 2.887 inches were received in 24 hours and most of this fell within 2 hours. But little effect had been produced on the Imperial Junction beach. The strand of 1908 at that place bore only four species, Atriplex (two species), Sueda, and Spirostachys. The steep clayey and gravelly slopes of Obsidian Island had been deeply eroded, and the strand of 1908, on which nothing had been seen previously, now bore Heliotropium, Sesuvium, Sptrostachys, Cyperus speciosus, Isocoma veneta var. acradenia, Baccharis glu- tinosa, Aster exilis, Distichlis spicata, Parosela emoryt, Eclipta alba, Sonchus asper, and Cucurbita palmata. This heavy census suggests both delayed germinations and also possible wind dissemination, especially of the winged compositaceous seeds. SALTON SEA PLATE 20 A. Emersion of 1907, Imperial Junction Beach, as it appeared October 1911. Vigorous development of Atriplex, Spirostachys, and Sueda in open formation. B. Emersion of 1907, Travertine Terrace, as it appeared in October 1911. Dense mat of Distichlis, Astragalus, Prosopis pubescens, Pluchea sericea, Isocoma, etc. MOVEMENTS OF VEGETATION IN THE SALTON SINK. 133 The Imperial Junction beach in May 1910 bore Heliotropium and Sesuviwm in addi- tion to the four present in the previous October, the entire number being represented by widely scattered individuals. The area at Travertine Terraces was examined in September 1910, at which time Pluchea sericea, Distichlis, and Heliotropium were on the strip laid bare before midsummer of 1908, while the beach of the latter half of the year contained, in addition, Juncus coopert, Pluchea sericea, Astragalus, Prosopis pubescens, P. glandulosa, and Spirostachys. In the two years which had elapsed since the last visit, Bouteloua, Phragmites, Salix, and Populus had dropped out, while Juncus, Pluchea, Astragalus, and Prosopis had come in, probably by delayed germinations of seeds deposited by the water. The 1908 emersion at Mecca showed Distichlis, Spirostachys, Atriplex, Prosopis, and Sueda, the invasion by which could not be assigned with certainty to any given factor or agency in September 1910. The strand of 1908 at Travertine Terraces bore only Distichlis, Salix, Prosopis pubes- cens, and P. glandulosa, in October 1911, and the loss of more than half of the initial occu- pants was indicative of some heavy stresses. Similar unfavorable conditions must have affected Obsidian Island also, for the emersion of 1908, which had appeared so richly populated on the previous visit, bore only Cucurbita, Spirostachys, Heliotropium, and pos- sibly an Atriplex. The vegetation of this strand on Imperial Junction beach included only Sueda, Spirostachys, and the two species of Atriplex previously noted. Preliminary to the final visit made to all of the observational areas in 1912, the Traver- tine Terraces were visited in June 1912. Distichlis, Cyperus, Pluchea sericea, Prosopis pubescens, P. glandulosa, and Astragalus were present. Here, as in the strand of 1907, a single innovation was included at the last-—-Cyperus, which had not appeared at this place before, and for the presence of which no reasonable suggestion may be made. The inclusive closing observations made on all of the beaches in October 1912 were some- what disturbed at Mecca by the fact that reclamation operations and agricultural utilization of the land below the water level of 1907 had begun and the exact limits of the area laid bare by the recession of the waters could not be made out after the removal of the landmarks. Conditions in the other localities were much more favorable to a general examination of the behavior of the plants on the strands. The emersion of 1908 at Travertine Terraces bore abundant mats of Distichlis spicata, Juncus cooperi, Pluchea sericea, Prosopis pubes- cens, P. glandulosa, and Astragalus. The moist condition of the soil which had persisted in small spots indicated (see pp. 124, 139, and 140) a hemmed underflow. A brief visit to the place was made in February 1913, at which time Isocoma, Phragmites, Heliotropium, and Parosela emoryi were found, in addition to the above-named species. No plants could be definitely assigned to this strand on Obsidian Island; it will be recalled that the vegetation was very sparse on the last visit, but Atriplex lentiformis, A. hymenelytra, Spirostachys, and Sueda were probably within the limits uncovered by the recession of this year. The strand of 1908 at Imperial Junction beach appeared to be almost bare when visited in October 1912, but in two traverses three individuals of Pluchea sericea were seen, the only other vegetation being a few scattered examples of Sucda. The surviving invaders in the strand of 1908 are thus seen to number far less both as to species and individuals than the emersion of the year before, one having had six years of desiccation and the other five. A satisfactory analysis of the movements of vegetation on the emersion of 1908 is not easily to be made. The gently sloping alkaline surfaces of Imperial Junction beach showed nothing during 1908, 1909, 1910, and 1911 but Aériplex (two species), Sucda, Spirostachys, with an intrusion of Heliotropium in 1910 (which was not found in the fol- lowing year) and of Pluchea sericea in 1912. 134 THE SALTON SEA. The 1908 zone on the shores of Obsidian Island remained bare during that year, but had been occupied by 12 species in 1909, all of which but 3 had disappeared by 1911. Probably a fourth is to be added to the number in 1912. The Travertine Terraces were occupied or invaded at once by 8 species not identical with those found on the other strands of the same year. A year later 7 species were found, 2 of the original ones having disappeared and 2 new invasions having occurred. In June 1912 the plant population was represented by 6 species, which were identical with those noted in October except Juncus and Cyperus; but circumstances justify the assumption that both species were continuously present, so that the final census here may be taken to include 6 species. Three species came into the 1908 zone at Mecca soon after its emersion, being followed by three others, which were seen two years later. Subsequent history of this beach was complicated by economic operations, but it was evident that the number of species would remain small here, as at the other gently sloping beach, although the individuals would be more numerous. REOCCUPATION OF STRANDS OF 1909. The year 1909 was characterized by a net recession of 48 inches in the lake. In July and August there was heavy precipitation, amounting to 3.5 inches (July to October) at the Salton trestle station. Nearly 3 inches of this fell in one day at this place. The run- off and underflow were so great that the results of evaporation were met and the lake stood nearly at constant level for a period of about six weeks, a condition very favorable for sorting material in beach ridges and depositing seeds; also, the precipitation would result in some leaching of salts from the surfaces of all of the exposed beaches, while unusual value would be given to the washes as disseminating or invading agencies in bringing plants down the slopes in lines radial to the lake. The conditions of evaporation and the changes in the level of the lake from midsummer of 1909 to midsummer of 1910 are illustrated by the following data furnished by the U. 8. Weather Bureau from observations made on the northeastern part of the lake near the Station at the Salt Slough. Table 32 shows evaporation from pan 500 feet from shore and 2 feet above the surface of the water. TABLE 382. ANI! OOD ss cece Sides a visae vey oontavleus «Decent e,aue 14.41 January TIO 0. issececn wired wevttia ee soe 3.61 JULY 1909" ais cts agace ecu 4.8 ae See 14.77 | February 1910 .................. 000000 5.01 August31909 och cues oR es Melee. 12.63 |) Marobs 1910) 99s srascctses acdaedias omits 40 es 9.17 September 1909...............2..000 0005 1240+ |) April VOLO oc. cse.sccaseeuane cecereeu a d athars Sodas 10.96 October’ 1909 .ctiviere asrenies eres se eae ees S20 | Niay 1910 sae vy ceuvy ee Hewldv an Masog ex 14.01 November 1909 ..........ccseeeeeveeees 6.21 —_— December'2009 i.e cance: eines so estes ae ae 4.67 Weta £00 Fear os Vek ae gk esae he oA 116.95 The total evaporation from another pan, 7,500 feet from the shore, was 114.97 inches, and hence for convenience the yearly evaporation from the surface of the lake will be taken as about 116 inches. The level of the lake during this period, however, showed fluctuations of a character more widely divergent from the curve of evaporation than at any other time during which observations were carried on by the members of the Institution, as denoted by the monthly recession measured by the Weather Bureau showing the altera- tions in level of Salton Lake June 1909-May 1910 (table 33). TaBLe 33. Feet. Feet SUNG: VO0S is cee a ahiien aos sisaten nx Naveed eects Ge O24 | Saomry VOU sc. os scmed va eda ee vied hs 0.0 PU MO oss te dh siseadve hie eisai ee ebianinn ak waa ee | SOOT SOY oe oe aia. a boleh a aw ames Se Al PURE TOUS 5 9c ieee eee kie ds 4 deel aaa aed 2A Maroh YOO i... slap abate ehianisheee exetne 3 eptember VION cst soe x tury ne Reedy eeeoe oe {6 | April 1 910 eects catenins eaneeien fateooes ‘2 October 1909 ............- 0s. ss esee sees eB IM By A910 cocoa «Fadia 4 2 csssnieecenenect aise ‘5 November 1909 cos «905 4 6k bcka-e d kaw oo iiilinek “thal te ak pokey th vantemn cena Ne heg a December 1909 .......... 0. ccc cece cc accuee 5 Total lor Feat. .ceusaws vsciayau ve uyaaas 4.3 The total net recession of the level during twelve months was therefore less than half the possible evaporation. SALTON SEA PLATE 21 A. View of Shore of southeast point of Obsidian Island, October 1912. A large shrub of Prosopis pubescens on the strand of 1907 at left. Scattered clumps of Sesuvium, Spirostachys, and Atriplex in distance. B. View of strands of East Bay, Obsidian Island, October 1912. Dense formation of Baccharis and Pluchea sericea on Emersion of 1907 in distance. Clumps of Sesuvium, Spirostachys, and Atriplex in foreground. MOVEMENTS OF VEGETATION IN THE SALTON SINK. 135 The dissolved salts in the water had reached a proportion of over 0.5 per cent, or 520 parts in 100,000, in June of this year, and some lowering of the rate of increase in the calcium was also noticeable. The conjoint effect, however, of all of the conditions enumerated was a greater number of successful invasions of the emersed strip. The occupation of the newly bared beach at Imperial Junction was characterized by a zone of plants of indefinite width coming to within 300 feet of the margin of the water late in October 1909. This position was one indicative of midsummer germinations in mud moistened by the summer rainfall. Leptochloa imbricata, Scirpus paludosus, Typha angustifolia, Heliotropium curassavicum, Pluchea sericea, Sesuvium sessile, Sueda torrey- anum, and Spirostachys occidentalis were present, the last two being the most abundant and evidently coming from seeds, while the single plant of Typha was growing from a rhizome washed ashore. The two species of Atriplex on this shore were present in the midwinter zone with Sueda and Spirostachys. The collection of precipitation water had furnished a supply of moisture for the plants growing in the filled channels of the washes above the 1909 zone. A visit a few days later to the beaches of Obsidian Island resulted in the detection of vegetation on the emersion of 1909, which was the only time in which plants were seen to appear here during the first year of the emersion. The sowings were composed of Helio- tropium, Amaranthus palmeri, Sesuvium, Spirostachys, and two species of Atriplex. The other observational areas were not inspected during this year. In May 1910 the strand of 1909 on Imperial Junction beach included Helrotropium, Sueda, Atriplex (2 species), Spirostachys, Sesuvium, and Cucurbita, the last-named being an innovation, while Leptochloa, Scirpus, Typha, and Pluchea sericea had perished. (Plate 28 a.) The strand of 1909 at Travertine Terraces bore only Distichlis, Heliotropium, and Pluchea camphorata in September 1910. A year later the entire emersed zone at this place was heavily carpeted with Distichlis, which was also noted in abundance on the soil laid bare during 1909; scattering clumps of Juncus, Salix nigra, Prosopis pubescens, and Astraga- lus were also found. The entire population was thus seen to be different from that of Imperial Junction beach. The 1909 strand on Obsidian Island showed only Spirostachys and two species of Atriplex in October 1911. The heavy carpet of Distichlis persisted at the Travertine Terraces and was examined in June and October 1912. Cyperus, Astragalus, Pluchea camphorata, P. sericea, Prosopis pubescens, and a single Salix were seen in June. Cyperus was not found in October 1912, but Isocoma and Heliotropium had come in. A supplementary visit was made in February 1913, at which time Atriplex lentiformis, A. canescens, and Juncus coopert were added to the above list. (Plate 22 a.) The limits of the emersion of 1909 at Mecca were not to be accurately determined, but a photograph was made of a large Atriplex lentiformis and some Spirostachys, both of which were under very favorable conditions, in October 1912. The only plants which could be identified at this time as positively on the emersed strip of 1909 on Obsidian Island were Parosela emoryi, Cucurbita palmata, and Spirostachys; Atriplex of two species and perhaps also Heliotropium might be safely included. In 1912 only Spirostachys and Sueda appeared to be surviving of the large number of invaders which had found a foothold on the moist strand at Imperial Junction beach in 1909. The first of these was making a maximum shoot development. An analysis of the movements of plants on the strands laid bare in 1909 similar to that made for the two years previous brings out the following: The invaders of the bared strip at Imperial Junction beach included 10 species at first, which was reduced to 7 in May 1910, by four fatalities and one additional intruder, while the close of the observa- tions was made with only the two halophytes, Spirostachys and Sucda, persisting. 136 THE SALTON SEA. The 1909 strand at the Travertine Terraces bore only 3 species late in 1910; a year later the number had been increased to 8, while the closing part of the year 1912 was marked by the advent of 2 more; 3 additional species were seen in February 1918. The steep beaches of Obsidian Island were populated in the strand of 1909 by 6 species during the first year, which two years later was reduced to 3; 5 or possibly 6 species may be assigned here at the close of 1912. The strand of 1909 on the gently sloping beaches near Mecca appeared to be populated only by Atriplex and Spirostachys. It was obvious that the ranks of the invaders were being depleted on the gentler alkaline slopes and accessions were being made on the steeper gravelly slopes of the Travertine Terraces and on Obsidian Island. A very important factor in determining this aspect of the reoccupation of the bared areas is to be attributed to the saline content of the soils. REOCCUPATION OF STRANDS OF 1910. The lowering of the level of the lake during 1910 amounted to over 59 inches, and the dissolved salts amounted to about 0.6 per cent on June 1 of that year. A visit to the Impe- rial Junction beach in May was made and Heliotropium, Leptochloa, Scirpus, Atriplex (two species), and many young plants of Spirostachys were found near the margin of the water. Travertine Terraces were not inspected until September, but at that time a rank of Saliz and Populus was found at the upper margin of the strand, while Distichlis and Helio- tropium were scattered over the shelf. In October 1911 a heavy rank of Saliz had formed at the upper margin, in which were seen a few Populus, while Distichlis, Scirpus olneyt, Pluchea sericea, P. camphorata, Juncus, Prosopis, and Heliotropium were represented. At this time the flora of the strand of 1910 on the Imperial Junction beach was composed of Spirostachys, Sueda, and Heliotropium. The only note at. Mecca in June 1912 was to the effect that Heliotropium in this emersion was going under, while in October only a few straggling specimens of Spirostachys and Sueda were seen, although some of this might be due to agricultural operations. (Plate 22 A.) A few Atriplex lentiformis and Populus may be assigned to this zone also. (Plate 23 7) The strand of 1910 at Travertine Terraces bore Pluchea sericea, P. camphorata, Heliotropium, Scirpus olneyi, S. paludosus, Salix, and Populus in June 1912. The last two named had by this time made a dense row of young trees with a height as great as 15 feet. In October the rank of trees (Salix and Populus) had made additional growth and included some Pluchea sericea, while Pluchea camphorata, Isocoma, Scirpus olneyi, Distichlis, Heliotropium, Prosopis pubescens, and Juncus coopert were variously distributed. The strand of 1910 bore only a straggling growth of Sueda and Spirostachys on Im- perial Junction beach when examined in October 1912. At the same time a visit was made to Obsidian Island, which had not been seen since this beach was bared, and Pluchea sericea, Atriplex lentiformis, Baccharis glutinosa, Heliotropium, and Spirostachys were found. The advent and fate of the intruders on the areas laid bare in 1910 may be briefly summarized as follows: Six species germinated on the strand at Imperial Junction beach in May 1910, of which two survived when seen late in 1912. Four species had come into the bare area at Travertine Terraces in September 1910, and five additional ones were seen a year later. Seven species were seen in June 1912, but probably the missing two, Juncus and Prosopis, were overlooked, since they were found in October. On this last date the full complement of 1911 was identified, with the addition of a tenth in the form of Isocoma. The contrast of reduced numbers of the invaders on the alkaline gently sloping beach of Imperial Junction with the increase in the occupants of the steeper gravelly slopes of Travertine Terraces is again presented, the constituency of the flora of the strand of Obsid- ian Island lending support to the presumption that a similar movement must have taken place there also. SALTON SEA PLATE 22 bah 5 aha Gib A. Strand of 1909, Travertine Terraces, October 1912. Dense formation of Distichlis with row of scattered individuals of Isocoma along middle of zone. Rank of Populus and Salix marking lower margin of strand. One Prosopis pubescens in middle of emersion; also several clumps of Juncus cooperi, Astragalus, Pluchea, and Heliotropium present, but not showing. B. Establishment of a Ranch on Emersion of 1910, Mecca Beach, October 1912. Maize in foreground and recently planted cuttings of dates on left. Ranch house surrounded by trees of Populus and Salix MOVEMENTS OF VEGETATION IN THE SALTON SINK. 137 REOCCUPATION OF THE STRANDS OF 1911. The total recession during this year was about 49.6 inches. The soluble salt-content of the water amounted to about 0.7 per cent at the beginning of the year. A visit was made to the Imperial Junction beach in April 1911 by Mr. E. E. Free, who reported a band of vegetation, about 200 feet in width, extending up the slope from within 50 feet of the margin of the water. In following this for half a mile, Scirpus palu- dosus, Leptochloa imbricata, Heliotropium, Distichlis spicata, Sesuvium, Rumezx, Spirostachys, and Sueda were seen. An examination of the zone in September of the same year brought to light only Spirostachys, Sueda, and Distichlis. It was notable that Atripler had ceased to be a pioneer on this beach. (See Plate 24 8.) The strand of 1911 at Mecca bore Typha, Scirpus, Atriplex lentiformis, and Spiro- stachys when examined in June 1912. Most of these appeared to be surviving when the place was visited in the following October. The strand of 1911 at Travertine Terraces bore only Distichlis, Salix, and Populus in September of that year, which was also a notable departure from the history of previous strands. In June 1912 Distichlis, Heliotropium, and Spirostachys were seen on the shelf, while the remaining vegetation consisted of Salix, Populus, Typha, Pluchea sericea, P. camphorata, and Scirpus paludosus, largely collected in a dense rank at the upper margin. In October the rank at the upper margin of the strand included Saliz, Populus, Prosopis pubescens, Typha, and Pluchea sericea, while Pluchea camphorata, Scirpus olneyi, Spiro- stachys, Sesuvium sessile, Heliotropium, and Distichlis were now found on the floor of the terrace. (See Plate 24 a.) The recently bared strand of Obsidian Island was vacant. REOCCUPATION OF THE STRANDS OF 1912. The water of Salton Lake had increased in concentration to such an extent that it contained nearly 0.9 per cent of dissolved material by June 1, 1912. The deposition of calcium as a carbonate had continued at such rate that emersed objects, such as twigs and branches of trees, were heavily coated with lime. The toxic action of the water would probably be greatly increased by this disturbance of the balance. The strand of 1912 at Mecca was well dried out in places and offered much efflorescence when examined in October. Scattered over it and collected in favorable spots were Atriplex lentiformis, Pluchea sericea, P. camphorata, Spirostachys, Populus, Heliotropium, Typha, Salix, Distichlis, and Scirpus paludosus, an invasion which for numbers was foreshadowed by the plants seen on the emersion of 1911. No plants were found on the ground laid bare by the recession of 1912 at Travertine Terraces in this year, in June, except Sesuviwm. In October, however, the angle at the foot of the distinctly cut bank, which may be taken to mark the upper margin of this strand, included Salix 2 or 3 feet in height, Populus, Atriplex lentiformis, Spirostachys, Heliotropium, Prosopis, and Pluchea camphorata represented by active young plantlets, while Sesuxiwm was found both at the top and foot of the low bank. (Plate 25 3.) A visit on the following day showed that the strand of 1912 on Obsidian Island was still bare. Two days later the inspection of the Imperial Junction beach was made. The uppermost portion bared early in the year bore a dense zone of Atriplex lentiformis, A. fasciculata, Sueda, Spirostachys, Scirpus paludosus, and Sesuviwum. A second zone near the margin of the water, representing germinations of the midsummer, included Distichlis, Leptochloa, Scirpus paludosus, and Atriplex lentiformis. It is to be seen from the above that the successful invasions of 1912 number far more than those of the preceding two or three years on most of the beaches. So far as the Im- perial Beach is concerned more favorable conditions may well be ascribed to the silt thrown 138 THE SALTON SEA. into the lake by the current of the Alamo River, which formed a deposit on the strands laid bare in this part of the lake, when visited in September 1911. In October 1912 Airi- plex lentiformis, Pluchea sericea, Baccharis glutinosa, Heliotropium, and Spirostachys were present, the last-named in abundance. The detail given again exemplifies the fact that depletion of the ranks of the original invaders follows quickly on the gentler alkaline slopes, while on the steep gravelly and sandy beaches the number of species soon increases. Thus the original census of the strand of 1911 at Travertine Terraces included only one species, presumably by reason of delayed germinations. x a oS os i En ae oO - 63 Deel 8 ne oa y—~ g 2 ce » a or . At 2 —s—_# 33 . < = 450 D : t H “o a a w 60 SG 2S : : a S ‘ 3 8 ze a = 2 ae PS = wo c o c= as 50 = a t— 2 ao 30 £ S E 25 |L—&——|350” 2 BB See 2 40 £8 5 : € 3 5 300‘C e se £30 5 a L : = f 250 8 10 = o-{ -2735US68.Datum -280.3 S.P.Datum , i flow of Alamo and New Rivers grobable ncrease in Depth due to flow ~— ol Ince Second &10* 158 Nov. Jan.1905 Jan.1906 Jan.|907}-—. Jan.1908 Jan.1909 Jan.I9!0 Jan.19it Jan.1912 Fic. 3.—Curves showing changing level of Salton Sea, 1905-1912, with estimated inflow of water through Alamo and New Rivers. (After H. T. Cory.) RECESSION OF 1913. The only examinations of the strands of 1913 which were made previous to the com- pletion of the manuscript of this book were devoted to the emersions at Travertine Terraces. On February 8 the water was standing at the foot of a bank from 20 to 30 inches in height x i. & a3 9 8 ¢ g | g1 8 9 3 te ea ER Le 3 a 1 eee i ia ” | I : | ei | nf a | r | ies ! ha 1 | 4 | 1 J b-2/- f= - 54 SSSR TS aS HB SEES pS SS } i SSS ETE_EE Fea aa PSS Se SO RE EVEL ===>- a = 36-7 WATER LOVE Och. 16,1912 Qo 2% 50 190 290 5 Fia. 4.—Diagram showing recession of water at Travertine Terraces, 1907-1912. (which was of the maximum steepness), masses bein i g constantly undermined by wave action. Some of the flotsam accumulated at the base was taken for a test to meen what seeds might have already been deposited here. About a couple of pounds of this material was placed under cultural conditions and numbers of seedlings representing Atri- SALTON SEA PLATE 23 A. Emersion of 1910, Mecca Beach, October 1912. Large hemispherical clumps of Atriplex lentiformis and trees of Popu- lus macdougalii. Dead Prosopis in distance. B. Emersion of 1911, Mecca Beach, showing conchoidal mud-fractures slightly weathered. Dead Prosopis with Nests of Water Birds used only while the Trees were surrounded by Water. Scattered plants include Atriplex lentiformis Isocoma, Heliotropium, Spirostachys, and Pluchea camphorata. SALTON SEA PLATE 24 A. Uppermost portion of Strand of 1911, Travertine Terraces, October 15, 1912. Rank of Vegetation at margin includes Salix, Populus, Typha, and Pluchea sericea, while Pluchea camphorata, Scirpus, Spirostachys, and Helio- tropium are scattered in the open. B. Strand of 1911, Imperial Junction Beach, as it appeared on October 4, 1912. Scattering Plants of Spirostachys, Suda, and Distichlis are present MOVEMENTS OF VEGETATION IN THE SALTON SINK. 139 plex, Heliotropium (?), Sesuvium, and a grass appeared. Measurements, of the level of the lake, besides those compiled into the curve shown in Fig. 4, were now received, from which it appeared that the surface had fallen 9.6 inches in August, 7.2 inches in Septem- ber, 3.6 inches in October, 2.4 inches in November, and 1.2 inches in December 1912. The next measurement (made on February 1, about the time of the first observation in 1913) showed that the level of the water had been raised 2.4 inches during January, clearly due to the lessened evaporation and the increased inflow, and that the level had fallen an equal amount during February, so that the level on March 1 was the same as at the begin- ning of the year. A fall of 3.6 inches took place during March and 2.4 inches during April. The recession was probably at a slightly higher rate during May. On May 24 the rank of seedlings at the base of the bank included Aériplex, Cyperus, Heliotropium, Prosopis pubescens, Salix, Populus, and a grass. The experiments described on page 162 would indi- cate that the Saliz and Populus seedlings were from wind-borne fruits trapped by the bank, and that the Prosopis was from plantlets germinated in the water. The last-named, how- ever, was found only among thick bunches of twigs and other buoyant débris, and it is possible that the pods or seeds did not float independently (see Plate 26). INFLUENCE OF THE LAKE UPON THE VEGETATION OF THE DRY SLOPES ABOVE ITS LEVEL. The flood waters of the Colorado River which flowed down the slopes of the Cahuilla Basin into the Salton Sink caused an enormous amount of erosion, as described elsewhere in this volume. With the widening of the channels, bars and sand-spits were formed on which a vegetation characteristic of a river margin was established. The tops of the banks and the gentler slopes of the same were seen to support species of a halophytic character, the presence of the stream having but little effect on the plants indigenous to the region. More marked effects were to be seen on the bajadas or detrital slopes leading down from the mountains directly toward the lake. The waters of the lake were gently pushed up over these slopes and doubtless penetrated slowly to some depth. The infiltrating water would meet the slow underflow from the mountains with the result that the latter would be dammed and would be brought nearer the surface. A more luxuriant growth of halophytic and spinose forms on the steep slopes contiguous to the Travertine wash on the westward may be ascribed to this cause rather than to any increase in the relative humidity. Although every effort was made to find effects due simply to the influence of increased humidity nothing of the kind was seen, and all such appearances above the level of the strands were in places in which the structure was such as to present conditions for retarding the underflow. The amount of water evaporating from the lake was compara- tively enormous, yet the vapor formed was carried away so rapidly that it would have but little effect in checking the transpiratory activities of land vegetation and thus increas- ing the physiological value of the scanty supply of soil moisture. The relative humidity with no wind may show an increase for a few hundred yards away from the shore of the lake, but such conditions prevail only during comparatively brief periods. Some observa- tions on this matter were made near the western shore of the lake in February 1907, and the following extract is taken from my notebook. “On February 11, the wet and dry bulbs at 40 feet from the shore read 73 and 62 at 225” p. m., indicating a relative humidity of 68 per cent and at 3° 34", 7114 and 67, indicating a relative humidity of 87 per cent. In order to ascertain the extent of this humid zone the instruments were taken directly west, three-eighths of a mile from the shore, where a reading of 82 and 61 was obtained, indicating a relative humidity of 49 per cent at 3°30” p. m., and 15 minutes later a second reading showed 81 and 6014, indicating a relative humidity of about the same. It is thus to be seen that the zone of greater saturation is a comparatively narrow one, as on returning to one-fourth mile from the shore at 4 p. m. the readings were 80 and 7314, indicating a relative humidity of 81 per cent. 140 THE SALTON SEA. ‘On the evening of June 20, 1912, the relative humidity at a distance of 3 or 4 miles from the shore was 32, which showed no increase at sunrise on the following morning. It had already been previously noted that contiguity to the Delta or bodies of water in an arid region does not necessarily result in increased humidity, as evidenced by the fact that the amount of moisture in the air a few yards from the banks of the Colorado River was but little different from that of the drier parts of the desert widely separated from the Delta.” * ENDURANCE AND SURVIVAL OF SEEDS AND PLANTS IN PLACE. The principal mechanical and chemical features of the Salton Sink now having been fairly presented, it will be pertinent to turn attention to the mechanical relations and life histories of the species which compose the aggressive hosts of plants which press upon the bared areas as invaders and possible occupants of the empty areas. So far as the islands in Salton Lake are concerned, it may be assumed with fair cer- tainty that their submersion in the saline water of the lake for a few weeks or a few months would result in the destruction of all of the living plants except such forms as Phragmites, Distichlis, Juncus, and Typha, species about which there is but little to discuss in the present connection, since they are not numbered among the pioneers on islands. Populus estab- lished in boggy locations near flowing wells endured submergence of its roots to the depth of a foot or two for a few weeks without material injury. The mounds of organic material surrounding saline springs 2 miles to the westward of the railroad station of Salton, which had been submerged by the waters of the lake a depth of a yard or more during the years 1907 and 1908, were found to show living rhizomes and stems of Phragmites, Typha, Juncus cooperi, and Distichlis spicata. The water which had covered them showed a total salt content of about 0.3 per cent at the beginning of this period, which had increased to 0.5 per cent a year later. The species noted generally grow in soils more or less highly charged with salts and their survival was in no wise un- expected. The changes resulting in dead stems of shrubs and trees have received extensive and detailed treatment by Professor Brannon in another section of this volume. The tops of many of the larger plants, such as Atriplex, Sueda, and Spirostachys, with maturing seeds, the bases of the stems of which were surrounded by the waters of the lake, were held aloft in place and a portion of the crop was not dropped until after the ground around them was laid bare, with the result that a sowing was made without the aid of other distributional agencies in freshly bared strands. This effect was most marked at the extreme northwestern corner of the lake in 1908, and also to some extent on the north- eastern shore, where the water covered a gentle alkaline slope. On gently rising slopes, like those near the northwestern end of the lake near Mecca and on the eastern shore near Imperial Junction, the water flowed about the bases of the shrubby vegetation, giving the spectacle of xerophytic forms growing under aquatic con- ditions. Efflorescences and other deposits and accumulations of salts on the surface were dissolved at once, so that samples of water like that taken near Travertine Point in May 1906 showed a total solid content 1,152.8 parts in 100,000, or over 1 per cent, of which 884 parts were common salt. This is about three times as much as in the main body of the lake. This fringe of highly saline water would exert a stronger toxic action on seeds lifted from the soil than the water of the body of the lake. The shallow layers of water near shore attained a much higher temperature in the midday with the thermometer over 100° F. than that farther out in the lake. Both the higher concentration and the higher tempera- ture would increase the sterilizing effects of the rising lake. (See p. 121.) After both of these conditions have been allowed for, however, the supposition remains that the submergence of the arid soil of the isolated islands might not destroy all of the seeds present. The analysis of the manner in which a plant begins existence on a newly emersed island should take this fact into account, although the consideration is largely a 1 MacDougal, The Delta of the Rio Colorado. Bull. Amer. Geog. Soc., vol. xxxvul, p. 4, 1906. SALTON SEA PLATE 25 A. View down course of Travertine Wash across Emersions of 1911 and 1912. The steep banks of the shallow wash were obliterated during submergence. Dense formation of Distichlis, with scattered plants of Spirostachys, Atriplex, and Heliotropium. Numbers of Prosopis pubescens killed by water are seen in various stages of emersion. October 1912. B. Cut Bank marking the mid-winter of 1911-12, Travertine Terraces. Salix, Populus, Atriplex lentiformis, Heliotropium, Pluchea camphorata, and Prosopis pubescens in rank at foot of bank on emersion of 1912. Sesuvium is also included and its rounded clumps are seen on the strand of I9I1 above MOVEMENTS OF VEGETATION IN THE SALTON SINK. 141 theoretical rather than a practical one, since the upper layers of hills which became sub- merged and then arose as an island from the receding waters had invariably been so eroded and worn by the action of the waves that there remained but little question as to the presence of any plant by invasion rather than by endurance of the flood in place. POSSIBLE INVADERS OF THE STRANDS. Practically all of the species inhabiting the Cahuilla Basin, including those native to the alpine slopes of the San Jacinto Mountains, are to be included among the forms the seeds of which might be carried by run-off streams, winds, or other agencies down to the unoccupied areas around the receding lake. The differences in climatic conditions and in the soil, however, would obviously constitute an effectual barrier to the greater number of the plants native to the rocky slopes of the mountains. Some of the barriers affecting the dispersal of species on mountain slopes are much too subtle to be detected by available methods of geographic survey. This is well illustrated by the cultures made at the Desert Laboratory, in which many species abundant on the higher slopes of the Santa Catalina Mountains in positions from which their seeds must have been carried to the lowlands in myriads for centuries are not found below a certain limit, although when the seeds are transported by man to the lower lands the plantlets survive and in some instances, such as that of Juglans, outstrip the lowland species in vegetative activity. The species inhabiting the bajadas or detrital slopes of the basin, or of the lowermost part included in the Salton Sink, would be the most important elements in any invasion of surfaces left bare by the receding waters of the lake. The census of these forms is to be found in the section of this paper compiled by Mr. 8. B. Parish. A number of introduced species and weeds would constitute another element, and plants of this kind would be carried along the line of the Southern Pacific Railroad, which runs at varying distances from the shore of the lake for about three-fourths of its length. The water actually washed the ends of the ties for many miles of the line at the maximum level. It will be recalled also that a long stretch of the track previously ran below the maximum level and was moved up the slope to evade the rising waters. The other element to be considered would be the species native to the valley of the Colorado River. The entrance from this region would be principally by flotation. The number of species included would be too large to be discussed in detail. Only those with seeds which would be uninjured by long immersion would constitute potentialities in invasion from this source. The census of the invasions from 1907 to 1912 inclusive includes the following species: List of Species appearing on the Strands of Salton Sea. Amaranthus palmeri Aster exilis var. australis spinosus Astragalus limatus Atriplex canescens fasciculata hymenelytra lentiformis linearis polycarpa Baccharis glutinosa Bouteloua arenosa Chamesyce polycarpa hirtella Chenopodium murale Coldenia plicata Conyza coulteri Cryptanthe barbigera Cucurbita palmata Cyperus speciosus Eclipta alba Eleocharis sp. Encelia eriocephala frutescens Eriogonum plumatella thomassii Franseria dumosa Heliotropium curassavicum Hilaria rigida Hymenochloa salsola Isocoma veneta var. acradenia Juncus cooperi Lepidium lasiocarpum Lippia nudiflora Leptochloa imbricata Oenothera scapoides aurantiaca Oligomeris glaucescens Olneya tesota Parosela emoryi spinosa Phragmites communis Pluchea camphorata sericea, Polypogon monspeliensis Populus macdougalii Prosopis glandulosa pubescens Psathyrotes ramosissima Rumex berlandieri Salix nigra Scirpus americanus olneyi paludosus Sesuvium sessile Sonchus asper oleraceus Spheralcea orcuttii Spirostachys occidentalis Suseda torreyana Typha angustifolia Wislizenia refracta 142 THE SALTON SEA. BIOLOGICAL AND PHYSICAL CONDITIONS OF DISSEMINATION AND REOCCUPATION. The sterilization of an area of land surface and its exposure to invasion of plants and animals start a series of distributional activities which may not come to a condition of sta- bility or quiescence until after an extended series of adjustments and successions have been displayed. The initial rush of immigration to an unoccupied area would make visible to the observer the action and relative value of the disseminating agencies which are ordinarily hidden from view, demonstrate the selective action of habital complexes, and give scope to any of the possibilities usually associated with the rapid multiplication of the individuals in the successive generations of any lineal series. The inrush of forms into such an unoccu- pied area unbalanced by the competition and resistance of resident forms would permit species to exist under physical conditions near the limit of endurance and under the influ- ence of environic complexes with which they do not usually come into contact. Some special inductive effects expressed in the architecture or in the qualities of the biotic ele- ments might be expected under such circumstances. The sterilization of the Salton Sink by the making of the lake and its gradual recession leaving a continuously widening strip of beach or strand around the decreasing body of water has given unexcelled opportunities for the determination of some of the more im- portant phases of the reactions denoted above. In this instance the bared area lay in the midst of one of the most arid areas in the western hemisphere. When the fresh waters of the Colorado River were poured into the Sink of the Salton, rapidly at first and then in a diminished stream which continues through the irrigating system to the present time, there is but little doubt that the seeds, rhizomes, and propa- gating bodies of many hundreds, perhaps thousands, of species were carried down into the lake. The vegetation which might thus contribute to the swarm of invaders comprised the various types of desert plants native to the arid regions in Arizona, California, Utah, Nevada, Colorado, and perhaps Wyoming, as well as the species native to the slopes of the Cahuilla Basin. The total list of invaders which survived and made some occupancy of the beaches includes less than 60 species, as detailed on page 141. These played most unequal parts in the pioneering. Some were represented by single individuals, others by two or three, while others formed dense ranks on the beaches which must have included hundreds or thousands of individuals annually. The difference between the vegetation of the first emersion and that of succeeding seasons was far greater than that indicated by the difficulties of invasion or by the toxic action which might be ascribed to the simple and direct concentration of the lake water. In fact, a comparison of the data obtained by the chemist (see page 39) shows that a simple concentration of the lake water did not take place. Among other things the amount of potassium in the water showed no increase, although the amount present was probably adequate for the nutrition of all of the plants concerned. Next it was found that the mag- nesium lagged behind in the increase of the elements present in the water. The most striking feature of the chemical analyses, however, was illustrated by the calcium determina- tion, this element soon beginning to show a marked divergence from the main course of con- centration. The possible significance of this deviation may be best understood when it is recalled that sodium in large proportions is toxic to plant protoplasm. This toxic action does not occur in the presence of calcium when the two elements sustain certain propor- tions to each other. Whether or not this masking or antagonizing effect is produced in the complex solution of Salton water is not known.) It is not possible to predicate with any exactness the probable compounds of the ele- 'Osterhout, W. J. V. The permeability of protoplasm of ions and th th i i pp. 112-115, Jan. 1912. See also vol. xxxvI, pp. 575, 576, 1912. SERA aetibe, el ee, SALTON SEA u fe Q& A m ho a A. Cut Bank at upper limit of Strand of 1913, Travertine Terraces, as it appeared February 1913 B. Bank at upper margin of Strand of 1912, Travertine Terraces, in May 1913, after a fall of about 8 inches in level of water. Seedlings of Heliotropium, Salix, Populus, Prosopis pubescens, Atriplex, Cyperus, and a Grass were established at foot of bank MOVEMENTS OF VEGETATION IN THE SALTON SINK. 143 ments found in the Salton Lake water and it does not seem profitable therefore to reduce the chemical results to terms of normal concentration. Sodium was present in the water of the lake to the extent of 112 parts in 100,000 in 1907, and calcium to less than 10 parts, the calcium therefore being but one-eleventh of the amount of the sodium. The ratio had changed but little by June 1908, but by June 1909 the sodium was as 13 parts to 1 of cal- cium, and the period immediately preceding may be taken as being the time when the sulphur bacteria began to multiply and take on an increased aggregate activity, resulting in the reduction of the sulphates and the deposition of calcium carbonate. Although the water coming into the lake from various sources was charged with calcium in small pro- portion, yet in 1910 the ratio of sodium to calcium had increased to 14 to 1, in 1911 to 14.5 to 1, in 1912 to 14.62 to 1, and in 1913, 16.4 to 1. Many physical agencies would operate to carry seeds toward the sterilized beaches, and it will be profitable to consider some of the possibilities upon which only inferential evidence may be offered. The Colorado River at the time that it poured the greatest volume of water into the lake was in a state of flood and had spread out over the lowlands along its course, lifting millions of seeds from their resting-place on the ground; these would be carried toward the lake. Some kinds, however, because of their specific gravity, would sink, and these would have very little chance of reaching the strands of the lake, as they would soon be covered with silt. Others which might float at first would soon become softened and by the imbibition of water would swell as if for germination, with the result that these also would be destroyed before they passed through the long overflow channel leading from the main channel into the Salton Sink, and some might actually proceed to germination, the plantlets being carried still further, as is suggested by the results on pages 144-153. Some of the heavy seeds might adhere to trunks of trees and other floating objects and be earried with them to a final resting-place on the strands. In all cases, however, seeds of any kind would be subjected to the action of the saline water as soon as they were thrown into the waters of the lake. Seeds with indurated outer coats, such as some of the species examined in the flotation tests described on page 164, might be picked up by the water and carried the entire distance, or the seedlings might endure aquatic condi- tions for extended periods (see page 150). The heavy winds from the southeast, southwest, and northwest would lift and carry a large variety of seeds (such as those of Baccharis and Pluchea) long distances. Some might fall to the ground directly on the bared strand exactly at a time when the germina- tion conditions were most favorable. Much greater numbers would tend to fall on the surface of the stream and of the lake itself. Once in the water, their further transportation would depend upon wave-action, which would be produced by the wind. Traveling animals, especially birds, might bring seeds attached to their feathers or imbedded in mud clinging to their feet. In the present instances the principal birds were the pelican, cormorant, and various species of ducks. The first two feed upon fish and habitually alight either on rocks or branches of trees rising above the level of the water, or upon the arid rocky soil about their nests, as noted on the various islands. In one case the seeds would be dropped in the water and their origin would escape analysis, while in the second instance the soil would be too dry to induce germination, which if it did take place would end disastrously, since the surface about the nesting-grounds was trampled and worn so that green plants had but little chance of survival. Pelicans also alight and stand in ranks along muddy shores, and might in this manner deposit seeds in rows or lines. The lake occupied the bottom of the bowl of the Cahuilla Basin and drainage channels from all of the converging slopes led toward it. In none of these, however, were there constant 144 THE SALTON SEA. streams. The run-off or surface flow resulted from the precipitation accompanying the heavier storms, and when the streams were formed in this manner the water rushing down the steep slopes of the bajadas undoubtedly would pick up the accumulation of seeds lying in the shallow channels and carry them down into the lake, or perhaps cover some with sand and silt which with the included moisture would be very favorable to germination and survival. The invasions thus facilitated would result simply in the advance of species down the slopes crossing the strands at right angles, generally in places perhaps occupied previous to the making of the lake. Such action was of course noticeable on the upper- most beaches left bare during the first year’s gecession of the lake, and transportation by this method became less and less efficient as c. distance from the maximum-level shore- line was left behind. a The flotation by run-off streams becomes :j more effective method for carrying plants onto the bared strands, especially from the fat that no prolonged subjecting to soaking would be endured. The subjection of the seeds to the brief action of the ephemeral stream and its subsequent contact with the moistened sand or soil would be of a character highly favorable to survival. It is evident that the soil conditions of the beaches during the first year of recession were different from those offered in any following year. Briefly stated, the surface layer had been leached to a slight depth by water containing the lowest pro- portion of salts. Another feature favorable to the development of the comparatively heavy vegetation of the emersion of the first year was the fact that the rapid rise of the lake would have lifted seeds from the ground and thus at the theoretical moment when the lake was at its maximum level the number of floating seeds which might be driven ashore by wind and wave action was greater than at any subsequent time in the history of the lake. The toxic activity of the water itself was least at the maximum level and began to increase at once. As was noted in the history of the earliest strand, the dead stems of some species remain- ing erect in place held aloft fruits which were not all cast off until after the waters had receded. FLOTATION AND GERMINATION. Since it appeared that the greater number of the species were finally carried to the emersed zones around the lake by water it was deemed important to make tests of the behavior of some of the forms when placed in water taken from the lake. The conditions were so specialized that it will not be profitable to review the literature of the general sub- ject farther than to call attention to the chapter on dissemination of seeds in “Origin of Species”’ and also to Guppy’s more recent observations.! These tests would have had their maximum value if they had been begun when the concentration of the water of the lake was lowest and the reactions of the seeds year by year correlated with the occurrences on the beaches. The overweening importance of flotation as a final means of dispersion was not evident, however, until the work was well under way. A supply of Salton water was taken from the lake in October 1912 and shipped to the Desert Laboratory at Tucson. The nearest analysis was that made in June 1912 (see page 47) and the concentration had in all probability slightly increased, so that the solid matter dissolved amounted to about 0.9 per cent. A series of glass dishes was provided on a long table in a room with a north light. No artificial heat was supplied and the range of temperature, due to the fact that the room was kept closed during the first two months of experiments, was between 40° and 60° F. At the close of this period all of the preparations were removed to a glass house without artificial heat, where the temperature ran as low as 50° F. at night and up to 88° F. during the daytime. The range of temperature and the alterations thus provided would be so +See Darwin, Means of Dispersal, etc., Origin of species, pp. 324-330. Reprint of six iti ~ople. 187 Guppy, Observations of a naturalist in the Pacific between 1896 and 1899, ol a 10H ata Daan MOVEMENTS OF VEGETATION IN THE SALTON SINK. 145 nearly similar to the conditions on the shores of the Salton as to make the results of direct value in a discussion of the action of the species treated on the shores of the lake. Seeds of species which would represent the main features of the occupation of the beaches and also the principal anatomical types were procured, and these were thrown into dishes containing about 400 c.c. of Salton water, after which a loosely fitting cover was provided to prevent evaporation and consequent concentration of the solution. After standing for some time, the water would be poured off and a new filling would be made in order that the effects of heightened concentrations might not introduce a vari- ation into the results. The preparations were examined at intervals of a day or two during the first two months, and each time the beaker would be lifted from the table and shaken violently to simulate disturbances of the water due to wave-action. It was realized that this did not give the entire effect of disturbances of the lake where the wave-action might continue for a week without intermission. On the other hand, the shaking resulted in many instances in sending numbers of seeds to the bottom which had been buoyed by adherent bubbles“of gas attached to their coatings, and while some of the seeds probably floated a few days longer in the experimental vessels than they would have done in the lake, yet the prolongation was probably not more than one day in ten. Amaranthus palmeri is a rapidly growing plant of short cycle and widely varying habit. It may reach maturity with a short unbranched stem a few inches in length, while on the other hand great branched stems may reach a height of 6 feet in the summer in the Delta lands of the Colorado. Here it is known as ‘‘quelite,” and the green stems are eagerly eaten by grazing animals. The Cocopah Indians collect the seeds in late summer and pre- serve them to be used as food in the winter. The interrupted flower spikes terminate the stems and branches and the indehiscent utricles hold the seeds so that they are cast from the plant during an indefinite period. The shining dark lenticular seeds are not over 0.8 mm. in diameter and less than half this in thickness, with an average weight of 0.3 mg. 100 seeds were thrown into a beaker of Salton water on November 16, 1912. The vessel was shaken daily at each examination and a few would go to the bottom, with the result that by January 10, 1913, all but one had sunk. The rising temperature started imbibi- tion with the result that swollen seeds were brought to the surface, being buoyed by gas formed by the end of January (January 25). A week later three seedlings were floating and two more were seen on February 17. The total number of germinations reached seven, and the seedlings had a combined length of root and hypocotyl of over 2 cm. As these seemed to come to a standstill all were ‘‘stranded”’ on March 10. Amaranthus was found on the muddy shores of Obsidian Island and Imperial Junction Beach in 1908 and 1909. Its failure to appear later can not be connected with anything except the increasing salinity of the water. Although germinations ensued in Salton water of June 1912, yet the proportion was very small and the development definitely limited, while the stranded plants failed to survive. The flotation period of the seeds, 40 to 50 days, was one which would allow it to be carried long distances by water. Its small size and weight would likewise render it liable to be carried by birds or winds. The actual locations of the individuals in the filled washes and on the shores was highly suggestive of flotation. Atriplex fasciculata is abundant in the Salton region, where it forms a small shrub, reaching a maximum height of 2 or 3 feet on various beaches. It is native to the higher parts of Big Island and Obsidian Island as well as of the slopes about the lake, and was a prominent feature of the first group of species appearing on emersed beaches. The germination of seeds within a few feet of the margin of the water suggests that the fruits might have floated to the place or have been carried there by birds, although no actual tests could be made which would offer evidence on the subject. It is probable that the species, like A. canescens, may survive progressive desiccation to the general condition 10 146 THE SALTON SEA. of the slopes of the Sink, although the observations on this matter have not been made with the exactness with which the matter was treated with regard to A. canescens and Parosela emoryt. A lot of seeds were cast into a vessel of Salton water on February 13, 1913, with the result that all floated for a few days, then began to sink, so that nearly all were on the bottom in three weeks. No germination ensued. The failure of the seeds to germi- nate may be ascribed to their age, and no opportunity was obtained for making a test with fresh seeds. Atriplex lentiformis is a shrub which attains huge dimensions, the larger individuals making a dome-shaped mass which may be 15 to 25 feet high with a diameter nearly as great. The flattened utricles weigh 7 or 8 mg. and mature in the autumn. A supply was collected by Mr. 8. B. Parish in October 1912 and 100 of these in an air-dried condition were thrown into a vessel of Salton water on November 16, 1912. These seeds showed a marked partial buoyancy. Nine sank on the second day after shaking, but the greater number remained afloat and germination ensued in both lots within a week. The radicles elongated so rapidly that a length of 1.5 to 2 cm. was reached within a week, and the plantlets remained in indefinite positions in the water, some resting on the bottom at a depth of 10 cm., others midway, others floating on the surface, all changing position to some extent with illumination and temperature variations (see Plate 27). The total number of germinations amounted to 85, and with warmer weather all of the plantlets rose to the surface, where they were noted on January 10, 1913. Lots of these were placed in the shallow water covering loam in a glass dish, but none of them struck root and sur- vived the ‘‘stranding,” although it is possible they might have done so if they had been tested as to this matter earlier; some attenuation and weakening had already been mani- fested when the trials were made. This species is abundant in the Mecca region and the seeds might readily be carried down the beach slopes by run-off streams, birds, or other agencies. The introduction of the plant on the Travertine Terrace and on Obsidian Island, however, might have been due to birds or to the flotation of seedlings or to seeds, although the distance to be bridged renders the last-named method the least probable of the three. Among these introductions is the notable instance of a single individual on Cormorant Island, in 1912, its only known introduction to this place in five years. An occurrence of no less interest was that noted in October 1912, when some plants were seen on the emersed archipelago near Big Island. Flotation of seedlings or transportation of seeds by birds is suggested in the last-named instance (Plate 23 a). Baccharis glutinosa is a woody composite (reaching a height of 2 meters or more, with lanceolate glutinous leaves) which blooms all through the summer, and the compressed ribbed fruits, with a long silky pappus, become free from the receptacles in November and December and are supposedly carried about by the winter winds, many of course falling on the water. It proved difficult to free the fruits from chaff for weighing, and although they were seen to be not the lightest seeds tested, yet the surfaces offered by the pappus caused them to be carried long distances by air-currents; 100 of the fruits taken from an herbarium specimen collected in November 1908 were thrown into Salton water November 16, 1912, and floated without being wetted for some time. Shaking caused two to sink to a slight depth on the following day, and these had descended to the bottom on the 26th; five were down on December 9; eight were at the bottom on December 14 and no further changes were noted except that mould had formed on the floating fruits. This state of affairs prevailed as late as January 20, 1913. Shaking did not sink any more of the fruits, which were found to be nearly all imperfect; 35 which had sunk seemed to be sound. _No germinations having ensued by March 9, or 115 days after the seeds had been put into the water, the preparation was discarded, as at this time the seeds had SALTON SEA PLATE 27 cS | 4 | 3 XG AE Flotation of Seedlings of Atriplex lentiformis in Salton Water, concentration of 1912 MOVEMENTS OF VEGETATION IN THE SALTON SINK. 147 decayed. Baccharis appeared as a pioneer on the muddy Imperial Junction beaches in 1907 and 1908 and also on Obsidian Island, where it became strongly established. A single individual was one of the first two plants on Cormorant Island in 1908, but only a single individual was added to the place in the following four years. It also appeared on the archipelago near Big Island as found in 1912. All of the introductions were in places suggestive of wind deposition, and in the southeastern part of thelake. The locations are those most liable to receive material by southeast winds from the Delta, in which the species is especially abundant. The original introduction on Obsidian Island was still flourishing in October 1912, and the single plant on Cormorant Island had survived until the same date. Oligomeris glaucescens is the sole representative of the Resedacee in this region. It is a glaucous herb, branching at the base, with the branches ascending; its terminal flower spikes become mature around the Salton in April and May, while the ripe capsules con- taining seeds are found in place as late as the following February. The seeds are very rounded and slimy and their weight and dimensions appeared to be about the same as those of Sesuvium. Dried capsules from herbarium species collected in February 1908 were broken up and the mass of chaff and seeds was thrown into a vessel of Salton water November 25, 1912. No exact account was made of the number of seeds, but about three- fourths sank within a day. No further change was noted until the preparation was placed in a sun-lighted glass house on January 20, 1913, when all of the remainder of the seeds sank within a day. A single germination was noted 86 days after being wetted, and a dozen more were noted March 10. Preparations were made for ‘‘stranding”’ the lot, but all began to degenerate and were dead within a week. These reactions and the occurrences of the species only in 1907 and 1908, on the shores at Imperial Junction Beach, Mecca, and Obsidian Island, suggest that the condition barring the species from later emersions consisted of the increased salinity of the water. The small size of the seed would make it liable to be carried about by birds, while the period of flotation indicated would cause it to be carried for distances between Obsidian Island and the shores to the southward. It is notable that it did not appear on the steeply sloping beaches of Travertine Terraces. Isocoma veneta var. acradenia is a perennial composite especially abundant on the slopes and bajadas west of the lake, and generally found on very arid locations. The pappus bristles are stout and it does not appear liable to be carried long distances by the wind. Fruits taken from an old dried specimen were put into a dish of Salton water on Novem- ber 25, 1912, and a few sank every day, so that all were on the bottom a month later. The proportion of perfect seeds is probably small and only two good embryos were found in the entire lot on January 25, 1913, at which time no indications of activity had been shown. This plant was found on the emersions after they had been bare for some time, two years generally, and then only on the western shore, in positions into which they might have been carried by the wind or run-off streams from the parent plants which were only a few hundred yards away. Juncus cooperi is a rush of limited distribution in the deserts of the Colorado and Mohave, occurring around saline springs. The reddish-brown cylindrical seeds weigh about 0.015 mg. and are furnished with an appendage which readily absorbs water. Despite this fact the seeds floated from November 16 until December 4, at which time only 10 of the original 100 sank, and the remainder followed on the 23d, showing a flotation period of 37 days. Germination began soon after sinking, and by January 10, 1913, the hypo- cotyls were seen emerging from nearly all of the seeds on the bottom of the vessel. The seedlings became entangled by the extremely long trichomes which arise from the base of the hypocotyl, and rose to the surface in clumps with illumination and warmer weather late in January. Some of these clumps were “stranded” on January 29 and had taken hold a week later. The schedule of the species is thus seen to include three weeks of flota- tion of some of the seeds, which is extended to nearly twice that time in others. Germina- 148 THE SALTON SEA. tion begins about six weeks after contact with the water and seedlings which had floated three weeks were capable of taking root when stranded. All of these periods would probably be shortened in seeds liberated in late summer during warmer weather. The observations of this plant show that it came on the beach only at the Travertine Terraces, where it appeared on the bared zones during the first year of emersion. Its ap- pearance soon after the emersion of an old accretion mound of a saline spring in 1909 suggests that the stem-bases and perhaps the roots are capable of prolonged submergence. The restricted establishment of the plant, with its comparatively long flotation period of seeds and plantlets, is probably a matter of soil; both seeds and plantlets must have been carried to many shores both by water and by birds. Leptochloa imbricata is an annual grass which appears to be abundant throughout the Delta of the Colorado, where fruits consisting of the minute grains inclosed in the glumes were collected by Mr. 8. B. Parish in October 1912; 100 of these were placed in Salton water on November 16, 1912, with the result that one germinated within 10 days, while the remainder were still afloat and inactive until the end of January, when all of the sound embryos awoke with the rising temperatures and increased illumination. After allowing the plantlets to float for two weeks, the entire lot were ‘‘stranded,”’ with the result that a fair proportion took hold and sent roots down into the moist saline soil. The sole observations of this plant concerned its appearance on the emersed zones of Imperial Junction beach and on Obsidian Island in 1907, 1908, 1909, 1910, and 1911. Both places are directly affected by the inflowing current of the Alamo River from the Delta, and in both the recession of the water is followed by a desiccation of the soil which soon terminates the existence of the plant. It is notable that Leptochloa was not seen in any position in which it might not have been deposited by waves and that the radius of its dispersal does not extend more than 8 or 9 miles from the mouth of the fresh-water stream which might bring it into the lake. The small fruits might readily adhere to the feet of birds or to floating driftwood. Pluchea camphorata is a stoutish annual with its clumps of stems reaching a height of about 2 feet in late summer. The achenes are furnished with capillary bristles and these would be cast loose from the plant in time for germination in the warm weather of autumn. Seeds collected near Mecca in mid-October 1912 were placed on the water on February 11, 1918, and remained on the surface. The first germination was noted on March 1 and during the next 10 days it appeared that practically the proportion that might be expected had awakened. The fruits had not been separated, as a result of which small tangled clumps of bristles and seedlings floated on the surface or were suspended at various depths in the water. A test was now arranged by which the water in which the plantlets were floating was poured into a glass dish filled half-full with packed soil. After this had been done the dish was tilted at such an angle as to give a miniature beach and a half dozen were ‘‘stranded”’ in the process. Representatives of this species were found on the emersion of 1907 in May 1908 at Mecca, and it was seen to be distributed down the slope, following the recession of the lake at similar intervals as late as 1913; but seeding did not always follow the first occu- pation, and the plant was noted on the 1907 strand at Travertine Terraces, from which it afterward disappeared, but it was noted here, on the emersions of 1909, on which it persisted until the preparation of this manuscript, and it came in on the emersion of 1910 at the same place (Travertine Terraces). The emersion of 1911 was also occupied by this plant at Travertine Terraces, and it was found on the islands which rose from the water near Big Island in 1909. It is plain, therefore, that fruits or plantlets of the species are transported across stretches of open water. The instances in which the occurrence of the plant could be best correlated with the phenomena of emersion were at the Travertine Terraces, and here the plant was not seen within a year’s recession of the margin of the MOVEMENTS OF VEGETATION IN THE SALTON SINK. 149 water. The only place in which the plant was found near the margin of the lake was on newly bared areas in the archipelago of Big Island and the weight of probability lies with the conclusion that the fruits were carried to this place by winds. The fact that the fruits do not wet readily would operate against their being carried by adhesion to the legs or feathers of birds. The stranding test was arranged to bring the plantlets in contact with a substratum much like that of the shores of the lake and at a time when the seedlings had not been weakened by prolonged flotation. Prosopis glandulosa is the ‘“‘mesquite’”’ of the Colorado desert. It is widely variant in habit, being no more than a shrub on the upper parts of the bajadas or rocky slopes, while it becomes a tree of some size in the Delta, where it is the dominant woody plant. Its abundance in the alluvial lands and along streamways gives especial significance to its relations to water. The pods are compressed and indehiscent, 10 to 12 em. long, ripening in the autumn, and the seeds are surrounded by a mass of dry, spongy tissue which seems to contain a fermentable sugar. A large share of the pods are so completely bored by some insect that perhaps not more than one or two perfect seeds may be found in each. Flota- tion tests were made by placing whole pods in Salton water and others were broken into fragments of various lengths. The first trials were begun in January 1913, with the species native to the Tucson region. Naked seeds sank like pebbles. Pods and fragments, however, sustained various positions in the water, some floating, some a short distance below the surface, and some on the bottom, according to the buoying capacity of the gases resulting from the fermenta- tion of material in the pods. No germinations were obtained in the first lot of seeds, prob- ably by reason of the unfavorably low temperatures; some of the seeds became swollen, but the decay of the pods made it advisable to discard the preparation. Next, a lot of seeds was put into Salton water on January 25. A month later some were removed to a vessel containing only one-fourth Salton water added to tap water. This seemed to make no difference in their activity, however, as both lots germinated in about equal numbers by March 10, at which time the temperatures were rising. The seedlings did not become free from the seed-coats within a fortnight, but remained at the bottom until deteriora- tion set in. Next, a lot of pods collected in the Salton region, early in March 1913, by Mr. 8. B. Parish, was put into water. The fermentation of the material in the fleshy part of the pods made a strong odor, but no germinations ensued. These tests make it clear that the mesquite may not be disseminated by flotation of the seeds or seedlings to any great distance. The fleshy pods are attractive as food to many animals, perhaps to the larger birds, although the author knows nothing as to the extent to which they are so used. The mesquite appeared on the beaches of 1908 and 1909 at Travertine Terrace, but has not formed a member of the beach communities of later date. The single specimen which was among the pioneers on Obsidian Island was transported to that place as a living branch, which was half buried in the sand of the beach and gave rise to the tree, the appearance of which has been described. (Plate 21 a.) Prosopis pubescens, or screwbean, has played the most prominent part of any tree on the beaches of the Salton. The indehiscent pods are rolled up into compact cylinders, which mature in late summer, but remain attached to the tree indefinitely. The first tests with the seeds were made with material taken from herbarium specimens of the University of Arizona. These weighed about 6 mg., and sank to the bottom of a vessel of Salton water in which they were placed on January 19, 1913. Swelling began within a week, and one was germinating by the 31st; fifteen more, representing the entire number of perfect seeds, germinated within the next ten days. The cotyledons were released from the seed-coats somewhat slowly; and four rose to the surface of the water on February 15, two weeks after germination began. These carried on a slow but normal development for 150 THE SALTON SEA. a month, and at the end of this time some generation of the radicles of two was noted, and the lot was “stranded” in soil saturated with Salton water of 1913. None survived, and it is probable that the time of actual flotation without injury is somewhat less than a month and that the conditions for establishment were unfavorable in any one of a dozen features. The fact that the seedlings float at all and come to the surface would make this plant more liable to be carried about than the mesquite, which, as described on page 149, does not free the seedling in the manner of this species. Pods taken from a tree in Feb- ruary 1913 sank to the bottom. Neither the seeds in these, nor in others freed from them, showed any action in a month; and no suggestion is to be made as to the agencies which might be cooperative with flotation in facilitating dissemination. The screwbean was a marked feature on the steeper beaches laid bare on Obsidian Island and at the Travertine Terraces, where it has been found since 1907. Plantlets were seen late in the autumn on strands laid bare during the summer, presumably from seed of the preceding season. The greater number were immediately off the cut bank, marking the mid-winter stage of the water, and as the seeds are too large to be carried by the wind, and as they also sink, the presumption is in favor of the suggestion that seedlings are carried to the place by the waves. (Plate 29 a.) Rumezx berlandieri is a perennial which appears to reach maturity during the first season in the Salton region and plants may be found in bloom from May to October or November. The fruits, consisting of ovoid achenes inclosed in the sepals, may be held on the stems for several months, and their average weight is about 1.5 mg., of which the achene makes at least two-thirds; 100 of these fruits, from a dried specimen collected near the lake in February 1908, were thrown into a dish of Salton water on November 25, 1912; 5 sank on the next day and a few more went down daily, until half were on the bottom of the dish a week after being wetted. Germination began a week later with those still remaining afloat and continued for over a month, at the end of which time the plantlets formed a tangled mass. Near the end of this time the seeds at the bottom began activity and the young plantlets rose to the surface, as a result of the formation of gas in the green tissues. The germination of the entire lot extended over a month. A lot of twenty achenes was separated from the calices and thrown into the water to ascertain whether or not flotation was due to the seed or the adherent structures. All but one sank within three days and this one was probably defective. Nearly all had germinated within a week from the time they were wetted, although the temperature was now higher than when the origi- nal lot was tested. The plantlets rose to the surface as they attained some size, although three that were entangled with some débris on the bottom of the dish remained there without injury for a fortnight. Two months after the first germinations had taken place a tangled mass of plantlets was ‘‘stranded”’ on loam saturated with Salton water and many survived and sent down roots into the soil. The soil, however, was moistened with Salton water taken from the lake in 1912 and the plants soon perished. The long flotation of the plant is a condition that would allow it to be carried great distances by wave-action or by currents, and doubt- less introductions would result from the stranding of water-borne seedlings which might be grounded on soil the water of which was not too high in salts. Confirmation of these suggestions is afforded by the behavior of the plant on the beaches. Up to 1908 it was met frequently in various localities, but after this time it did not appear until 1911, when the shore being laid bare at Imperial Junction had been overlaid by silt from the Alamo River in the Delta of which it was very abundant. Scirpus paludosus did not appear on the shores of the lake until 1908, when it was seen on the Imperial Junction beach. The plant was apparently carried the full length of the lake during the year and came onto the beach at Mecca in 1909, while it did not figure among the pioneers on the Travertine Terraces until 1911, although Scirpus olneyi MOVEMENTS OF VEGETATION IN THE SALTON SINK. 151 had been seen there in the two preceding years. Nothing is known as to the part played by propagative bodies in this dispersal. It is to be recalled that the species was not noted on the shores of any of the islands. Scirpus paludosus is a bulrush first known from the upper waters of the Colorado River, and since it was not included in many collections along the lower courses of the river more than twenty years ago, it seems reasonable to assume that its flotation down stream and into the Delta (see Parish, page 106, this volume) is a comparatively recent matter. Nothing may be said as to the causes which may have set up the distributional movement. No question remains, however, as to the fact that it is now much more widely disseminated than a few years ago. The three-angled flattened achenes weigh 3 mg. and they come free and clean when mature. Mr. 8. B. Parish collected a supply in the Sink in October 1912 and 100 of these were placed in Salton water November 16, 1912. All floated at first and none showed signs of wetting or of absorption of water until the 25th, when a few were seen adhering to the walls of the glass vessel. A day later some were water-logged and remained immediately beneath the surface. Shortly one sank to the bottom, and this process continued so that 14 were on the bottom on December 14, a month after being thrown into the water, while the total number on the bottom was but 24 on January 19, 1913; about 31 in all went to the bottom, where they remained for ten days, when some of these, as well as some on the surface, germinated and the number increased so that 17 seedlings were counted on March 17, and these were all on the surface. Some which had been floating a month were now “‘stranded”’ in a dish in which the water was that taken from the Salton in February 1913, and it was seen from the first that their survival was assured. Sound plantlets were seen on the water as late as April 4, six weeks after germination. A large number of sound seeds remained, some floating and some on the bottom. It is to be seen from this that the flotation properties of either the seeds or of the seedlings would be an important factor in the dissemination of the plant, and would easily account for its presence in the Delta or any place in the Sink. Some unusual charac- ters of fruits were displayed by the plants of this species collected from the Salton beaches. Sesuvium sessile is the prostrate sea-purslane of the beaches and saline areas of widely extended territory. It blooms during the entire warm season and seeds may be collected at almost any time in the Salton region. These are extremely small, weighing only 0.04 mg. A supply was procured in October 1912 and 100 placed in a dish of Salton water on Novem- ber 16, 1913. All floated until December 9, at which time (23 days after being put into the water) 1 had germinated and 13 sank upon shaking. All sank by January 13, 1913, and the original seedling was still alive, no other germinations having taken place. With the coming of warm weather the outer coats were shiny with a thin layer of collected gas and they began to rise to the surface, where all were floating on January 26. Germinations now began and apparently all of the sound seeds had sprouted by February 10. A lot were ‘“‘stranded”’ on February 15 and a large number sent roots down into the saline soil and were quickly established under terrestrial conditions. It is evident that the behavior of the seeds is such that they might be carried in almost any manner about the lake, and the species appeared on emersed beaches soon after the soil was bared in the main observational areas; it is also included among the small number found on sterilized islands. Like nearly all succulent halophytes, it does not survive long in desiccated soils. Spirostachys occidentalis is a succulent halophyte which occurs in saline areas in this region. The seeds are retained indefinitely after maturity, and hence it was possible to secure a supply of the crop of 1912 from plants near Mecca as late in the season as Feb- ruary 7, 1913. Some were thrown into a vessel of Salton water on February 14 and all had sunk within 24 hours. Germination began within the second day, and all of the perfect seeds had germinated within 100 hours after wetting. The seedlings soon began to rise to 152 THE SALTON SEA. the surface and most of them floated there until March 17, when they were dredged out in clumps and placed in a tilted dish (containing soil and Salton water of 1913) in such man- ner that they were stranded. Two days later it was seen that some were beginning to erect the plumules and to show indications of survival, which was followed by complete estab- lishment. The capacity of these seedlings for flotation is a matter of especial interest in connection with the fact that the seeds sink so readily. Perhaps no other species has occurred so numerously upon the beaches as Spirostachys. It was notably abundant in the strip of beach at the uppermost level of the lake, where the soil containing a supply of the seeds was moistened for only a short time, and it also arose from seeds falling from stems which had maintained an erect position in the mud. The ready germination of the seed suggests that the species may not be carried far by the flotation of these bodies, although germination might occur soon after immersion at lower temperatures; yet the seeds sink readily. On account of the long period of possible flotation and survival, the plantlets may be carried great distances by currents, and the arrangement of plantlets on the beaches at Imperial Junction and at Mecca suggests the stranding of seedlings which may have arisen by germinations almost any- where about the margin of the lake. Spirostachys was notably lacking on the emersion of 1910 at Travertine Terraces for a period after the beach was laid bare and it generally followed the recession of the lake at some distance at this place. Flotation would be excluded as a possible means of dispersal in all such cases, including its appearance on the higher level of Cormorant Island some years after the locations had been left behind by the water. Here and elsewhere the only reason- able supposition would be that seeds had been carried by strong winds or by birds. Spirostachys appeared on Travertine Terraces in the emersion of 1912 by the middle of the summer of that year, and during the same season it also arose at other places, in a manner suggesting that the conditions for dispersal were highly favorable, although no reasonable conjecture may be made as to what these may be. Plantlets have been found in the mud near the margin of the water, yet the plant survives progressive desiccation without serious changes in the shoot. It is still a member of the formation on the uppermost levels of the Imperial Junction, where it first appeared in 1907, and it has not been lost from any location in which it came as a pioneer with the exception of Travertine Terraces, where seepages from an underflow were seen to furnish a supply of water which possibly contained too little salt. Sueda torreyana is an halophyte with soft subterete leaves, and the plant has a some- what indefinite length of existence, maturing its fruits in late summer. With Sptrostachys it is the most abundant species of the Sink, and it is especially abundant in places with saline soils. Seeds were collected near Mecca early in March 1913, and as they are retained indefinitely on the stems the supply of fruits lasts throughout the year and renders the species continuously subject to the action of dispersing agencies. Half of a lot placed in Salton water of February 1913 sank within three days and the remainder soon followed. Germination began within two days of the wetting. Free seedlings floated, though most were held at the bottom by portions of the utricle from which they had not become freed. A number were stranded in the usual manner on March 20, after they had been subjected to flotation for two weeks. The seeds must have been present in the soil in immense numbers and those which were simply wetted but not moved near the upper limit reached by the lake in February 1907 doubtless constituted the majority seen on the ground laid bare by the recession of that year. It has also been noted that many of the stems remained erect although sur- rounded by water, and that the seeds which later fell around their bared bases constituted a part of the pioneer flora of the gently sloping beaches, such as that of Mecca. The species is to be included among those near Travertine Wash which were benefited by the hemmed SALTON SEA PLATE 28 ee eS ee A. Rounded Clumps of Sesuvium, and single Plants of Spirostachys and Atriplex on the Mecca Beach, Strands of 1909 and 1910. Photographed October 1911. B. Dead Stems of Sueda, Atriplex, and Spirostachys holding fruits from which the flooded ground of the Emersion of 1907 might be re-seeded, Mecca Beach, February 1908. C. Rank of Sueda torreyana, upper margin of Strand of 1908, near Salton Station. Photographed June 1908 MOVEMENTS OF VEGETATION IN THE SALTON SINK. 153 underflow, so that a more vigorous growth was made on the ground some little distance back from the recent high beach line. It is notable that although it was abundant here it did not at any time come onto the beaches of Travertine Terraces, which were less than a mile away, except in 1908, in which year it was also found on the shores of Obsidian Island; it did not survive in either place. It formed along the shore line near the Salton Slough in 1909, but has not maintained itself there. Sueda was growing on the higher slopes of Big Island when the lake was formed and continued to survive there, yet it did not come onto the beaches of the outlying archi- pelago less than a half mile to the southward, which emerged from the water in 1909 and 1910, although the northwest winds would be in a direction favorable for carrying them to this distance. It would be allowable to suppose that some seeds were carried by birds, yet the greater majority of the occurrences outside of the special cases noted above might be attributed to the action of run-off streams which carried the seeds down the slopes to the beach, and those remaining at the margin or in the shallow water germinated in the zonal form dis- played in Plate 28 c. Seedlings of such origin or from fruits carried to the water of the lake by any agency might endure sufficient flotation to cause them to be carried to some distance, although the failure of the plant to appear on two beaches near maturing crops of fruits does not support such an idea very strongly. It seems very probable that with the increasing concentration of the water of the lake the species may appear on other slopes not yet sufficiently saline to offer suitable habitat conditions. OCCURRENCE AND BEHAVIOR OF VARIOUS SPECIES ON THE BEACHES. Aster exilis was represented in the emersion of 1907 at Mecca by one plant which was seen in May 1908. It was next seen in the corresponding zone on Obsidian Island in late November 1908. It also appeared on the emersion of 1908 with A. spinosus. It played no further part in the occupation of the beaches, although growing in abundance in the Delta and on flooded lands along the inflowing streams. Myriads of fruits must have been carried over the lake and lodged on the moist strands. The failure to germinate may only be attributed to the concentration of the soil salts, although this has not been tested. Aster spinosus appeared on the shore of the southeast point of Obsidian Island in the emersion of 1907, as was observed for the first time in May 1908. The plants were in fruit when examined in November of that year. No further occurrence of the species in emersed zones was reported. Its transportation to the locality mentioned was probably due to birds. Astragalus limatus is a native of the arid slopes and bajadas in the western part of the Sink and its appearance on the beaches was confined to those to which it might have been carried by run-off streams. It occurred on the Travertine Terraces on the emersions of 1907, 1908, 1909, where it has maintained itself. The first appearance of the plant in any place was not on soil that had been bared less than a year. Several clumps of the plant on the Travertine Terraces were observed in full bloom on February 8, 1913. Atriplex canescens forms a small shrub about 3 feet in height and is very abundant in the Sink. It is therefore up the slope from almost all of the observational areas. Old plants fell down with the crumbling banks upon the highest beach of Travertine Terraces, although they did not survive; but seedlings arose here and upon terraces of the following years. The plant was also a constituent of the first lot of plants on bared beaches at the other observational areas. It is to be noted that while it came onto the moist soil within a comparatively short distance of the water, it survived the progressive desiccation. In this behavior it is comparable with Parosela emoryi, which comes on recently bared gravelly and sandy beaches and persists throughout the changes resulting from progressive desicca- tion, although in soils with a much smaller salt-content. 154 THE SALTON SEA. Atriplex hymenelytra is native to the upper slopes of Obsidian Island and it was seen to follow down the slopes on the steeply sloping beaches. It is not known whether the fruits could be transported by flotation or not. The suggestion lies near that its survival may be determined very largely by the composition of the soil. Atriplex linearis was one of the four species of this genus which appeared among the earliest occupants of the alkaline strand at Imperial Junction beach; but if present after that year, it was represented by so few individuals as to escape observation. Atriplex polycarpa, which ordinarily forms a shrub less than 3 feet in height, was observed but three times on the emersions, one on the beach of 1907 at Mecca and on the beaches of 1907 and 1908 at Travertine Terraces. It did not survive beyond the season and its general behavior here may be correlated with the fact that it ordinarily is found on rather dry soils, not very highly charged with salts. Bouteloua arenosa appeared on the emersion of 1907 on the western side of the lake, and “also upon the strand of 1908 at the Travertine Terraces, but it soon disappeared from both places. Chamesyce polycarpa hirtella appeared on beaches on the western side of the lake in 1908 in places in which it might readily have been carried by run-off streams. The rapid widening of the beaches upon the gentle slopes, however, was not followed by its appear- ance a second time. , Chenopodium murale was found on the gentle slopes of the beach laid bare at Mecca, as seen in February 1908, but it could not be found there three months later. A second occurrence was noted on Imperial Junction Beach in October 1909, but the plant did not become established in either location. The seeds seem to be subject to the action of many carrying agencies, as it is to be seen in many widely separated parts of the world. Follow- ing these observations, it was next noted by the author along Khor Adit, leading down into the Red Sea, at an elevation of 1,500 feet, in northern Sudan. Conyza coulteri was found in the emersion of 1907 at Mecca, during the visits of February and May 1908, but did not appear as a constituent of the beach flora on any subsequent occasion. Coldenia plicata was found on the ground laid bare in 1907 on the western side of the lake, but no further participation in the occupation of the beaches was noted. Mature plants fell with the masses of soil onto the beach of 1907 at Travertine Terraces, but appar- ently all died; although seeds of this plant must have been sown on other beaches here, none survived. Cryptanthe barbigera was found on the southeast shore of Obsidian Island in such position with respect to the high level of the water as to give rise to the inference that it may have been carried there by birds. The burr-like fruits might readily become attached to feathers or carried on muddy feet. Although maturing fruits were seen, no reproduction occurred. Cucurbita palmata inhabits the dry gravelly slopes of the Sink and has been seen on the bajadas west of the lake, maturing fruit in late summer or in mid-winter. The globular fruits have a hard outer shell and are very light, in consequence of which they may be rolled about by the wind, carried by small run-off streams, or floated by the lake. This species was introduced more than once at Imperial Junction beach, but did not survive long there. Neither here nor on the gravelly beaches of Obsidian Island did it come ashore later than 1909. The defection might be partly due to two causes: the supply of fruits which would be brought onto the waters of the lake would decrease with the recession of the shore line, and the increasing salinity of the water would be injurious to the seeds; this last point was not tested. The plant survives in the emersion of 1909 on the gravell me ; : = y beach of Obsidian Island, which furnishes the conditions of its usual habitat The size and character of the fruit are such that it may be said with certainty that it has been trans- ported only by flotation to the islands or to Imperial Junction Beach. (See Plate 29 a.) SALTON SEA PLATE 29 A. Prosopis pubescens at upper margin of Strand of 1907, west bay of Obsidian Island. Fruits of Cucurbita palmata matured from introductions by flotation are also seen. November 1908. B. Rank of Vegetation including Atriplex canescens, Coldenia palmeri, Franseria dumosa, Hymenochloa salsola, Paro- sela emoryi, and Petalonyx thurberi on ancient strands below sea-level near Travertine Rock. Line of Erosion of high beach level of Blake Sea on Rocks in distance. MOVEMENTS OF VEGETATION IN THE SALTON SINK. 155 The introduction of this species by flotation and its subsequent survival on the beach is in contrast with the conclusions reached by Guppy with regard to the transport of gourds by ocean currents. Gourds are known to be carried by such means to points widely distant from their place of origin, but Guppy maintains that establishment takes place only by the aid of man.!_ The constantly receding water in the Salton offers conditions different from those of sea-shores, and the gourd carried by its waters is a desert species which encounters accustomed conditions when stranded on a desiccating gravelly or sandy beach. Cyperus speciosus is a sedge, the significant occurrence of which in this region is along high-water channels of the Colorado River. It was found in the Delta lands by Mr. Parish; also on the emersion of 1908 of Obsidian Island, which is directly in the way of inflowing currents of the Alamo, which is in effect an effluent of the Colorado. It also grew upon the 1908 emersion at Travertine Terraces as well as those of 1909‘and 1910. It may be probable that the later appearances are from seeds of the first introduction. No tests were made of the flotation of the fruits, but it is probable that their small size would render them liable to become attached to the feet of birds. Distichlis spicata was one of the most abundant of the pioneers and it appeared on all of the beaches, although it soon perished under the stress of desiccation, especially in such places as the slopes at Imperial Junction beach. The dispersal by seeds is probably one of the most important means of dissemination, and in this both birds and waves might play a part. Its appearance on a strand area was generally early in the year succeeding emersion, but in some instances it followed the water closely, as it did one year at Imperial Junction beach. The session of 1909 was characterized by heavy mid-summer precipita- tion, but this resulted in no increase in the number of pioneers of this species during that year. Beginning with -1910, however, it was seen that the beaches of 1909 everywhere about the lake bore a heavier mat of Distichlis than the ground laid bare in any other year. The disparity was evident as late as October 1912. The vigorous growth of the rhizomes is such that the plant soon spreads from its original location to such an extent as to cross the boundaries of emersion zones. No ex- perimental proof can be adduced, but the conditions suggest that detached rhizomes might readily be carried long distances by the water of the lake. It is possible that the immediate appearance of the species following hard on the recession may be due to the “stranding” of such living stems. Eclipta alba was represented by one individual in the emersion of 1907 on Obsidian Island, when examined in November 1908, and a year later the species was represented in the emersions of 1908 in the same locality. Further observations of it in this place are lacking and it probably did not reproduce itself on the invaded area. A third occurrence was noted in 1909, on a newly emersed area near Big Island, which likewise had no result. The fruits of Eclipta are known to be capable of flotation for many months, and the manner of occurrence on the islands of the Salton suggests that the plant was carried to the beaches in this manner. Encelia eriocephala made an appearance on the Travertine Terraces following the recession of 1908, where it failed to survive, and a single fruiting individual was found at the same time on the eastern shore of Obsidian Island. It failed to reproduce itself and, although fruiting specimens were seen up the slope from Imperial Junction beach, it found no further place on the bared areas. Eriogonum thomasii was first seen at flood level and slightly above it on Obsidian Island in February 1908, as if the seeds had been carried to their places by birds or had been moved short distances by them after being washed ashore by the waves. These plants were seen three months later, and again in November 1908 on the southwestern 1 Guppy, H. B., Observations of a naturalist in the Pacific, vol. 1, pp. 570, 571, 1906. 156 THE SALTON SEA. side of the lake, but it disappeared within a year after this date. It is to be noted that the species was found on the higher parts of islands not reached by the flood and that it inhabits localities of extreme aridity. Franseria dumosa appeared in the emersion of 1907 on the western shores of the lake, to which place it was probably carried by run-off streams. The soil conditions would favor its survival. This plant is a constituent of the final formation on old beaches. Heliotropium curassavicum was one of the earliest pioneers on almost all of the beaches, being found in bloom on the 1907 emersion as early as February 1908. It is notable, how- ever, that with the progressive desiccation of the beaches it perished within two years, except in such places as the Mecca area, where the low, gently sloping beach had a high soil-moisture content, and on Obsidian Island, where the beaches were so steep that the actual distance from the margin of the water of the receding lake and the original strand was comparatively small. It was consistently included among the pioneers in the successive emersions studied, and likewise appeared to be injured very quickly by any notable decrease of soil-moisture. Heliotropium came in on the islands of the archipelago south of Big Island in 1908, but Cormorant Island, the most isolated of all of the emersed areas, did not receive it until 1912, when two plants in bloom were found on its upper slopes. The transportation to this place may be attributed to birds, with a high degree of probability, especially as the seeds are very small and sink within a short time. No opportunity was afforded for testing the flotation of the seedlings. The two plants found were on a spot that must have been above the water for at least two years and they had germinated and come to bloom within the year, after the manner of freshly deposited seeds. Hilaria rigida was noted but once, and then on the zone laid bare by the recession of 1907 on the western side of the lake. The species is native of the higher slopes to the west- ward and the seeds might have been carried down the slopes by run-off streams. Nothing is known as to the survival of early introductions, but the species might be expected ulti- mately on such slopes. Lepidium lasiocarpum appeared in the silt-filled channels of the washes on Imperial Junction beach following the recession of 1907, but apparently did not survive long or withstand the progressive desiccation, the effects of which were very marked in this locality. Lippia nudiflora came onto the beach of Obsidian Island at the same time as Psathrotes and Spheralcea and formed distinct mats on the southeastern shore, which persisted during all of 1908, but had perished when the place was visited in October 1909. The weight of inference would suggest flotation to the place, either in the form of seeds or living stems. Ocenothera scapoidea aurantiaca came down onto the strand of 1907 with Franseria and Wislizenia and also appeared on Obsidian Island, to which place it may have been carried either by waves or by birds. It is surmised that its seeds would not endure long immersion in salty water. It disappeared from the ranks of invaders on the shore of the island within a year, but probably follows down the gravelly slopes on the western side of the lake by irregular seeding and establishment. Olneya tesota occurs with some frequency up the slopes of the Sink in all directions from the lake, but its seedlings were seen but once on the emersed zones. This was in the emersion of 1907 and the observation was made in November 1908. It seems probable that the beans were washed down the slopes by run-off streams. Parosela emoryt was in many respects one of the most important species which appeared on the beaches. It came on the emersion of 1907 on Obsidian Island, which it must have reached by flotation, and was seen there in 1908. A second invasion in the same year was noted on the southwestern side of the lake. The original location on Obsidian Island was still held in October 1912. Occurrences in the emersion of 1908 were also noted on the southwestern side of the lake and on Obsidian Island below the introduction of 1907. A MOVEMENTS OF VEGETATION IN THE SALTON SINK. 157 third introduction took place in the zone of 1909 at the same place on the west bay of Obsidian Island, which was in such position as to receive objects from the west or north- west driven down the lake by northwestern winds. An occurrence was also noted on the lands emerging from the water south of Big Island. The occupations seem permanent. As has been shown elsewhere, this plant is a prominent constituent of the formations on ancient beaches near Travertine Point, 200 feet above the recent high level of the lake. The spectacle is therefore presented of a species coming onto the moist strand and becom- ing established in the moist saline soil in such manner that the intense and progressive desiccation does not injure it. The only other plant which appeared in the present work with a similar capacity of endurance was Franseria dumosa. (Plate 29 B.) Parosela spinosa attains the dimensions of a small tree along numerous washes in the Sink and is especially abundant near the Travertine Terraces, but its single appearance on the beaches was upon the gently sloping shores to the southward, in the emersion of 1908. Nothing is known as to its reactions to saline solutions and no tests were made of the seeds. Phragmites communis is found at numerous places all through this region where springs and seepages afford sufficient soil-moisture, and many thousands of rhizomes must have been carried into the lake by the inflowing current. The first occurrence of these was noted at the Travertine Terraces in 1908, where two were on the shelf laid bare in 1907, and it also appeared on the emersion of the following year. Since that time Phragmites has not appeared among the species occupying the beaches, although no cause for this absence may be assigned, unless it be the increasing toxicity of the lake water. No doubt exists as to its endurance of the lake water during 1907 and 1908; for an accretion mound of a spring near Salton station, which had been submerged during 1907 and 1908, bore rhizomes of this plant which had survived the immersion. It is possible, however, that the toxic limit of concentration accentuated by the loss of calcium was reached in 1909 and that living plants were not carried by the lake after 1908. All of the known introduc- tions were from rhizomes floating in the water. (Plate 18 a.) Pluchea sericea, or arrow-wood, is a shrub which reaches a height of 3 to 7 feet in the Delta of the Colorado. The fruits bear copious pappus, on account of which it is readily borne aloft by the winds. Maturing fruits were seen in May and consequently this plant found a foothold on the strip which had been laid bare at this time and was still moist. It appeared on all of the beaches except that it was more sparsely represented during 1909 and 1910 than at other times. Introduction appeared to be followed by establishment in nearly every case. No well-attested records are at hand as to its extermination on any beach, although it was recorded as missing in certain instances, perhaps by being over- looked. The heavy stand on this plant on the strip laid bare on Imperial Junction beach in 1907 was still represented by a great number of individuals in October 1912, although it was seen that many were slowly dying, presumably from the lack of moisture in the soil. The actual margin of the lake had by this time receded over a mile from where they stood. Pluchea was also one of the first plants carried to the sterilized surfaces of the islands, and its extreme buoyancy in the air renders it liable to be sown thickly everywhere. Mil- lions of the fruits, like those of all of the compositaceous plants of the region, would un- doubtedly fall into the water, but it can not be said whether or not these would germinate there or in the moist mud at the margin when they became stranded. Polypogon monspeliensis was seen but once, in May 1908, when mature plants with maturing seeds were found on ground recently laid bare by Imperial J unction beach. The single introduction may have been by birds or by wave action. Populus macdougalii is found in moist locations at various places in the Cahuilla Basin and some trees were below the maximum level of the water near Mecca. The water, which at the time of the high level contained only about 0.25 or 0.33 per cent of salt, covered the 158 THE SALTON SEA. ground around the bases of the trunks, but no measure of the penetration could be made. Some of these trees survived. The fruits are probably capable of enduring even a higher proportion of salts, since a rank of plantlets was found at the upper margin of the shelf laid bare at Travertine Terraces in 1908, also at the upper margin of the emersions of 1910, 1911, and 1912. The low rate of recession during the winter, coupled with the violent wave- action, is responsible for the small cut bank which separates the emersion of one year from that of the next. It was just under this bank that the dense row of this tree was established, which is indicative either of a dispersal of fruits before mid-winter and their germination early in the year, or of early maturation of fruits in the spring and their immediate lodg- ment in the angle at the foot of the upright bank. The flying fruits fill the air at the time of their ripening and many might fall directly on the ground, or upon the water a few feet or a few hundred yards away and be washed ashore. (Plates 24 a and 26 B.) Psathyrotes ramosissima appeared on the beach laid bare in 1907 on Obsidian Island. It was seen first in February 1908 and was fruiting in May of the same year. Despite this production of seeds, it did not reproduce itself and was not seen again here or elsewhere in the emersed zones. The only assignable cause of its defection would be the toxicity of the soil salts increasing with desiccation. Salix nigra like Populus survives the submergence of its roots in water slightly saline, as was shown by the behavior of many trees on the Mecca beach. The seeds likewise appear to germinate in the presence of some concentration of salts. The occupation of the beaches by seedlings was confined to the western shore of the lake. Some were seen at the foot of the long bajadas southeast of the Travertine Terraces and this plant formed the most prominent feature of the dense rank of perennial plants which was formed at the upper margins of the emersions of each year. It appears on the records for 1907 and 1908, but disappeared from both within a year. But one Salix was present on the 1909 emersion when it was examined in 1912. The invasion began to be heavier after this and a crowded row of Salix as well as of Populus was found at the upper margin of the emer- sion of 1910, in September of that year, which had persisted and by June 1912 had made a rank of trees 10 to 15 feet in height which have continued to grow since that time. Another heavy rank was formed at the upper margin of the emersion of 1912 and plantlets were seen as early as June of that year. The species did not appear so early in 1912 and no plantlets were found at the upper margin of the emersion in June 1912, but in mid-October young trees 2 or 3 feet in height were present. (Plates 24 a and 25 sz.) The dispersal of the seeds must be under much the same conditions as those of Populus. The species was found only on the beaches of the western shores of the lake and no instance was found of its transportation across the open lake to the emerging islands. Trees are present at springs near Travertine Terraces and fruits may have lodged directly on the moist strands or have fallen in the water and been washed ashore. Germination took place in soil high in salt-content. Spheralcea orcuttii was seen but once and then only in a single plant on the shore of Obsidian Island in the path of the direct inflow current of the Alamo River, under the same circumstances as Psathyrotes, but failed to reproduce itself. Both of these plants probably floated to the place. _ _ Typha angustata was found in great abundance in the Delta and around many springs in the Salton Sink. Its fruits must be carried widely by air-currents, yet its appearances as a pioneer on the emersed strips were very few. It was first seen in the moist filled channels of the old washes on Imperial Junction beach emersed in 1907, but these plants succumbed to the increasing aridity within a year. Next it was a pioneer in the emersion of 1908 on the ‘Travertine Terraces, but these individuals perished with the desiccation following recession. A single plant coming from a rhizome washed ashore was found on the Imperial Junction beach in 1909 and much material of this plant was found at various MOVEMENTS OF VEGETATION IN THE SALTON SINK. 159 places around the lake, although it is not known how many plants arose in this way. It is certain, however, that the duration of such individuals would be brief unless they were near a spring or seepage. Typha has appeared in the emersions of 1911 and 1912 near Mecca, but here the ground is high in moisture and numerous seepages and overflow channels from wells make conditions widely different from those at the lower end of a desert bajada. Numerous colonies of mature plants fruiting abundantly are also present up the slopes from the Mecca location, so that its dissemination here may be considered simply as a spreading down a slope already occupied by it. The seeds are known to sink within a very short time after they fall upon the water and almost all cases of their dispersal must be attributed to air-currents. Wislizenia refracta was seen but once and then on the emersion of 1907 on the western side of the lake, along with Franseria, Hilaria, and other species of arid habitats. It played no further known part in the initial occupation of the beaches. THE MOVEMENTS OF PLANTS INTO STERILIZED AREAS. The main thesis of this paper has been that of the manner in which seed-plants were carried into moist zones or strands around a receding lake which had been completely sterilized by immersion in the salt water. The submerged area of nearly 500 square miles was part of an extremely arid region, the native flora of which is fairly represented by the census of the Sink above the present water-level. The number of species is small and includes 8 trees, 23 shrubs, 10 semi-shrubs, 30 perennial herbs, and 51 annual herbs. These plants being present on the slopes of the Sink, their seeds would be most liable of all plants to be carried down to the beaches. The agricultural operations on the slopes to the south- ward of the body of the lake brought into this region 31 additional species, which were also especially liable to be carried to the bared areas. Besides those included in the pre- ceding number, every species within a long radius to the southward and eastward might be considered as being liable to be carried to the strands of the receding lake. The high mountain wall to the westward would form a fairly effectual barrier in that quarter. The culture ground or emersed strips to which attention was directed were not in a static condition, however. The ground was arid and the soil variously charged with salts varying in composition and concentration. The observations of Mr. EK. E. Free, recorded in this paper, show that the soluble content of the soils was not materially modified to any great depth in places which have been laid bare by the receding waters during the first six years’ recession of the lake. The moment that the waves failed permanently to lave any part of the strand, therefore, it began to pass from a saturated condition back to the desiccated state induced by the arid climate of the region. The surface layer, a few inches in depth, would at first contain a soil-solution approximately equivalent to the water in the lake, but evaporation rapidly concentrated this, and the plant which might have found a foothold in the comparatively fresh soil would be subjected to the action of solutions of increasing concentration. The salts dissolved in the water of the lake increased from 0.25 to nearly 1 per cent during the period of the observation, so that even the initial conditions of emergence of the soil changed progressively. Sixty species were found on the beaches examined during the six years in which the observational areas were examined closely. These beaches amounted to about 5 square miles only, but since they were chosen to include diverse soils and slopes they probably represent more than a majority of the species introduced. Five trees, 17 shrubby species, and 38 herbaceous forms were included among those which appeared upon the beaches during the period mentioned. Two of the trees inhabit moist areas, while the other three are of xerophytic habit; 9 of the shrubs are halophytes and survive in soils varying in salt content, this habit being most pronounced in Sueda and Spirostachys and least per- haps in Baccharis or Pluchea. Eight of the herbaceous pioneers are halophytic, while the 160 THE SALTON SEA. remaining 30 are usually found in places in which the concentration of the salts in the soil-water is very low, and some of these are of xerophytic habit while others inhabit moist areas. It is evident that the species which may be expected to appear in greatest number on the beaches laid bare in the further recession of the lake will be drawn from the 17 pronounced halophytes; the presence of seepages and springs will afford small areas in which others may appear. The secondary occupations or the species which might be expected on the older and uppermost strands would be of the xerophytic trees, shrubs, and herbaceous xerophytes. It is interesting to note that included among the earlier forms on the beaches was Sonchus oleraceus, which was also one of the first plants to be found on the lava flows of Mauna Loa, Hawaii, as observed by Forbes.! The seeds of this widely disseminated weed are undoubtedly carried long distances, and as they are produced in great profusion may be expected almost anywhere in the tropics or temperate zones. HISTORICAL. The strand flora of Krakatau included but 67 species in 1906, as reported by Ernst,? although the strand flora of the Indo-Malayan region comprises about 320 species. It is thus to be seen that not one-fourth of the plants, the habits of which were suitable for such conditions, had been carried to the shores of this island sterilized by volcanic action in a quarter of acentury. The total flora in 1906 included but 92 seed-plants and 22 ferns and mosses. Ernst estimates that 39 to 72 per cent of these may have been brought to the island by sea-currents, 10 to 19 per cent by birds, and 16 to 30 per cent by air-currents. This author attributes the transportation of the 16 ferns to this last agency, also all of the compositaceous species. Britton, in a recently published discussion of the derivation of the flora of Bermuda,? points out that 80 per cent of the higher plants are identical with forms native to the West Indies and to the mainland of North America. The analysis of the habits and occurrence of these species leads to the conclusion that 41 (or over 18 per cent, including all of the halophytes) have been carried to Bermuda by flotation and 88 per cent by air-currents. The grasses and composite, as well as ferns and mosses, are supposed to have arrived in this manner. The presence of nearly 44 per cent is attributed to birds. It is to be noted that the island of Krakatau is separated from other land by only a few miles of open water, while Bermuda lies far out in the Atlantic and nearly 600 miles of ocean separate its surface from Cape Hatteras, the nearest land. It is also important to consider that the revegetation of Krakatau has been in progress but a few years, while the carrying of seeds and spores to Bermuda has gone on without interruption for an extremely long period. These differences may well account for the much greater importance attached to birds as carriers in the ease of Bermuda and the lessened effectiveness of flotation. It is notable that these authors estimate the part played by winds as practically the same. THE PHYSICAL CONDITIONS OF THE SALTON SINK. The general setting of the biological stage around and on the Salton is much different from the complexes furnished by Bermuda or Krakatau. The islands in the Salton Lake were submerged in salty water and emerged to pass by rapid desiccation to an arid condi- tion. Their surfaces are separated from the mainland by distances nearly as great as those of Krakatau, however. Fresh-water inlets constantly pour currents laden with silt ’ Preliminary observations concerning the plant invasion on some of the lava flows ii i io) the f of Mauna Loa, Hawaii. Occa- mtb tbo a the Bernice Pauahi Bishop Museum of Polynesian Ethnology and Natural History, vol. v, No. * The new flora of the volcanic island of Krakatau, p. 57. Translated by A * Botanical explorations in Bermuda. Jour. N. Y. Bot. Garden, vol. zh, Dp. ie ae MOVEMENTS OF VEGETATION IN THE SALTON SINK. 161 and seeds into the saline lake which is increasing in salinity and toxicity as against the constancy of composition of sea-water. The initial vegetation on beaches laid bare by the recession of the lake appears on strands down-slope from localities bearing a halophytic and xerophytic vegetation. These desert slopes receive but little rainfall and have nothing but ephemeral streams, run-off currents (which rush down the shallow channels for a few minutes or a few hours) being made by the precipitation of vagrant storms. Such run-off currents furnish a special method of flotation of seeds and propagule which would not be encountered in any place except in arid regions. NATURE OF THE EVIDENCE. It is obvious that most of the conclusions with regard to the dissemination or dispersal of plants in cases like those under discussion must rest largely upon circumstantial evidence and that the reasoning employed is largely inferential. To emphasize this point it need only be said that no one has actually followed an individual plant in its movement over a wide distance with identification at all stages of the journey. The only part of the matter which is amenable to direct observation and controlled experimentation is that which may be brought into tests of the length of flotation and germination of seeds, fruits, and propa- gative bodies. The importance of results bearing upon this subject is attested by the num- ber of workers who have devoted attention to the matter. The positive evidence at hand as to the relation of the size of seeds, form and structure of fruits, shape of appendages, etc., is also of some value, while the presence of seeds in mud on the feet of birds, or attached to their plumage, is highly suggestive of possibilities and allows the observer to account for the presence of plants in localities otherwise unexplainable. The seeds of one-fourth of the species which were found on the newly emersed beaches were treated experimentally in flotation and germination tests, and these represented fairly well the size, structure, and general anatomical features of the entire lot, although the untested remainder might well be expected to present some novel reactions. GERMINATION, FLOTATION, AND SURVIVAL. The first and main series of tests were made in equalized temperatures slightly lower than a similar treatment would have secured in the Salton region itself, but it is doubtful whether much difference might be found in the minima encountered in the two cases. Furthermore, the march of the seasons in the two instances was parallel, since the seeds were put in Salton water in the middle or latter part of November at about the same time as if they had been cast from the plant in the Salton area. The plantlets derived by the germinations would have corresponded fairly well with others allowed to germinate on the beaches where the seeds might have been deposited by the waves or carried by the winds or birds. The seeds which might be expected to fall into the water in a free condi- tion weighed from 0.013 to 6 mg. The fruits inclosing the seeds of Atriplex lentiformis weigh 7 or 8 mg. Other fruits not tested, such as the large globular melons of Cucurbita palmata, may weigh as much as 30 to 50 grams, even when dried, but by reason of their comparatively great size are very buoyant on water. It seems to be generally conceded that structures or appendages are of small con- sequence in dispersal of seeds over any great distance, and it is worthy of note that fully 80 per cent of the species occupying the beaches are liable to be moved by the wind. The initial plants on the bared strips were generally arranged in well-defined zones or bands, which might give the observer the impression that they could have been deposited in this strict order only by waves, but the recession of the water is followed by a concomi- tant desiccation which leaves a moist strip which varies in width according to the slope and texture of the soil in the separate localities and affords the only suitable conditions 11 162 THE SALTON SEA. for germination of deposited seeds. Next it is to be recognized that the crop of seeds, such as those of Populus, Salix, or Pluchea, may mature and be freed from the plant all within a comparatively brief period, during which time the number carried about by the wind would be distinctly noticeable. Thus the few days in which the fruits of Populus are in the air serve to distribute them in such number that they may become an annoyance by clogging wire mosquito screens, or by passing into houses through open doors or windows. It is evident that strands would be thickly strewn during the fortnight in which these seeds are so thick in the air and that this plant might be entirely absent from the remainder of the emersion of the year unless seeds or fruits falling on the water were capable of sur- viving the effects of such flotation and saline action. Still another phase of wind-deposition of seeds was presented by the Travertine Ter- races, which are characterized by a strict rank of vegetation arising at the foot of the higher well-defined cut bank formed in mid-winter as well as the base of the more indefinite ledge which may be made in mid-summer. The vertical surface of the cut bank, being composed of moist earth and of the flotsam collected at the base on a slope that drops away slightly (making an angle greater than a right angle), a mechanical trap is formed causing eddies and other effects and resulting in deposition of wind-borne bodies including seeds. (Plates 258 and 26.) A visit was made to this place on February 8, 1913, when an unusually high-cut bank was observed which might have been formed by wave-action of more than average violence or to a greater recession of the lake level. The face of the vertical wall of soil which was caving and falling down ranged from a few inches near the margins of the observational area to a height of 26 or 28 inches at the crest of the slope which they cut across. The water was comparatively quiet, but the wavelets were still lapping against the base of the mud wall, and here was massed flotsam consisting of feathers, shells, and fragments of wood of more than one kind and other fine material, among which some seeds might be expected. About two pounds of this material were taken up from a line extending about a yard along the bank, put into a waterproof covering, and carried to Tucson. Some glass jars were filled with a commercial quartz sand and this was saturated with Salton water taken from the lake in October 1912. The débris was now spread on the surface and enough water added to float it. The preparations were placed in a glass house at a temperature of about 50° to 80° F. The first germination appeared 11 days after the preparations were made, and the first plantlet developed was an Atriplex; a grass started into activity a week later and a second grass was included in the final list of germinations. By March 20, a month after the first seedlings had appeared, 20 germinations had taken place. The rising temperature made it necessary to water the preparations and in this process the fragments of wood on the surface were moved about by the water in such manner as to destroy six of the seedlings; a second watering resulted in the death of two more. The sand in which the plantlets were growing would have been moistened by water from under- neath on the beach, and the fatalities noted would have probably been less at this stage of their development than in the experiments. It is quite probable that the greater num- ber of the seedlings which were killed before identification were Sesuvium or Heliotropium. A visit to the place late in May showed a large number of seedlings of Saliz and Populus besides a few of Heliotropium, Cyperus, Atriplex, Prosopis, and a grass. The results of these tests suggest that the willows and poplars which come into the rank at the base of such vertical walls made by the waves must be wind-borne. While about 50 of the 60 species appearing on the Salton beaches might owe their presence to the action of winds, nevertheless other agencies display great effectiveness: a rough eval- uation of the balance of forces will show a number in which the seeds might be carried by the wind only, and in such a list there would be included Aster spinosus, A. exilis var. australis, Baccharis glutinosa, Encelia eriocephala, Isocoma veneta var. acradenia, Pluchea MOVEMENTS OF VEGETATION IN THE SALTON SINK. 163 camphorata, P. sericea, Populus sp., Salix nigra, Sonchus asper, S. oleraceus, and Typha angustifolia, a total of 12 species, or about 20 per cent of the total number. Three-fourths of this number were carried to the sterilized islands, but not Populus, Salix, and Isocoma, or rather it should be said that successful germinations of these were not seen. The beaches on the islands offered conditions not at all suitable for any of these species and they were lacking from the Imperial Junction beach perhaps for similar reasons. Positive evidence of the action of birds in carrying seeds is extremely difficult to secure. The seeds of a large number of the species coming onto the beaches are very small, and nearly all of those tested for flotation and germination would adhere to a rod of wood or glass or to the finger when thrust into the water in which they were floating and then withdrawn. Such adhesions would be greatly facilitated by even the smallest particle or layer of mud and it seems reasonable to assume that many thousands of seeds must have been carried about in this manner. Cryptanthe barbigera occurred on Obsidian Island in such manner in 1908 as to suggest that the burr-like fruits may have been carried to the shore by birds, while Spirostachys was found near nests of pelicans and cormorants in posi- tions in which seeds may have been detached from the legs or feathers of a bird. In addition to the species the seeds of which appear to be borne about by the wind only, Astragalus, Chamesyce, Coldenia, Eleocharis, Encelia, Franseria, Hilaria, Hymenochloa, Oenothera, Olneya, Parosela emoryt, Prosopis glandulosa, and P. pubescens would not be liable to be carried by the birds. The 34 remaining species offer many probabilities of introduction by birds, although profitable surmise may not be made as to the actual importance of the plants carried about in this way. (Plates 318 and 328.) Attention is to be called again to the fact that the zonal arrangement of the initial individuals on a beach would not be incompatible with the idea of seeds carried by birds, since it was observed that flocks of alighting birds ranged themselves in long lines at or near the water’s edge in positions fairly correspondent with those later occupied by ranks of vegetation; but nothing more definite may be said, in concluding the discussion of this phase of the subject, than that one species was found where it must have been carried by birds, another near nests in positions which it might have reached by wind or flotation, and that 32 others had seeds liable to be carried about by adhesion to the limbs or plumage, or in the crops. In the transportation of seeds and plants which finally lodged on the sterilized areas flotation played a part much more important in the Salton region than in any other region that has yet been studied, paradoxical as this may seem in a desert region. Furthermore, the number of instances in which seeds and plants could have been introduced by water only are fairly large. Among these are to be included the deposition of rootstocks of Typha, of rhizomes of Phragmites, of a living branch of Prosopis glandulosa, of stems of Distichlis, and of fruits of Cucurbita on various beaches, while the tubers of Scirpus palu- dosus are readily moved in this manner. The species tested in the Laboratory may be taken as offering the general conditions of the entire lot of plants with which this paper is concerned. Reference to previous pages will confirm the statement that Atriplex lentiformis does not sink, but the floating seeds germinate within a few days on the surface. Juncus floats 37 days, then germinates quickly after sinking. Leptochloa germinates 78 days after being wetted and while still afloat. Sesuvium remains afloat 23 to 50 days, then sinks and germinates. Jtwmezx seeds float only 3 days when stripped, but supported by the calices or other parts of the inflores- cence may remain afloat until germination ensues. Jsocoma remains afloat less than a month. Oligomeris remains afloat about 60 days and germination ensues, both at surface and among the sunken ones to some extent. Amaranthus floats 30 days, then the seeds sink and a few germinate within the next fortnight. Baccharis floats less than a month and Pluchea about the same length of time. Spirostachys sinks within a day and begins germi- 164 THE SALTON SEA. nation at once. Prosopis pubescens and P. glandulosa sink at once, and the seeds of the first begin to germinate within a few days. Scirpus paludosus had a large number of seeds afloat after 120 days, although more than half had sunk while a few had germinated. Seeds of Sueda behave much like those of Scirpus, except that a greater proportion sink within a day or two and germination ensues more quickly among these at the bottom, beginning within two or three days. It is thus seen that 9 of the 15 species tested remain afloat for a length of time which would permit them to be driven by the action of the waves a dis- tance equal to the long axis of the Salton Lake. It is evident, however, that a consideration of the flotation of the seed alone affords evidence of but little conclusive value in the study of the revegetation of a sterilized area such as that under discussion. The fate of the seed at the end of its flotation period or of the seedling is quite an important consideration. The results of floating seedlings has probably not hitherto had adequate evaluation as a bio-geographical factor in dissemination and dispersal. The evidence of the flotation tests makes it plain that in 10 of the 15 species tested the flotation period of the seed, which in the different species may be from a day or two to four months or more, may be followed by the survival of the plantlet for periods of such length as to render them liable to be carried about by many agencies. Buoyant plantlets may remain on the surface or in the surface layers of the water for a period as great or in some cases much greater than that of the seed, and their sur- vival would depend upon the nature of the beaches on which they might become stranded. The successful tests with Sesuviuwm, Leptochloa, Pluchea, Sptrostachys, Prosopis, and others showed that plantlets which had floated for two or three weeks were sound and healthy and sent roots down into the substratum when ‘“‘stranded,” in imitation of their probable action in the lake. Plantlets of Rumex remained sound and made a slow growth after a flotation period of two months, but their stranding test was made with soil saturated with the Salton water of 1912, in which they have not survived around the lake. The latest introductions of the species on the beaches was upon soils saturated with water of the concentration of 1908, which had but little more than one-third of the salt-content of that of 1912, but it came on the Imperial Junction beach again in 1911, when the alkaline soil was being overlaid with silt from the Alamo inflow. There is but little in this behavior to suggest special adaptation or peculiar fitness, since the seedlings of almost any land plant will float, especially in the sunlight with the copious liberation of free gases. Writers on plant-dispersal have been disposed to regard ready germination as preju- dicial to wide dissemination, although it is admitted by Guppy that the species included in the mangrove formations are widely distributed by the action of currents.’ Some of the species encountered about the Salton present the condition of the radius of possible dissemination being greatly lengthened by flotation of seedlings. Another phase of the movement of seeds and propagative bodies into the sterilized areas of the beaches peripheral to the lake is that by which the ephemeral run-off streams following rainstorms carry material down the slopes, effecting a radial penetration and arrangement of the introduced species. Some slight action of this kind was observed on Obsidian Island, where one or two species were carried down from the summits which rose above the water even at its recent highest level, but it was most marked on the western shore of the lake, where long bajadas come down to the lake from the mountains several a distant, and similar action was also indicated by some of the movements of vegetation at Mecca. The actual weight or size of the seed or body would have but little influence, since the force of these streams which may run in any given “‘dry”’ channel but a few hours in a year is such that the finest sand and rocks weighing several pounds are tumbled along 1 Guppy. Observations of a naturalist in the Pacific, vol. 11, pp. 76-87, 1906. See Chapter IX, “Abortive germina- tion in warm seas.” ; MOVEMENTS OF VEGETATION IN THE SALTON SINK. 165 and deposited on the margins or in bars, and the seeds which survive the attrition they receive during such movement would, some of them, come to rest in layers of moist sand, in which germination might speedily ensue. Species with a long resting period might lie dormant until a favorable condition of moisture or temperature were encountered. Among the species observed which may be definitely assigned to the class moved by run-off are Astragalus, Bouteloua, Chamesyce, Coldenia, Conyza, Encelia frutescens, Hilaria, Hymenochloa, Oenothera, Olneya, Parosela emoryi, and Wislizenia. The distances traversed were not in any instance over a mile or two and it is evident that, when the movement of a plant this distance down a gentle slope is considered, many forces (including winds and animals) might play a part, although perhaps a minor one. Obviously the action of the run-off streams would contribute to give the recession of the first year from the maximum level of the lake a flora far richer in number of forms than that of any succeeding year. SPECIES APPEARING EARLIEST ON THE BEACHES. A profitable comparison of the initial species on newly emersed beaches may be made by a review of the observations of the plants on Imperial Junction Beach with a slope of 1 to 300 and an alkaline adobe soil, and on the Travertine Terraces with a slope of about 1 in 20, upon which the water cut terraces and shelves. Amaranthus, Atriplex canescens, A. fasciculata, A. linearis, A. polycarpa, Cucurbita, Baccharis, Heliotropium, Lepidium, Leptochloa, Oligomeris, Pluchea sericea, Rumex, Sesu- uium, Spirostachys, Sueda, and Typha appeared on the emersion of 1907 at Imperial Junc- tion Beach, as noted in the early part of 1908, while only Atriplex canescens and Phragmites were included in the first plants on the beach of 1907 at Travertine Terraces. Atriplex fasciculata, A. polycarpa, Leptochloa, Scirpus paludosus, Spirostachys, Typha, Sueda, and Polypogon were found on the emersion of 1908 at Imperial Junction Beach and Typha, Heliotropium, Distichlis, Pluchea sericea, Phragmites, Bouteloua, Salix, and Populus on the beach of the same year at Travertine Terraces. Leptochloa, Scirpus, Typha, Heliotropium, Pluchea sericea, Sesuvium, Sueda, Spiro- stachys, Atriplex fasciculata, and A. canescens came on the emersion of 1909 at Imperial Junction Beach. Distichlis, Heliotropium, and Pluchea camphorata comprised the entire census of the Travertine emersion of 1909 when visited in 1910. Heliotropium, Leptochloa, Scirpus, Atriplex fasciculata, Spirostachys, and Sueda were the first species to come on the emersion of 1910 at Imperial Junction Beach; Distichlis, Heliotropium, Populus, and Salix made up the entire pioneer flora of the beach of 1910 at Travertine Terraces. Scirpus, Leptochloa, Heliotropium, Distichlis, Sesuvium, Rumex, Sueda, and Spiro- stachys came first on the emersion of 1911 at Imperial Junction Beach, while the emersion of 1911 at Travertine Terraces (when the census was taken as in previous years, late in the autumn) showed only Distichlis, Salix, and Populus. Atriplex fasciculata, A. lentiformis, Sueda, Spirostachys, Scirpus, Sesuvium, Distichlis, and Leptochloa came into the emersion of 1912 during that year at Imperial Junction Beach, while Sesuvium, Salix, Populus, Atriplex lentiformis, Spirostachys, Heliotropium, Prosopis pubescens, and Pluchea camphorata appeared on the shelf of 1912 at Travertine Terraces. Numerically as to species, the pioneer flora of the alkaline beaches in the series of six years ran 15, 8, 10, 6, 8, and 8, while that of the steeper sandy and gravelly benches of the Travertine Terraces formed a series of 2, 8, 4, 3, and 8. The average number of species appearing on Imperial Junction Beach was nearly twice as great as that on the Travertine Terraces, the low gradient of the slope of the beach being favorable to the introduction and establishment of plants, while the alkaline soil appeared to afford con- ditions for halophytic forms superior to those of the Travertine Terraces. The range of 166 THE SALTON SEA. variation in the number of species introduced year by year is seen to be large, a fact which suggests that the lack of a wind in a certain direction during the fortnight in which seeds are falling from the parent plant might materially restrict their dispersal, while an oppor- tune storm would carry light or appendaged seeds to a much greater distance. The number of events which might influence any agency in moving seeds are almost countless. Thus the feeding habits of the fish which are preyed upon by pelicans and cormorants might result in variations in the part played by these birds. Rainstorms moistening the surface layers might not only furnish favorable conditions for germination, but also carry a large supply of seeds of various species down the slopes, etc. Mention has already been made of the possible effects of rapid and slow evaporation by which stranded seeds would be left behind on soil from which evaporation was very great. The analyses of the annual census shows that Atriplex linearis, Lepidium, and Cucur- bita did not find a place on the beaches after the first recession in 1907, and that Atriplex polycarpa and Phragmites dropped out of the pioneer class after 1908; Pluchea sericea, while coming on many of the beaches later, did not figure as an innovator after 1909, and Atriplex canescens and Typha stopped in 1910; Pluchea camphorata began appearing among the pioneers on Travertine Terraces in 1910, although it was one of the first plants to be seen on one of the sterilized islands. Six and probably seven species, therefore, which originally came with the first occupants of the two special areas noted above, had dropped out within the first four years of the recession of the lake. Such defection suggests at once a restricting factor in the concentration of the water of the lake saturating the soils of the beaches, together with the proportions of certain constituents. The only features which might have a bearing on this matter are shown in table 34. Tapie 34.—Concentration and proportion of certain constituents of Salton water (per cent). 1907. 1908. 1909. 1910. 1911. 1912. Sodium) 5 cscs eantj ie eins hans Sie 111 134 160 189 228 271 Ca leium 235 35.84 aharnes eae Ones 4 > 10 12 13 14 15.6 aA POtASHIUINY vee scaled AG 8 eaada Ge nAS 2.3 2.8 3.2 3.5 3.8 3.8 Chlorine... 2... cece eee eee 170 204 241 281 339 395 Total of soluble solids ....... 364 437 519 604 718 847 The differentiation of the action of the elements in such a complex mixture as that presented by Salton water is not easily made. While the increasing disproportion between sodium and calcium might offer itself as a feature accounting for toxicity, yet it is to be seen that even as late as 1912 the amount of the latter present would suffice to balance the sodium. The great amount of chlorine suggests that it is to this substance that the increasing toxicity might be ascribed. SUCCESSIONS AND ELIMINATIONS. The successions or transitions in the vegetation of arid shores of bodies of either salt or fresh water are very abrupt, as has been found by the examination of great stretches of the coast of the Gulf of California. The tidal zone may bear such plants as Laguncularia and other tide-marsh plants, but immediately above the action of the waves the vegetation of the desert finds place. The ephemeral character of Salton Lake with its rapidly sinking level called into action a set of conditions entirely different from those to be met on the shores of a body of water 1 Niklewski Bronislaw. Ueber den Austritt von Calcium- iumi ae Pe ee a ee ee ene coreg MOVEMENTS OF VEGETATION IN THE SALTON SINK. 167 fluctuating about a fixed level. In the Salton the water receded at such rate that during the time of maximum evaporation in May or June a strip more than a yard in width would be bared permanently, and seeds of all kinds in motion at that time might fall on it and germinate. All other physical conditions now were minor to the fact that the soil began to desiccate toward a soil-moisture content equivalent to that of the surrounding desert. Occasionally small flat places or shallow depressions in the soil would be occupied by a growth of Spirulina, which with the drying of the soil would, with the surface layer of the soil a few millimetres in thickness, break into innumerable concave fragments, but this was not followed by any definite procedure. The main facts of interest on the shores centered about the survival of the initial sowings on the beaches, the later introductions being for the most part only of minor im- portance. The chief features of the endurance of the initial forms and of the appearance of additional species on the beaches after the first year may be best illustrated by a recapitu- lation of the observations on the two beaches taken for the discussion of initial occupation, the Imperial Junction Beach and the Travertine Terraces. The emersion of 1907 at Imperial Beach bore Atriplex canescens, A. fasciculata, A. linearis, A. polycarpa, Amaranthus, Baccharis, Cucurbita, Distichlis, Lepidium, Leptochloa, Heliotropium, Oligomeris, Pluchea sericea, Sesuvium, Spirostachys, and Sueda early in 1908. Late in 1908 Amaranthus, Baccharis, Distichlis, Heliotropium, Oligomeris, Rumex, Sesuvium, and Typha had disappeared. Late in 1909 Aériplex canescens, A. fasciculatus, Sueda, Pluchea sericea, Spirostachys, and Cucurbita still survived, while a secondary introduction of Baccharis and of Chenopodium had taken place, both being represented by only a few individuals, and these did not maintain themselves. Atriplex canescens, A. fasciculata, and Spirostachys had multiplied and thrived in 1910, while Sueda seemed to have not multiplied; Pluchea was losing a large share of its individuals as a result of the desiccation. The census in 1911 was practically that of 1910 with the added losses of Pluchea, and no change in the balance was visible late in 1912. The original sowing of this place included 17 species, 8 of which had disappeared within a year; one of the remaining 8 was lost in the following year, and two of the original pioneers were reintroduced only to disappear quickly. The census showed only 5 species in 1910, all of which were still in evidence in 1912, but with Pluchea sericea losing ground. The full return of the area to the conditions prevailing up the slope might bring in F’ranseria or an occasional Larrea or Olneya, while the number of individuals of the other species would be reduced on account of the diminished soil-moisture supply. It is to be noted that the changes here are wholly and directly connected with the water supply, and that the survivors are halophytes, one of which was undergoing deterioration by reason of the inadequate supply. The original introductions on the emersions of 1907 of the Travertine Terraces com- prised two species, Atriplex canescens (from seeds which had fallen down a caving bank) and Phragmites (which had washed ashore as a rhizome). The cut bank may have figured as a mechanical trap for grounding wind-borne seeds, or some other condition may have come in, for now Atriplex polycarpa, A. canescens (re-introduced), Bouteloua, Astragalus, Dis- tichlis, Heliotropium, Juncus, Pluchea camphorata, Prosopis pubescens, Phragmites, Sesurium, Spirostachys, and Sueda (13 in all) were present. The place was not seen again until October 1910, when only Distichlis, Prosopis pubescens, Phragmites, and Astragalus remained. Late in 1911 Astragalus was not found, although as an annual its seeds were probably present, while Pluchea sericea and Salix nigra had come in, making 5 species with another probably present. Late in 1912 all of the above elements had come in except Salix and Isocoma, while a single small plant of Prosopis glandulosa was recognized, which had prob- ably been confused with P. pubescens, up to that time. The surface was fully occupied, and of the six species present it seemed likely that Phragmites, Distichlis, and Pluchea would 168 THE SALTON SEA. soonest perish on account of the increasing aridity. The final condition of this beach would probably be one in which Isocoma would endure, although no estimate of the behavior of the other species can be made except to point out that they are not on the slope just above the high level of the lake, which is of the extreme desert type of this region. The two beaches which have thus been analyzed display different types of behavior. The gently sloping alkaline Imperial Junction beach received a sowing of 17 species during the first year of emersion and no secondary introductions. The stress of increasing aridity has depleted the number of the pioneers until but 5 species remain, of which one, Pluchea, will soon fail for lack of sufficient soil-moisture. The final flora of the slope will probably consist of species now occupying it, but with greatly reduced number. An occasional indi- vidual of one or two other species may come in. The Travertine Terrace of 1907 was probably subject to wave action during the greater part of that year and its original occupants may be taken to include 13 coming on in 1908; these were quickly reduced to 4 species two years later, when secondary introductions began, of which Pluchea sericea and Isocoma have played an important part. The last named may be regarded as a plant which would be suitable for endurance of the final condi- tions of the desiccated slopes of this locality. ANCIENT STRANDS. The strands of the Travertine Terraces were on the crest of an arched slope or bajada, which would ultimately be subjected to the maximum action of the wind, which is the more important meteoric feature in this region. In consequence of its action it was not possible to find the ancient beach ridges or the edges of terraces which would correspond in position to the vertical banks which marked the mid-winter level of the Salton during the recent period of the lake. But a number of well-marked strands were to be recognized, lying at various levels within 100 feet of the level of the ancient high beach-line. These beaches owe their preservation to the fact that they were formed on the concave part of the slope and in places sheltered from the prevailing wind and with no run-off. These were well marked to the southward of Travertine Rock and also to the westward. The character of these strands is such that they may not be safely taken for seasonally formed strands, but each one might be considered as marking the maximum level of the lake at some previous filling. This assumption is supported by the fact that such well-marked beach ridges were not found anywhere near the present level of the lake (see Plate 29 8). Ancient strands of well-marked structure are to be seen on the steep slopes westward of Salton Slough, where a hill rises to such a height that its summit was covered at the highest level of Blake Sea; 83 well-marked beach ridges were seen on the slopes of this hill in 1910. An examination of a strand south of Travertine Rock was made in October 1912, and a photograph was taken (see Plate 30 a). The plants marking its position was a compara- tively dense desert formation inclusive of Atriplex canescens, Coldenia palmeria, Franseria dumosa, Hymenochloa salsola, Parosela emoryi, and Petalonyx thurberi, all of which were restricted to a band or zone which varied from 12 to 18 feet in width. A second examination of another strand in February 1913 included the above except Atriplex canescens. It is notable that Parosela and Atriplex, which are members of this formation, which dates the beginning of its development back for at least a century or two, also appear on similar strands on Obsidian Island and elsewhere shortly after emersion. THE REOCCUPATION OF STERILIZED ISLANDS. The slopes of the Sink were interrupted at various places by small elevations consisting of rock in place, the most notable elevation being in the southeastern part of the submerged area. Three of these hills rose above the highest recent level of the water, and detailed SALTON SEA PLATE 30 A. Ranks of Vegetation on Strands formed by water of Blake Sea probably three or four centuries ago. View to northward near Travertine Rock with Salton Sea on right. B. Dense mat of Sesuvium as first occupant of sterilized area on outlying portion of Big Island. October 1909. MOVEMENTS OF VEGETATION IN THE SALTON SINK. 169 observations on the beaches and shores of one of these, Obsidian Island, have already been described. All but two or three of the numerous species which appeared on the emersed strips were, in all reasonable probability, carried here from places outside of the lake, and reference will be made to their action in the following pages; but some of the hills were completely covered by the water at rest, or by wave-action, and these deserve special atten- tion, since as experimental settings they constitute much stricter tests of the theories of seed-dispersal than others in which the summits remained above the water (see Plate 31 8). The greater mass of the hills consisted of rock with surface layers and pockets of wind- blown soil before the inundation. The action of the water resulted in washing much of this soil from the hills and the saturated salt-solutions which steadily increased in concen- tration may be taken to have killed or set in action all seeds present. If germination ensued the seedlings would of course be floated away. Some of the species of seed plants which thrive in the vicinity of salt springs have already been noted as surviving after a year of submergence, and a small shrubby species, among the Carrizo dunes, showed similar endurance, yet none of these species was seen on the islands, which appeared to be completely sterilized, except perhaps as to minute organisms such as bacteria, and it seems probable that those present in the dry soil were extirpated and those found after emersion are to be regarded as introductions. The first example of colonization of a sterilized island was met in October 1908, when a visit was made to Cormorant Island, 2 miles to the northward of Obsidian Island, near the location of some salses or mud volcanoes now deep under water. This island was sepa- rated from the mainland by about 6 or 7 miles of water, which reached a depth of about 34 feet at the maximum, and a greater depth between it and the nearest island. (See Plate 318.) The top of the island showed a rocky surface of whitened rhyolitic obsidian in November 1908. No exposure of fine soil was visible and it was evident that the waves had been driven across its summit repeatedly during a large part of 1907. A great number of nests of aquatic birds were placed among the rocks and the surface all about them showed signs of having been trampled by both young and old birds. Pluchea sericea and Baccharis glutinosa were each represented by one plant. As both of these species are compositaceous and the fruits are carried long distances by the wind, the probabilities are largely in favor of their deposition here by the action of air-currents. Both species grow in abundance on the mainland 12 or 15 miles to the southward, and the summer winds may have brought them here late in 1907. The seeds of both appear to germinate in the presence of some proportion of salt. Cormorant Island lay away from the main routes of travel about the lake and its rocky shores could not be gained in any kind of rough weather. The second visit to the place was made in October 1912. The original individuals of Pluchea and Baccharis had survived. An additional individual of each was also present. That of Pluchea was some yards from the original plant, and may have arisen from one of its seeds or from a newly introduced one. The second individual of Baccharis was near the original, so near that it might have come from offshoots or seeds, but probably did not represent a second intro- duction. That the total increase of the two species was limited to a single individual each in four years was a very astonishing fact, as both of the original plants had fruited probably three times and the annual crop must have included a large number of seeds, while the later emersions down the slopes of the island showed many clay exposures in which germination might have taken place. The island was dome-shaped, being about 1,200 feet in its long axis with half that width, and rising about 30 feet above the water. The entire surface was now carefully scanned, with the result that two individuals of Sesuviwm and one of Atriplex lentiformis were found in a place probably emersed in 1908, and one Atriplex was lower down in the emersion of 1909 and one in that of 1910 and one in 1911. Heliotro- pium curassavicum was represented by two individuals, both of which were in bloom, 170 THE SALTON SEA. while between 15 and 20 examples of Spirostachys were scattered over various emersions, their age suggesting germinations during 1912 or late in 1911. Six species were thus seen to be present, with a total representation of about 30 individuals, and plants of 3 of the species were mature and either were fruiting or had done so; 2 of the 6 were compositaceous with fruits abundantly furnished with pappus. It would seem probable that these two species were carried here by the wind. The seeds of Atriplex lentiformis germinate very quickly after being wetted, but the plantlets are capable of floating for periods of several weeks. The occurrence of the four individuals seen here might well be attributed to such action. The fruits as noted (see page 146) are fairly large and are not so readily carried by either birds or the wind as some of the others. The minute seeds of Sesuvium float for two or three weeks after falling in the water, and the seedlings rise to the surface and are capable of flotation for extended periods. Seeds of this size are also very liable to be carried about by violent winds, while they readily adhere to any object thrust in the water in which they are floating and might well be carried to the island by any one of the methods mentioned. The seeds of Heliotropium might likewise be carried about by several agencies, although it is to be noted that despite the enormous number of seeds and the abundance of the opportunities only two individuals of Sesuvium and of Heliotropium grew on the island in six years. This, however, would be a very brief period in the conditions of transportation of a species to new lands in practical geographical work. Spirostachys is also subject to the conjunction of many conditions which might bring it to the island, but the appearance of young plants far above the present level of the water suggests that the seeds were carried to the place either by winds or birds. The indefinite retention of the seeds on the parent plant, making a crop which falls to the ground over a long period, furnishes material which might be acted on by either of the agencies mentioned. The positions of the young plants render of comparatively little importance the fact that the seedlings are capable of flotation for an extended period. All of the species mentioned were also present on the beaches of Obsidian Island, 2 miles to the southward, and also on the smaller islands to the southwestward, described below. Immediately after the examination of Cormorant Island in November 1908, a visit was made to a small island which had risen above the surface or rather had been laid bare at some time late in 1907 or early in 1908; a sand bar, barely above the level of the water, now ran from it to the larger land mass, but had not been bare long enough to figure in the history of the pioneer species which must have reached the place several months earlier. Many square yards of sandy and gravelly soil furnished suitable conditions for germination, while the higher part of the emergence consisted of a lot of blocks of Obsidian, some a meter in diameter, which would have served very effectually in causing the deposition of wind-borne seeds. A single plant of Sésuvium in bloom, Heliotropium also in bloom, Pluchea camphorata, Spirostachys, Parosela emoryi, Atriplex lentiformis (?), and single plants of two unrecognizable species were present. A second small island to the westward and south of Big Island showed a surface consisting of extremely rough and broken rock masses with small level pockets of moist soil, bearing a few plants of Pluchea sericea, Sonchus asper, Heliotropium, Atriplex lentiformis (?), and Eclipta alba. These islands were much frequented by pelicans and cormorants nesting among the rocks. The considerations noted above apply to the presence of the same species here. Eclipta probably floated to the place, and Sonchus might have been carried by the wind; the position of Parosela on Obsidian Island (as described previously) suggests flotation, but nothing in its location here could be taken as evidence of the manner in which it had been carried to the spot. This would apply also to the unknown species. These islands were visited again a year later (October 1909) and showed only Pluchea camphorata, Spirostachys, Sesuvium, Heliotropium, Atriplex, Scirpus paludosus, and Pluchea sericea; from which it is to be seen that Sonchus, Eclipta, and the two unknown species SALTON A SF PLATE 31 A. Pluchea sericea and Spirostachys occidentalis as first occupants of a sterilized area on Big Island, October 1908 B. Baccharis glutinosa, one of the first occupants of Cormorant Island, October 1912 SALTON SEA PLATE 32 A. Islands and Clay Bars emerging on southwestern Shore of Salton Lake, October 1912. The sterilized areas are being first occupied by Spirostachys, Sueda, and Sesuvium. Larger islands out in the lake to left. B. Spirostachys occidentalis established near a Cormorant’s Nest on small island near western shore of Salton Lake. Photographed October 1912. MOVEMENTS OF VEGETATION IN THE SALTON SINK. 171 had disappeared, while Scirpus, with a seed especially liable to be carried by the currents, had come in. An additional island had been laid bare a short distance north of Big Island some time during 1909, and in October of that year the restricted bit of gravel and sand which might offer conditions for plants bore a few individuals of Pluchea camphorata and Sesuvium, both of which might have been wind-borne, though Sesuwtwm may have come by any of the methods mentioned above. Although the lake abounded in fish and these were used as food by the birds, no direct evidence was obtained that they were effective, directly or indirectly, in the transportation of seeds. The next inspection of the surface of these islands was made in October 1912, at which time all had become connected with the main part of Big Island. The slopes of this bore only Atriplex fasciculata, Sueda, and Eriogonum deserticola, not identical with the pioneers on the beaches of the lower lands. The northernmost islet now bore Sptrostachys and Heliotropium on a clayey flat, while Atriplex lentiformis, Pluchea sericea, and Baccharis glutinosa were on the higher rockier ridge. The original introductions had thus disappeared and the later flora was like the pioneers in not being derived from the slopes of Big Island, but was composed of species which were probably wind-borne, or had come from seeds or plantlets deposited by the waves, while birds may have been responsible for some; all had crossed a water barrier at least 10 miles in width. The lower areas in this archipelago were much frequented by aquatic birds, while on all of the visits tracks of a number of small quadrupeds were seen. On the final visit in October 1912, the tracks of a raccoon were seen; burrows of rodents were numerous; the round foot-prints of some member of the cat family were in evidence, and small lizards abounded. No estimate could be made of the numbers of any of these animals. Two rabbits were seen, however; these and the rodents feed upon plants and seeds and their activity may have had a very direct connection with the disappearance of the species noted above. This island has been continuously isolated since some time in 1904 or 1905, a matter of some importance in the heredity of these short-lived species. In October 1912 a small rocky island, separated from the western shore of the lake by a stretch of open water 500 feet in width, was encountered near the Carrizo District. Some clayey silt was found among the rocks. Spirostachys was represented by several young plants and two individuals of Heliotropium in bloom were seen. The rocks and clay near the margin of the water were coated with a blue-green alga. Some of the Spiro- stachys grew up beside a cormorant’s nest and may well have been brought there by these birds, although this species is such a ubiquitous invader that its movements can not be ascribed entirely to any oneagency. (Plate 328.) A shoal with a clayey and sandy surface, lying a few yards off the shore at the extreme southwestern corner of the lake, was inspected on the following day and bore only Spirostachys and Sueda. (Plate 32 a.) GENERAL DISCUSSION. By D. T. MacDovaat. The Cahuilla Basin is a structural trough lying immediately to the eastward of an abruptly rising mountain range which separates it from the Pacific. The depression is shaped much like the bowl of a spoon, the tip of which comes to within a short distance of the Gulf of California. An ancient beach-line, lying a few feet above present sea-level, incloses an area of about 2,200 square miles; and this lowermost portion of the basin has been designated the Salton Sink in the present paper. The original depression was of unknown depth; and it has been filled to within 284 feet below the mean tide-level by the alluvial outwash from the slopes of the mountains which bound it on three sides. Borings to a depth of 1,700 feet show a series of interbedded sands and clays, such as might be encountered in any one of the similar troughs in central Arizona, which had been filled by material worn down by wind and water from the moun- tain slopes. The bottom of the basin obviously lies far below the present level of the gulf; yet it is clear that ‘“‘Blake Sea,” the ancient body of water which filled the basin to the level of the highest beach, was composed of fresh water, at least during a part of its exist- ence—as evidenced by the heavy layer of travertine formed on the rocks beneath its surface at the highest level, by the presence of fresh-water shells, and by the composition of the saline layer on the bottom of the Sink, evidently derived by condensation of fresh water rather than from evaporation of sea water. (See discussion of this point by Free, pages 22 to 34.) Essential differences of great importance to the naturalist are to be recognized between the history of the undrained basins of California and Nevada and that of the Cahuilla. At times the former contained bodies of water, the formation, fluctuation, and disappearances of which were directly connected with climatic variations in the region in which they lie. The Cahuilla Basin, on the other hand, has probably been the scene of incursions from the ocean in earlier times; and the formation and existence of Blake Sea may not be directly connected with the climate of the immediate region. Likewise, the modern gathering of the waters in the Sink is not to be attributed to climatic effects. The earliest existence of the Salton Sea within historic times is that shown on Rocque’s map (1762, see Plate 5). Collated reports give the presence of flood-water in some volume in the Sink in 1828, 1840, 1849, 1852, 1859, 1862, 1867, and 1891. These occurrences do not seem to be correlated with any climatic measurements, and are perhaps directly connected with the oscillations of asilt-laden stream in alluvium. This view is strengthened by the fact that the fillings of the Pattie Basin have not been synchronous with those of the Cahuilla. The study of the behavior of the soils of the emerged strands and uncovered beaches has yielded some points of interest and value in the interpretation of lacustrine action, especially with respect to the evaporation from saturated soils, and the surprisingly small leaching effect resulting from prolonged submergence in water, the dissolved content of which is lower than that of the soil solutions. It has been customary with geologists and geographers to assume that the waters of ephemeral and fluctuating lakes are reduced and their salts dropped out as in a test tube. The action of iron and sulphur bacteria has been well known for some time, but no good opportunity had hitherto occurred for controlled observations of their possible influence 173 174 THE SALTON SEA. on the composition of disappearing lake waters and on the deposition of calcium and other substances from these waters. The reduction of sulphates in the lake water by Spirillum and other forms resulted in the formation of hydrogen sulphide as one product, some of which would escape into the air, while a portion would be oxidized by Beggiatoa, with the separation of free sulphur and the liberation of sulphuric acid. Oxides of iron might also be formed and all of the above recombinations might take place in stages of concentration in which no such action would occur in biologically sterile solutions. The intervention of algal and bacterial organisms in the formation of a type of travertine characteristic of the higher levels of Blake Sea has also been demonstrated beyond reasonable doubt. These processes seem to be carried on much more vigorously in the presence of organic matter, and the results of such activities are particularly in evidence in the vicinity of decaying stems and woody structures of all kinds. Furthermore, the hydrolyzing action of organisms of the Amylobacter group upon the cortical tissues of the plants submerged by the waters of the lake is a matter of some interest in the study of the formation of coal deposits and in the making and preservation of fossils. The Cahuilla Basin is subject to a mixed type of climate. It lies far enough inland so that overheating should result in a continental type of climate, particularly with respect to the rainfall. Its great bowl, however, lies immediately in the lee of a great mountain range which rises abruptly from its southwestern side, with the result that fringes of moun- tain storms reach out over part of its area at times, while the topographical conditions favor the development of the intense and localized precipitation known as cloudbursts. The general features of the precipitation are shown in table 35, furnished by the U. S. Weather Bureau. TABLE 35.—Rainfall record, in inches, of Indio, California (elevation, 20 feet). 3 .| S| 8 s |B : : ee Q Ln] as 3 Season. i 3 a 2 g q a B 3 8 a Sele) So) Bee Bd ee) ee) ee |e Slale@leilela)erela|e)elelaie | a 1877-78. |....]..../.../...] 0} 198/010] 0 | o | o | o | o | .... |1s78| 1.10 1878-79. | 0 | 0 | 0 | 0 | 0 | 1.00] 0.60] 0.30) 0 | o | o | 0 | 1.90] 1879] 1'30 1379-80. | 0 | 0 | 0 | 0/040} 0 | 0 | 0 | 0 | 0 | 0 | © | 0.40] 1880! 0:70 1880-81. | 0 | 0 | 0 | 0 | 0 |070/ 3.45! 0 | 050] 0 | o | o | 4.65|1881| 3.95 1981-82, | 0 | 0 | o | o| o| o | 150] o | o | 0 | 0 | o | 1150] 1882] 2'50 1gs2-83. | 0 | o | o | 0 | 100] 0 |o80}113]/011| 0 | 0 | © | 3:04] 1883] 296 1883-84. | 0 | 0 | 0 | 006] 0 | 086] 0 | 3.16] 0.62/0.46/0.46] © | 5.60] 1884| 5.38 1884-85. | 0 | 0 | o | 0 | 0 Jaz! o | o | o |o10] 0 | © | 0:80) 1885] 1.00 1385-86. | 0 | 0 | 0 | o jos} 0 | o | o | o | o | o | © | 090] 1886] 012 1986-87. | 0 | 0 | 0 | 0 |o12] 0 | 0 J 093] 0 jo30) o | © | 1.35] 1887] 1143 1887-88. | 0 | T |0.05}015| 0 | 0 |o75) 0 | o | o | o | © | 0985/1888] 296 1888-89. | 0 | 0 | 0 | 0 |1.30]111] 0.57) 0 | 105! 0 | o | o | 3:83] 1889 647 1889-90. | 0 | 0.95] 0 | 0.60 0.01] 3.29] 0.65] 0.06] 0 | o | o | o | 5:56| 18901 1.03 1890-91. | 0 /o.10}020} 0 | 0 | 0.22} 0/190} 0 |] o | o | o | o42\iso1| 331 1891-92, | 0 [1.16] 0 | 0 | o | 0.25} 2.00] 0.43] 0.22/0.04/014) 0 | 4124/1892] 2:83 1892-93. | 0 | 0 | 0 | o| of} 0/003] 0 |160} 0 | © | 0 | 1.631893] 264 1893-94. | 0.05|0.75/0.07} o |o14/ T!] 0} o| o| o | o | 0 | Vo1lisos) tT 1994-95. | T | 0 | 0 | of] o| of 601] 0| of o | o | & | 6017/1895) 601 1895-96. | 0 | 0 | 0 | o| 0 | o jos2/ 0 | o | o | o | o | o92| 1896 o92 1so6-07, | 0 | 0 | 0 | o | o| 0 | 210/019] o | o | o | o | 1.29] 1907] 3.39 1897-98. | 0 | 0 |210/ 0 | 0 | 0 Jo10/ 0 | 030| o | o | o | 2501189081 1:70 1g98-99. | 0 |0.30| 0 | 0 | 0 | 100/040; 0 | 0 | 0 | o | © | 1:70| 1809! 130 1899-1900. | © | 0 |0.10/ 0 | 0.60] 0.20] 1.00] 0 | 030/015| tT | 0 | 2'35| 1900! 274 1900-01. | 0 | © | 0.08} 104/017) 0 | 029/146) 0 | 0 | o | © | 3:04| 1901! 1:75 1901-02. | 0 | 0 | 0 | 0 | 0 | 0 |o40/020/ 0 | 0 | o | o | oo! i902! 2:00 1902-03. /0.10/ 0 | 0 | 0 | 050/080} 0 | 0 | 020/075} 0 | 0 | 85/1903! 1158 1903-04. | 0 [0.10012] 0 | 0 | 041/087] 0.35]0.20} 0 | Tt | o | 205| 19041 2:43 1904-05. ) T |0.33/ © | 0.08) 0.19] 0.41| 087/200] 1.30] 0 | T | o | 518|1905| 537 1905-06. | 0 | 0 | T | T | 106/014] T | 0.97] 206/047] 0 | 0 | 4:70|1908| 719 1906-07. | T | 1.07} 0.04] T | 0.60/ 1.89] 0.59/ 063| 0.96) 0 | 0.05! 0 | 5.831907 sas 1907-08. ; 0 | © | 0 | 1.60] 0.05] T | 0.95] 0.57| 0.01) 0 | o | o | 318] 1908 | ga 1908-09. | T |0.45/160] 0 | 0 | 0.60| 0.28) 0.29|045| 0 | o | 0 | 3:13! 1909| 2.07 1909-10. | 0.00 } 0.87 | 1.12 | 0.00 0.20 | 0.86 | 0.47 | 0.00 0.08 | 0.00 | 0.00 | 0.00 | 3:.60/1910| 1.06 1910-11. | T_ | 0.08 | 0.00 | 0.12} 0.30) 0.00} 0.66] 1.06| 0.22} T | 0.00 | 0.00 | 2'44| 1911 | 2:53 1911-12, | 0.25 | 0.00 | 0.34 | 0.00 0.00 | 0.00 | 0.00 0.00] 1.66 | 0.35 | 0.53 | 0.02 | 3.15 | 1912| 260 1912-13. | 0.04 | 0.00 | 0.00] 1.90] 0.00] T | 0.12] 0.931 0.02 : " : GENERAL DISCUSSION. 175 The annual average from data covering 36 years is seen to be 2.74 inches and the char- acter of the precipitation phenomena suggests a high degree of aridity. The maximum amount received in one year was 7.10 inches (1906) and the lowest, a “trace” (less than 0.01 inch) in 1904, giving a variation as 1 to 1000, a proportion occurring in deserts of a pronounced degree of aridity only. In addition to this expression, the ratio of possible evaporation from a free-water surface to the annual amount of precipitation has been useful in characterizing deserts. About 116 inches of water would evaporate from the surface of a small vessel on the ground in the open in the Cahuilla during a year; this is 15 times the amount which has fallen in any one year, 43 times the average, and many thousands of times the minimum. (These estimates are to supersede the figures given in a statement made before the Royal Geographical Society of London in December 1911, and published in the Geographical Journal, vol. 40, p. 105, 1912.) The possible evapora- tion in northern Africa at Algiers is about 60 inches annually, but in the interior of the Sahara it is much greater. The figures in table 36, compiled by Dr. W. A. Cannon, give data from four points, of which Laghouat is far inland and Ghardaia the most arid. The measurements were made in millimeters with a Piche evaporimeter and should be multiplied by 0.74 to obtain the evaporation from a free-water surface. TABLE 36. Jan. Feb. | Mar. | Apr. | May | June | July | Aug. | Sept. | Oct. | Nov. | Dec. | Year Laghouat........ 889 | 1023 | 1434 | 2031 | 2892 | 3739 | 421.2 | 379.6 | 264.8 | 173.9 | 153.8 | 159.2 | 2753 Ghardaia.,...... 172.4 | 233.4 | 340.7 | 528.7 | 611.7 | 699.1 | 749.2 | 693.0 | 468.8 | 329.4 | 257.7 | 225.5 | 5309 Touggourt....... «ee» | 163.9 | 240.5 | 274.3 | 385.2 | 459.4 | 565.8 | 487.3 | 329.4 | 222.6 | 166.1 | 142.3 Algiers,......... 84.2 96.8 90.5 | 118.3 | 151.8 | 158.6 | 175.0 | 205.0 | 165.5 | 129.0 | 136.1 | 143.2 | 1654 The possible evaporation for a year at Ghardaia is about 160 inches, and as the average precipitation is 0.5 inch the ratio would be as 1 to 330 as compared with 1 to 43 in the Cahuilla. While this basin is by no means so arid as the Sahara, the desicca- tion results in the reduction of the soil moisture to a minimum during the warmer season, and organisms would not only be subjected to intense aridity, but also to the action of soil solutions of a high concentration. This effect is highly accentuated in places where the evaporation results in the deposition of soil salts, given a charged substratum in which only halophytes can survive. Seepages and flowing wells afford a supply of ground water in places which furnish conditions for vegetation not essentially different from those in any warm climate, although the species which inhabit them are subjected to a high rate of water-loss. The occasional rains call into action a few plants of short cycle, which like- wise may be of more or less mesophytic habit. The number of such ‘‘annuals,” however, is very small, as it is in the Libyan desert in Northern Africa.’ Outside of these types the specialized conditions of soil moisture, soil salts, evaporating capacity of the air, and aspect of the precipitation offered by the more arid portions of the basin and of deserts in general might be expected to be accompanied by the presence of specialized types of vegetation, especially since it is reasonably certain that the desert of the Colorado, in which the Cahuilla lies, has been arid since Tertiary times. It seems highly probable that the entire number of characteristic desert plants have undergone their evolution within the period during which this region has been arid. No fossil material has been recovered which suggests reduced spinose shoots, such as those of Zizyphus, Con- dalia, or Parkinsonia, on the one hand, or the succulent stems of the Cactaceze, Euphorbi- aces, Asclepiacez, or Crassulacez on the other, although some of the ancient Bennettiales and Cycadales suggest that they were suitable for endurance in regions in which the water- 1See MacDougal, The Deserts of Western Egypt. Plant World, vol. 16, p. 291. 1913. 176 THE SALTON SEA. loss would be great. Similar structures, however, are offered by the swamp xerophytes of the present time, and the ancient species with a habit and structure of sclerophylly were characterized by reproductive mechanisms which male a supply of moisture a necessity. If other desert plants were included in the earlier floras their remains have not yet been found. It is to be conceded, of course, that the conditions in deserts are not favorable to the making or preservation of fossils. The comparative fewness of the individuals, the destructive effects of organisms in brackish water, the fierce grinding action of the torrential stream flow, and the looseness of the material laid down, would all act to destroy plant remains and give a sparse record. The wind-blown material deposited as loess would be equally unfavorable for the preservation of plants. Probably the best set of conditions would be those in which a desert stream-way originating in the highlands would carry material from its middle dry course and bury it in the alluvium, as must be taking place in the Colorado Delta at present, yet nothing of the kind has been examined which would justify the assignment of any great age to the xerophytic habit among plants. Not a little interest attaches to the conifers in this connection, as the members of this group probably represent the earliest structures which might be suitable for existence in arid climates. The group as a whole is widely distributed with regard to climate and geological history. Some of the species live in cold, others in warm humid climates, while others not widely dissimilar in structure are found in regions of great aridity. Survival under such circumstances has been attributed by Groom to architectural features.1 The size, outer membranes, and longevity of the leaves may show a wide range of variation with consequent modification of use and loss of water. It seems highly probable that the trees of this group were the first plants of size to inhabit arid areas, although the xerophytic habit has not been carried so far as to allow them to endure the conditions of the intensity offered by the Sahara or the Cahuilla. Furthermore, it can not be said that the structural features of the coniferze are the ones which have been developed or modified to meet the conditions of greater aridity by other plants. In other words, the most marked and the most pronounced xerophytic developments have taken place by other methods than those which find expression in the conifers. It may make toward a better understanding of this matter if it is recalled that the major movement in phylogeny and habit in plants has been away from types with separated gametophytes in which the sporophyte and the gametophyte have an independent exist- ence, toward forms in which the sporophyte vastly outranks the gametophyte in size and carries the bisexual generation within its tissues, living under more or less aquatic con- ditions, toward preponderantly sporophytic forms, capable of effecting reproduction and of carrying on the reproductive functions under progressively arid conditions. The logical development of such an evolutionary tendency would of course produce the xerophytic types last of all, and while it is not easy to fix upon a time when these appeared, yet the weight of inference lies in favor of the assumption that desert vegetation is composed chiefly of modern vegetative types. A large proportion of the species which might be included in the flora of a desert region may be distinctly of a mesophytic habit requiring a large supply of moisture for their development. These forms are either localized around the sources of ground water or they may be annuals or of short cycle coming to maturity rapidly within seasons of periodic rainfall. As has been pointed out elsewhere, the number of types of this kind is very small in the Salton Sink and in the Libyan Desert where the rainfall is highly un- certain as to its amount and time of occurrence. The features of such plants which seem to have a direct connection with their habitat are those of rapid development and the ability of the ripened seeds to endure long desiccation at high temperatures. 1 Remarks on the ecology of the conifers. Annals of Botany, vol. xxrv, p. 241. 1910. GENERAL DISCUSSION. 177 Three main types of plants, however, are distinguishable as being characteristic of arid areas and of the included localities in which the soils are highly charged with salts: the spinose xerophytes, which are geserally woody shrubs, with reduced branches and small and indurated leaves, although some herbaceous forms are to be included; the succulents, in which roots, stems, and leaves have been variously modified by the exaggeration of parenchymatous tracts, so that they may hold a large amount of surplus water; and the halophytes or species inhabiting saline areas, many of which also carry a large balance of water. The spinose xerophytes are perhaps the most widely prevalent of all of the plants characteristic of arid regions. Alhagi, Parkinsonia, Holocantha, Condalia, Zizyphus, etc., are examples of the group and these plants show many morphological and physiological qualities having the force of adaptations or accommodations to the environment, though a too direct causal connection is not to be taken for granted. Plants of this type are found in all regions in which arid climates of any type prevail—in the Cahuilla, in the Sahara where the rainfall is low and uncertain, as well as in regions like that around the Desert Laboratory, in which well-defined periods of precipitation come with a certain regularity. Examples of this type are also found in localized dry habitats in regions with a comparatively moist climate. It is notable that spinose types make up the few species which are to be found in the most arid regions in which the supply of moisture all comes from underneath and the evaporating capacity of the air is highest. The roots are in constant absorbing contact with the soil at all times, and although the actual amount, or rate of acquisition of water may be low, it is true that once absorbent contact with the soil is broken it is not easily restored; hence these forms are not easily transplantable. The sap of the stems and leaves of some of the spinose xerophytes has been found to show a concentration of over 100 atmos- pheres, which is far higher than that of ordinary mesophytic plants, and the suggestion lies near that this condition in the shoot might be a means of damage to the absorbing organs of uprooted plants and that the transpiration stream once broken is not to be restored readily under such extreme conditions, which do not extend to include the roots under ordinary circumstances. The succulents are most abundant in regions in which the precipitation comes within well-defined seasons with fair regularity, and hence are not to be expected in the Cahuilla or in such climates as those of the Sahara. They are, however, a very important element of the flora of the region about the Desert Laboratory, parts of South America, and eastern and southern Africa. Cacti, several genera of the Asclepiadacez, Euphorbiaces, the Yuccas, Amaryllidacez, etc., furnish examples in which tracts of parenchyma in various members become greatly exaggerated, and in which a great balance of water accumulates. The root-systems of such plants, so far as they are known, appear to have an absorbent contact with the soil only during the rainy season, when the moisture content is high, and the con- centration of the soil solution is lowest. The inequality between the concentration of the sap of the shoot and the roots, which was noted in the spinose plants, is not present, as the juice of the cacti does not usually show an osmotic pressure of more than 10 or 12 atmospheres. In contrast with the continuous absorbent action of the roots of the spiny types, many of the succulents actually undergo a decortication of the root-system during the dry seasons, so that the taking in of water in any quantity is impossible until new rootlets have been formed. This usually takes place very quickly with the coming of the rains. It is to be taken for granted that the succulents are subject to the same high evapora- tive action of the air as the spinose forms of the regions which they inhabit. It is therefore to be expected that their external structure and morphology would include some of the features characteristic of the spinose forms. As a matter of fact many of these are exhibited in more pronounced or highly developed stages. Induration, thickening, or heavy coatings 12 178 THE SALTON SEA. of wax and resinous material on the epidermis, retardation and restriction of the develop- ment of branches and shoots, are among the more striking of these features. Spinosity in general results from the atrophy of branches, but this failure to develop has become so complete in such cacti as the Echinocactus that the entire shoot is cut down to a single cylindrical or globose stem, bearing only spines. The halophytes, many of which are cosmopolitan, inhabiting sea-shores in moist climates, comprise both the spinose and succulent types. The latter are represented by Spirostachys and Sueda, and the former by Distichlis, Atriplex, and Pluchea. The con- centration of the sap of the halophytes is midway between that of the spinose forms and of the cacti. A distinction is to be made between the cosmopolitan succulent types and those of the salterns and bitterns of the desert. The species from the shore wilt quickly when placed under high evaporative conditions and are therefore much different from the types named above. It is probable that the type as a whole originated in charged soils in arid regions. It does not seem to be going too far to assume also that these conditions are identical with or fairly similar to those now prevalent in the Salton Sink and elsewhere. The xerophytes and halophytes of the Salton Sink are therefore to be seen living under the conditions coincident or causal to their evolutionary development. It will not be profitable to venture an analysis of the environmental complex in an effort to detect possible causality. The only hypothesis hitherto put forward in explanation of this matter is to the effect that succulents, particularly the halophytes, owe their qualities to the inductive action of the chlorides generally present in the substratum, both in saline areas and along seashores. The chief argument seems to have been one of coincidence, and no note was taken of the fact that succulent halophytes also belong in the bitterns in which the salts are carbonates and sulphates. One condition of the environment is universal in deserts, the high evaporating capacity of the air, the effects of which are to be seen even in many plants with an abundant supply of soil-moisture and whose shoots are exposed to desert wind-action. This state of affairs, conducive to a high rate of water-loss, is generally manifested or accompanied by the restriction of leaves, the induration of surfaces, and the atrophy of shoots exhibited by spinose and succulent xerophytes and halophytes. A general parallel is offered by the behavior of a dish of hydrated gelatine which may be exposed to the air under high evapora- tive conditions. If the conditions of desiccation were of the intensity of the desert, the outer surface would soon become coagulated and the hardened surface would thus effectu- ally check the loss of water from the layers beneath. This in its final analysis is what takes place in the body of the plant. The expanding shoot is a structure of plasmatic colloids and the loss of water or other features which would tend to coagulation of expand- ing cells would check their growth and be followed by atrophies of various degrees. The actual form of the shoot of the plants so affected would of course be modified in the most serious manner according to the special morphological features presented, and it is to be understood that the stele and the general morphogeny of the type were determined pre- vious to their encounter with the conditions of aridity; yet with the widest diversity of actual surface, shape and size of organs, stomatal variations, and seasonal habit, results may be seen suggestive of the hardening or coagulatory effects of high evaporatory rates, and the coincidences are so sweeping and universal as to suggest that a causal connection is present. If it be assumed that the modification of the extent of development of the shoot may be connected with physical agencies directly external to this member and acting directly upon it, there remains the question of succulence to be accounted for. Succulence in general entails the exaggeration of parenchymatous tissue, cortical or medullary, and the accumulation in the resulting enlarged tracts of a surplus or balance of water, which physiologically forms a pocket or reservoir with respect to the transpira- tion stream. Such tracts may be developed in roots, stems, or branches—or even in leaves, GENERAL DISCUSSION. 179 as in the agaves and crassulas. The plasmatic units or cells are increased in both number and size, growth having been stimulated in both the phases of extension and multiplication. The concentration of the sap in such cells as noted above is generally not very high and in a large number of instances mucilages and gums are present, making the cells practically masses of indefinitely expanding water-absorbing colloids. The obvious direction in which to turn, therefore, in the search for agencies which might cause or accompany succulency would be one in which hydratation agencies which would cause the swelling of the colloidal cells would be taken into account. Not until recently would such an inquiry have met with evidence that would be valuable in the discussion of the subject. This evidence is now at hand, however, in the form of the recent writings of Borowikow, as to influence of certain substances upon growth expansions.! The experimentation of Borowikow seems to show that the acids (hydrochloric, sulphuric, nitric, acetic, and boric) within certain limits of concentration cause an increased growth of seedlings in a manner corresponding fairly with the swelling of gelatinous colloids treated with the same reagents in equal concentrations. The effect is not due simply to the hydrogen, but is the summed result of both ions. The alkaline metals exert a similar hydrating effect upon gelatine and upon growth. Increase of the concentration beyond certain limits would be followed by coagulatory effects in gelatine and by diminished growth in the plant. Coagulatory effects and retardation of growth are produced by the action of salts of metals, of mercury, copper, strontium, calcium, barium, magnesium, lithium, potassium, and ammonia, which show a diminishing effect upon proteinaceous colloids in the order named. These results suggest a possible means of analysis of the conditions of succulency. The hydratation of the mucilaginous contents of the bodies of certain cacti tested at the Desert Laboratory does not take place in the manner of the growth-effects described by Borowikow, however, and it is evident that the problem is a complicated one. The disturbances in the Salton Sink following the making and the gradual desiccation of the lake were of course attended by the exposure of many species to unusual intensities of concentration of some of the substances active in coagulation and in hydratation. But little attention could be given to such environic reactions, yet (as described in a previous section of this paper) noticeable deviations in leaf and stem characters were found in Aster exilis, Prosopis glandulosa, and Atriplex canescens, in addition to variations in the fruits of Scirpus paludosus. The thicker stems and heavier shoots of the first-named species are strik- ingly suggestive of effects such as those ascribed to hydratation action of acids and alkalies. The whole possible range of hydratation and coagulatory effects seem to be included in the seasonal changes in the brines described by Professor Peirce in the present volume. Dilute solutions with sodium chloride and calcium salts present seem to be accompanied by hydratation and growth, which are checked with a concentration supposedly increased beyond the optimum. The crystallization out of the sodium and of the calcium and the subjection of the alge to the influence of the magnesium and potassium appear to be fol- lowed by encysting, which may be taken as being due to the neutralizing effect of these substances upon the proteinaceous colloids of the cells. The fact that 4 out of a total of 60 species which found place on the strands exhibited modifications of structure not observed elsewhere led directly to a consideration of the endemic species of the Sink. Adriplex saltonensis Parish, Spheralcea orcutti Vasey and Rose, Cryptanthe costata Brandegee, Calandrina ambigua Howell, Astragalus limatus Sheldon, A. aridus A. Gray, and Chamesyce saltonensis Millspaugh are to be included in this category. It is true that Calandrina is found a short distance beyond the limits of the Sink, but the remaining 6 species are not known to occur beyond and above the high 1 Borowikow, G. A. Ueber die Ursachen des Wachstums der Pflanzen, Biochem. Zeitschr., vol. xLiv, p. 230. 1913. 180 THE SALTON SEA. beach line marking the level of Blake Sea which filled the basin to this shore line within comparatively recent time. The situation suggests that these species originated in the Sink since it was last filled and the inference is strongly in favor of such a conclusion. According to Mr. Parish Astragalus limatus is so closely related to A. preussit A. Gray of the Mohave desert to the northward that it has been considered a variety of it, but the other endemic species noted have no near relatives in the immediate region. If the pos- sibility of the origination of these forms within the Sink be allowed it is also suggested that other species might have originated in like manner but become disseminated over a wide area in such manner that their nativity is undiscoverable. The origination of a species within the limits of the Sink or in any basin such as the Cahuilla would probably be followed by its dissemination over surrounding arid regions. The recent formation of the lake, in which about 7 cubic miles of water were poured into the Sink, covering an area of 450 square miles, constituted the beginning of an experiment in which exact studies could be made as to the movements of plants on such sterilized areas. Submergence, of course, resulted in the extermination of the vegetation; then the slow recession of the waters due to evaporation exposed a peripheral strip from a few feet to a half mile wide every year. The total area thus laid bare every year was 8 to 10 square miles, and this afforded the means by which the action of various disseminational agencies might be tested and compared; furthermore, the test areas showed a slight but measurable progressive modification year by year. The concentration and composition of the water with which each area was moistened was characteristic, and the constitution of the vege- tation on the strip nearest above newly bared strips differed every year from that nearest the new beaches of the preceding year. The disseminational agencies which were to be taken into account included those which might be active in the dry phase of the basin and others which might have been called into action by the formation of the lake itself. Winds carried seeds about the region in a manner probably but little changed. The run-off streams coming down the slopes continued their characteristic action in the incidental transportation of fruits, seeds, and propagative bodies, much as in the dry periods of the Sink, but the flood streams of the Colorado would constitute an unusual feature, while the wave and current actions of the lake would be features directly connected with the existence of the lake. The activities of birds in distributing seeds has been the subject of a voluminous literature, although their actual efficiency in carrying seeds to new areas has not been adequately tested. Their possible intervention in this region, however, was enormously increased in importance, since vast numbers of cormorants, pelicans, and ducks were attracted by the possibilities of food in the form of fish offered by the lake. It is obvious, therefore, that the phenomena connected with the making and desicca- tion of the lake did not include simply the extermination of the species on an area and the simple re-entrance of these and other species into the bared area. The things which exter- minated the vegetation brought many new conditions, including new disseminational agencies and a whole series of soil conditions different from those prevailing in the area previous to flooding. The progress of a seed from its place of origin on the slopes of the Cahuilla toward the bared beaches might be imagined as across or over a number of barriers, which would be of such character as to stop a large number and thus form the sieves of natural selection. Thus, for example, a free seed being carried down the shallow channel of a run-off stream during the few minutes or hours in which water would run down its shallow channel at intervals of a year or two, would be liable to be buried under débris, crushed by rolling rocks, or deposited where it might be eaten by rodents—or, being wetted, would be killed by the high temperature. It is as if the moving seeds were thrown against a series of sieves the meshes of which offered openings differing not only in area but in form. The screening GENERAL DISCUSSION. 181 action of the environic sieves was such that out of the hundreds of species which must have been thrown against the barriers only 60 reached the beaches and germinated. The last of the series of screens would be the conditions of the beaches themselves, which varied from year to year. To complete the figure it would be necessary to imagine the last of the series of sieves in the great selecting machine as being constantly changed in pattern and size of mesh. Furthermore, the selective action does not cease with the germination of the seed which has been carried to the beach. Here a new series of sieves was to be encountered, con- sisting chiefly of the phenomena of progressive desiccation, which would soon bring many species to the limits of endurance and to death. Others surviving under extreme tension would offer suggestive possibilities as to responses of possible importance in their evolution- ary development, such as have been suggested by the behavior of Prosopis, Aster, Scirpus, and Atriplex. It became obvious during the course of the work that the origination of qualities or structures upon which dissemination would depend might, in many instances at least, have no possible connection in a causal way with the agencies themselves. Thus, for example, the desert gourd which was carried about the lake and deposited on various beaches owes this dissemination to structures which could hardly be attributed to any excitation action on the part of water or to any previous selecting action. The same mechanical qualities of flotation and dissemination are displayed by the fragments of pumice which were carried about at the same time. The communal life and successions on the beaches showed some interesting diversities. The manner of the occupation of the zones laid bare by the receding waters was chiefly determined by the water, and hence the first communities of plants were of the nature of strand-steppes. The history of such formations showed two distinct phases, both also determined chiefly by edaphic conditions. The strands of the more gently sloping alkaline beaches were at first occupied by a greater number of species than the bared strips on the steeper gravelly shores, and the pioneers gained a foot-hold earlier. This difference may be attributed in greater part to the fact that the narrower, more steeply sloping beaches were subject to the action of storm waves for a longer period than the broader zones on the gentler slopes, and also to the fact that the latter actually presented a greater area of soil for the reception of seeds. The arrangement of the pioneers on any beach would be characterized as open, and on the gentler alkaline slopes the tendency in general was toward a decrease, both in the number of species and individuals, with few or no secondary introductions, thus making a direct change toward the true open or desert formations. The gently sloping beach at Mecca, however, receiving some seepage water, did not exhibit such simple results. Steeply sloping beaches, as represented by the Travertine Terraces and the shores of Obsidian Island, showed two phases of succession, differing chiefly in degree. The open formations on the shores of Obsidian Island showed some tendency to become closer or denser in irregular areas, which soon began to thin in the final change toward the open desert forma- tion. The benches of Travertine Terrace were characterized by the development of dense ranks comprising a half dozen species or less at the upper margins of the annually bared strands, which were soon thinned in accordance with the general tendency toward desert formations, but at the same time the greater part of the surface of the terrace was knit together in a close formation by a mat of Distichlis. This closed formation, however, soon began to show the effects of desiccation, and the progression toward the desert formation with the introduction of xerophytes would be seen within three or four years after the zone had been laid bare by the lake. The transition is so rapid, or so abrupt, that species appearing on strands two years old are also included in ancient beach ranks marking the positions of strands 300 to 400 years old. The revegetation of the area submerged by the 182 THE SALTON SEA. waters of the lake in Salton Sink is therefore seen to be chiefly influenced by soil-water or other edaphic conditions during the first year or two after emersion, after which the forma- tions become increasingly open in the progression toward the extreme desert type. The evaporating capacity of the air may be taken as being an agency of increasing importance with the age of the formation. Submergence and consequent extermination of the flora of portions of the Salton Sink have occurred many times in the last few centuries, and the reoccupation of the bared strands has taken place with the complex interplay of biological and mechanical agencies partly suggested and partly described in the preceding pages. Hee igi Hi eves as a ey Paige: meen é; iy Sarwan Vint fe y fy iti ii Jot Kyiv G Burs ray ea its Minter, aie Tata a in ery Da ase veNe ana Hage ry Nit lt a ay ili Bane at NI ad ier ORT eo EATERS “ Pilati ! aL, KV Aah es t HPs) ae oo ata h an) yf S : RANE te br) i yo OUND Gi ey 3 7 yy Ny a Nes ah