owia la ee ee nateeels oe LA ass aieNiadleliedle ed “ha Se Ad ah fed Mian lta el ety - tt Br Ha i “¢ valnoriws c ree we oeren ve Cer ele PH ‘? Oa ee wee Pr died voeleate oeheie er avi git ewe vt aot arereieteiatel? ~ s'e* ‘aneet Oe Lt iis pif dies ete verwvever® veeeeesoee ‘ vrrre aren’ ned " eer - 4 vre'eyy re a wieneiete o 2 tino Vere ‘geavese ven ”» mT es ee we Vy rie rie-ns sie surrey as vevvrer re Ssvy's ee et abe hu's eV 89 aw vie's eRe’ R ply a a ld “aererery geerrvererrrs . de Th ah send hh eh AA ee. AL date tee ie” vrenets wre COOP ess wr EOE ROT FV ddd he ee wwe es POV ISP eee We er eww Lillh-ds ee ee oes * : VVC rw gees eee own ee pee’ vw ewnv we oe ve ayy ee PVUERTEU EV OVE tear’ Ae dk Sie lb hd Ah i db dette RS PE ee dedh Sil th hdl eda te tts vel cd dilicth fiend WOR TES Ferry ele ovetews ere err eres err oer o'v'vtgiee'e'e'¥ wesvenee ‘ ‘ ¢ vietetp ero onms gett eee POY Tere eVPV Ver ewe rey Te eee ews ews'ey - renew wy Ayo Vere Tree a ‘ ‘ Ln a oe ee eveevurwre ‘va ‘a ron evn errs CVF VOW WV ess oe eee eee ere CO eV Cow ew Y fry EP ww nie re” SPREE PS oT RRP ee Ce PCOMTT VT CRE NE eNO PPE PEE rr , vee ve vite ers VEEP eR we ei ovrrtous evevereare eome . vee ere ree eee eet tere Pera TUT wT TEP Rey ‘ were oe vere a sutviele vin'y Pe OCH HF ae eee " eww Fala ete wees € oT RU wre POO er Pee Ue 8 eee ww eve eee we Tee OPC eee ee ee ee ee ‘yy UO ewes so ' ea ee he eB “oo OW Vie WT ge , ee i ht bern t-9 1th ey ee Ps nestled tote Pliner eet Fee ee ee ea ya al ele! ar ery erg WF FIT Pellyl yy oie ote at a deh COT VP Vereen e A Ce Re ee ee ee ata Pye pale Witle ee elev vis es ee de ee a de Fee ieee Nyheter tere tpi wet 2 ee Or te) ire a es-c> Oa A OE re tary preieanaee : Poe te ee ae i ere eliete ighgi tgle a? of! eee! ae rawr Peer ee ee tote! rut eet Par A ae ik Ae de ve “ igh etary eraraly ae Wrelere’ ee ee ee ee ae Pra a be a Ce hahycotuloteyig’ * orely ale lag ale aye eer ee ee Five ee" ee rf ofuPyre? Pyle ee raete oe ati ve etee ere i ane Ok Me a ie See PORE wiels Pre eRe Liha A PaO PR Ur ewes Ryo's wile wiv ler or rey aelyete ee eh wle we rs e rw he ell aTNe FFE R Ye + + Witla winwiete a ee RE WWD ; Ci a PETE OV eh ow eure es eleiete ed we oa ee avwve serve i ee ee ee a ory er Peer wee ‘ ‘eerwinet aera ee ee eee +8 ‘ OVV VE BW ee ele ene nisi see wie d ta n Pe eb eee Se a weer eee eles rw haly ¥ HOV ERR cele e are pei Sw lee 88 alee vive pis wld elre 4 piel ete! ee ee ee ee et is ‘ ee Feiss Bapsi thd 6 2 weve ae te Peres ‘ ‘ eeatehe TEU UUREG CA ee ee : TOOULORMEL LER Uy EAR eM OM ek oe ' ™ ‘een eo wary ieee a) rth eet : ve vietels'e be vee ee vive 4 ae i ee : ee ereetete a Wt 3 eases te Be add rgtae do's . getdate esp etv erate eae eee ee pwee vie wi kl ae ee eee 4, “al avy sip eer he bene. > ’ Sere et eee ‘ eeereeemen wee ar tiake roe ry Maem comet Te ih A A dl Pahang hea he hp itn hg! pom eel a lee ere RTE Re CP pre! SUM oe a ee Le ek air etwie ia tet ete ene Hew ee ee Bs ei ohn lelene! shee iele ip etety Tae eee eels ry ah enwret Tieteteete ete er rehete tere inte al alate tetate! Feta a heals eine eee hee neh wel e a iewhy Dele a ik haley belle eels Se ae heteral aretetetwieten estate taser i tok he ee ele! ie woe ae ee ONE Meee Tee vi erere’ v erenewrre ey eet eee ey t Reterel range Pee eee —_—— a ot. =? We is SS he Ger eed ee 0 SIN re ae dN Od Phy cmmieit ii iwunrcOOecnbbeeey” cau. 8 we AEE RCH Pep hr pe Ld a owt j 2 ee ' eT 4 ri « . TOR - { 4 |. an Mth A anit oNAn ok PPL -)) i a Ck. Nees 2 d were matte my Avr a | ifr 4 vi fas. A 2, aa 3 - ial . | way hel Move sd “ Play : SS Nae PAN veniam el “sae gt iH peeeee 4] sean ene: waeng Ae itil a TE ERO | HHL FEEL i Ad dt) TLL 4 bubba ded Yi y rg va rinargll eon En WAN hale, Pam eere 4 dela) COT Fae (thn ad ; ing PER de wihy Athy Hine wi ‘een, Vee Way. \ VA - svantennnrstee Tut | si 5 | ap oe "watt AE YU Vy Hey AAA AD | ath Na dng’ a NW 54> Wa ATR ASF Om Avi POURRA) Ey ak PULL T Wie ao Mf nail i * Ts |S 7% ms . : Wich ARN AAA AT Y ep EO v tai | ivi LEGER AAS | Nis Wot im wre Pad | &. NSS Iga Ser lela eh |, Ip Ca wet ha Ue A Ne, ie om ye". 4 Sr m e4 WED 4, ey NG pte Wi tat A w~ \ “Wty “A Goote, A. \ pinta ibis ery Vid mee Ny: er AZ RA Lie auch byt \ CF ; > ae ab. > Dp a- a } : 7% TA “yas Ad | Thad Pilhice A | “ehh e 1H ae ae yy Tl Lak LALLA ROY tow. PIA TTT AS 4p Nant) ||) ies | | ne’ “7 in| “wedeneny nn & Pre ~ “ela yrs SHS, Lb qe SOUTMA SUIT Rp ae ATO RNIy OUT cen Hi sti ney” At yit! ONY eels ea fini sve eT NNW \y¥u * 3 | | ms TEL TT “mazes? é aed Oe op ye Oy ow al : | that A ed oat te Ps AANA ra Le 4p, ¢ APE i teenie Very ie@ een ‘sre pol ‘| ot] aa Pe ' | jin, ] 3 \@ sat iy es A | | 3 4m. we aities watt | “0h WADA frre... ser MM RANA AAS) Ny . — : s Oy. ONEDs- - l . » HY ef 1 ey IP ng vr + = wer. Bas Bat ps 8 AA ee tp p< a. vw Oe ate Wr Hie t Wt ae Eee! | |S : WEN wennees vis 4 Meare Vayaatirereti nt i beh - 7 AL a he | revel, | hla ‘ . be ad . ee ; : & cA yr ey . ‘ t. vw Le TENS ith ng bh es OECD Rca ew crac dey cette cmc see ae Pe "ed PAD. or A an “YO pe hes at | ~ ty ee ty A | Ming sag a mah Ss fat cx (ua “6¢ ee eat | 4 ss - wm ta ‘@,. yn it Mit th ML $5 wy eT; rs : t. ‘ a] w & rss \ Rr e “ty a Vo. EET. eo eS oe ad w Tries 2“ \ Tea ew HMA tt Le Wee wee. ee REE we aw for eNed s A | te wd th AR PTL Rh hk EOP ) ode ra tog ee ct ped acePaba hehe Mua ah Rau -* Wiens AA Ladiad da use ‘a pas Were iM ' My Gees re SNAP AS nh Within’ Ty) Pratt se RUNNA DOA aipaendor 8 ydy. PEt lan jte ipewit c Sec eny } «> Aah Aah vit aie ; ; past peta est ree pO , aS? sAnbensticel x pest at Y Stet al Len \ ee tee VWs idle tea Oe oh 040. we Perel) W ail Wis ing Lett WMounmeses See etn a ial Uee | - J, a > | iy nett Vey HHT +) MM a & ie add AY fl I yuk Vanda Se wt | ee eles! aA ae eee Le Newer | ye eee Me at Suto cone Pp aye oe Tid etl. TG a id! | 1) 11 | pep taba Pe etal a “2 NUS ANE et Pekin wil ddA] ov at | we ay. roe wa ach Sits eeuheh poee ey TY AAS Sap sieht wr hh ve 8 “Ws i , OM Wasncee ay | . “ CA Aly; w w Sam Nil 4 1 w ; Sy 2 *e., Z Juin, ay: a'r polls. well! ee ae. rT Ah Raach] we Oley "oy “v wt AWS A ‘ 2 ra A; Pe, “aq \ ty wit SsShky le we at rer N ‘i “Vv iy 45 sd - of 1 Ne Fe ig : ww Pn a a” wealtoes, nn MS Na “My My SUC sip 2 WN, | \ iit qf Her ws bets anti? on a by be v p oe eaaene™ hy My SST e Tyabb heh Nag eR Seesaw iaeMl waetiog gs OEMINE LMT GT ahaa Qh IN oR a ANd Le SAL ALT L) Leena Ts ee COTY vig ey AI rs) ™~ 5 THE AMERICAN JOURNAL OF SCIENCE, Epnitorn: EDWARD 8S. DANA. ASSOCIATE EDITORS Prorrssors GEO. L. GOODALE, JOHN TROWBRIDGE, H. P. BOWDITCH awp W. G. FARLOW, or Camprin6GE, Prorrssors O. C. MARSH, A. E. VERRILL anp H. 8. WILLIAMS, or New Haven, Prorrssor GEORGE F. BARKER, or Putwaperpuia, ProFrressor H. A. ROWLAND, oF Battimorge, Mr. J. 8S. DILLER, oF Wasuinerton. FOURTH SERIES. VOL. II—[WHOLE NUMBER, CLIIL] Nos. 13-18. een Boy WOU NE. 1897. oY WITH SEVEN PLATES. fi Ss On \\ LS ¥os! '@flenal Miseut: 2 we NEW HAVEN, CONNECTICUT. “32% So ARNT AE SOT. ( Bee a } "eee eee THE TUTTLE, MOREHOUSE & TAYLOR PRESS, ov Fc ea ee ag NEW HAVEN, CONN. 4 “a ae 4 {, i, — CONTENTS TO VOLUME III. Number 13. Page Art. I.—Worship of Meteorites; by H. A. Newron.-.--- a II.—Spectra of Argon; by J. TrowsripcEe and T. W. _SESTICOTETUN Teel SI hl Oe I Bs Ieee pa ne eS EA Rt ed Bel Ze 15 III.—Some Queries on Rock Differentiation; by G. F. TSSTY DUROTAR, SESS STS Ml aN ani eee er ee cy. MIR A et a IV.—Igneous Rocks from Smyrna and Pergamon; by H. S$. BURA SETTING Gall QING eee re eh wt es A se ay 41 V.—Revision of the Genera of Ledidze and Nuculidz of the Atlantic Coast of the United States; by A. E. VERRILL SAD’ Tees ic) BOE] BN OTs teelcliah setae, wkd ita ap ateaN ade) an Eins Nese canara 51 VI.—Experiment with Gold; by M. C. Lea... .__---.---- 64 VII.—Note on a new Meteorite from the Sacramento Moun- tains, New Mexico; by W. M. Foorr. (With Plates I PICU eer een nens yt ts cuban fy | 9 apyret ty Dare Larceny uy WS 65 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Absorption Spectra of Iodine and Bromine solutions at Temperatures above the Critical Temperature of the Solvent, Woop: Boiling Points in a Crookes Vacuum, KRAFFT and WEILANDT, 67.—Formation of Per- sulphuric Acid, ELBS and SCHONHERR, 68.—Reaction of Silver oxide upon Hydrogen Peroxide, RIEGLER: Silver peroxynitrate, SULC, MULDER and HERING, 69.—New Hydrocarbon, SCHICKLER: Fractional Distillation of acids of the Acetic series, SOREL, 70.—Rontgen Rays, J. MACINTYRE: Rotation in Constant Electric Fields, QurncKE: Interferential refractor for electric waves, O. WEIDEBURG: Cadmium normal element, W. JAEGER and R. WACHSMUTH, 71. Geology and Mineralogy—Geological Survey of Canada, Annual Report, 1894: Pleistocene glaciation in New Brunswick, Nova Scotia, and Prince Edward - Island, R. CHALMERS, 72.—Notes sur la flore des couches permiennes de Trien- bach (Alsace), R. ZEILLER, 74.— Artificial Production of the Mineral Northupite, DE SCHULTEN, 75.—Genesis of the Tale deposits of St. Lawrence Co., N. Y., C. H. Smytu, JR. Handbook of Rocks for use without the Microscope, J. F. Kemp, 76. Botany and Zoology—tillustrated Flora of the Northern United States, Canada, etc., N. L. Brirron and A. Brown, 76.—Notes on the Flora of Newfoundland, B. L. Roprnson and H. VON SCHRENK: Survival of the Unlike, L. H. Batter: Sphagna Boreali-Americana Exsiccata, D. C. Eaton and E. Faxon, 77.— Analecta Algologica, Continuatio III, J. G. AGARDH: Phycotheca Boreali- Americana, F. §. Conuins, I. HoLpEN and W. A. SETCHELL: Ueber das Ver- halten der Kerne bei den Friichtentwickelung einiger Ascomyceten, R. A. HARPER, 78.—Gigantic Cephalopod on the Florida coast. Miscellaneous Scientific Intelligence—History of Elementary Mathematics, F. Casori, 79.—Eclipse Party in Africa, HE. J. Loomis: Elementary Meteorology for High Schools and Colleges, F. WaLpo, 80.—The Meteor of December 4, 81. Obituary—BENJAMIN APTHORP GOULD, 81. Catalogue of the Collection of Meteorites in the Peabody Museum of Yale University, 83. 1V CONTENTS. Number 14. Arr, VITI—Outline of a Natural Classification of the Trilo- bites; by C. HE. Bexcuer.. (With Plate IIL) .-__.__- 89 IX.—Preliminary Trial of an Interferential Induction Bal- ances .by.C) BARUS4 = 22.2. 6 107 X.—The Multiple Spectra of Gases; by J. Trowsripe@s and 2. - W;- RICHARDS = =_.:.¢222s22-222442¢:201: 225 117 XIJ.—Studies in the Cyperacee; by T. Horm. (With Plate PV) So Soe Sa OS RS) 121 XIJ.—Simple Instrument for inclining a Preparation in the Microscope; by T. A. JAccAr, JR." 92) eee 129 XIII.—Nocturnal protective coloration in Mammals, Birds, Fishes, Insects, etc., as developed by Natural Selection ; by ALE.) VERRILL. oo. oe so ceie oe 132 XIV.—Nocturnal and diurnal changes in the colors of certain fishes and of the squid (Loligo), with notes on their sleeping habits; by A. E, ‘VanRRinu. 222-2) 22a 135 XV.—The Stylinodontia, a Suborder of Eocene Edentates ; by OC) Marsa! ys uiveiget Jens) 20). ee ee 137 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Diffusion of Metals, RoBERTS-AUSTEN, 147.—Optical Rotation in the Crystalline and the Liquid States, TRAUB, 148.—Electrolysis of Water, SoxoLorr: Electrolytic Production of Hypochlorites and Chlorates, Oxrt- TEL, 149.—Action of Nitrous acid in a Grove cell, IaLE: Spectra of Fused Salts of the Alkali Metals, DEGRAMONT, 150.—Preparation of Lithium and Beryllium, BorcueErs: Light of the glow beetle, H. MuraoKA, 151.—Ré6ntgen Rays, KEL- tN: Electric hght in Capillary tubes, O. ScHotr: Temperature of the sun, W. K. WILSON and P. L. Gray: Argon and helium, LOCKYER, 152. Geology and Natural History—U. S. Geological Survey, 153.—Geological reconnaissance in Northwestern Oregon, J. S. DILLER, 155.—Underground water of the Arkansas Valley in Hastern Colorado, G. K. GiuBertT: Geological Society of America, 156.—Pre-Cambrian rocks and fossils, 157.—Antiquity of man in Britain, W. J. L. Apporr: Age of the Lower Coals of Missouri, D. Wuitk, 158.—Relation of the fauna.of the Ithaca group to the faunas of the Portage and Chemung, E. M. KinpLEe: Phosphate-Deposits of Arkansas, J. C. BRANNER, 159.—Die Leitfossilien, ein Handbuch fir den Unterricht und fur das Bestimmen von Versteinerungen, KF. KOKEN: Ueber die neue geologische Uebersichtskarte der Schweiz, C. ScHmipt: Ancient volcanic rocks of South Mt., Penn., F. Bascom, 160.—Geology of the Fox Islands, Me., G. O. SMITH: Cell in Development and Inheritance, E. B. WiLSon, 161. —Tables for the Deter- mination of Minerals by Physical Properties, ascertainable with the aid of a few Field Instruments, P. Frazer: Der Lichtsinn augenloser Tiere, W. A. Nace: Additional information concerning the giant Cephalopod of Florida, A. E. VERRILL, 162. Obituary—GEN. FRANCIS A. WALKER, 164. CONTENTS. V Number 15. Page Art. XVI.—Crater Lake, Oregon; by J. 8. Dinter. (With 1> HEM: Ys ee cls sc a ele pe eld a il gates tech 165 XVII.—Origin and Relations of the Grenville and Hastings Series in the Canadian Laurentian; by F. D. Apvams, A. Be SREON Ann. Wier LEB oes core tt oe See 173 XVIII.—Outline of a Natural Classification of the Trilo- bites by O.n., bEEBCHER, }'(Paroubl,)-2. 2% (3% se eeo) 181 XIX.—Scoured Bowlders of the Mattawa Valley; by F. B. iecnoRie tt) G01 Tee ara 2 a, ee ito e 208 XX.—Excursions of the Diaphragm of a Telephone; by C. MEW Ss pee Taree hilt mess tube tire glee ae oe 219 XXI.—Arctic Sea Ice as a Geological Agent; by R. 8. Tarr 223 XXII.—Contribution to the Geology of Newport Neck and Wouament tlaids by -W. O. Crospy 2). . 20 ).52! 230 XXIII.—Estimation of Molybdenum [odometrically; by F. PUROMOCH 222 2 in = bees Sek ec es ae seo PENA ON 237 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Density of Helium, RAMSAY: Expansion of Helium and Argon, KUENEN and RANDALL: Electric discharge in Argon and Helium, CoL- LIE and Ramsay, 241.—Gases obtained from Uraninite and Eliasite, LOCKYER, 242.—Preparation of Lithium and on a quick Nitrogen Absorbent, WARREN: Artificial Production of Diamonds, Moissan, 243.—Preparation of Metallic Hy- droxides by Electrolysis, Lorenz: ‘ Excited” Metals, and on Excited Alumi- num as a reducing agent, WISLICENUS, 244.—Constants of Nature: Compact apvaratus for the study of electric waves, J. C. Bose: New formula for spec- trum waves, J. J. BALMERS, 245.—Discharge rays and the connection between them and the Cathode and Rontgen rays, M.. W. HorrMann: Large storage battery: Outlines of Electricity and Magnetism, C. A. PERKINS, 246. Geology and Mineralogy—Principal features of the Geology of Southeastern Washington, I. C. RUSSELL, 246.—Geologic Atlas of the United States Yellow- stone National Park, U. S. Geol. Surv., 1896, 248.—Report of the Director of the United States Geological Survey for the year 1895-96, C. D. WaLcorTt, 249. —Geology of the Castle Mountain Mining district, Montana, W. H: WEED and L. V. Prrsson: Eocene deposits of the Middle Atlantic Slope in Delaware, Maryland, and Virginia, W. B. CLark: Catalogue des Bibliographies Géolo- giques rédigé avec le concours des membres de la Commission bibliographique du Congrés, EK. DE MARGERIE, 250.—Yellow Limestone of Jamaica, R. ‘[. HILL: - Handbuch der Mineralogie, Dr. C. H1nTz#, 251. Miscellaneous Scientific Intelligence—Annals of the Astronomical Observatory of Harvard College: Smithsonian Physical Tables, T. Gray: North Carolina and its Resources: National Museum, 252. v1 CONTENTS. Numbernd. Art. XXIV.—Experimental investigation of the equilibrium of the forces acting in the flotation of disks and rings of metal :. leading to measures of surface tension; by A. Me Mayerige te oi) 0 SO Aa te ee ens 253 XXV.—Computing Diffusion; by G. F. Beckrr--_---.----- 280 XXVI.—Acid Dike in the Connecticut Triassic area; by E. Op Oviny, Oe oo C2 PI ee es 287 XXVII.—Application of Iodic Acid to the Analysis of Iodides; by F. A. Goocu and C. F. Waker -.-...--- 293 XXVUIL—Granitic Rocks of the Pyramid Peak District, Sierra Nevada, Cal.; by W. LinpGREN..------------- 301 XXIX.—Difference in the Climate of the Greenland and American sides of Davis’ and Baffin’s Bay; by R. 58. PARR o's soe 2 ee yt eee |e 315 XX X.—Foramina perforating the Cranial Region of a Per- mian Reptile and on a Cast of its Brain Cavity; by E. Co CAS Bog) yee ee ee, or ree ee 321 XX XI.—Temperature and Ohmic Resistance of Gases dur- ing the Oscillatory Electric Discharge; by J. TRow- Page BRIDGE and T. W. Ricwarps. (With Plate VI.).---- 327 XXXII.—Does a Vacuum conduct Electricity?; by J. TROWBRIDGH 2 D4 oa ee 343 XX XIII.—‘ Plasticity ” of Glacial Ice; by I. C. RussEtu. 344 XX XIV.—Affinities of Hesperornis; by O. C. Marsu----- 347 SCIENTIFIC INTELLIGENCE. Geology and Mineralogy—Vertebrate Fossils of the Denver Basin, O. C. MARSH: Geology of Minnesota, 349.—Geological Survey of Alabama, H. McCatiey: Catalogue and index of contributions to North American Geology, N. H. DAr- TON: Description géologique de la partie sud-est de la 14-me feuille de la vii zone de la carte générale du gouvernement Tomsk (Feuille Balachouka), P. VENUKOFF, 350.—Catalogus Mammalium tam viventum quam fossilium, EH. L. TROUESSART: Congres géologique international: Geology of Santa Catalina Island, W. 8. T. SmitH: Elementary Geology, R. 8. Tarr, 351.—Catalogue of the Minerals of Tasmania, W. F. PETTrERD: Production of Precious Stones in 1895, G. FE. Kunz, 352, Botany and Zoology—Manual and Dictionary of the Flowering Plants and Ferns, J. CU. WILLIS: Studien ueber den Hexenbesenrost der Berberitze, J. HRIKSSON, 353.—Phycotheca Borealis-Americana, COLLINS, HOLDEN and SETCHELL: Ana- lytic Keys to the Genera and Species of North American Mosses, C. R. BARNES and F, DEF. HEALD: Index Desmidiacearum citationibus locupletissimus atque Bibliographia, C. F. O. NorpsTtEDT, 354.—Die Protrophie, A. Minks: Supposed great Octopus of Florida, 355. Miscellaneous Scientific Intelligence—Lehrbuch der Allgemeinen Chemie, W. Ost- WALD: Inorganic Chemical Preparations, F. H. THorp: Tutorial Chemistry, G. H. Battey: Laboratory Manual of Inorganic Chemistry, R. P. WILLIAMS, 357. Mountain Observatories in America and Europe, EH. S. HOLDEN: Essays by George John Romanes, C. L. Morgan: North American Birds, H. NEHRLING: Das Klima von Frankfurt am Main, J. Zie@LpR and W. KOnIG, 358. Obituary—Professor JAMES JOSEPH SYLVESTER, 358. SS — CONTENTS. Vil Number 17. Page Husert Anson Newton; by J. Wittarp GIBBs-.-.-.----- 359 Art. XXXV.—Means of producing a Constant Angular MELOCHV Aye IN. «Gs WEBSTER toe eS. 379 XXXVI.—Rapid Break for large Currents; by A. G. (AT TRYST fy Ste Re aie a eee ee Rae en eee 383 XXXVII.—Electrical Conductivity of the Ether; by J. SRrzonyE iMG) et hee EOD IN Me ee ReU ae Psu 387 XXXVIII.—Effect of Great Current Strength on the Con- ductivity of Electrolytes; by T. W. Ricuarps and IMB OMGER EDGE eased se o2 oe ee oe ts BO) XX XIX.—Southern Devonian formations; by H. 8. W11- TL LLANES Gos Aa NU oe 05S MR EA A ak Je ae, Rk ee Roe Tee Ee a 393 XL.—Genus Lingulepis; by C. D. Watcotr _.-____.--.--- 404 XLI.—Seiches on the Bay of Fundy; by A. W. Durr-_---- 406 XLII.—Reeblingite, a new Silicate from Franklin Furnace, N. J., containing Sulphur Dioxide and Lead; by S. L. PeINEIRED and. Ei. Wh Kh oOTn es hae oS |e Lk |: ALS SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Rotation of Circularly Polarizing Crystals in the State of Powder, LANDOLT: Oxidation of Nitrogen Gas, RAYLEIGH, 416.—Pres- ence of Nitrites in the Air, DEFREN: Solubility of Lead and Bismuth in Zinc, SPRING and RoMANOFF, 418.—Traité Elémentaire de Méchanique Chemique fondeé sur la Thermodynamique, P. DunEM: Elements of Theoretical Physics, C. CHRISTIANSEN, £19.—Theory of Physics, J. S. AMES: Outlines of Physics, an Elementary Text-Book, E. L. NicHors, 420. Geology and Natural History—Geological Survey of Canada, 421.—Boletin del Instituto Geolégico de México, J. G. AGuImLERA: Introduction to Geology, W. B. Scort, 422.—Glaciers of North America, I. C. RussEtu: Treatise on Rocks, Rock-Weathering and Soils, G. P. MERRILL, 423.—Elemente der Mineralogie begriindet von Carl Friedrich Naumann, F. ZirKEL: Introduction to the Study of Meteorites. 424.—Pseudomorphs after halite from Jamaica, W. I., E. ©. Hovey: Flora of the Southern United States, A. W. CHAPMAN: Neural Terms, International and National, B. G. WiupEr, 425. Miscellaneous Scientific Intelligence—National Academy of Sciences: Microscopic Researches on the Formative Property of Glycogen, C. CREIGHTON: Tutorial Statics, W. Brices and G. H. Bryan: Royal Society of London, 426. Obituary—EDWARD DRINKER Cops, 427.—MatrHew Carey Lea: Josepa F. JAMES, 428. Vill CONTENTS. Number 18. ‘ap Page Art. XLIII.—Studies in the Cyperacee; by,T. Houm.__--- 499 XLIV.—Bacteria and the Decomposition of Rocks; by J. Cu BRANNER, ci¢- bo 22 Soils 4. Jb XLV.—Wellsite, a new Mineral; by J. H. Prarr and H. W. Foor. .. 2.2 oS 2ee BE 2 i 443 XLVI.—Magnetic Tnerement of. Rigidity in Strong Fields; by: H. -D. Dav cose: Se eee 449 XLVII.—Geologic Fault in New York; io P. F. ScHnEempEerR 458 XLVIII.—Certain Double Halogen Salts of Cesium and Rubidium; by H. L. Wetts and H. W. Foore_..... .- 461 XLIX.—Double Fluorides of Zirconium with Lithium, So- dium and Thallium ; by H. L. Weis and H. W. Foorr 466 L.—Broadening of Sodium Lines by Intense Magnetic Fields ; A. StC. Dunstan, M. E. Rice and C. A. Kraus ‘a 472 LI.—Relative Motion of the Earth and the Ether; by A. A. MICHELSON |... 5-25 3-2 bee es eee eee 475 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Relation of Refraction to Density, TRAUBE: Properties of Free Hydrazine. DE Bruyn, 479.—Nitrogen Pentasulphide, MUTHMANN and CLEVER, 480.—Direct Union of Carbon and Hydrogen, BoNnE and JORDAN, 481. —Preparation of Rubidium and its Dioxide. ERDMANNand KOTHNER: Pyrogenic Reactions of Aliphatic Hydrocarbons, HABER, 482.—Explosive Properties of Acetylene, BERTHELOT and VIEILLE: Form of the Atoms, C. FOURLINNIE, 483.—Use of rapid Electrical oscillations for determining dielectric constants, W. NERNST: Ultra-red rays, H. Rupens and A. TROWBRIDGE: Production of X-rays of different penetrative values, A. A. C. Swinton: Physiological effects of the X-rays, M. SoREL, 485.—Influence of magnetism on the nature of light emitted by a substance, P. ZEEMAN: First Principles of Natural Philosophy, A. E. DOLBEAR, 486. Geology and Mineraloyy—Congres géologique international. 486.—Geological Sur- vey of Canada. 488.—A guide to the fossil invertebrates and plants in the department of geology and paleontology in the British Museum: Sea mills of Cephalonia, F. W. Crospy and W. O. Crossy: Geologischer Wegweiser durch das Dresdner Elbthalgebiet zwischen Meissen und Tetschen: Lawsonite: Preliminary report on the Corundum deposits of Georgia, 489. Botany—Practical Botany for Beginners, F: O. Bower: Guide to the Study of Common Plants, an Introduction to Botany. V. M. Spaupine: Hlements of Botany, F. Darwin: Laboratory Practice for Beginners in Botany, W. A. SETCHELL: Lessons in Elementary Botany for Secondary Schools, T. H, Mac- BRIDE, 490. Miscellaneous Scientific Intelligence—Researches on the Evolution of Stellar Sys- tems T. J. J. Sen, 49]1.—Annals of the Astronomical Observatory of Har- vard College: Geographische Abhandlungen, A. PENCK, 492.—Beitrage zur Geophysik : Zeitschrift fir physikalische Erdkunde, G. GERLAND: Lavori esequiti nell’ Instituto di Fisica dell’ Universita di Pisa, "HL. Bavretti: Ostwald’s Klassiker der exakten Wissenschaften: La canse premiére d’apres les données expérimentales, E. FerrrhreE: Den Norske Nordhavs-Expedition, 494. InpDEX TO VOL, III, 495. Chas. D. Walcott, U. S. Geol. Survey. Established by BENJAMIN SILLIMAN in 1818. AMERICAN JOURNAL OF SCIENCE. Epvirorn: EDWARD S. DANA. ASSOCIATE EDITORS Prorrssors GHO. L. GOODALE, JOHN TROWBRIDGE, H. P. BOWDITCH anv W. G. FARLOW, or Camsprines, ‘Proressors O. C. MARSH, A. E. VERRILL ann H. S. WILLIAMS, OF NEw Haven, Proressor GEORGE F. BARKER, OF PHILADELPHIA, Proressor H. A. ROWLAND, oF Battimorgz, Mr. J. S. DILLER, or Wasurneron. FOURTH SERIES. VOL, IN—[WHOLE NUMBER, CLIII.] ‘No. 13.—JANUARY, 1897. WITH PLATES I-II. NEW HAVEN, CONNECTICUT. . f Soe TUTTLE, MOREHOUSE & TAYLOR, PRINTERS, 125 TEMPLE STREET. Published monthly. - Six dollars per year (postage prepaid). $6.40 to foreign sub- - seribers of countries im the Postal Union. Remittances should be made either by me orders, registered letters, or bank checks. Scientific and Medical Books — | (After N ovember woth address all communications to 13 1 is Sr es Arch Street.) y y $ . Bye pees eae lie se MINERALS. We take pleasure in announcing that a long lease has just been taken on the large building No. 1317 ARCH STREET, PHILADELPHIA — where our central offices and salesrooms will be located after NOVEMBER 25, (896. The necessity of having a central store within easy reach of patrons visiting the city, has grown within recent years, and while the great bulk of our | stock will remain on storage, there will be on sale in the large and ie: lighted show rooms all of our best minerals and books. > The display of minerals will be of especial importance, being unquestionably _ more extensive and finer than can be found in any similar establishment in the world. = THWE NEW LOCATION isin the heart of the business section of the city, and being midway between the two great railway depots is readily accessible. It is less than three minutes walk from either Broad St. (Penna. R. R.) or Twelfth and Market (Phila. eas Reading R. R.) and is. adjacent to all the great retail stores, YOU ARE CORDIALLY INVITED to call whether you expect to pur- ae chase or not, as we take especial pride in showing visitors through our ae establishment. ie 8 ——séDr. A. E, FOOTE, WarRREN M. Foote, Manager. 1884-26-28 North Forty-First Street. PHILADELPHIA, PA., U.S. A. J a of Jf ig wig, — —£ 7 7 4, ips £ * P)'s on iy id; er’ i ee! f THE AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES. ] Se ioe Art. I1-——The Worship of Meteorites; by HusBeErt A. NEWTON. [A lecture delivered* in New Haven, Conn., March 29, 1889.] HERE is a small fragment of iron that has a curious history. It is a portion of a mass of meteoric iron found upon a brick altar in one of the Ohio mounds. Along with it were various objects—a serpent cut out of mica—several terracotta figu- rines—two remarkable dishes carved from stone mto the form of animals; pearls, shells, copper ornaments, and nearly three hundred ankle bones of deer and elk. There were but one or two fragments of other bones, and one animal furnished but two of these ankle bones; hence they must have been selected for some special, important reason. The figurines had been apparently broken for some purpose, and the whole col- lection had suffered in the fire not a little. In a like altar of another mound of the same group were found nearly two bushels of like objects. It must have been in some ceremony of a religious, possibly one of a funereal, character that the mound builders collected here on the altar their ornaments and other valuables, and after burning them buried the charred debris in the huge earthern mound that was built over them and the altar. * This lecture has not hitherto been published; perhaps the author regarded it as hardly falling within the sphere of a scientific journal. Even if this be the case, however, ‘the general interest of the subject is such as to justify its being printed now. Further, it seems due to the author that the scientific public should have the benefit of all his contributions to the subject to wnich he gave so much study. It may be added that the biographical notice of Professor Newton, it is expected, will appear in the May number.—EDITOoRS. Am. Jour. Sci.—FourtH Series, Vou. III, No. 13.—January, 1897. i 2 H. A. Newton—The Worship of Meteorites. What would we not give if this fragment could be endowed with the power of repeating to us its experience,—chapters in the history of that people? But nearly all that we can say is that it was found among objects held by them in peculiar esteem, and used by them in some serious, probably religious ceremony. F There was formerly, and so far as I know there is still, in the collection of meteorites in Munich, a stone that weighs about a pound. It fell in 1853 in the region north of Zanzibar on the East African coast, and was seen and picked up by some shepherd boys. The German missionaries tried to buy it, but the neighboring Wanikas, because it fell from heaven, took it to be a god. They secured possession of it, anointed it with oil, clothed it with apparel, ornamented it with pearls, and built for it a kind of temple to give it proper divine honors. The agents of the missionaries were not allowed even to see the stone, far less could they purchase the Wanika’s tutelar deity. Neither entreaties, nor arguments, nor offers of the missionaries, nor of the officials were of any avail. But when three years later the wild nomad tribes of the Masai came down upon the Wanikas, burned their village, and killed large numbers of them, the Wanikas thought very differently of the stone’s pro- tecting power. In fact they lost all respect for it. A famine having meanwhile arisen, the elders of the tribe were quite ready to exchange their palladium for the silver dollars of the missionaries. Among the Buddha legends is one of two merchants who offered food to the Buddha, which was accepted, and in conse- quence of their request for some memorial of him the Buddha gave them a hair and fragments of his nails, and told them that hereafter a stone should fall from heaven near the place where they lived, and that they should erect a pagoda and worship these relics as though they were Buddha himself. The nations of India have always been specially superstitious about stones fallen from the skies. In 1620 an aerolite fell near Jullunder, and the king sent for a man well known for the excellent sword blades that he made, and ordered him to work the lump into a sword, a dagger and a knife. The mass, however, would not stand the hammer, but crumbled in pieces. By mixture with iron of the earth the required weapons were — made. in 1867 a shower of stones fell, some forty in number, at Saonlod. The terrified inhabitants of the village, seeing in them the instruments of vengeance of an offended deity, set about gathering all they could find, and having pounded them into pieces they scattered them to the winds. In 1870. a meteorite fell at Nidigullam, and the Hindoos at H. A. Newton—The Worship of Meteorites. 3 once carried it to their temple and worshiped it. The same has been repeated in India on the occasion of several other stonefalls in the present century. One native ruler refused to allow a stone to be carried across his territory for fear of the injury that might come to his people or his lands. Two Japanese meteorites, formerly the property of a daimio family, were long kept and handed down as heirlooms, being in the care of the priests in one of the family temples. They were among the family offerings made to Skokujo on her festi- val days. They were connected with her worship by the belief that they had fallen from the shores of the Silver River, Heavenly River, or Milky Way, after they had been used by her as weights with which to steady her loom. One of these stones was presented by its late owner to the british Museum, and it is in its collection of meteorites. There is a curious institution among the Chinese that has existed, according to Biot, from a time more than one thousand years before Christ. The Chinese attributed to different groups of stars a direct influence upon different parts of the empire. Some of these groups correspond, for example, to the imperial palaces, to the rivers, the roads, and the mountains of China. By reason of this belief regular observa- tions are made by the imperial astronomers of all that passes in the heavens, especially of the groups of stars in which comets and meteors originate, or across which they travel. The interpretation of what is seen in the sky forms part of the duties of these very important officials. These observations have been carefully written out, and are preserved in the archives of the empire. Upon the ending of a dynasty, by change of name or otherwise, these comet and meteor records have been published as a special chapter of the chronicles of the dynasty. The existing dynasty began in 1647, since which date the records are, therefore, unpublished. 3 In 13492 a stone of 300 pounds weight fell at Ensisheim, in Alsace. The Emperor Maximilian, then at Basel, had the stone bronght to the neighboring castle, and a Couneil of State was held to consider what message from heaven the stonefall brought to them. As a result the stone was hung up in tie ehureh with an appropriate legend and with strictest command that it should ever remain there intact. It was held to bean omen of import in the contest then in progress with France and in the contest impending with the Turks. Nineteen years later a shower of stones fell near Crema, east of Milan. The pope was at war with the French and the stones fell into the French territory. Before the year had passed the French, after a long possession of Lombardy and serious threatening of the States of the Church, were forced to retire from Italy. At — © a | AOA Fees. > “Ss a’ } — a. | ain ea |. Pa oe? 2D ee ef =. 4 H. A. Newton—The Worship of Meteorites. this time Raphael was painting for an altar-piece his magnifi- cent Madonna di Foligno now in the Vatican. Beneath the rainbow in the picture, indicating Divine reconciliation, Raphael painted also this Crema fireball apparently to set forth Divine aid and deliverance. I have thus rapidly gone over some selected facts showing how the mound-builders, the wild Africans, the Hindoos, the Japanese, the Chinese, the modern Europeans have been ready . to revere these myfterious bodies that come from the skies. But it is in the Greek and Latin literature that we have reason to expect the more numerous and full accounts, both legendary and historic, of this reverence and worship. It is now I believe admitted by the best scholars that both in Greece and in Italy, there was a period earlier than the age of images, when the objects worshiped were not wrought by hand. Men worshiped trees and caves, groves and mountains and also unwrought stones. Even after men began to make their objects of worship, these were in many cases mere hewn stones, not images. The earlier Greek term ayadpa, an object of worship, stands apart from the later term ecw, image. What would be more natural in that age to the affrighted witnesses of the most magnificent of spectacles, the fall of a meteorite, than for them to regard the object which had come out of a clear sky, with terrific noise and fire and smoke, as something sent to them by the gods to be revered and wor- shiped ? It was nobler to worship a stone fallen from the sky than one of earthly origin. The worship of an unwrought stone once established has wonderful vitality. For example, the Greek writers speak of such a worship in their day among the Arabian tribes. When Mohammed with his intense iconoclasm came down upon Mecea and took the sacred city, he either for reasons of policy, or from feeling, spared the ancient worship of this black stone. Entering into the sacred inclosure he approached and saluted it with his staff (where it was built into the corner of the Kaaba), made the sevenfold circuit of the temple court, returned and kissed the stone, and then entered the building and destroyed the 360 idols within it. To-day that stone is the most sacred jewel of Islam. Toward it each devout Moslem is bidden to look five times a day as he prays. It is called the Right Hand of God on Earth. It is reputed to have been a stone of Para- dise, to have dropped from heaven together with Adam. Or, again, it was given by Gabriel to Abraham to attest his divinity. Or, again, when Abraham was reconstructing the Kaaba that had been destroyed by the deluge, he sent his son Ishmael for a stone to put in its corner, and Gabriel met Ishmael and gave him this stone. It was originally transparent hyacinth, but H. A. Newton—The Worship of Meteorites. 5 became black by reason of being kissed by asinner. In the day of judgment it will witness in favor of all those who have touched it with sincere hearts, and will be endowed with sight and speech. The color of this stone, according to Burckhardt, is deep reddish brown, approaching to black; it is like basalt, and is supposed by some to be a meteorite. It is not important for my purpose to separate the history from the myth. Eusebius quotes from an old Pheenician writer, Sanchouniathon, that the Goddess Astarte found a stone that fell from the air, that she took it to Tyre, and that they worshiped it there in the sacred shrine. We have reason to question whether that Phoenician writer ever lived. What matters it? The existence of the story in Eusebius’ time has to us a significance not greatly unlike that of the existence of the worship itself in the earlier years. Vergil describes a detonating meteor in such terms that I feel reasonably sure that either he had seen and heard, or else he had had direct conversations with others who had seen and heard, a splendid example of these meteors. The passage is in the second book of the Aeneid. The city of Troy was cap- tured and was burning. All wasin confusion. The family of Aeneas was gathered ready for flight, but Auchises would not go. An omen, lambent flames on the head of his grandson, began only to shake his purpose to perish with his country. He prayed for more positive guidance. It is Aeneas who describes the scene: ‘“ Hardly had the old man spoken when across the darkness a star ran down from the sky carrying a brilhant light torch. We saw it go sweeping along above the roof of the house. It lighted up the streets and disappeared in the woods on Mount Ida. A long train, a line of light, remained across the sky, and all around the place was a sulphurous smell. A heavy sound of thunder came from the left. Overcome now, my father raised his hands to heaven, addressed the gods and wor- shiped the sacred star. Now, now, he cried, no longer delay.” This story is, of course, all legendary, but Vergil’s descrip- tion of the scene is true to life as conceived by pagan Rome in his day. The images that fell down from Jupiter, or that fell from the skies, are often spoken of by Greek and by Latin writers. I mention three or four cases only where this allusion points to . a meteoric origin as possible or probable. The earliest repre- sentative of Venus at old Paphos on the island of Cyprus, was one of these heaven-descended images. It was not the Venus of the Capitol, nor the Venus of Milo, but as described was a rude triangular stone. Cicero, in the grand closing passage of his oration against a a a ; | 6 HH, A. Newton—The Worship of Meteorites. Verres, calls upon Ceres, whose statue he says was not made by hands but was believed to have fallen from the skies. The earliest of the images of Pallas at Athens was said to have had a like origin. Pausanias saw at Delphi a stone of moderate size which they anointed every day, and covered during every festival with new shorn wool. They are of opinion, he adds respecting this stone, that it was the one given by Cybele to Saturn to swallow as a substitute for the infant Jupiter, which Saturn after swallowing vomited out on the earth. There is a marvelous story of a peculiar stone in the poem Lithika by the apocryphal Orpheus. Phoebus Apollo gave the stone to the Trojan Helenus, and Helenus used it in soothsay- ing. It was called Orites, and by some Siderites. It had the faculty of speech, and when Helenus wished to consult it he performed special ablutions and fasts for twenty-one days, then made various sacrifices, bathed the stone in a living fountain, dressed it and carried it in his bosom. The stone now became alive, and to make it speak he would take it in his. arms and dandle it, when the stone would begin to ery like a child for the breast. Helenus would now question the stone, and receive its answers. By means of these he was able to foretell the ruin of the Trojan state. Whoever framed that story had, I believe, before him a real stone, and the description is very like that of a meteorite, saying nothing of its having come from Apollo. The Orphic writer says that it was rough, rounded, heavy, black, and close-grained. Fibers. like wrinkles were drawn in circular forms over the whole sur- face above and below. Here I show you a stone such as was described—rounded,. black, heavy, close-grained, and having fibers like wrinkles in circular forms over the whole surface above and below. The name Sitderites was at a later date applied to the load-. stone, but by this writer the two stones are separately described, and are apparently distinct, If this name was of Greek origin it seems to be allied to s¢deros. vron, and this heavy stone, like nearly all meteorites, probably contained iron. If, however, this name came from a Latin source (for it is used both by Greek and by Latin writers) it has affinities with Szdus,. a star, and its meteoric character is still more clearly indicated. One of the most interesting of the stories about images that _have fallen from heaven, is the basis of that beautiful tragedy of Euripides, * Iphigeneia i in Tauris.” To many of you the story is familiar, but it will bear repetition. The goddess Diana detained at Aulis the Grecian fleets by contrary winds, and required the sacrifice of. Iphigeneia, the daughter of Agamemnon, before the Greeks could set sail. The father consented ;—and the daughter, apparently sacri- H. A. Newton—The Worship of Meteorites. i ficed, was really rescued by Diana, and borne to the Tauric, or Crimean peninsula on the north shore of the Black Sea. She was then made a priestess in the temple of the goddess. At this shrine the barbaric inhabitants used to sacrifice before an image of Diana, that fell from heaven, all strangers that were shipwrecked upon the coast. The unhappy Iphigeneia, forced to take a leading part in these human sacrifices, laments her sad lot :— ‘‘ But now a stranger on this strand, ’Gainst which the wild waves beat, I hold my dreary, joyless seat, Far distant from my native land; Nor nuptial bed is mine, nor child, nor friend. At Argos now no more I raise The festal song in Juno’s praise ; Nor o’er the loom sweet sounding bend, As the creative shuttle flies, Give forms of Titans fierce to rise, And dreadful with her purple spear Image Athenian Pallas there. But on this barbarous shore Th’ unhappy stranger’s fate I moan, The ruthless altar stained with gore, His deep and dying groan; And for each tear that weeps his woes, From me a tear of pity flows.” Orestes, the brother of Iphigeneia, had avenged upon his mother the murder of his father. For this he was driven by the Furies. While stretched before the shrine of Phoebus he heard the divine voice from the golden tripod, commanding him to speed his way to the wild coast of the Taurians, thence to take by fraud or by fortune the statue of Diana that fell from heaven, and earry it to Attica. Doing this he should have rest from the Furies. He was captured, however, along with his friend Pylades, and brought to the altar to be sacrificed. The relationship of the brother and sister became here revealed, and they together fled, carrying with them the image. It was not without a struggle that they reached the shore, but finally, “On his left arm sustained Orestes bore his sister through the tide, Mounted the bark’s tall side and on the deck Safe placed her and Diana’s holy image Which fell from heaven.” Neptune favored the Greeks, Minerva forbade pursuit, and the image was borne to Halae (or as some said to Brauron) in Attica. Sans aa Cicero spoke of the Trojan Palladium as something that fell! from the sky;—gund de coelo delapsum. Other classical writers, notably Ovid, speak of it in similar terms. The story oe gn 8 HT. A. Newton—The Worship of Meteorites. in its various forms points toward a stone-fall as its basis. One form of it runs thus :— Pallas and her foster sister Athena were wrestling with each — other, when Pallas was accidentally killed. In grief Athena made an image of Pallas and set it up on Olympus. When King Ilus was about building his city on the Trojan plain he prayed for afavorable omen. In response to his prayer Jupiter cast this image down at the feet of the suppliant king. In the new city it was set up in a temple specially erected to con- tain and protect it. So long as Troy could keep safely this image, the city, it was firmly believed, could not be taken by its foes. : According to one story the Greeks stole the image before capturing the city. As many cities afterwards claimed to pos- sess the treasure as claimed to be the birthplace of Homer. According to the Romans, Aeneas carried the Palladium to Italy, and the image was regarded as the most sacred treasure of the Roman State. For centuries even in historic times it was so carefully kept by the Vestal Virgins that the Pontifex Maximus was not allowed to see it. We naturally have doubts about the nature, or even the existence, of an object so kept ont of sight. What it was that the Vestals thus guarded, or whether they had anything to represent the image of Pallas, will probably never be known. But it is far otherwise with another famous object of Roman worship. To the east of the Trojan plain on which the Pal- ladium fell rise the mountains of Phrygia and Galatia. In Pessinus, near the border line of these two countries, and in the caves and woods near Pessinus, the goddess Cybele, the mother of the great gods, Jupiter, Neptune, and Pluto, was specially worshiped. This worship may not have been more degrading than the worship of many other Asiatic divinities. But it was wretched and unmanly almost beyond our possible conception. It furnished to Catullus the theme for the most celebrated of his poems, one of the strongest pictures in all literature. The Grecian, athlete entered her service with joy- ful music and dancing. Too late he looks back from the Asiatic shore, out of his hopeless degradation, on the noble- ness of his former Grecian life. The lion of Cybele drives him in craven fear again into the wild woods, to spend his days in the menial servitude. The Roman poet exclaims, “ O god- dess, great goddess Cybele, goddess queen of Dindymus; far from my house be all thy frenzies; others, others, drive thou frantic.” At some unknown early time a meteoric stone fell near to Pessinus. It was taken to the shrine of Cybele and there set up and worshiped as her image. This image and its worship H. A. Newton—The Worship of Meteorites. 9 very early attained a wide celebrity. About two hundred years before Christ, in the time of the second Punic war, the stone was transported to Rome. The detailed history of the transfer is given by several writers in varied terms. It forms one of Livy’s charming stories, it is told in poetic terms by Ovid, it is given as a tradition by Herodian. Tor every detail of the history I do not ask confiding belief, but the principal event is, I suppose, historically true. In the year 205 before Christ Hannibal had, since crossing the Alps, been holding his place in Italy for more than a dozen years, threatening the existence of the Roman State. The fortunes of war were now somewhat adverse to the Cartha- genian general. A shower of stones alarmed the Romans. The decemvirs consulted the Sybilline books, and there found certain verses which imported that whensoever a foreign enemy shall have carried war into the land of Italy he may be expelled and conquered if the Idan mother be brought from Pessinus to Rome. These words were reported to the Senate. Encouraging responses came at the same time from the Pythian _ oracle at Delphi. The Senate set about considering how the goddess might be transported to Rome. There was then no alliance with the. states of Asia. But King Attalus was on friendly terms with the Romans because they had a common enemy in Philip II. of Macedon. The Senate, therefore, selected an imposing embassy from the noblest Romans. A convoy of five quin- queremes was ordered for them, that they might make an appearance suited to the gr andeur of the Roman people. The embassy landed on their way and made inquiry of the oracle at Delphi, and were informed “that they would attain what they were in search of by means of King Attalus, and that when they should have carried the goddess to Rome they were to take care that whoever was the best man in the city should perform the rite of hospitality to her.” The king received them kindly, but refused their request, whereupon an earth- quake tremor shook the place and the goddess herself spoke from her shrine, “It is my will, Rome is a worthy place for any god; delay not.” The king yielded; a thousand axes hewed down the sacred pines, and a thousand hands built the vessel. The completed and painted ship received the stone and bore it to the mouth of the Tiber. It was the spring of the following year before the ship arrived. Meanwhile new prodigies frightened the people. A brilliant meteor had crossed Italy from east to west, a little south of Rome, and a heavy detonation followed. From this, or from some other meteor, another shower of stones had fallen. In expiation, according to the custom of the, country 10 H. A. Newton—The orship of Meteorites. in case of stonefalls, religious exercises during nine days were ordered. The Senate after careful deliberation selected one of the Scipios, deciding that he of all the good men in the city was the best, and they deputed him to receive the stone. The whole city went out to meet the goddess. Matrons and daugh- ters, senators and knights, the vestals and the common people all joined the throng. But a drought had reduced the water of the Tiber so that the vessel rounded upon the bar. All the efforts of the men pulling upon the ropes failed to move it. A noble matron who had been slandered stepped forward into the water. Dipping her hands three times into the waves and raising them three times to heaven, she besought the god- dess to vindicate her good name if she had been unjustly slan- dered. She laid hold of the rope and the vessel followed her slightest movement, amid the plaudits of the multitude. Scipio, as he had been ordered by the Senate, waded out into the water, received the stone from the priests, carried it to the land, and delivered it to the principal matrons of the city, a band of whom were in waiting to receive it. They, relieving each other in succession and handing it from one set to another, carried it to the gates of the city, and thence through the streets to the temple of Victory on the Palatine Hill. Censers were placed at the doors of the houses wherever the procession passed, and incense was burned in them, all praying that the goddess would enter the city with good-will and a favorable disposition. The people in crowds carried presents to the temple. A religions feast and an eight days’ festival with games were established to be celebrated thereafter each year in the early part of April. Before another year had passed Hannibal, after having maintained his army in Italy for fifteen years, was forced to withdraw again to Africa. From the liberal offerings of the people, in gratitude for deliverance, a temple was erected to Cybele, long known as the Temple of the Great Mother of the Gods, so that twelve years after its arrival at Rome the stone was taken from the Temple of Victory and set up in its new home. A silver statue of the goddess was ‘constructed, to which the stone was made to serve in place of a head. Here, in public view, for at least five hundred years that stone was a prominent object of Roman worship. Its physical appearance is described by several writers. It was conical in shape, end- ing ina point, this shape giving occasion to the name Weedle of Oybele. It was brown in color, and looked like a piece of lava. Arnobius, a Christian writer just before the accession of Constantine, and over five hundred years from the date of its arrival at Rome, says of the stone : “Tf historians speak the truth and insert no false accounts H. A. Newton—The Worship of Meteorites. EY into their records, there was brought from Phrygia, sent by King Attalus, nothing else in fact.than a kind of stone, not a large one, one that could be earried in a man’s hand without strain, in color tawny and black, having prominent, irregular, angular points, a stone which we all see to: -day, having a rough irregular place as the sign of a mouth, and having no promi- nence corresponding to the face of an image.” Arnobius goes on to ask whether it was possible that this stone drove the strong enemy Hannibal out of Italy—made him who shook the Roman State, unlike himself, a craven and a coward. Just when this stone disappeared from public view I do not know. In directing the recent excavations on the Palatine Hill, Prof. Lanciani was at first in great hopes of finding it ;— becanse it had no intrinsic value to the many spoliators of Rome, nor to the former excavators of Roman temples. But the place in which he expected to find it was absolutely empty. At a later date, however, he found in a rare volume an account of excavations made on the Palatine Hill in 1730, in which the private chapel of the Empress was found and explored. In this we perhaps have an account, and itis to be feared, the last account of a sight of the Cybele stone. The writer says: “T am sorry that no fragment of a statue, or bas-relief, or inscription has been found in the chapel, because this absence of any positive indication prevents us from ascertaining the name of the divinity to whom the place was principally dedi- cated. The only object which I discovered in it was a stone nearly three feet high, conical in shape, of a deep brown color, looking very much like a piece of lava, and ending in a sharp point. No attention was paid to it, and I know not what became of it.” This description is almost identical with that given by Arnobius and others of the stone from Pessinus. Another stone of meteoric origin was brought to Rome, and there for a brief period was most fantastically wor- shiped. This was near the beginning of the third cen- tury after Christ. It came like the other stones of which I have spoken from Asia. In the City of Emesa, on the banks of the Orontes about midway between Damascus and Antioch, there was in those days a magnificent temple of the Sun. A gorgeous worship was maintained before a stone that fell from heaven, that served as the image of the Sun-god. The description of it is not very unlike that of the Cybele meteorite. Herodian, who probably saw it, says: “It is a large stone, rounded on the base, and ovadually tapering upward to a sharp point; it is shaped like a cone. Its color is black, and there is a sacred tradition that it fell from heaven. They show certain small prominences and depressions in the stone, and those who see them persuade their eyes that they are seeing an image of the sun not made by hands.” 12 H. A. Newton—The Worship of Meteorites. This Sun-god was named Heliogabalus, and before the altar a boy of nine years of age began to serve as priest. Such a Syrian service did not make the boy grow manly nor virtuous, and when at the age of fifteen he became emperor through the money and intrigues of his grandmother, and the murder of the Emperor Macrinns, we have for three years at Rome the view of the sorriest scrapegrace that ever sat on-a throne. He assumed with the name of Antoninus also the name of his god Heliogabalus. To the great disgust of the Roman Senate and people, he brought with him from Syria the image of his god, the sacred stone, and himself continued before it his priestly service with all its fantastic forms and gesticulations. He built within the city walls a grand and beautiful temple, with a great number of altars around it; he repaired thither every morning, and sacrificed hecatombs of bulls and an infinite number of sheep, loading the altars with aromatics, and pouring out firkins of the oldest and richest wines. He himself led the choruses, and women of his own country danced with him in circles around the altars, while the whole Senatorian and Equestrian orders stood in a ring like the audi- ence of a theater. But now he must havea wife for his god. So he broke into the apartments guarded by the vestals and carried to the palace the Trojan Palladium, or what he supposed was that object, and was intending to celebrate the nuptials of the two images. His god, however, he concluded would not be pleased with a warlike wife like Pallas, therefore, he ordered brought from Carthage an ancient image of Urania, or the Moon, which had been set up by Dido when she first built old Car- thage. With this image he demanded the immense treasures in her temple, and he also collected from every direction im- mense suins of money to furnish to the Moon a suitable mar- riage portion when married to the Sun. He built another temple in the suburbs of Rome, to which the Emesa stone, the god (?) was carried in procession every year, while the populace were entertained with games, and shows, : and feastings and carousings. Herodian thus describes this per- formance : “The god was brought from the city to this place in a chariot glittering with gold and precious stones, and drawn by six large white horses without the least spot, superbly har- nessed with gold and other curious trappings reflecting a vari- ety of colors. Antoninus himself held the reins—nor was any mortal permitted to be in the chariot; but all kept attendant around him as charioteer to the deity, while he ran backward leading the horses, with his face to the chariot, that he might have a constant view of his god. In this manner he performed H. A. Newton—The Worship of Meteorites. 13 the whole procession, running backwards with the reins in his hands, and always keeping his eyes on the god, and that he might not stumble or slip (as he could not see where he went), the whole way was strewn with golden sand, and his guards ran with him and supported him on either side. The people attended the solemnity, running on each side of the way with tapers and flambeaux, and throwing down garlands and flowers as they passed. All the effigies of the other gods, the most costly ornaments and gifts of the temples, and the brilliant arms and ensigns of the imperial dignity, with all the rich fur- niture of the palace, helped to grace the procession. The horse and all the rest of the army marched in pomp before and after the chariot.” The reign of a foolish boy at this period of Rome’s history was necessarily a short one, and at the age of eighteen the soldiers knlled him and let the Roman populace have the body to drag through the city streets. The worship of the Sun-god at once ceased, and, no doubt, the stone also was thrown away. The Cybele stone, however, remained an object of public worship, since the quotation from Arnobius, which I have given, was written nearly a century later than the reign of Heligab- alus. , I propose to speak briefly of one more meteorite whose wor- ship has had a world-wide fame ;—the image of the Ephesian Artemis. This worship had its center at Ephesus, but was widely extended along the shores of the Mediterranean. Tem- ple after temple was built on the same site at Ephesus, each superior to the preceding, until the structure was reckoned one of the seven wonders of the world. As a temple, it became the theatre of a most elaborate religious ceremonial. As an asylum, it protected from pursuit and arrest all kinds of fugi- tives from justice or vengeance. As a museum,it possessed some of the finest products of Greek art, notably works of Phidias and Apelles. As a bank, it received and guarded the treasures which merchants and princes from all lands brought for safe keeping. In its own right it possessed exten- sive lands and large revenues. The great City of Ephesus assumed as her leading title that of vewxdpos, or temple-warden of Artemis, putting this name on her coins, and in her monu- mental inscriptions. The image, which was the central object in this temple, was said to have fallen from heaven. Copies of it in all sizes and forms were made of gold, of silver, of bronze, of stone and of wood, by Ephesian artificers, and were supplied by them to markets in all lands. What a lifelike picture is given us in the 19th chapter of the Acts of the Apostles, of the excited crowd of Ephesians, urged on by the silversmiths, who made for sale — S61. Se 14 H. A. Newton—The Worship of Meteorites. the silver shrines of the goddess, and who saw that their craft was in danger if men learned to regard Artemis as no real divinity, and to despise the image that fell down from the sky. We cannot suppose that the Ephesian Artemis image of the first century was a meteorite, though we have the distinct appel- lation, Diipetes, fallen from the sky. But I believe that there was a meteoric stone that was the original of the Ephesian images, and it seems not at all improbable that in some one of the destructions of the temple it disappeared. Or, in the progress of time, there may have been a desire to represent the goddess in a more artistic form than the shapeless stone afforded. Many forms of the Ephesian Artemis are still preserved, and they have, amid all their variations, a certain peculiar character in common. That common character seems to me to confirm the statement that the original image fell from heaven. This goddess is regarded, let me say, as different from the Grecian Artemis, the beautiful huntressso well known in Greek art, and I am speaking only of the images of the Ephesian Artemis. ’ . There is one peculiarity in the outward forms of the meteor- ites that is characteristic of nearly all of them. I mean the molded forms, and the depressions all over the surfaces. They are better appreciated by being seen, than by any description I can give you. They are common to meteorites of all kinds, from the most friable stone to the most compact iron. (I show you one, a stone from Iowa—also the plaster cast of another, a stone from some fall, I know not which one.) Those who have recently visited the collection in the Peabody Museum may recollect the model of an iron that fell two or three years ago in Arkansas, which displays most beautifully these depres- sions. Let now an artist attempt to idealise any one of these molded forms, and to make something like a human shape out of one of them. Hemust necessarily set it upright, and he must give it a head. You have then a head surmounting one of these molded forms. Let now the convenience and the taste of the artificers of the images have some liberty to act—and we know that they did act, for we have considerable variety in these images—and a development in the conventional representation of the image is sure to follow. [The lecture closed with the exhibition of a series of lantern pictures showing the forms of some typical meteorites. ] Trowbridge and Richards—The Spectra of Argon. 15 Art. II.—TZhe Spectra of Argon; by JoHN TROWBRIDGE ‘ and THEODORE WILLIAM RICHARDS. Ir is well known that argon possesses at least two marked spectra—one, termed the red, which is chiefly characterized by red lines ; and another called the blue, which, as its name sig- nifies, is strongly marked by blue lines. In studying these spectra by means of a high tension accu- mulator, we have been led to observe carefully the electrical conditions necessary for producing them. It is obvious that a battery of a large number of cells is the most suitable source for the study of discharges of electricity through gases. Especially is this true of a storage battery ; for the readiness with which it can be charged by a dynamo, the constancy of the electro- motive force (about 2-1 volts per cell), the ease with which it ean be coupled for quantity or tension; and the steadiness of the discharge afforded by it, make such a battery far superior to an induction coil or to an electrical machine. With an induction coil the discharge is not undirectional and is affected by the necessary irregularities of the break. These irregularities make themselves felt in a marked degree when a condenser is used in the secondary circuit. The elec- trical machine gives an intermittent current, and has a varying capacity. The advantages of a battery for the study of the discharge of electricity through gases have been pointed out by De la Rue and Miller and by Hittorf.* These investigators worked with voltaic cells which were not constant and which required great oversight and continual renewals. In our investigations we are using a lead accumulator of the Planté type; and we find it highly advantageous for spectroscopic work; for by means of the steady current afforded by it, one can study the spectra of gases under especially good conditions. Our battery consists of five thousand cells, so arranged that they may be disconnected and wholly reconnected in any desired manner in less than a minute. The electromotive force of the complete series is somewhat over ten thousand volts, but when the cells are connected for quantity, they may be readily charged by means of a dynamo giving a tension of only sixty volts. The insulation of the terminals of this bat- tery was a matter of some difficulty, for even dry wood allows considerable leakage from one case of cells to another ; but by the plentiful use of paraffin, mica, and vulcanite, the problem was solved with reasonable success. The discharge from only avery small fraction of the battery produced a most uncom- * Ann. der Phys. und Chem. (N. F.), vol. vii, 1879, p. 553. _ 2) Ste ek SSE = OOS ens EE 16 Trowbridge and Richards—The Spectra of Argon. fortable shock, and it is probable that the discharge of the whole battery would be instantly fatal. The great heat of this full discharge immediately shatters a Geissler tube, the glass being splintered throughout the whole length of the capillary. Hence a resistance of several million ohms was usually inter- posed between the battery and the rest of the apparatus. This resistance was also of service in protecting the experimenters from serious accidental shocks. Ordinary distilled water con- tained in long tubes with movable electrodes was the most convenient resistance for our purpose; dilute solutions of cadmic iodide in amyl alcohol and of cadmic sulphate in water between cadmium electrodes, were also sometimes used. Unless these liquids are contained in tubes of rather large diameter, they are likely to cause irregularities by boiling under the influence of the heat of the current. Graphite resistances are too combustible for the purpose. The argon used in our experiments was very kindly given to one of us vy Lord Rayleigh. It was a portion of the purest preparation which had been used in his finai determinations of the density of the gas; and our tubes were carefully filled with it by the kindness of F. O. R. Gétze, of Leipzig. The preliminary work described in this paper was chiefly done with a single tube containing gas at a pressure of about 1™”. The tube had a wide capillary and was about 15° in total length. In such a tube, the red glow of argon is readily obtained with a voltage of about two thousand, but not with much less. A higher tension of gas demands a higher tension of electricity in order to start the discharge, no matter how much or how little other resistance is interposed; but when the glow has once started it is continued by means of a much smaller electromotive force. This is shown by the fact that a Thomson electrostatic voltmeter, connected with the terminals of the Geissier tube, indicated differences of potential between the ends of the tube ranging from six hundred and thirty volts upward. De la Rue and Miller, who found no potential dif- ference between the ends of Geissler tubes, must have been working with discontinuous discharges. Crookes’s estimate that 27,600 volts are necessary to produce the red spectrum is evidently excessive. The introduction of a capacity between the fone: of the Geissler tube, for example, two plates of metal sixteen hun- dred square centimeters in area separated by plate glass one centimeter thick, made no difference in the red glow, so long as the connections were good and the condenser quiet.* As soon as a spark gap was introduced, or the condenser began to * Sir W. Thomson (Lord Kelvin): Papers on Electrostatics and Magnetism. MacMillan, 1872, p. 236. Trowbridge and Lichards—The Spectra of Argon. igh emit the homming sound peculiar to it, the beautiful blue glow so characteristic of argon immediately appeared. If this light is examined by a revolving mirror it is seen to consist of intermittent discharges. The battery charges the condenser to the potential necessary to produce a spark, between the terminal of the spark gap. The discharge of this aceumu- lated electricity is produced in the tube and then the operation is repeated. The time interval between the discharges is evi- dently a function of the capacity of the condenser, as well as of the electromotive force of the battery and the resistance between it and the condenser. | The accurate determinations of the potential and current strength of the intermittent blue discharge 1 is a matter of some difficulty ; and at present we feel hardly in position to make a definite statement regarding these measurements. However, the potential required certainly cannot be greater than two thousand volts,—the electromotive force of the battery which will easily produce the blue glow. Here again, Crookes’s esti mate of far above 27,000 volts was. very much too large. Since it is necessary to employ a condenser to produce the blue spectrum of argon, we were led to examine the electrical conditions which are necessary for the discharge. In the eir- cuit with the tube containing argon, between the tube and one of the plates of the condenser, we first interposed a small coil of about eight ohms resistance, having a self-induction of -015 of a henry. The blue glow changed to the red glow. We then modified the self-induction and discovered that even the self-induction of the leads to the tube, which consisted of a few feet of uncoiled wire, undoubtedly modified the blue discharge, for an amount of induction equivalent to 000051 henry had a - marked effect in diminishing the brilliancy of the blue glow. A comparatively small ohmic resistance substituted for the impedance of the self-induction between the tube and one plate of the condenser produced precisely the same effect as this .coil, causing a complete transformation from blue to red. he change from blue to red is so marked that a tube of argon may well serve as an inductometer of some sensitiveness, as well as a means of comparing the influence of self-induction with ohmie resistance. The effect of impedance or resistance must be to prolong or to damp the oscillations of the condenser discharge. Indeed, the resistance of the tube itself may be so great as to damp the oscillations without the need of the intro- duction of outside resistance or self-induction ; therefore argon at high tension gives the red glow with a condenser and rate of oscillations which are quite capable of producing the blue glow in a tube of lower tension. Am. Jour. Sc1.—Fourta Szries, Vou. III, No. 13.—JANvUARY, 1897. 2 : 5 18 Trowbridge and Richards—The Spectra of Argon. Kayser* criticises Crookes’s statement that a condenser and a spark gap are necessary for the production of the blue spec- trum. He finds that with a lower pressure in the tube than 2™" the blue spectrum can be readily obtained without con- densers and spark gaps. He also states that it is much easier to produce the pure blue spectrum than the pure red. In order to obtain the red spectrum the strength of the current must be adapted to the gas pressure. Kayser employed an induction coil. The condenser, however, in the primary of an induction coil sends an oscillatory discharge through the sec- ondary. Although Kayser did not employ a condenser in the manner recommended by Crookes, he still had a condenser in his electrical system and the resistance of his Geissler tube was probably so proportioned that the secondary circuit was in resonance with the primary circuit. To prove this we placed a tube containing argon across the terminals of the secondary of an induction coil, and having removed the condenser attached to the primary, we sent the discharge through the tube by means of a break in the induction coil. The light of the discharge was red, and when it was examined by a revolving mirror no trace of blue was seen in the capillary portion of the tube. An ‘adjustable condenser and a variable induction were then placed in the primary circuit. By varying the amount of the capacity together with the self-induction in the primary sys- tem, the discharge in the secondary, when examined by a. revolving mirror, was seen to consist of both red and blue dis- charges. The red glow was evidently due to a unidirectional discharge and the blue to an oscillatory discharge. The unidi- rectional discharge was caused by the failure of the breaks to charge the condenser to the primary, or by increased resistance in the tube. When, however, the condenser was charged, it immediately discharged in an oscillatory manner; and the sec- ondary coil responded by resonance. ‘The rarified argon thus shows in an interesting manner what is the function of the condenser in the primary of an induction coil. It serves to send oscillatory discharges through the primary circuit ; and the greatest effect. is obtained in the secondary cireuit when it is In resonance or in tune, so to speak, with the primary. The presence of a condenser was necessary to form the blue glow in Kayser’s work, only when the resistance of his tube and the self-induction of his coil together were enough to damp the discharge of the small capacity of the coil. He could have obtained the pure red in any case by interposing, as we have done, a resistance or self-induction between the con- denser and the tube, although our other resource for obtaining * Sitzungsberichte der kOniglich preussischen Akad, der Wissenschaften zu Berlin, May 7, 1896. Trowbridge and LRichards—The Spectra of Argon. 19 the red glow in any tube from the continuous discharge of a constant battery was apparently not open to him. By taking out all resistances except the spectrum tube, and sending an exceedingly strong current through the tube for very brief intervals of time, we have been able to cause the blue glow; but it seems probable that under these conditions the capacity of the battery itself engenders oscillations which are no longer damped by interposed resistance. Whether the blue glow with its accompanying change of spectrum is due merely to the great quantity of electricity discharged in a very short space of time, or to some property intrinsic in the to-and- fro motion of the oscillatory discharge, is a question, which we hope soon to answer. The red glow, if caused by oscillations at all, must be caused by oscillations within the Geissler tube itself ;* for all outside oscillations are cut off by the large resistance between the battery and the tube. The effect of the oscillatory discharge in producing the blue spectrum of argon can also be shown by the use of an electri- eal machine. If the terminals of the tube containing argon are connected with the terminals of an electrical machine, the pure red spectrum is obtained. [If a spark gap is interposed in such a manner that a condenser charged by the machine can discharge through the tube, the blue discharge immediately results. The condenser discharge oscillates through the gas. The oscillatory discharge of the condenser is evidently an important factor in producing the blue spectrum of argon. According to Lord Kelvin’s law, if R denotes the resistance of the circuit, L, the self-induction, and C the capacity of the cir- cuit, the discharge of the condenser becomes non-oscillatory when R a / a It may be, therefore, that an estimate of the resistance in the tube can be obtained by measuring the self- induction which is required to change from the blue discharge to the red. When the tube containing argon at a suitable pressure is brought near a Hertz oscillator, giving a rate of about 115,000,000 oscillations per second, it immediately shows the blue color. In this case the oscillator consisted of two zine plates about 40° square with a spark gap between them. The capacity and impedance of the circuit was extremely small. The unusual sensitiveness of an argon tube to oscillatory discharges leads us to believe that it will be of great use in the study of wave motions of electricity. As we have seen, it is competent to show when the Hertz oscillator is working prop- erly, that is, sending forth electrical oscillations and not unidi- * Ann. der Physik und Chemie, 1893, vol. xlviii, 549, Ebert and E. Wiedemann. 9) SRST Se- Sey Si _ s >. ni. Seat 1 Ff Wm 1 A} ee ee: 20 Trowbridge and Lichards—The Spectra of Argon. rectional discharges. The change of color in the tube from red to blue is so marked that an argon tube reveals what is not shown in a conspicuous manner by other gases. We have thought that this remarkable property of an argon tube is worthy of being distinguished by a name which might describe it and we have, therefore, called an argon tube fitted for the study of electrical waves a talantoscope (taXavTwots). In an oscillatory discharge the molecules receive powerful electrical impulses of opposite sign. ‘These impulses are sepa- rated, it may be, by millionths of a second. It is significant that the shorter wave lengths of lght accompany these electrical oscillations. It is our purpose to extend our study of the effect of electrical oscillations through more highly rarified media in which arise the Rontgen rays. These rays are probably highly modified by the oscillatory discharge. A battery of a large number of cells now at our command will afford the best means of studying this subject: for its dis- charges, as we have pointed out, are free from the fluctuating effects produced by induction coils, transformers and electrical machines. Our present paper is, therefore, only preliminary ~ to a more exhaustive study of the discharges of electricity through rarified gases, by means of a storage battery of ten thousand cells, which will give an electromotive force of about twenty thousand volts. : Harvard University, Dec. 1st, 1896. Becker—Some Queries on Rock Differentiation. 21 ArT. IIT. — Some Oucries on Rock Differentiation ; by Gro. F. BECKER. Hypothesis of differentiation.—As I understand the theory of what is called the differentiation of rock magmas, now so generally held by lithologists, its outlines may be expressed in the following terms: In some extensive districts the massive rocks are found to possess similarities of composition, and such rocks have been called consanguineous. This consanguinity might be accounted for by supposing an originally homogene- ous magma to have undergone partial segregation into fluid portions of distinctly different yet allied composition, prior to eruption, and this is the process called differentiation. It is also known by observation in the laboratory, that if a more or less complex, homogeneous solution be exposed to certain physical conditions, segregation into distinct portions may take place. It is hence inferred that such is actually the history of consanguineous rocks.* By a slight extension of this inference most massive rocks are regarded as resulting by differentiation from a generalized magma. The existence of distinct though allied rocks locally asso- ciated with transitional varieties is undisputed. The validity of the explanation offered by the modern school for these occurrences is another matter. The hypothesis of differentia- tion is extremely attractive and if it were substantiated would lead to a well organized system of rock investigation. It may also be correct ; but there are respects in which it appears to be in need of much explanation, and it does not seem certain that the fundamental postulate of originally homogeneous magmas of vast volume is well established. Possible modes of segregation.—The segregationt of a homogeneous fluid into distinguishable portions under the influence of varying temperature or pressure may take place by different methods. An increase or decrease in the concen- tration of certain components may occur in the cooler part of the finid or in that portion which is under greatest pressure. ‘There are also cases in which solutions which are homogeneous * The early literature bearing on this subject. together with fresh contributions, is given in Mr. Arnold Hague’s work on the Geology of the Kureka district, U.S. ‘Geol. Survey, Mon. 20, 1892, p. 267 et seq. Some of Prof. J. P. Iddings’ papers bearing on the subject are: the Crystallization of Igneous Rocks, 1889; Electric peak and Sepulchre mountain, 1891; the Origin of Igneous Rocks, 1892, ete. The latest of Prof. W. C. Brogger’s contributions is: die Eruptionsfolge der triad- schen Hruptivgesteine bei Predazzo, 1895. ; + The term differentiation is ambiguous. In the older, and as it seems to me the more proper sense, differentiation is the discrimination of existing differences. ‘One differentiates lime-soda feldspars by their angles of extinction. 22 Becker —Some Queries on Rock Differentiation. - at one temperature tend to separate at some other temperature into two or more immiscible fluids, or into a fluid portion and a solid one. There seem to be no other conceivable ways in which segregation or differentiation under purely physical influences can take place, and it is one of the purposes of this paper to examine the mechanism of these processes in the light of the modern conclusions of chemical physics. It will be most convenient to consider first those cases in which only miscible liquids, or liquids present only in miscible proportions, are concerned, reserving consideration of immisci- bility and insolubility for subsequent discussion. The differentiation of a homogeneous magma, or its segrega- tion into distinct though related miscible fluids, involves rela- tive movement of the particles of the magma. ‘This movement cannot take place in visible streams or currents, such as would — result from convection; for if a tendency to segregation existed, stirring would neutralize or overcome it. Segregation might, however, be accomplished by what may be called molecular flow, this term being understood to mean the pro- gressive translation of portions of a liquid, molecule by mole- cule, among the similar or dissimilar molecules of the remainder of the liquid. Instances of molecular jflow.—Molecular flow is exhibited in ordinary diffusion, in osmosis, and in some cases of the segregation of fluids. Although these are seemingly very dif- ferent manifestations, they are all reducible to the tendency whieh fluids exhibit to attain a condition of stable equilibrium, throngh an equalization of the partial pressures of each com- ponent in different parts of the fluid. If two liquids which are miscible in all proportions but are chemically indifferent to one another are placed in a small closed vessel and are maintained at a constant temperature, each diffuses into the other, and the flow of molecules will never cease until the mixture is uniform throughout, so that each liquid occupies the volume formerly occupied by both. If two such liquids are each soluble in a third, each will diffuse into the solvent at its own peculiar rate, and in such cases mere diffusion produces partial separation of the dissolved sub- stances. At any given time the substance which diffuses more rapidly will be found in greater relative abundance at any point at all distant from the original surface of diffusion. Thus Graham in experimenting on the diffusion in water of a solution which contained 5 per cent of common salt and 5 per cent of sodium sulphate found that the upper layer of water after a certain interval contained ten times as much of the chloride as of the sulphate. Becker—Some Queries on Rock Differentiation. 23 In this case any one layer of the water may be regarded asa kind of septum through which the chloride diffuses at a higher rate than does the sulphate. There are many septa more efficient in separating solutions than is water. Especially familiar is bladder, through which one class of compounds (the erystalloids) passes very readily, while another (the colloids) passes with difficulty. Here indeed the material of the septum perhaps has some molecular action on the solutions, and if so the explanation is thereby complicated. Nevertheless osmosis is regarded by physicists as a case of ordinary diffusion compli- cated, as some think, by the molecular action of the septum.* There is another class of septa which seems to be without molecular action and behave as mere “atomic sieves.” They are wholly impermeable to some solutions (bladder is not) and easily permeated by others. By means of these “semi-perme- able” septa, which are produced by precipitation, it has been found that the molecular flow of a given solution continues until a certain definite pressure exists on one side of the sep- tum, this pressure being characteristic of the substance experi- mented upon and independent of the nature of the membrane. This is the “osmotic” pressure of the dissolved substance, and Mr. van’t Hoff has shownt that it is equal to the pressure which would be exerted by the substance ina gaseous state when occupying the same volume at the same temperature. Evi- dently then there is a very close analogy between gases and substances in a state of solution, and it is in fact now well recognized that, as van’t Hoff pointed out, they obey several of the same fundamental laws when osmotic pressure and simple pressure are considered as interchangeable terms.{ Mr. W. Nernst§ has discussed the phenomena of simple dif- fusion in their relation to osmotic pressure and is led to the conclusion that osmotic pressure is the force immediately con- cerned in the diffusion of liquids, just as the pressure exerted by a gas in confinement is the cause of the diffusion of gases. The most important case of molecular flow for the purposes of this discussion arises when a homogeneous solution is heated at the top. Molecular flow then takes place from the top * Tait: Prop. of Matter, 2d ed., p. 275. It is probable that the osmotic action of animal membranes is exactly the same in principle as that of precipitated ones. They are both ‘‘atom sieves,” only the ‘‘ meshes” are of a different size, per- haps. If any ‘‘molecular action” (whatever that may be) exists in one case, it probably does in the other also. The evidence of such action is not distinct. + Zeitschr. phys. Chem., vol. i, 1887, p. 481. : ¢ The osmotic pressure is inversely proportional to the volume of the fluid in a given space. The osmotic pressure at constant volume is proportional to the absolute temperature. Solutions.of equal volume of different substances which contain equal numbers of molecules at equal temperatures exert equal osmotic pressure. These laws correspond to those of Boyle, Gay-Lussac and Avogadro. § Zeitschr. phys. Chem., vol. ii, 1888, p. 613. a ee ee = #4 = oS a OY —S- SS * 24 Becker—Some Queries on Rock Differentiation. towards the bottom, so that a concentration of the substance in solution occurs in the lower portion. This fact appears to have been observed first by Mr. C. Ludwig in 1856* and was subse- quently studied by Mr. Ch. Soret.t The phenomenon has been explained by Mr. van’t Hoff.t The osmotic pressure of a substance in solution is proportional to the absolute tempera- ture. Hence there is a resultant pressure in an unequally heated fluid which is directed towards the cooler portion, and this will not be equilibrated until the osmotic pressure in the cooler portion is appropriately increased. Now the osmotic pressure is proportional to the concentration as well as to the temperature; and the condition of equilibrium is therefore that the concentrations should be inversely as the absolute temperatures. Thus if a magma were heated to T = 1500° at the top while the bottom of the mass were kept at T = 1400°, the concentration of the substances dissolved in the magma would increase until that at the bottom were 15/14 = 1:07 of that at the top. Mr. van’t Hoff’s theory of this case agrees somewhat roughly but substantially with Mr. Soret’s experi- ments. The solution of any substance in a finid is attended by a change in the total volume. When this change is a decrease in volume, solubility increases with increasing pressure; and vice versa. Hence in a deep mass of solution of constant temperature there is believed to be a tendency to concentration through pressure at the bottom or the top of the mass. The change of concentration, however, would be so small that physicists are not hopeful of demonstrating it experimentally. The thermodynamical discussion of Messrs. Gouy and Chap- eron§ seems to show that even in vessels 100 meters in depth the effect of gravity on sodium chloride solution would influ- ence concentration only to the extent of a fraction of 1 per cent, the concentration being greater at the bottom. It is evident that the process by which concentration is effected is molecular flow. Further remarks on the variability of solu- bility with pressure will be made in diseussing the properties of immiscible fluids. There is no question that molecular flow does play a part in lithogenesis. One may often see blebs or smears of matter in either granular or porphyritic rocks which have manifestly undergone at least superficial solution, the rock around the bleb showing an aureole of diffusion. So, too, crystallization in many cases is explicable only by the molecular flow of a cer- * Wien Sitz. Ber., vol, xx, 1856, p. 539. + Comptes Rendus, Paris, vol. xci, 1880, p. 289. t Zeitschr. phys. Chem, vol. i, 1887, p. 487. § Ann. de ch: phys. (V1), vol. xii, 1887, p. 384. Becker—Some Queries on Rock Differentiation. 25 tain ingredient to one spot from the adjacent mass. There is absolutely no theoretical reason why such processes should not oceur, for at very short distances molecular flow is a very rapid process, as will be explained presently. On the other hand, it is questionable whether masses of rock of hundreds of meters in thickness could be thus separated, even if the time allowed for completion of the process were equal to an entire geological period. Character of diffusion.—lt has been explained above that all the processes of molecular flow are reducible to the same elementary action, viz: movements. due to differences of osmotic pressure. This kind of flow is most simply manifested in ordinary diffusion, and it is also in the ordinary diffusion of concentrated solutions that molecular flow takes place most rapidly. It is possible to bring an indefinitely large mass of an absolutely and permanently concentrated liquid in contact with another liquid in which the first is thoroughly soluble. Under these conditions, the resultant osmotic pressure being proportional to concentration, must have its highest value, and molecular flow (measured by the amount of substance passing through a given area in a given time) must be greater than it otherwise can be. Think, for example, of a tall vessel at the bottom of which is a layer of solid sulphate of copper, the rest of the vessel being full of pure water. Then a concen- trated solution of the sulphate will form in contact with the solid sulphate and this layer will continue concentrated until solution is complete. Diffusion will then proceed as rapidly as it can do at the temperature of the experiment. Under such conditions the amount of a dissolved substance which diffuses through an area of one square centimeter in one second, when the gradient of concentration (perpendicular to the area) is one gram of substance per cubic centimeter of fluid per centimeter of distance, is a constant called the “ diffusivity ” of the sub- stance in water.* In the case of gases Maxwell showed that a simple numerical relation exists between the diffusivity of sub- stance, the diffusivity of energy or heat, and the diffusivity of momentum which gives rise to viscosity.t In the case of fluids the a priore determination of these constants is not yet possible and they must be found by experiment. When this constant called diffusivity of substance is deter- mined, the process of diffusion can be accurately predicted under uniform conditions of temperature and pressure at least for weak solutions. In 1855 Prof. A. Fickt advanced the hypothesis that the time rate at which a salt diffuses through a *Tait: Prop. of Matter, 2d ed., 1890, p. 271. + Theory of Heat, 1894, p. 332, and Nature, vol. viii. t Pogg. Ann., vol. xciv, 1855, p. 59. f 26 Becker—Some Querves on Rock Differentiation. stated area is proportional to the difference between the con- centrations of two areas infinitely near one another. This is, mutatis mutandis, the same law which underlies Fourier’s treatment of heat conduction.* Ver y extended researches by many physicists have confirmed Fick’s law, excepting for very strong solutions, and therefore also the applicability of Fourier’s mathematical developments concerning conduction to the elucidation of diffusion. It can be shown that Fick’s hypothesis would be strictly applicable to solutions of any concentration if Pfeffer’s law, that osmotic pressure and concentration are proportional, were exact. But this law corresponds to Boyle’s law that gaseous pressure and volume are inversely propor- tional, and this, as every one knows, is exact only when the number of molecules per unit volume is not too great. At least two influences tend to render Pfeffer’s law and Fick’s hypothesis inexact for high concentration. There is a tendency to change of molecular weight with increasing concentration, a species of polymerization, and this would be attended by decreased diffusibility. When there is no such aggregation there is an increase of diffusibility, due, it is thought, to the attraction between the solvent and the dissolved substance. While it is proper to point out the deviation of very strong solutions undergoing diffusion from the law of Fick, this devia- tion is of importance only near the contact from which diffn- sion takes place, and not always then. For if a solvent dissolves a salt only to a limited extent, even a saturated solu- tion may be a weak one, and dissolved molecules will not be so crowded as to show irregular behavior. Thus in the case of magmas undergoing molecular flow, Fick’s law will be valid at least to within a short distance from contacts and may hold absolutely up to the contact. In the differentiation of a homo- geneous magma into consanguineous portions it is hardly sup- posable that the molecules undergoing transfer are densely crowded. Consanguineous rocks do not differ very greatly in composition, so that no extensive transfer of material is called for. Furthermore, magmas must be regarded as solutions of a series of very similar substances, and it is known that in such cases the solubility of each is diminished by the presence of the others. This was first pointed out by Mr. Nernst and has been confirmed on experimental and theoretical grounds by Dr. * Tf v is the quantity of substance in solution per unit volume in Fick’s case (or the temperature in Fourier’s), 1f w is the distance measured in the single dire2tion in which diffusion is supposed to take place, and if « is the diffusivity regarded as constant, then in either problem dv ye ay Gt as Becker—Some Queries on Rock Differentiation. 27 A. A. Noyes.* Thus the solubility of lead chloride is reduced by the presence of other chlorides to something like one-half of its separate solubility. It appears substantially certain therefore that a series of silicates in solution must restrict the solubility of each. Consequently conclusions drawn from the assumption that Fick’s hypothesis is exact will be applicable to the process of differentiation. It is interesting and extremely important to observe that the problem of determining the distribution of a diffusing lava is formally the same as that of finding the distribution of tem- perature in a cooling globe of large radius. This last is a sub- ject which has become familiar to geologists through Lord Kelvin’s application of the results in estimating the age of the earth. The diffusivity of a substance is inversely proportional to the molecular friction which the molecules or the ions experi- ence. Thus if 7, and 7, are the frictional resistances of the ions of an electrolyte, O the osmotic pressure and D the dit- fusivity, then, as was shown by Nernst: ees) Cee This internal friction is usually known in English as vis- cosity. The viscosity of liquids so far as is known increases with the pressure to which they are subjected, with the excep- tion of water, which at ordinary temperatures becomes less viscous with increasing pressure. Of course water at tempera- tures approaching its freezing point is an anomalous liquid. According to Amagat, the anomalous expansion of water ceases at 50° OC. and Tammann suggests that the exceptional relation between viscosity and pressure is probably confined to the same limit. Late of diffusion.—lf two miscible solutions are brought in contact and a time z, measured in seconds, is allowed to elapse, the fluid at a distance of x centimeters from the contact will contain a certain amount of the diffusing fluid per cubic centi- meter, which may be called s. One of the fluids may be sup- posed kept at constant composition, as in the case of a solid dissolving in a solvent. Then this same concentration s will be found at mw centimeters after the lapse of n’¢ seconds. For example, if common salt is brought in contact with water, the water in immediate contact with the salt will soon become saturated. At the distance of 1° the solution will be half satu- rated in about one day ; at 2°" it will be semi-saturated in four days and at 100™ in 100* = 10,000 days; at 100 meters the * Zeitsch. phys. Chemie, vol. ix, 1892, p. 623. a oe > > ACS eee Re —_ ia 2 eh 28 Becker—Some Queries on Rock Differentiation. water would contain half as much salt as it could dissolve after 10,000 x 10,000 days, some 270,000 years. It is because of’ these relations that stirring is so efficaceous in assisting solution. If a mass of fluid consisting of equal but separate parts of concentrated brine and water were so stirred that the streaks of each were not more than a couple of centimeters in thickness, a day would complete by diffusion a homogeneity which unassisted diffusion could accomplish only after hundreds of thousands of years. Now salt (in common with other haloids) is a highly diffu- sive substance. The oxygen salts, such as the sulphates of zine, copper and magnesium, are more analogous to the silicates which compose magmas. Of these the magnesian salt diffuses fastest and the zine salt the slowest. As an illustration, the copper salt may be taken. Its diffusivity has been determined by Mr. T. Schuhmeister at 000000243 in square centimeters per second. With this datum it is easy to compute the distri- bution of the diffusing salt at any time.* The following table is computed for one year and for 1000 years from the com- mencement of diffusion. The distances are measured in centi- meters from the contact, and under “saturation” the quantity of sulphate of copper per unit volume of the substance at cor- responding distance is given, the strength of the original undif- fused sulphate solution being taken as unity: Distance Distance Saturation. in centimeters ; in meters; time 1 year. time 1000 years. 1°000 0° 0. "750 3°96°" 1°25™ "500 8°36 a "250 14°24 4°50 "100 20°37 6°44 "050 24°45 7°73 "025 27°76 8°78 "010 31°91 10°19 ‘005 35°04 11°08 Diffusion of CuSO, at the end of 1 year and of 1000 years [or of hypothetical lava at the end of 50 times these periods; or of heat in underground strata after 2h 'gm and of 89 days]. The last number in this table shows only half of 1 per cent of the original solution and this may be taken as the limit of sensible diffusion. It is easy to derive from this table the fivures for any other time. Thus after a hundred years the distances answering to the given saturations will be ten times as great; sensible diffusion will cease at 350™ and semi-saturation * For the necessary Information on this computation see Kelvin’s Math. and Phys. Papers, vol. iii, p. 432, or Brit. Assoc. Rep., 1888. Becker—Some Queries on Rock Differentiation. 29 will occur at 84™ from the original contact. After 10,000 years the distances just stated will each be multiplied by 10. At the expiration of a million years the water would be just sensibly discolored by bluestone at 350™ and semi-saturation would have reached to a distance of some 84". Viscosity of lavas.—Lavas are assuredly far less diffusible and far more viscous substances than sulphate of copper solu- tion.* There is no means of determining with any approach to accuracy what the diffusivity of lava really is, but there is some reason to think that the viscosity of even the most fluid lavas is more than 50 times as great as that of water.t If one *Tt is of course needless to call attention to the difference between fusibility and fluidity. A mass may be easily fusible but very viscous when fused, or it may fuse with great difficulty and when fused be very fluid. That a lava is very fluid does not indicate that it is considerably superheated. Water at 0° C. is sensibly as mobile as at 100°, though refined experiments reveal a difference. + The viscosity of lavas is evinced by the slowness with which lava streams advance. Thus in the Kilauea eruption of 1840 the lava flowed eleven miles down a declivity of 1244 feet in two days, yet according to Wilkes and Dana (Characteristics of Volcanoes, 1890, p. 63) in this case the stream was fed from several fissures along its whole course instead of being an overflow from a single opening. The heat of the stream must have been pretty well maintained by such accessions. The average rate of flow of this lava down a 2 per cent slope was about 4 mile per hour or 22 feet per minute. Now water ina stream of such a eross section on such a grade would flow at about 6 miles per hour or about 24 times as fast. Since lava is about 2°5 times as dense as water, these data roughly indicate for the kinetic viscosity 2'5 x 24 or 60 times the viscosity of water. In the case of gases Maxwell shows that the diffus vity of mass is 1°5435 the kinetic viscosity (Theory of Heat, chapter 22) and that the ratio of diffusivity to viscosity in the case of liquids is much smaller than in gases. Hence it seems safe to assume the diffusivity of lava as not more than 4 of that of a solution like that of bluestone. In the more recent literature I have not met with investigations whieh throw light on the relations between diffusivity and viscusity. The resistance which molecules, atoms or ions meet when undergoing diffusion Ostwald illustrates by the slow subsidence of pulverized solids in air (Lehrbuch der Allgem. Chemie. vol. i, 189!, p. 698). This slowness is due, at least in part, to the viscosity of the air, and Stokes in 1851 showed that the resistance of spherical particles is pro- portional to the radius. As has been mentioned, viscosity is a resistance due to the diffusion of momentum. That viscosity impedes diffusion of matter appears evident, for example, from the behavior of sealing wax, which is an ultra-viscous fluid. Sticks of wax of different colors which have become adherent during hot weather do not diffuse into one another sensibly even after months of contact. On the other band diffusion of crystalloids takes place in quasi-solid gelatine jelly at the same rate as in a fluid (Graham). This. however, I take to be not comparable to the action of aviscous fluid, but to the behavior of a colloid septum such as bladder, but of great thickness. The jelly seems to me to have a structure similar to a sponge of very fine grain preventing convection but not the diffusion of crystalloids. Colloids do not diffuse in such a jelly. In discussing lavas it should not be forgotten that high temperature accelerates diffusion, which adds to the difficulty of making any estimate of the diffusivity of rock magmas. In choosing as an illustration of diffusion an hypothetical magma with 54, of the diffusivity of bluestone, I have been guided in part by observation on lavas. Lavas with this diffusivity mingled in thin layers, like banded rhyolite, would dif- fuse into approximate uniformity in a few hours. No one can doubt that the rhyolitic bands have been in contact for at least a few hours in the fluid state and that they must, therefore, be less diffusible than my hypothetical lava. Similar banding is not infrequent in andesites though it is less common than in rhyolite. — S SIS = sy SE z am 3 30 Becker—Some Queries on Rock Differentiation. supposes the diffusivity of such a lava to be »j, that of sulphate of copper solution, then the time needful to givea certain satura- tion at a certain point will be 50 times as great. It would take 50 years instead of 1 to establish the conditions given in the second column of the table. A million years is 20,000x 50 or 141°X50 and consequently in this vast period sensible impregnation of the lava would have extended to only about 49™ from contact (i. e. 141x385™) and semi-saturation to some 12”. It may seem to some readers that I have exaggerated the viscosity of lava. Certainly literature contains some accounts of lavas said to “run like water,” but I have been able to find no approximately precise data indicating such fluidity. That most lava streams, those of Vesuvius, for example, advance even on steep declivities at a small fraction of a mile an hour is certain. It will be seen, however, that the arguments of this paper would not be essentially changed if lava were sup- posed no more viscous than a bluestone solution. but it may also be asked why even greater fluidity may not be assumed in lavas prior to eruption, a fluidity sufficient to allow of segrega- tion in a moderate time. There seem to me abundant grounds for refusing assent to such an assumption. Hypogeal magmas must be under great pressure and they must be close to their melting points; for if they were considerably superheated the surrounding rock masses would melt and the temperature would fall to the melting point. To bring about considerable superheating would be almost as difficult as to boil water in a vessel of ice. The less viscous the magma the more difficult would superheating be. Now liquids which, like lava, condense in solidifying are most viscous at the melting point, and pres- sure increases their viscosity. Hence hypogeal lavas must be more viscous than they are when they reach the surface. The relief of pressure is equivalent to superheating. It is, there- fore, irrational to assume that lavas prior to eruption are at all more fluid than they are at eruption. All indications point to the opposite conclusion. These illustrations show that diffusion of fluids, particularly viscous ones, is an excessively slow process. it is instructive to compare its rate with that of the diffusion of heat. Accord- ing to Lord Kelvin, the time required for the diffusion of com- mon salt in water to be represented by a given curve is more than 870 times as long as that needful to diffuse heat through underground strata in such a way as to be represented by the samme curve.* Copper sulphate requires five times as much time *The diffusivity of heat in underground strata has an average value of 0°01. In Lord Kelvin’s article on heat, Enc. Brit, 9th ed., vol. XI, p. 582, Table B (auto- graph issue), the time required for the diffusion of heat in underground strata should be given as 3,170,000 years instead of as 54, of this period. Becker—Some Queries on Rock Differentiation. 31 as salt and the hypothetical lava 50 times as much as the sul- hate. Diffusion of matter in the lava, therefore, takes over 200,000 times as long as the diffusion of heat in solid rock at ordinary temperatures. Even if melted lava conducts heat many thousand times worse than solid rock, so that the con- ductivity of the fluid might be neglected, the temperature in an unequally heated mass of melted lava would be sensibly equalized by the conduction of the solid walls of the reservoir before any tendency to molecular flow which difference of tem- perature might have induced would have had time to produce sensible effects. Now in any case of the segregation of homogeneous, mis- cible, fluid magmas by molecular flow, the available osmotic force is only the difference between two osmotic forces, and the transfer of a given quantity of matter to a given distance will be much slower than in simple diffusion. Hence, so far as I can see, a mass of lava of volume, say 1 cubic kilometer, would not have had time to segregate into distinctly different rocks by molecular flow if it had been kept melted since the close of the Archzean, even if the temperature of the top could have been kept sensibly above the temperature of the bottom. But | it is very difficult to imagine how a mass of lava could be more highly heated at the top than at the bottom, since in general temperature increases with depth. If the bottom were more highly heated than the top, of course convection currents would be set up and those would effectually prevent any segregation on Soret’s method. I do not think this method should be invoked in explanation of rock differences unless it can be shown how heating from the top can occur. No such difficulties present themselves in such cases as that of the growth of a erystal. A supersaturated solution of a given substance cannot exist in immediate contact with the solid form of the same substance, but a solution may be super- saturated at a very short distance from the solid mass. In eases of erystallization there is thus an osmotic pressure-dif- ference directed toward the growing crystal, and molecular flow results. The molecular flow attending the formation of phenocrysts is usually confined to distances of a few milli- meters, or at most a few centimeters, and is clearly a process involving no unreasonable amount of time. Similarly the formation of aureoles of diffusion around blebs to distances of a few centimeters is not a very lengthened process. Thus if such a bleb had the properties of the hypothetical lava dis- cussed above, an aureole 5™ in depth might form around it in about a year, and sensible diffusion would extend to a distance of 1™™ in the comparatively short period of 34 hours. Convection unavoidable.—It has been shown above that 32 Becker—Some Queries on Rock Differentiation. segregation of magmas by the method of Ludwig and Soret would occupy a stupendous time even if a mass of melted lava could be kept free from convection currents. This freedom, however, could only be secured by permanent, regular decrease of temperature from the upper surface of the magma down- wards. In any fluid of only moderate viscosity even a very small rise of temperature at the bottom would cause convec- tion currents which in a day would undo the segregation it had taken thousands of years to accomplish by Soret’s method. Lavas of very great viscosity would also mingle by convection far more rapidly than the most diffusible solutions could segre- gate through differences of temperature. Mingling again might occur in the absence of bottom-heating by any mechan- ical disturbance of surrounding rock masses. | That in general the temperature of the globe increases with depth is perhaps the best established generalization of geology. Hence even if it be granted for the sake of argument that in some particular locality the temperature decreases with depth, it is clear that such a thermal distribution is a case of unstable equilibrium. It can, therefore, only be temporary and it would surely be a strange exception were such an abnormal distribution of temperature to last for 1000 years. Yet in that time no segregation worth mentioning as an origin of rock - differences could occur. The normal condition of a hypogeal molten magma must be that in which temperature increases with depth and in which convection effectually precludes any process of segregation by molecular flow. Immiscible fluids.— Another method of segregation, which is quite distinct from that discussed above, depends upon changes in the mutual solubility of fluids. Some fluids which at certain temperatures mingle in all proportions dissolve one another only in certain proportions at other temperatures. Thus benzol and acetic acid mix without limit at 15°, but below this temperature separate out into two layers, one of which con- tains nearly twice as much acetic acid as the other. So, too, phenol and water mingle freely above 69°, but not below this. temperature ; and there are many similar instances. The phe- nomena were studied by Mr. Alexejew, who concluded that in all cases where the solutions do not react chemically upon one another they become miscible above a certain temperature.* Though Mr. Alexejew studied some of the physical relations of solutions of fluids in fluids, he did not determine whether the passage from complete to partial miscibility is accompanied by expansion or contraction. This step, however, has been taken by Mr. Herman Pfeiffer in Prof. Ostwald’s laboratory. He finds that this change is accompanied by a sudden sharp * Wied. Ann., vol. xxviii, 1886, p. 327. 4 Becker—Some Queries on Lock Differentiation. 33 contraction of volume.* It appears to follow of necessity that at the temperature of complete miscibility under a given pres- sure an increase of pressure would resolve the homogeneous fluid into immiscible portions. In close relation to this separation of a homogeneous fluid into different layers is the precipitation of a solid from a fluid. The process of solution of a solid is one involving the absorp- tion of heat, and in general the solubility of solids increases with the temperature. Anomalous cases appear to be referable to changes in molecular aggregation, the formation of hydrates and like causes. The influence of pressure on the solubility of solids was first carefully investigated by Dr. Sorbyt, and Mr. F. Braunt has more recently made a very thorough study of the subject. Experiments have naturally dealt almost exelu- sively with aqueous solutions at ordinary temperatures, and it must be borne in mind that as water is a fluid of very excep- tional properties the direct results of experiments on solutions in water are not immediately applicable to other fluids such as lavas. Most substances dissolve in water under contraction of volume and only about half a dozen compounds are known which undergo dilatation during solution. Now when con- traction takes place increase of pressure will and does assist solution. If contraction were a universal concomitant of solu- tion the interior of the earth would be fluid. But Mr. Braun gives apparently sound reasons for believing that even in aque- ous solutions under high pressure and temperature, dilatation and not contraction would attend solution. The investigations of Prof. Carl Barus§$ and others and some observations of mine on dikes|| show that lavas contract in solidifying. The fre- quent corrosion of phenocrysts is seemingly due to increase of solubility attending relief of pressure. Thus for magmas it appears that increase of pressure promotes precipitation of solids as well as segregation into distinct fluids. Segregation by immiscibility.—In fluids which, though origi- nally homogeneous, tend to break up into two or more immiscible parts, two distinguishable modes of separation may be followed. As the temperature of separation is approached, the walls of the vessel being cooler than the fluid, any component about to separate out will separate to some extent on the containing walls much as frost or dew forms on good conductors. It does not appear that any large part of a large mass of even a mod- erately viscous fluid could be segregated in this way, for the process involves molecular flow from the interior of the fluid. * Zeitschr. phys. Chemie, vol. ix, 1892, p. 469. feroc..ROS.,/ vol. xii, 1863, p. 538: t Wied. Ann., vol. xxx, 1887, p. 250. S This Journal, vol. xliii, 1892, p. 56; vol. xlv, 1893, p. 1. || North Amer. Rey., April, 1893. Am. Jour. Sci1.—Fourts Series, Vou. II, No. 13.—Janvary, 1897. 3 , V, 34 Becker—Some Queries on lock Differentiation. Thus in a spherical mass of 100™ radius, if half of a compo- nent were thus to be deposited on the walls, a portion of this deposit would have traversed a distance of nearly 21™ by molecular flow, which would take thousands of years even in the case of a solution of bluestone in water. If the separating fluid does not condense on the sides of the enclosing cavity, it must condense somewhat like fog in the mass of the fluid. Now in a very fluid mass, like water, the larger drops of such a fluid will rise or sink more rapidly than the smaller ones, coalescence will occur and the lighter fluid may separate out in a layer. But even in the case of a foreign material suspended in air viscosity greatly delays such separa- tion. The clouds are substantially aggregates of small water drops which, because of the viscosity of the air, fall so slowly that the shghtest current sweeps them along. So, too, dust remains suspended in the atmosphere because air is viscous.* In fluids such as lava it scarcely seems credible that any exten- sive separation of a precipitated immiscible liquid should occur. It may be that spherulites and perhaps some phenocrysts are crystallized from drops of such liquids. However this may be, it is certain that many of the phenocrysts form before eruption and remain suspended in the magma in spite of densi- ties differing considerably from that of the medium.t Thus even in the process of the separation of fluids into immiscible or partially miscible fractions I can see no adequate explana- tion of rock segregation. Furthermore, if, as seems to follow from the law of fusion, magmas are not heated much above the melting point, there is but a small range of temperature within which such separations could occur and they would be correspondingly rare. Heterogeneity of the earth.—lf the physical theory of solu- tion fails to account for rock segregation, two alternatives are left. Either segregation takes place in accordance with some principle of physics as yet undiscovered (2gnotwm per ignotius), or the facts which have led to the hypothesis of segregation are capable of a different interpretation not at variance with the known properties of matter and compatible with reason- able limits to geological time. So faras I know, all geologists and astronomers are in unison * The viscosity of media is probably only one of the influences affecting the sub- sidence of disseminated fluid or solid particles + It is well known that phenocrysts in fresh surface-flows are often bent and eyen broken. Sometimes black borders have formed about hornblendes thus frac- tured. Such fractures must have happened during eruption. The lithologist will not require to be reminded that phenocrysts of augite and of amphibole with a density of say 3°25 are often of about the same size as those of feldspar with a density of say 2°65, yet there is as a rule no tendency to the separation of the lighter and heavier phenocrysts into distinct layers. Becker—Some Queries on Rock Differentiation. 35 in the belief that the earth has been fluid, not indeed at any one time from center to its present surface, but at least to a great depth from the temporary surface of the growing globe. Yet the earth is clearly not a homogeneous mass, nor is it a system of concentric shells, each homogeneous. , The mere fact that one hemisphere is almost entirely covered by water shows that the globe is of greater density below this great ocean than beneath the opposite continental surface. Were the shells homogeneous no continents could protrude above the sea. Were the earth of uniform composition no mountain ranges could stand above the plains. Were the material below the plains uniformly distributed there could be no anomalies of gravity such as occur near Moscow, in Kansas, and else- where. The distribution of feldspars in the western part of this continent shows lack of homogeneity, for on the Pacific slope potash feldspars are narvelously rare. No trachyte and extremely little typical granite is known from the Wahsatch range to the Pacific ocean. The distribution of metallic ores shows heterogeneity. Much more than 90 per cent of the known tin ores of the world lie in a belt stretching from the straits of Malacca to Tasmania. There has been deposition of mercurial ores in this belt also; but their quantity is insignifi- eant. A belt of quicksilver deposits extends from British Columbia to Chili. In this belt there is tinstone at many points, but the total product of tin on this belt is scarcely worth mentioning. I can only infer from these facts that quicksilver is an extremely subordinate component of the earth in the Australasian region and that the globe contains little tin along the Cordilleran belt. It is needless to observe that in almost any small area the rocks show marked variations or that two hand specimens from the same locality are rarely indis- tinguishable. Deserving of special mention, however, are the striped rhyolites, the banded gabbros studied by Sir Archibald Geikie and Mr. Teall, and the ribbon gneisses so abundant, for example, in the southern Appalachians. The rhyolite at least has been fluid, and most geologists consider gabbroitic and granitic magmas as fluids. The diffusion exhibited in these cases is slight and sometimes hardly perceptible to the naked eye, yet it is scarcely supposable that these bands were not in contact for days at least in their mobile state. Now my hypo- thetical lava would diffuse to the depth of a millimeter in three or four hours. Hence these sharply banded rocks must be much less diffusible than my assumed lava and the diffusivity of the granular rocks and their fluidity would scarcely exceed zero.* Thus from the surface of the globe to its minutest * Banding such as is found in gueiss and rhyolite couid not result directly from segregation on Soret’s methods or by difference of pressure, for these processes 36 Becker—Some Queries on Rock Differentiation. portions there are clear indications of heterogeneity, notwith- standing that similar rocks and similar series of rocks occur in widely dispersed localities. Is there any valid indication that uniformity ever reigned ? It used to be thought that the Archzan rocks were uniform, but it is well known now that they are not so. The early eruptions and intrusions seem quite as diverse as the modern ones, excepting so far as original differences are masked by metamorphism. The theory of the permanence of continental areas has many very strong supporters, and that land areas have existed since the Cambrian seems certain. The mountain sys- tem of the world in its larger features appears to have been outlined during the Archeean, and there are observations indi- cating a highly accentuated topography even in those days. Were the indications of heterogeneous composition confined to the immediate neighborhood of the earth’s surface, it might be maintained that these inequalities had been brought about since consolidation, but everything tends to show that only the shell of the earth next to the surface and a few miles in depth par- take sensibly in orogenic movements, while several of the evi- dences of heterogeneity point to inequalities at great depth. Uniformity unattainable.—lf the earth condensed from a nebulous ring, it is fairly inconceivable that the successive shells of the growing mass should each have been uniform in composition ; and if the origin of the earth is a ring thrown off from the sun, the coalescence of this ring to a globe can- not have resulted immediately in uniform distribution of mat- ter. The sun spots show that the sun is not even yet an ageregate of shells each uniform in composition. The exterior layer of the globe must have retained such fluidity as it pos- sessed for a very long time, and must have passed by insensible gradations through every temperature between the initial one and that of consolidation. Had the various component por- tions of this layer been of large size, of low viscosity, and not miscible with one another, they would have arranged them- selves in the order of density quite irrespective of chemical composition. If the masses were viscous, however, nothing like a perfect separation according to density could occur in” can lead only to very gradual transitions. Banding might conceivably result from such a segregation followed by active stirring, but only on condition that stirring was immediately followed by solidification, for otherwise diffusion would restore homogeneity. Separation of a magma into immiscible portions followed by active stirring might also produce banding, but again only on condition of immediate consolidation, since otherwise separation into two layers would again take place and much more rapidly than at first. Miscible substances in contact which do not diffuse at a finite rate can have no sensible vapor tension and must be solids or ultra viscous fluids. Immiscible fluids must have a perfectly sharp contact like that between a lead button in an assayer’s crucible and the enveloping slag. Becker—Some Queries on Rock Differentiation. 37 this way and only a rude approximation to regularity would be attained. If the melted masses were partially or wholly miscible, much the same arrangement would take place at first because diffu- sion on any large scale is at best a relatively slow process. Then diffusion and convection would come in play, tending to equalize the composition. : On the other hand, unless the originally heterogeneous masses had very different properties from those of remelted rocks, whether of Algonkian or modern time, they cannot have diffused on any large scale, and uniformity along equipo- tential surfaces cannot have been attained even if fifty million years were allowed for the process, unless the convection cur- rents were so powerful and universal as to break up the original masses into streaks of a few meters in width. I see no cause for convection so active as this. In the nebulous state the material of the earth must have assumed some approach to convective equilibrium of temperature, and though here and there the solidifying globe may have been affected by dis- turbances of frightful intensity, analogous to sun spots, a general diversity of temperature sufficient to stir the whole or most of the melted layer into uniformity seems utterly improbable. What is known of the properties of matter seems to me to point to the hypothesis that the material of the earth is rudely arranged by density irrespective of chemical composition, the different masses mingling for a few meters or scores of meters along their common boundaries, this structure being due to original heterogeneity. If segregation took place at all, prior to the consolidation of such a globe, this process would be lim- ited to particular masses and would tend to still greater hetero- geneity. Hypogeal refusion.—Consider now the effect of the refu- sion of any portion of the earth’s mass. Unless the tempera- ture of the magma were raised essentially above the initial temperature of the molten globe, or unless it were melted at a very different pressure, the magma would simply be restored to the condition in which it existed before the primal consolida- tion. There is no indication that lavas prior to eruption are really raised to temperatures greatly above that of fusion, for almost all of them bring solid phenocrysts to the surface, nor is 1t easy to see how they could be heated much above the melting point, for so long as there was unfused material of a similar character in the neighborhood of the subterranean mass undergoing fusion, any heat increment would of course melt more rock instead of raising the temperature of that already fused. Thus it is substantially certain that in the molten globe > — ~- 38 Becker—Some Queries on Rock Differentiation. a given magma passed very gradually through the temperature at which it has more recently been remelted. As for the pressure, it seems possible that under continental areas a given fusing subterranean mass may exist under a somewhat smaller load than that to which it was subjected at primal consolida- tion, tor the general tendency of continents is to upheaval and degradation by erosion. If this change of pressure is of any consequence at all, it will tend to make the refusing mass more fusible and more miscible. If now the mass were both homogeneous and in molecular equilibrium before the primal consolidation, it may be in equi- librium after refusion. If the pressure is smaller than the original one, this difference would have no tendency to promote segregation. If by some almost inconceivable coincidence, the upper portion of the refused mass were heated to a higher temperature than the lower part, this temperature would be equalized by conduction through the walls, if not through the liquid, before any sensible segregation could occur. If the mass were heterogeneous in consequence of primal segregation, fusion would again tend to restore molecular equi- librium and the only chance of a new segregation would lie in the possible difference between primitive and ultimate pressure, which, if positive, would tend to mixture rather than to separa- tion. If the mass were heterogeneous because’ the primal fusion had not continued long enough to bring about homogeneity, refusion would be accompanied by a tendency to continue the process of molecular flow and to decrease the heterogeneity of the mass; but even if the refused mass were kept molten for a million years, this tendency would probably have only insig- nificant results. Mixture by eruption.—Little or nothing is known of the process of refusion of subterranean masses to eruptive magmas.. Supposing a mass which is fused and near its melting point to remain in its subterranean reservoir, it must in general receive or lose heat. In the latter case it will reconsolidate, in the former the mass of melted material will increase. In the ease of rocks, fusion is accompanied by expansion and the magma must have more space than it occupied in a solid con- dition. Any elastic strains in surrounding masses will also tend to expel it and it would seem to me most probable that magmas are expelled as soon as the mass of melted material had increased to a certain limit dependent upon local condi- tions. If so, there must be little time for the fulfillment on a large scale of a process so slow as molecular flow. Doubtless fusion may be confined to a nearly homogeneous portion of the earth’s lithoid shell. If the hypothesis explained above of Becker—Some Queries on Rock Differentiation. 39 primitive, unmingled, unsegregated masses is correct, a good many of these must have the composition of augite- andesite : for this as well as several other simple rock types | has issued at most distant points with almost constant characteristics. Fusion may, however, also affect two or more diverse masses and then eruption tends to mingle them. Ejection through pipes or fis- sures must indeed be a most efficient stirring process, and since relief of pressure is accompanied by depression of the melting point, different magmas thus ejected are superheated and may mingle to an obser vable extent by diffusion before they finally consolidate. In such cases one would probably find two (or more) rock types accompanied by mixtures of variable compo- sition. Again, a fissure through which different types were extruded successively or in mixture might at the close of the eruption be filled with a single type or with a mixture. If such a mixture were at all intimate, diffusion would mask the original differences and the case would be one of apparent tran- sition.* Possibly some of the observed occurrences which have led to the hypothesis of differentiation are really of this character, for I think it has been shown in the foregoing pages both that transitions can be explained on the hypothesis of primitive heterogeneity and that the explanation of differentiation itself presents formidable difficulties. I do not see why it should be necessary or desirable to assume that in the early history of the globe the vast shell from which eruptions issue was reduced to substantial uniformity. Experience affords no analogy in sup- port of such an assumption nor has any theory been propounded which will account for it. If primitive heterogeneity is still an important feature in the earth’s structure, and if unmingled magmas represent primitive differences, ‘the labors of lithologists would naturally be directed to detecting these original types. These would probably be recognizable by their wide distribution and con- stant character. Then areas of rapid variation would be regarded as representing mere mixtures and it might be possi- ble to reduce instead of increasing the number of rock species. Abstract.—All known processes by which the segregation or differentiation of a fluid magma could take place involve * The order of mixture and extrusion would seem to depend on many circum- stances, among others on the shape of the subterranean reservoirs. If this were a cone with its vertex nearest the surface, the disposition of the ejecta would be very different from that which would be observed if the reservoir were a flat- tened lens with its edge horizontal and a vent ou one surface. If each eruption represents a separate meiting, still other dispositions will result. It appears to me anything but remarkable that different observers find eruptions in different areas taking different orders. Gradual solidification from fissure walls of dike magmas circulating by convection may lead to preponderance of less fusible in- gredients near the edges of a solid dike. ‘ 40 Becker—Some Queries on Rock Differentiation. molecular flow. This is demonstrably an excessively slow process excepting for distances not exceeding a few centimeters. Soret’s method of segregation, even if it were not too slow, seems inapplicable because it involves a temperature unaccount- ably decreasing with depth. The normal variation of tem- perature, an increase with distance from the surface, would be fatal to such segregation. ‘The least objectionable method of segregation would be the separation of a magma into immiscl- ble fractions; but this seems to involve a superheated, very fluid magma, while the law of fusion and the distribution of phenocrysts in rocks indicate that magmas prior to eruption are not superheated to any considerable extent and are very viscous. The homogeneity of vast subterranean masses called for by the hypothesis of differentiation is unproved and improbable. The differences between well-defined rock types are more prob- ably due to original and persistent heterogeneity in the compo- sition of the globe. Hypogeal fusion and eruption tend rather to mingling than to segregation, and transitional rock varieties are not improbably mere fortuitous mixtures of the diverse primitive, relatively small masses of which the lithoid shell of the earth was built up.* Washington, D. C., October, 1896. *T owe thanks to Dr. Arthur A. Noyes of the Mass. Institute of Technology for kindly examining the manuscript of this paper. Dr. Noyes’s reputation as an investigator in osmotic questions gives his approval of my argument great value. Washington—Igqneous Rocks from Smyrna, ete. 41 Art. 1V.—On Igneous Rocks from Smyrna and Pergamon; by Henry 8. WASHINGTON. Or the specimens described in the following brief notes those from Mt. Pagos near Smyrna and from Pergamon were collected by the writer in the spring of 1892. Those from the other Smyrna localities were collected by Mr. J. 8. Diller in the summers of 1881 and 1882, who most kindly presented them and his notes for use in the preparation of this paper. It may be mentioned, in justice to him, that these form but a small part of those he collected, the rest not being available in time for publication. The writer gladly avails himself of this opportunity of expressing to Mr. Diller his heartiest thanks for his kindness and generosity. It must be premised that these notes are but fragmentary and that their publication seems only justified by the scantiness of our knowledge of the rocks of Asia Minor. Augite-andesite, Snyrna.—The city of Smyrna lies at the head of a deep gulf in the west coast of Asia Minor and is surrounded on three sides by igneous rocks. To the northwest are the hills of Phoceea, probably of andesites and their tufts ; to the north and northeast the andesitic masses of Yamanlar Dagh and Mt. Sipylos; and immediately to the south and southwest a ridge of igneous rock formed of the hills of Mt. Pagos, Kara Tash and Giéz Tepé. Since the early descriptions of Hamilton* and Tchihatcheff,t little or nothing has been written on the geology of this region, with the exception of a short paper by vom Rath.t Of the rocks of Phocea and the Sipylos ridge [ have no specimens, so that we must confine our- selves to those from the immediate vicinity of Smyrna. From — the brief descriptions of vom Rath$ and the notes of Mr. Diller, it seems that the Sipylos rocks are closely similar to those of Mt. Pagos, and Mr. Diller also compares them in places with the “ hypersthene-bearing andesite by Assos.” Mt. Pagos, 185™ high, is the eastern end of a ridge, up the northern and eastern flanks of which the modern city of Smyrna extends, and which formed in antiquity the ancient acropolis or citadel, the summit being still crowned with ruined walls. It is composed of a mass of andesite which, according to Hamilton and Tchihatcheff, has been forced up through underlying beds of Cretaceous (?) limestone. In many places *W. J. Hamilton: Researches in Asia Minor, London, 1842, i, 54; also Hamilton and Strickland, Trans. Geol. Soc., London, ii, 293. + P. de Tchihatcheff: Asie Mineure, Géologie, i, Paris, 1867, pp. 69-73. t Vom Rath: Sitzber. Niederrh. Ges zu Bonn, 1882, pp. 16-26. § Quoted in Roth, Geologie, ii, 326, 1883. 42 Washington—ILgneous Rocks from Smyrna, ete. the andesite is overlaid by masses of detritus containing numer- ous shells, and here and there accumulations of shells are found, principally of oysters and scallops (murex); snail shells also abound, which closely resemble the numerous land snails of Greece, so far as I was able to judge. Fragments of pot- tery, mortar and charcoal are also found in the detritus, and these deposits seemed both to Mr. Diller and myself to be in reality “ kitchen-middens ” rather than deposits from water or old beaches. Mr. Diller’s notes, however, show that along the —Meles River the andesite is overlaid by soft travertine-like lime- stone, apparently a lacustrine deposit. In places also andesitic conglomerates were observed. The general mass, however, is very compact and shows, espe- cially toward the west, a beautifully banded structure, the bands being red and black. The greater part of the mass visible is reddish, the color being due, as we shall see, to decomposition. The freshest rock has a dark gray groundmass, carrying numerous phenocrysts of feldspar. and augite and fewer of biotite. Its texture is harsh and in cavities are found small crystals of biotite and acicular hornblende, with globular masses of a zeolite which vom Rath regards as a natrolite. Fine mammillary hyalite is also abundant. The specific gravity of a fresh gray piece was found to be 2°640 at 17° C. Under the microscope the rock is seen to be composed of plagioclase, diopside and biotite, with accessory magnetite, apatite and zircon, lying in a glass base. Neither hypersthene, hornblende, olivine nor quartz was seen in any of the speci- mens. The structure is eminently the vitrophyric of Rosen- busch and is well shown in fig. 4 of Tafel v of his Mikro- scopische Physiographie, vol. ii. The plagioclase phenocrysts are often well shaped, but in many cases fragmentary. They are clear and carry only few inclusions of glass, apatite and zircon. Zonal structure is com- mon, the interior being, as usual, the more basic. Nearly all show well-developed twinning lamelle according to the albite law, and here and there pericline twinning is seen. Examina- tion of sections on 6(010) by Michel Lévy’s method* showed extinction angles of 20° and 25° on each side of the twinning plane, indicating a labradorite of the composition Ab,An,, or somewhat more basic. The few feldspar phenocrysts that showed no lamelle could not be identified with certainty as orthoclase. The diopside phenocrysts are often well-shaped, though fragments occur. They are almost colorless and clear, and carry only few inclusions of brown glass and magnetite. Some twinning parallel to a@(100) is seen, but zonal structure is uncommon. Biotite phenocrysts, in stout hexagonal prisms, * Michel Lévy : Détermination des Feldspaths, Paris, 1894. Washington—Igneous Locks from Smyrna, ete. 43 are not as common as diopside. They are deep brown, strongly pleochroic, and basal sections show a slight dispersion of the axes parallel to 0(010). They are quite fresh and only slightly altered on the edges to a narrow opacitic border, with rounded angles. The groundmass in general is typically hyalopilitic, the microlites being mostly of labradorite, with fewer of what seems to be orthoclase, some small colorless diopside prisms, many magnetite grains, and a few small crystals of apatite and zircon. These generally show flow-structure. The glass base is usually colorless, though sometimes light brown. The preceding description applies especially to the fresh gray specimens. The red ones show much the same features, the differences being due to decomposition. Thus in these the augites are colored red or black on the edges, the biotites are a dark red brown, and the groundmass is much decomposed with limonitic products very abundant. An analysis by the writer of a typical, fresh, dark gray specimen from near the top of Mt. Pagos is inserted here. Its sp. gr. 1s 2°640 at 17° C. SIC eee ere eens ate mene ree 60°68 Jet ORES 8S TERE Ra sr een eee 16°19 Shel Og ee ain ea pe Shales ou ie Oe a tee NR, ge ets 1°58 MoO tee ee Te 2°96 CAO ree et he ea oe 5°88 Nat OPN eee ee ee 3°11 JEG NG SE Ste i pig Nel apa 3°95 RIOT LL See LE es OVERS OI 100°70 Jt it high in silica and alkalies for an andesite and with rather low alumina. The potash is considerably higher than soda, which would account for the presence of some orthoclase in the groundmass, since biotite is not very abundant, though part of it may belong to the glass base. On account of the vitreous character of the rock it is scarcely practicable to calen- late from the analysis a possible mineralogical composition. There is, however, more SiO, than enough to satisfy all the other constituents, even assuming that all the K,O is in ortho- clase, which, of course, is not the case; so that quartz would have separated out if the rock had been formed under condi- tions allowing of a holocrystalline development. On this account the rock might be properly called a dacite. The rocks of Kara Tash (Black Rock) 2*™ west of Smyrna, three specimens of which were sent me by Mr. Diller, differ 44 Washington—Iqneous Rocks from Smyrna, ete. considerably from those just described. Rosenbusch* describes rocks from this locality as biotite-hypersthene-andesite. I was unable to find any hypersthene in the specimens examined, and the microphotograph he gives resembles very much my sec- tions of the Pagos rocks, and is of a totally different character from all of those of the Kara Tash rocks sent by Mr. Diller, so that it seems probable that the locality of his specimen is incorrectly given. The Kara Tash rocks are all very dark and compact, show- ing numerous small glassy feldspars, and a few augite and biotite phenoerysts in a very dark brown or black, highly vitreous groundmass. Some evidences of flow-structure are seen in the hand specimen and are even more marked in the mass, judging from Mr. Diller’s notes. Under the microscope the very well-shaped feldspar pheno- erysts are seen to be of labradorite a trifle more basic than Ab,An,. They are clear and fresh and show, almost without exception, twinning lamellae, while zonal structure is frequent. Inclusions are not very abundant, and consist mostly of small spots of dusty brownish glass, with a few apatites. The very pale green diopsides are highly automorphic, perfectly fresh and contain few inclusions. The not very numerous biotites are greenish brown and perfectly fresh, even incipient altera- tion not being seen. Some large, well-shaped magnetite grains may also be classed among the phenocrysts. The very abund- ant groundmass is highly vitreous, the glass being quite color- less, and only rare small crystals of augite, feldspar, magnetite and apatite being present. It is chiefly remarkable for the very great abundance of trichites, with which it is thickly crowded. By far the greater part of these are curved, some- times to a high degree, but some straight ones are seen. Though under low powers they seem opaque, yet high powers show the majority of them to be clear and colorless. Those which seem black and opaque under these conditions probably owe their appearance to their excessive tenuity. What the nature of these minute bodies may be it is impossible to say, since they exert no action on polarized light. The most natu- ral supposition is that they are either diopside or feldspar, and I am inclined to consider them the latter,-—probably orthoclase, —basing my opinion on the results of analysis and their rather remote analogies with forms seen elsewhere. Quite well developed flow-structure is observed, which is brought out more prominently by the presence of some narrow dusty gray streaks. The specific gravity was found to be 2°601 at 18° C., the lower figure as compared with that of the Mt. Pagos speci- men being chiefly due to the more highly vitreous character. * Rosenbusch: Mikr. Phys., ii, 890, 1896, also Taf. v, fig. 4. Washington—Igneous Rocks from Smyrna, ete. 45 An analysis of the rock of Kara Tash by the writer is given. As will be seen, it resembles in general that of the Mt. Pagos STAG AEA OD Se eae ae pee 61°93 ge ee et OAT [PEN RES Se eS ae ee 1°93 Rie sae eee as be 2°28 Wer Opts vero ee ty ee 2°66 (A i ats a ees Palle, ante age 4°31 IanO et Coe ge ne 2°92 One Foes Baba h NATL 3°92 MOMs ase SHE iia! Ay 2°28 100°65 rock, though silica is somewhat higher and lime a little lower, while alkalies, ferrous oxide and magnesia remain almost exactly the same. Alumina is, however, much higher and ferric oxide lower. Water also is much higher, which may be attributed to the abundant glass. One of Mr. Diller’s specimens from the coast 4*" west of Smyrna, between Kara Tash and Gidz Tepé, shows very clearly the red and black banding already mentioned. The black. bands resemble the rock just described, while the red look more like the Pagos specimens. Under the microscope both bands resemble somewhat more the Pagos rock than that of Kara Tash proper, though they differ from either. The phenocrysts, which are of labradorite, diopside and some biotite, are often fragmentary and the finely automorphic crystal boundaries of the Kara Tash phenocrysts are lacking. The red bands show a slight brown coloration of the diopsides and biotites, with limonitic spots here and there. The base is a clear colorless glass, and is quite hyalopilitic through the presence of numerous minute, strazght, trans- parent, colorless trichites, with some larger feldspar laths. The groundmass of the black streaks shows in addition large amounts of fine “ dust,” which gives it a dirty appearance and greatly interferes with its transparency. Im this it resembles the dusty streaks of the Kara Tash specimens. Another specimen from the coast 3°" west of Smyrna shows a fine-grained reddish groundmass containing very many glassy feldspars and some small brown spots representing original augite and biotite crystals. In thin section it greatly resem- bles the red Pagos rocks, and all except the feldspar is much decomposed. The diopsides are all deeply bordered with brown, the biotites are all entirely reduced to brown rusty masses, and the groundmass is much decomposed and dirty. Another specimen from near the same place shows a dark, f 7 46 Washington—Igneous Locks from Smyrna, ete. slightly greenish gray, waxy groundmass, with the usual pheno- erysts. It contains also an angular enclosure of a fine-grained reddish rock, showing only few phenocrysts of feldspar. This is separated in part “from the enclosing rock by a erevice, the walls of which are lined with crystals of prehnite. Seen in thin section, the main rock is seen to be much decomposed, but resembled originally the red bands just described. The labra- dorite phenocrysts are fresh, but the diopsides and biotites are decomposed as before, though the product is rather darker. The abundant vitreous groundmass is of a brown glass, rather dusty, and the numerous straight trichites are black thrauee decomposition. The reddish enclosure is so thoroughly decomposed as to render examination very unsatisfactory. It is evidently a vol- canic rock, probably more basic than the andesites previously described, and allied to the olivine-free basalts. The only phenocrysts visible in thin section are of colorless diopside, in stout prisms often with pyramidal terminations. These show the usual brown decomposition border. The groundmass is much decomposed in general to an indeterminate dirty mass, with spots of calcite here and there. In parts of the slide, however, it preserves fairly well its original structure—being composed vf long slender prisms, or perhaps sections of plates, of feldspar, with interstitial brown glass base. These feldspars show no twinning, and in many cases extinguish parallel to their length, or nearly so, while in others the extinction is oblique at angles up to 26°. We may then suppose them to be of orthoclase or oligeclase, and labradorite. In general they form groups of needles parallel to each other, but in a few cases compose sheaf-like, diverging forms. No evidence of the former presence of olivine or biotite could be made out. The only specimen of the rock of Gidz Tepé* (Eye Hill) shows a compact, greenish gray groundmass, with brownish feldspars and a few small augites and biotites. Under the microscope this also is seen to be somewhat decomposed, though in a different manner from the preceding. While the substance of the feldspar phenocrysts is fresh, yet they are traversed by numerous brown dusty cracks, and their glass inclusions are all similarly altered. The few augite phenocrysts present are pale green and quite fresh, without the usual dark borders. The dark green biotites are profoundly decomposed, though there is usually a core of unaltered sub- stance. There are also present many rather long stout prisms of a dull, opaque, brownish, granular decomposition product, * This is the name as given on Kiepert’s Map of Western Asia Minor (Berlin, 1892), and is not an uncommon one in the country. Rosenbusch renders it Yous Tepé. Washington—Igneous Rocks from Smyrna, ete. 47 the original mineral being entirely gone. It may have been augite, or perhaps more probably hypersthene. A few larger patches are reminiscent of magmatically altered hornblende. The groundmass is much decomposed, but not sufficiently so to hide the fact that it is holocrystalline and largely of feldspar, with numerous now rusty remains of long ferromagnesian microlites and laths. Serpentinous patches are also present. It must be noted that the rocks of Gidz (Yous) Tepé are referred by Rosenbusch to biotite-augite-andesites.* Finally, a specimen from a locality ‘‘3*" south of Smyrna in the railroad cut where the aqueduct crosses the Meles River,” may be briefly mentioned. This resembles the gray rocks of Mt. Pagos, but is rather decomposed. Under the microscope the phenocrysts of plagioclase and augite are rather less abundant, and only few of biotite are to beseen. The ground- mass is also gray and thickly sprinkled with dusty grains and microlites, and there is little or no evidence of flow. Greenish andesites are also reported by Diller from this neighborhood, whose color is apparently due to decomposition. Biotite-dacite, Pergamon.—The hill which formed the acropolis of the ancient Greek city of Pergamon lies 25*™ east of the coast and some 69 north of Smyrna. It stands at the junction of two small streams which unite here to form the ancient Kaikos. It is 310™ high—the southern end of a long ridge which rises steeply from the surrounding alluvial plain. In structure the hill apparently much resembles Mt. Pagos, but I do not feel competent to discuss this point, as my examina- tion was too cursory. The rock of which it is composed varies rather more than at Mt. Pagos, being generally a hornblende-free, but sometimes a hornblende-bearing, biotite-dacite, which also carries orthoclase in considerable amount. As the presence of hornblende does not affect the other characters, the varieties may be described together. A few tuff-like masses were also seen. The rocks of Pergamon have been described by J. Rotht and Lepsius.t The former briefly notes that sanidine, abund- ant plagioclase and biotite, and rare green augite, occur in a compact gray groundmass. No augite was seen by either Lepsius or myseif and was probably hornblende in reality. The description of Lepsius closely agrees with mine, though he notes no spherulites and speaks of the hornblende and biotite as brown. Over the greater part of the hill the rock is gray, but in places, as in some of the long slopes and near the Temple of * Rosenbusch, Mikr. Phys., ii, 889, 1896. + Roth: Geologie, ii, 248. t Lepsius: Geologie von Attika, Berlin, 1893, 168. 48 Washington—Iqneous Rocks from Smyrna, ete. Julia at the north end, it is reddish. It is highly porphyritie, resembling the Pagos rock, the dacites and andesites of Aegina and Methana, and some of our western porphyries. Pheno- erysts of feldspar, dark biotite, and fewer of hornblende (when present) are thickly scattered through the fine-grained ground- mass, without any evidence of flow structure. The specific gravity of the freshest specimen, which was also that chosen for analysis, was found to be 2°525 at 17° C. It is thus nota- © bly hghter than the Smyrna rocks, which is probably to be con- nected with its higher silica content and more vitreous struc- ture. When examined in thin section the feldspar phenocrysts are found to be chiefly of plagioclase, with a smaller number of orthoclase. In the former the extinction angles of the twin- ning lamellee indicate a labradorite of the composition Ab, An,. The sanidines are distinguished by their lack of maltiple twin- ning (even Carlsbad twins being rare), their lower refractive index and their parallel or nearly parallel extinctions. The feldspars are very clear and glassy, and contain only a few inclusions of glass and apatite, with very rarely a erystal of biotite. The biotites are, when fresh, of a slightly brownish olive-green, and show no signs of magmatic alteration. In the reddish specimens they are browner in tone and are somewhat decomposed. The hornblende phenocrysts are well-formed, stout prisms of a dark, olive-green color, pleochroic, and per- fectly fresh and unaltered. Augite phenocrysts are wanting entirely. The groundraass varies considerably in structure. In most specimens it is very abundant and highly vitreous. It is more or less hyalopilitic, but not very thickly so, through the presence of feldspar laths, which are mostly plagioclase, with fewer of orthoclase. In one or two slides small square sections of ortho- clase are abundant. Along with the feldspar laths are shreds of biotite and some magnetite grains and apatite needles, but no augite microlites. The glass base is colorless or slightly brown, and darker brown spherulites are often present. These are usually irregular in outline and possess a radiate structure, showing between crossed nicols an ill-defined cross. These spherulites are especially abundant in a specimen from near the great Zeus Altar. Perlitic cracking is also observable in many slides. In a few cases, as near the Temple of Julia and south of the Theatre, the base is microfelsitic and brownish in color, probably due to devitrification through decomposition. For the accompanying analysis (No. 1) by the writer, the freshest specimen was selected from about halfway up the southeast slope. It is gray and contains considerable horn- blende. Lepsius’s analysis 1s added for comparison (No. 2). Washington—Igneous Rocks from Smyrna, ete. 49 1! 2 Sia tetas fo) le hc 63°17 61:93 © ; i Oe ee eae 17°15 16°45 TSA oe Ri ae ae, 2°84 4°66 ire Gya a Bek he te LSA 0°40 Mis OM ca 2°17 2°94 CHO area 4°17 4°40 BIN Oa Brees ee 8 OS 4°03 REO 419 220 1D CO ye ietads Sale fa Bo 2°51 2°50 100°59 99°51 Sprormes. Bees t 2°525 2°539 oF Tae, to ©: The analyses resemble in general those of the Smyrna rocks, especially that of the Kara Tash specimens. Silica is, how- ever, considerably higher—so high, indeed, that the name dacite is justified, though no quartz has crystallized out.* Iron oxides, magnesia and lime are all lower thanat Smyrna. There is a discrepancy in the quantities of the two oxides of iron in the two analyses, but their total amount is about the same. There is also a discrepancy in the alkali determinations, though here again their total amount is about the same. Lepsius’ differ from the others in showing higher soda than potash, which is rather surprising when the abundance of biotite and the presence of orthoclase are considered. The high H,O is largely to be referred to the abundant glass base, though some of it belongs to the biotite molecule. General remarks.—Ilt would be of much interest to compare the rocks just described with the other volcanic rocks of western Asia Minor. At the present time, however, we are confronted at the outset of any such inguiry by the insur- mountable obstacle of almost total lack of data, i. e. of modern petrographical and chemical descriptions. An examination of the geological map of Tshihatcheff shows that a line of vol- eanic centers extends alorg the west coast from Smyrna north- ward, including the areas of Smyrna, Sipylos, Phoceea, Yund Dagh, Pergamon, Dalanlar (Kiepert, Doghanlar), the extensive district of the Troad, and Kapoudagh on the Sea of Marmora. As we have already seen in the case of the Assos rocks (page 41), and as may be inferred from the brief descriptions of Techihatcheff and Dillert the andesitic rocks of these centers much resemble each other. According to Diller, in the Troad some of them carry hypersthene, and they are associated with * Cf. Kiich: Vulk.-Gest. Republ. Colombia, Berlin, 1892, 19; also H. S. Washington: Jour. of Geology, iii, 21, 1895. + Diller: Quart. Jour Geol. Soc., xxxix, 632, 1883. Am. Jour. Sc1.—FourtH Surizs, Vou. III, No. 13.—Janvary, 1897. 4 | aelll atl ial foettlt fener Heh ide i H \ (eli {ie TA Feotilliyy hgett Ys tpl qi Testy ral ih el 50 Washington—Iqneous Locks from Smyrna, ete. basalts, nepheline-basalts and rhyolites. There is some evi- dence of the Troad rocks being on a line extending westward through Mytilene and Samothrace,* where quartz-trachytes, trachytes and basalts occur, to Thessaly, where basalt is found at Persufli.t Since the analyses of the andesitic “basalts” of Mytilene and Persufli offer certain analogies with the rocks of Smyrna and Pergamon, they may be here inserted. 1 2 SiO. One 56°58 53°61 TO Se: eee Onn 0°34 AOE mer Meee 14°88 16°11 Hei) ett ae 23M 3°05 Bie Q@ ye ee haces ais 3°04 4°45 Mn Opiiiie joe he 0°16 0°14 Mig Oi es Meron put as cyt 3°76 6°80 CaO) sereoeteaets 5 ee L869 7°00 BaQ@ diate hep. 0°07 ate NG 'Oen die Bact pax 3°36 3°95 TO) 2 feet Bee oes 2°18 3°08 PO. vices ete 0-15 a TO 2°12 1°65 CO oe ee a BBE Wa 100°39 100°18 No. 1{ is of the Mytilene rock, but I can find no description of it. Lepsius describes the Persufli rock (No. 2) as a quite typical olivine basalt, amygdaloidal, and somewhat decom- posed. The interest in these at present chiefly centers in the high alkalies, in which they resemble the rocks above, though soda is slightly higher than potash. The two resemble each other quite closely and the potash is high enough to lead one to infer the presence of orthoclase. It is of interest also to note that an augite-hornblende-ande- site is described by Becke§ from the Island of Chios, which lies west of Smyrna. This is propylitic in habit and is com- pared by him with the andesites of the Bosphorus. South of Smyrna no igneous rocks are noted on Tchihat- cheff’s map till the promontory of Budrum (Halicarnassos) is reached. These, however, as well as the rocks of Kos and Nisyros near by, are, as | have pointed out elsewhere,| proba- bly connected with the Aegina-Santorini line, the rocks of which show lower alkalies, with soda higher than potash. * Niedzwiedzki: Tsch. Min. Mitth., 1875, 89. + Lepsius, op. cit., 169. } Chatard anal., Bull. U. 8. G. 8., No. 60, 1890, 158. S In Teller: Geol. Beob. Insel Chios, Denksschr. Akad. Wiss. Wien, xl, 347, 1880. || H. S. Washington: Jour. of Geology, iii, 158, 1895. Verrill and Bush—Genera of Ledide and Nuculide. 51 Art. V.—Revision of the Genera of Ledide and Nucu- lide of the Atlantic Coast of the United States; by A. E. VERRILL and KATHARINE J. BUSH. (Brief Contributions to Zoology from the Museum of Yale University, No. L.) A SOMEWHAT extended study of the series of deep-sea bivalves belonging to these families, dredged off our coast by the U. S. Fish Commission, from 1872 to 1887, has compelled us to revise the known genera and subgenera and to propose several new groups. In view of an unexpected delay in the publica- tion of the report upon these families, which had been com- pleted and fully illustrated, it seemed desirable to publish a brief preliminary account of the classification adopted. These families are often united by modern malacologists - under a single family (Nuculide), while others regard them as distinct. They are certainly closely related anatomically, as well as by the structure of the shell. Thus all the members of both families havea single pair of simple “ foliobranchiate ” (or protobranchiate) gills; two pairs of large labial palpi, the outer ones furnished with long extensile labial tentacles; a large muscular foot with an expanded, concave, terminal disk, adapted for rapid motions in jumping and swimming, as well as for creeping; and all have two series of transverse teeth on the hinge-margin. The peculiar structures of foot and gills appear together elsewhere only in the family Solemyide, which is evidently a related group, though it lacks hinge teeth. As these three families have gills of a peculiar and simple structure, each one consisting of two rows of flat lamelle, attached to a single stem, they have recently been regarded as forming a special order (Protobranchiata). This group is of special interest because of its great antiq- uity. Large numbers of fossil forms, very closely allied to existing genera and species, occur even in Silurian and Devon- ian formations. | Thus the common living genera Vucula and Leda are repre- sented by numerous Devonian species, many of which cannot be separated from the recent forms, even as subgenera, by any tangible characters. Other species of the same age, referred to Paleoneilo, agree in nearly all essential characters with the living genus Zindaria. These fossil shells are generally larger and stronger than the corresponding living species. Many paleeozoic genera which are now extinct were as highly organized and as much specialized as their living allies. The thin-shelled, strongly siphonate genera, such as Yoldia, Yoldiella, etc., do not appear so early in geological time and 52 Verrill and Bush— Genera of Ledide and Nuculide. may be regarded as more modern specializations of the Leda- like forms. They are also the forms that swim and jump with the greatest activity. Therefore the thin and light character of their shells may be regarded as having been secondarily acquired, partly in consequence of their active movements, in which a heavy shell would be disadvantageous, and partly at because the development of long siphons enables them to live Kh concealed, much of the time, beneath the surface of the soft | mud in which they generally live. In Solemya the shell is ml still hghter and thinner, in accordance with more developed Ml swimming habits, combined with burrowing when at rest. | Such forms as Wucula and Tindaria, which have no siphon vet tubes, must live at or near the surface of the mud over which vi they creep with their large expanded pedal disk. (Fig. 15.) ie These have for their protection comparatively solid shells, \ similar to those of palseozoic species, in form, texture, and si sculpture. | af The family Nuculidee differs from Ledidge mainly in having | no siphon tubes, the mantle edges being completely disunited. ipl The Ledide are remarkable for the great variations in the ull structure of the hinge-teeth, ligament, cartilage, and mantle, Mi as well as in the form of the shell. The pallial sinus may be a wanting or well-developed. Some genera have long united Wi siphons ( Yoldia) ; some have shorter ones, more or less sep- bh arated (Leda) ; while in Z%ndaria there is no true siphon, Hl but only an efferent orifice differentiated. The ligament may be wholly external, as in MMalletia, Tindaria, etc., or it may be rudimentary and replaced by an internal cartilage or “ resi- lum”; or both may coexist in varying degrees of development 7 and degeneration. The hinge-teeth may be very numerous | and regularly v-shaped, in each series, or they may be com- i paratively few and irregular, sometimes becoming oblique and ‘| lamelliform (S2dicula). The beaks generally turn backward ‘li (Yoldia, Leda, Nucula), but in Malletia, Tindaria, and some other genera, they turn forward. On this account, when there is neither pallial sinus nor external ligament, it is often diffi- | cult, if not impossible, to tell which is the anterior end of the shell, without the soft parts. Hence many fossil and some recent species have possibly been reversed in the descriptions. Thus many of the palseozoic species referred to Vucula are described as having the beaks turned forward, the larger end of the shell being considered posterior ; but in modern Vucuwla, the beaks turn backward, and the shorter end is posterior. Many of the deep sea species with small, thin shells show no distinct muscular nor pallial scars, which increases this diffi- culty. When a differentiated external ligament is present, we have assumed that it is posterior to the beaks (opisthodetic), Verrill and Bush— Genera of Ledide and Nuculide. 538 though a narrow extension usually runs under and forward of the beaks in a groove. When the shell of a dimyarian bivalve gapes posteriorly, the existence of a siphon may generally be assumed; for otherwise the internal soft parts would be exposed to enemies. The existence of a posterior rostrum or a protrusion of the posterior margin defined by an inferior emargination indicates the existence of a siphon, or at least an anal tube, but these organs may exist without such modifica- tions of the shell. If these rules be applied to paleeozoic forms we must conclude that the rostrate and subrostrate forms of Paleoneilo, ete., had some sort of a siphon, and therefore were not Nuculide, as now restricted. Family Nucurip#. We have included Nuculina (D’Orb.) in the Nuculide with some doubt, because authors differ as to its structure. It is said that its ligament is wholly external; if so, it should, perhaps, form a distinct subfamily. Its anatomy is unknown. Fischer places it in the Arcide, near Limopsis, but it has no median ligamental area. Subfamily GLomIn.&, nov. Ligament thick, elongated, attached for most of its length to the inner surface of the posterior hinge-plate and running forward in a narrow groove beneath the beaks, so that its anterior portion is external and the thickened posterior portion is partly internal. No pallial sinus. Animal not known. This group includes, so far as known, only the genus Glomus Jeffreys, which has been referred by several writers to the Arcide, and by others to the Ledidz, from both of which it differs widely. Its relations to the Nuculidz are somewhat uncertain, owing to our ignorance of the soft parts. In the form and position of the ligament it differs entirely from all other genera of Nuculide and Ledidee. : Glomus Jeffreys. Figures 1, 2. Glomus Jeffreys, Annals, Mag. Nat. Hist., p. 433, 1876; Proc. Zool. Soc. London, p. 573, pl. xlv, figs. 5, 5a, 1879; Smith, Report Voy. Challenger, Zodl., xiii, pp. 248-249, pl. xxi, figs. 1-30, 1885. Shell thin, smooth, sub-equilateral, rounded at both ends, with the beaks turned forward. No lunule nor escutcheon. Hinge with two series of obliquely transverse teeth; a small lateral tooth. The following are described species: G. nitens Jeff., G. Jeffreysi Smith, G. simplex Smith, G. ine- qguilateralis Smith, G. Japonicus Smith. 54. Verrill and Bush—Genera of Ledide and Nuculide. Family Lepipa. Subfamily Lepin az. Leda Schumacher, 1817. Figure19. Type Z. rostrata (Mont.). This genus has been variously extended and restricted by authors, and several subgeneric and sectional groups have been proposed. In the more extended sense it is scarcely capable of a definition that will distinguish it from Yoldia, ete. We propose, therefore, to restrict it to the typical species, such as L. cuspidata Gld., caudata Donovan (fig. 19), pernula (Mill.), tenuisulcata (Couth.) and many others, closely related. These have a long, tapered, bicarinate rostrum, and well-developed siphon tubes, partially united. The palpal tentacles are long, flat, tapered, and arise external to the bases of the outer palpi, which are broad, with slender, acute posterior tips. Ledella, gen. nov. Figures 13,18. Type Z. Messanensis (Seg.). Junonia Seguenza, Nuculidi terziarie merid. d’ Ital., R. Accad. Lineei, i, p. 1175, 1877 (not of Htibner). This group includes a large number of small species, both living and fossil, in which the shell is rather short, usually ovate or swollen, with a small, acute, or subacute, unicarinate rostrum, situ- ated medially or submedially, and defined below by an emargina- tion or undulation in the postero-ventral margin. The postero- dorsal margin is usually convex. The escutcheon or ligamental area is distinctly defined by the carina, but is not sunken. The chondrophore is usually small but distinct. The siphon tubes are separate, at least in some species. The following species appear to belong here: I. seminula (Seg.), L. Messanensis (Seg.) (fig. 13), L. Nicotroe (Seg.), L. peraffinis (Seg.), L. rectidorsata (Seg.), L. confusa (Seg.), Z. solidula Smith, L. semen Smith, L. confinis Smith, LL. inopinata Smitb, ZL. prolata Smith, L. ultima Smith, and L. parva V. & B.* Portlandia Morch. Type P. aretica (Gray), 1819=Leda FPort- landica (Hitch.). We consider this a distinct genus, but would restrict it to the original type. In many respects this genus is intermediate between Leda and Yoldia. In its closed shell, definite rostrum, * Ledella parva, sp. nov. Figure 18. Shell minute, smooth, narrow ovate, inequilateral, obtusely rounded anteriorly, slightly rostrate posteriorly with a slight postero-ventral emargination; the short rostrum subtruncate at tip and detined by an inconspicuous ridge. Umbos some- what swollen, beaks a little prominent and turned slightly backward. Hinge- plate strong with fifteen anterior and nine posterior teeth. Chondrophore rather small, triangular with a distinctly projecting inner margin. Length, 3™™; height, 2™™. Station 2689, off Martha’s Vineyard, in 525 fathoms, 1886. anal le in Ae Ali 4 Bi a “ Res, Verrill and Bush—Genera of Ledide and Nuculide. 55 etc., it agrees more nearly with the former, but in general outline, with the latter. Yoldia Moller. Figures 12, 16. Type Y. hyperborea Torrell. We propose to restrict this genus to the typical forms, such as limatula (Say) (fig. 12), sapotilla (Gld.) (fig. 16), myalis (Couth.) and many closely allied foreign species. These have a nearly smooth, compressed, lanceolate, gaping shell, more or less prolonged and tapered posteriorly, with a poorly defined wide rostrum, generally without carinations. The external ligament is marginal, feebly developed, continuous under the beaks, and not much differentiated from the general epidermis. The chondrophore is large, concave, and projects within the margin. The pallial sinus is large and deep. The siphon-tubes and posterior pallial tentacle are long. The palpal tentacles are long and tapered; in life they may extend nearly to the end of the expanded siphon. Orthoyoldia, gen. nov. Type Orthoyoldia scapina (Dall). Shell oblong, gaping, blunt or rounded at both ends, without distinct rostrum; no carina. FPallial sinus large and _ broad. ‘Teeth numerousin bothseries. QO. scapina (Dall), from off Brazil, and O. solenoides (Dall), from the West Indies. Megayoldia, gen. nov. Figure 17. Type IM. thracieformis (Storer). Shell large, compressed, gaping, with a very short, blunt, indefinite, postero-dorsal rostrum. » a grand stock of CRYSTALLIZED ORPIMENT, TOPAZ, HEMA- be. TITE PSEUDOMORPHS AFTER PYRITE, QUARTZ ENCLOS- ING TOURMALINE, all the rare COPPER AND IRON ARSEN- ATES from the TINTIC District, WURTZILITE, UINTAHITE, OZOCERITE, TIEMANNITE, VARISCITE, WARDITE, etc. 124 pp. ILLUSTRATED CATALOGUE, 25c. in paper, 50c. in cloth. % 44 pp. ILLUSTRATED PRICE:LIST, 4c.3 Bulletins and Circulars free. Be GEO. L. ENGLISH & CO., Mineralogists, ai _ 64 East 12th St., New York City. matrix, 50c. to $5.00. LAZULITE in good loose crystals, both single and — Art. I.—Worship of Meteorites; by H. A. ‘Newron.- Ep II.—Spectra of Argon; by J. TrowsripcE and Des RicHaRps II.—Some Queries on Rock Diteioutiation ; gi G. Fe BECKER 1V.—Igneous Rocks from Smyrna and Pergamon; by H. S. Ze Wasameton 2/2000 0 a V.—Revision of the Genera of Ledide and Nuculidz of the Atlantic Coast of the United States; by A. E. Verity and K. J. Busu VI —Experiment with Gold; by M. C.-Lra VII._—Note on a new Meteorite from the Sacramento Moun- — tains, New Mexico; by W. M. Foorn. (With Plates I~ and i) eee SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Absorption Spectra of Iodine and Bromine chal at Temperatures above the Critical Temperature of the Solvent, Woop: Boiling Points in a Crookes Vacuum, KRAFFT and WEILANDT, 67.—Formation of Per-— sulphuric Acid, KLBS and ScHONHERR, €8—Reaction of Silver oxide upon Hydrogen Peroxide. RIEGLER: Silver peroxynitrate, SuLC, MULDER and HERINGA, — 69..—New: Hydrocarbon, ScHICKLER: Fractional Distillation of acids of the Acetic series, SOREL. 70.—Ro6ntgen Rays, J. Macintyre: Rotation in Constant Electric Fields, QuINCKE: Interferential refractor for electric waves, O. Weipesure: Cadmium normal element, W. JAEGER and R. WaAcHSMUTH, 71. Geology and Mineralogy—Geological Survey of Canada, Annual Report, 1894: Pleistocene glaciation in New Brunswick, Nova Scotia, and Prince Edward Island, R CHALMERS, 72.—Notes sur la flore des couches permiennes de Trien- bach (Alsace), R ZeI_tuer, 74.— Artificial Production of the Mineral N orthupite, pE ScHULTEN, 75.—Genesis of the Talc deposits of St. Lawrence Co., N. Y., ‘ C H Smyru, JR: Handbook of Rocks for use without the Mie J. Be Kemp, 76. Botany and Zoology—lllustrated Flora of the Northern United States, Canana, ete, N. L. Brirron and A. Brown, 76.—Notes on the Flora of Newfoundland, B. L. RoBtnson and H. Von SCHRENK : Survival of the Unlike, L. H. BAILEY: Sphagna Boreali-Americana Exsiccata, D. C Eaton and H. Faxon, 77.— Analecta Algologica, Continuatio III, J. G. AGarpH: Phycotheca Boreali- Americana, F. S Couwtns, I. Houpen and W. A. SETCHELL: Ueber das Ver- halten der Kerne bei den Frichtentwickelung einiger fag ie R. a Harper, 78 —Gigantic Cephalopod on the Florida coast. : Miscellaneous Scientific Intelligence—History of Elementary Mathemuttes By Cagort. 79 —Kclipse Party in Africa, H. J. Loomis: Hlementary Meteorsbaae for High Schools and Colleges, F. Watpo, 80.—The Meteor of December 4, 81. Obituary—BENJAMIN APTHORP GOULD, 81. Catalogue of the Collection of Meteorites in the Peay Museum of Yale J University, 83. ae Chas. D. Walcott, U. S. Geol. Suey. ee ae FEBRUARY, 1897. Established by BENJAMIN SILLIMAN in 1818. ae 3 r é : - 5 rt 2 Pi Svc ae ax AS ia ‘. es nate o5 a > 9 a te a wi 2 Ly eer. | 4 E : Fe ae ‘rivalled by anything short of the Cumberland Calcites, and the lovely wine color adds greatly to the clear quality. 3. Amber or honey yellow crystals of an ‘‘ Iceland spar” degree of transparency ; less brilliant than the former but with interesting modi-- fications and etched rhombohedral termination. Size varies from 2 to 6 in. diam.—50c. to $4.00. Price does not indi- cate relative ‘value, as the new types are not to be compared with the choicest of older specimens—possibly they are 75 per cent. cheaper _ than formerly. Ruby Blende scattered over Galena and in association with Chalco- pyrite and Pearl Spar. , Elongated Galena. Chalcopyrite, large and perfect crystals on Blende and Pearl Spar. The handsomest combinations of the kind ever seen. A few of many other recent accessions : Crystallized Calaverite from the famous Cripple Creek region. Altaite, Geikielite, Bromyrite crystals. Descloizite in large distinct crystals with Psittacinite. Microcline resembling Labradorite. Sunstone of extraordinaryly fine quality. CUNY Beta A INO: An importation of selected specimens. ‘‘ Golden Shadow” is the name given to a new and pretty type of Phantom Barite. Lustrous Hematites with Smoky Quartz, making as handsome and cheap speci- — mens as can be found, Delicate blue Barite on snow white Calcite. Butterfly twin Calcites, Clear Groups, Aragonites. PRICES LOWER THAN EVER. GEMS AND PRECIOUS STONES. Choice cut stones ready for setting, including Opals, Garnets, Topaz, — etc. Petrified Wood, Agates and Crocidolite cut for paperweights and ornaments. Dr. A. E. FOOTE, WaRREN M. Foorr, Manager. Removed to 1817 ARCH STREET, (Two minutes walk from City Hall, three minutes from Penna. R. R.) PHILADELPHIA, PA., U.S. A. ESTABLISHED 1876. {MAS igD AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES. ] Oe Art. VIII.— Outline of a Natural Classification os the Trilobites ; by CHARLES E. BEEcHER. (With Plate III.) Introduction. Wir the possible exception of the barnacles, no group of arthropods has received more varied treatment by specialists than the trilobites. This taxonomic uncertainty has been due mainly to a lack of knowledge of the structure, and to certain real or fancied resemblances to Limulus. The early references of trilobites to the mollusks, insects, and fishes need not be noticed, for since they have been made the subject of special study, they have been commonly classed with the crustacea, and placed near the phyllopods by most observers. Quite a number of naturalists, however, still divorce the trilobites and lmuloids from the crustacea and ally them with the arachnids. It is not proposed at this time to discuss the homologies of Zemuius, but the trilobites show the clearest evidence of primitive crustacean affinities, in their protonau- plius larval form, their hypostoma and metastoma, the five pairs of cephalic appendages, the slender jointed antennules, the biramous character of all the other limbs and their original phyllopodiform structure. They differ from Lemulus, not only in most of these regards, but also-in not having an oper- eulum. From this and all other arthropods, they are distin- guished by having compound eyes on free cheek pieces which apparently represent the pleura of a head segment that is other- wise lost, except possibly in some forms of stalked eyes and in the cephalic neuromeres of later forms. The most recent dis- cussions as to the affinities of trilobites are to be found in the papers by Bernard,” * * *° Kingsley, Woodward, and the Am. Jour. Sci.—FourtH Series, Vou. III, No. 14.—Fepruary, 1897. 7 ~ eo Ve wes Fin mF ae “im, 4oo. =e. aE << ae F 90 Beecher—Natural Classification of the Trilobites. writer,’ where from the facts presented, their intimate relation- ships with the crustacea follow as a necessary corollary. Previous Classifications. The various schemes of classification that have been applied to the trilobites since that of Brongniart,"* in 1822, have been enlarged and revised by various authors, until, at the present time, no particular arrangement of the families or genera can be said to be endorsed. The one which is generally recognized as manifestly faulty, that of Barrande,’ is the one most com- monly found in text-books and special memoirs. Barrande’s definitions and limitations of the generic and family groups were natural and accurate, showing a most complete knowledge of generic and specific values, but in the grouping and arrange- ment of the families, he selected characters of secondary rank. Of all the investigators who have attempted any classifica- tion of the families, J. W. Salter® seems to have had the clearest insight into the important value of certain characters, and to have approached nearest to a natural system. In zo6- logical research, the study of ontogeny and the principles of morphogenesis were then scarcely recognized as having any. direct application. It is quite remarkable, therefore, that Salter, as early as 1864, should have singled out, as the basis of his subdivisions, the characters which are the dominant variants in ontogeny. It is not necessary in this place to discuss all the classifica- tions which have been proposed. Barrande® gives a complete resumé down.to 1850, and shows in a very satisfactory manner the weak points of each, furnishing strong reasons as to why they are unnatural and therefore untenable. The underlying principles of these early attempts at a classification are here briefly summarized. (1) The first classification of trilobites was advanced by Brongniart* in 182%, in which all the forms previously known as Entomolithus paradoaus were shown to belong to five distinct genera. (2) Dalman,” in 1826, made two groups based upon the presence or absence of eyes. (3) Quenstedt,” in 1837, recognized the number of thoracic seg- ments and the structure of the eyes as of the greatest impor- tance. (4) Milne-Edwards,”* in 1840, considered the power of | enrollinent as of prime value. (5) Goldfuss,”” in 1848, made three groups, depending on the presence or absence of eyes and their structure. (6) Burmeister,” in 1848, accepted the two divisions of Milne-Edwards, and laid stress on the nature of the pleura and the size of the pygidium. (7) Emmrich’s first scheme,” in 1839, was founded on the shape of the pleura, the presence or absence of eyes and their structure. (8) The later classification of the same author,"* published in 1844, was Beecher—Natural Classification of the Trilobites. 91 based on whether the abdomen was composed of fused or free segments, and the minor divisions depended chiefly on the structure of the eyes and the facial suture. (9) Corda,”* in 1847, placed all trilobites in two groups, one having an entire pygidial margin, and the other with the pygidium lobed or denticulate. (10) McCoy,” in 1849, took the presence or absence of a facet on the pleura for a divisional character. As this isan indication of the power or the inability of enrollment, it does not differ materially from the schemes of Milne-Edwards and Burmeister. Zittel,*° in a historial review brought down to 1885, includes in addition the schemes of Barrande and Salter, and remarks that the basis of Barrande’s general grouping, namely, the structure of the pleura, has neither a high physiological nor morphological meaning. Both Barrande and Salter recognize nearly the same families, with slight differentes, and the latter adopts a division into two lines, based on the number of body ringsand the size of the pygidium. Theseincludeand are them- selves included in four groups, founded on the presence and form of the facial suture and the structure of the eyes. Haeckel* has recently given the trilobites their full value in a classification of the articulates. Although he has not advanced a detailed classification, still it is desirable to review the ordinal groups which he proposes. He considers the Z77/o- dita as a legion under the first class, Aspzdonia, of the crus- tacea, which is characterized as being without a nauplius larval form and as having a pair of preoral antennee. In this class is also included the legion Merostomata, the Trilobita being especially distinguished by the number and character of the legs. The writer® believes that it is now satisfactorily demon- strated that the protaspis, or early larval form of the trilobite, is a protonauplius, and homologous with the nauplius of higher erustacea. Therefore, the Z7dlobita cannot reiqain in the Aspidonia as here defined. Haeckel further divides the Z7zlobita into two orders, the first, the Archiaspides (or Protrilobita), and the second, the Hutrilobita (or Pygidiata). The Archiaspides is represented by the families Olenida and Triarthrida, and is distinguished by the absence of a real pygidium, and by the complete homon- omy of the numerous body segments and their phyllopodi- form appendages. The families are themselves distinguished by the semicircular or crescent-shaped cephalon and by the presence or absence of genal spines. The Lwutrilobita is rep- resented by the families Asaphida and Calymenida, and is marked by the heteronomy of the body segments as expressed in the functional pygidium. * Systematische Phylogenie der wirbellosen Thiere (Invertebrata). Zweiter Theil. 1896. 92 Beecher—Natural Classification of the Trilobites. Salter, Burmeister, and Emmrich have, as previously noticed, attempted to use the comparative size and development of the pygidium for dividing the trilobites into groups larger than families, and it seems evident from the present state of knowl- edge that it is impossible to make this character of more than family or even generic value. Many of the genera which must naturally be included in the Avrchzasprda have pygidia that cannot be said to be rudimentary, obsolete, or wanting in function. Even those genera having pygidia with few seg- ments, as JMesonacis, Holmia, Paradoxides, Selenopeltis, Dicranurus, Bronteus, Harpes, ete., show in many other more important characters that they are highly differentiated and specialized forms and that this feature is one expression of such development. The futility of the scheme is at once evi- dent when a comparison is made between allied genera which present marked differences in the size and segmentation of the pygidium, as Phacops and Dalmanites, Ceraurus and Enert- nurus, Calymene and Homolanotus, Harpes and Trinucleus, Mesonacis and Lecanthoides, Paradouides and Dikelocephalus. The last classification to be noticed is that of E. J. Chapman,” in 1889, in which four suborders or primary groups are pro- posed, differing considerably from any previous arrangement, and based upon arbitrary features of general structure and con- figuration, especially the form of the glabella, whether wide, conical, or enlarged. Twenty-seven families are recognized. In this scheme, Zrimucleus, Ampyx, and diglina form one section; Paradoxides and Acidaspis, together with Phacops and Hncrinurus, another; all under one suborder. Omitting the Agnostide, there are here considered in a single suborder the most characteristic representatives of nearly all the types of trilobite structure. Proétus, Cyphaspis, and Arethusina fall into three sections, under two suborders, although these genera, on account of their great similarity in essential points, are placed in a single family by most authors. A further analysis of this classification in its broader lines would be unprofitable. It is sufficient to state that the facts obtained from the study of the ontogeny of any species are completely in discordance with these classifications,.and clearly’ demand other interpretations. Rank of the Trilobites. As to the rank of the trilobites in a classification of the erus- tacea, there is also much diversity of opinion. They have long been regarded as an order, but any attempt to include them in this way under higher groups, such as the Lntomostraca, Mal- acostraca, or Palwocarida, results in such broad generalities and looseness of definition as to render these divisions of little Beecher —Natural Classification of the Trilobites. 93 value. Even the EAntomostraca, as restricted to the orders Phyllopoda, Ostracoda, Copepoda, and Cirripedia, seems hete- rogeneous and probably polyphyletic. Milne-Edwards,” Gegen- baur,’’ Walcott,” and others have considered the trilobites as belonging toa class of arthropods intermediate between the erustacea and arachnids. Some recent authors, as Lang,” have attempted to overcome the difficulty by attaching them as an appendage (“Anhang”) to the crustacea. Kingsley,” on the other hand, has placed them as a subclass of the crustacea, leaving all the other crustacea to come under a second subclass, the Hucrustacea. The present state of knowledge of their structure and development is in favor of giving the trilobites the rank of a subclass, but for purposes of comparison and cor- relation, the fullest results can be brought out by recognizing the old and well-known subclasses,—the Entomostraca and Malacostraca. The following tabular view of the leading points of the com- parative morphology of the three subclasses is introduced to show, first, the claims of the Z77z/obita as an equivalent group, and, second, the progressive differentiation of characters. In nearly every particular the trilobite is very primitive, and closely agrees with the theoretical crustacean ancestor. Its affinities are with both the other subclasses, especially their lower orders, but its position is not intermediate. Comparative Morphology of Crustacea. Subelass I. Trilobita. Subclass Il. Entomostraca.|Subelass Ill. Malacostraca. 1. All marine. ‘Marine and fresh water. |Marine and fresh water. 2. SEEEC. ‘Free, parasitic, and at-|Free and parasitic. | tached. | . 3. Body longitudinally Various. Various. triregional. 4, Larva a protonauplius. Larva almost universally a)Larva generally a zoea, a nauplius. nauplius stage being | often developed before | | hatching, except in -Hu- phausia and Peneus. 5. Number of segments Number of segments vari- Definite number of seg- variable. | able. ments. 6. Cranidium of five Head of five fused seg--Head of five fused seg- fused segments. | ments to which, rarely, ments to which one or a thoracic segment is; more, or all of the tho- added. . racic segments may unite, forming a more or less | complete cephalothorax. 7, Ocelli rarely present. |Ocelli present throughout!Ocelli absent in adult forms. fe tlhe: 8. Paired compound ses- Paired compound eyes /Paired compound eyes sile eyes on cheek pieces. usually present; stalked) usually present; stalked usually present, | or sessile. Absent in| or sessile. adult Cirripedia and some Copepoda. 94 Beecher—Natural Classification of the Trilobites. Subclass I. Trilobita. 9. Thorax distinct; num- ber of segments variable, all free. 10. Abdomen distinct ; variable number of fused segments. 11. All segments of cra- nidium, thorax, and ab- domen, except the anal segment, appendages 12. All appendages bira- mous except antennules | 13. Appendages typically, Exop- phyllopodiform. odite a swimming leg; endopodite modified into a crawling leg. 14. All appendages of the head except antennules pediform. 15. Thoracic appendages ambulatory and swim-| ming. carry paired! Subclass Il. Entomostraca. Thorax with variable num- ber of segments. Abdomen with variable number of separate segments. Some segments appendages. without Some appendages are modified and have lost biramous structure. Appendages generally greatly changed in most orders; phyllopodiform in young forms. and | throughout life in Phyl- lopoda. Some appendages of the/S head modified into row- | ing organs, mandibles, / or suckers. Thoracic appendages am- bulatory, swimming, and | seizing. Subclass IIT. Malacostraca. Thorax with eight seg- ments, some of which are generally united with the head. Abdomen of seven gen- erally free segments; eight in Leptostraca. All segments usually carry appendages except the last one or two. Some appendages have lost biramous structure. A ppendages typically phyl- lopodiform, but greatly modified in all but the lowest order (Nebalia). Some appendages of the head modified into man- dibles, or organs for seizing food. Thoracic appendages am- bulatory, swimming, and seizing. 16. Abdominal limbs on Abdominal limbs gener-) Abdominal limbs often re- all segments except the anal, phyllopodiform. 17. Coxal elements of all limbs forming gnatho-| bases, which become manducatory organs on) the head. 18. Respiration cuticular) and by fringes on exop-! odites. ally wanting. | | Coxal elements seldom cept on the head. lar and by the limbs and gill appendages. Coxal forming gnathobases ex-| Respiration mainly cuticu- duced except the last pair, which with telson frequently form a caudal fin. Chiefly branchial in some groups. elements seldom forming gnathobases ex- cept on the head; never on the abdomen. Respiration cuticular and by the limbs and epipo- dites. The more primitive characters of the trilobites as drawn from the foregoing table may be summarized as follows: They are all free ma rine animals ; (1) (2) the animal has a definite configuration; (8) the larva is a protonauplius-like form; (4) the body and abdomen are richly segmented, and the number of segments is variable; (5) the head corresponds to the typi- cal crustacean ; (6) the thorax and abdomen are always distinct, the number of segments in each being variable; (7) all seg- ments except the anal bear paired appendages ; (8) all append- ages are typically phyllopodiferm ; and (10) the coxal elements of all limbs form gnathobases, which become organs of man- ducation on the head. Beecher—WNatural Classification of the Trilobites. 95 It may be questioned by some whether the present state of knowledge of the ventral structure of trilobites warrants such general assertions as to details of organization. In the first place, it must be granted that there is a remarkable uniformity in the features of the dorsal crust, which naturally reflects to a degree the differentiation and variation of the organs and appendages of the ventral side. Furthermore, the actual ap- pendages have been observed in such diverse and characteristic genera as Zrinucleus, Triarthrus, Asaphus, Ceraurus, and Calymene, and found to conform closely to a single type, so that it seems safe to assume a like agreement throughout. Morphology of the Cephaton. The structure of the trilobite head suggests homologies which should be noticed here, and if these correlations are based upon true structural likenesses, they serve not only to emphasize the primitive character of the trilobite, but also aid in interpreting certain organs and structures in the higher erustacea. The five fused somites of the crustacean head are generally believed to correspond to the third, fourth, fifth, sixth, and seventh neuromeres, leaving the first and second unrepresented either by distinct segments or appendages. These two neuro- meres commonly constitute most of the cerebral mass above the cesophagus, and enervate the ocelli and paired eyes. In some, the antenne are enervated from supra-cesophageal gan- glia, while in other forms their ganglia are infra-cesophageal. It was formerly supposed that the stalked eyes of the higher crustacea represented appendages of a head segment, but this belief has been abandoned on account of the derivation of stalked out of sessile organs, as in enews, and also because the eyes do not always have a relatively fixed position, but may pertain to the first, second, or third head segments. Their structural position in the trilobites, however, is invariable, and it seems probable that in some families of higher crustacea, the eyes are in exact correlation, and may be similarly interpreted. The writer’ has previously discussed this question, and ad- duced reasons for considering the free cheeks in trilobites as “the pleura of an occuliferous head segment.” In many species (Dalmanites, diglina, etc.), the free cheeks are con- tinuous, forming one piece extending around the front of the head, between the cranidium and the hypostoma, while in others there is a separate piece, the epistoma, between the proximal ends of the cheek pieces holding a like position. These structures occupy the exact position of a true segment, and since, upon theoretical grounds, additional head segments 96 Beecher—Natural Classification of the Trilobites. are to be accounted for, the only satisfactory correlation is to consider them as such. Furthermore, the free cheeks are dis- tinctly separated from the cranidium by an open suture, and may be wholly converted into eyes, as in _4glima armata Barr., or the unfacetted portion may be reduced to almost nothing, as in Dezphon. In such eases, the parallelism is exact with true movable eyes. Bernard’ concludes from his studies of Apus that the hypostoma is homologous with the annelid pro- stomium. This would make the hypostoma represent the first, and the free cheeks the second of the obsolete segments. Thus the trilobite cephalon would fulfil the demand for addi- tional evidences of primitive head segments, and account for the development of eyes separate from the cephalothorax as commonly restricted. Supposed evidences of free cheeks or of facial sutures have been recognized in Limulus, Hemiaspis, and Bunodes, but these seem really to correspond to the lines on the dorsal sur- face of the cephalon of Harpes and some Zrinucleus, ranning from the glabella to the eye spots and to the margin, and are not the sutures marking the limits of the free cephalic ele- ments, as in Asaphus and Proétus. Limulus, however, has a suture comparable to that in Harpes and Trinucleus, extending around the ventral border of the cephalothorax nearly to the posterior angles, and partly separating the ventral plate. In the process of moulting, this suture opens and enables the ani- mal to free itself from its former test. These interpretations may be employed to some advantage in correlating the segmentation of the trilobite cephalon. As previously stated, the recognized plan in the nervous system of a generalized crustacean requires that there should be a brain or supra-cesophageal ganglion enervating (a) the unpaired eye, (b) the frontal sensory organs and stalked eyes, and (c) the anterior antenne; then a ventral nervous chord consisting of a succession of double ganglia enervating, respectively, the second pair of antennee, the mandibles, the first pair of max- illees, the second pair of maxillee, and lastly each of the paired thoracic and abdominal appendages. Altogether, there are seven neuromeres pertaining to the head, and on the basis that each neuromere corresponds to an original segment, as on the post-cephalic region, there would need to be this number accounted for. The anterior sezment, or number one in the trilobites, would be represented by the hypostoma; the second segment, by the paired eyes, free cheeks, and epistoma; the third, by the anterior lobe of the glabella and the first antennee ; the fourth, by the second lobe of the glabella and the second pair of antennee; the fifth, by the third lobe of the glabella and the mandibles; the sixth, by the fourth lobe of the giabella Beecher— Natural Classification of the Trilobites. 97 and the first maxilla; and the seventh, by the neck lobe, or occipital ring, and the second pair of maxille. The five annu- lations, or lobes, of the axis of the cranidium, since they pri- marily carry fulcra for the attachment of muscles supporting or moving the appendages, could thus be interpreted in terms of the ventral structure, making the first lobe the antennulary, the second the antennary, the third the mandibular, the fourth the first maxillary, and the fifth the second maxillary. No other group of crustacea furnishes such constant and well- developed structures’ representing the second theoretical head segment, which is obscure or obsolete in all the living groups, excepting probably the stalked eyes of some crustacea and the movable occular segment of the Stomatopods. Jor this reason, in addition to the many other important differences previously noted in the table of comparative morphology, the trilobites are regarded as a subclass, and the relative denomination and structural relations of this second segment, along with other characters, are considered as of sufficient physiological and morphological importance to determine the ordinal divisions. Principles of a Natural Classification. Most satisfactory and conclusive results have already come from the application of the law of morphogenesis, or the recapitulation theory, to various groups of animals, by means of which their natural classification and phylogenetic relations have been determined. Hyatt” says on this point (1889): “We have endeavored to demonstrate that a natural classification may be made by means of a system of analysis in which the individual is the unit of comparison, because its life in all its phases, morphological and physiological, healthy or pathologi- cal, embryo, larva, adolescent, and old (ontogeny), correlates with the morphological and physiological history of the group to which it belongs (phylogeny). It is also interesting to note that Agassiz’ recognized in ontogeny a standard of classi- fication. One of his strongest statements is as follows: “ Em- bryology [ontogeny] will in the end furnish us with the means of recognizing the true affinities among all animals, and of ascertaining their relative standing and normal position in their respective classes with the utmost degree of accuracy and precision.” These principles can be best applied in a group of animals which has a geological history more or less complete, and which is not wholly parasitic or greatly degenerated. It is of the greatest importance, also, to study the ontogeny of primi- tive and non-specialized species, because without very complete paleontological evidence, the development of a much later 98 Beecher—Natural Classification of the Trilobites. derived form may be so involved with larval adaptations and accelerated characters as to be misleading. The trilobites lend themselves to this treatment in fulfilling most of the necessary conditions. They have a known geo- logical history stretching through the entire paleozoic, from the beginning of the Cambrian to the Permian. ‘Their struc- ture is generalized and quite uniform, and no sessile, attached, parasitic, land or fresh water species are known. The ontog- eny of all the principal groups has been studied, including Cambrian, Ordovician, Silurian, and Devonian types. The trilobites necessarily furnish little information of the © stages of growth which may be classed as embryonic. The early embryonic stages are not preserved as fossils, and there- fore may be omitted. In this category are the protembryo, or the ovum in its unsegmented and segmented stages (the so- called “eggs of trilobites” may of course represent any stage of embryonic development before the escape of the young); the mesembryo, or blastosphere; the metembryo, or gastrula; the neoembryo, or planula-like stage; and the typembryo, when the first distinctive features make their appearance. The first embryonic stage recognized in the trilobites can be referred to the phylembryo as defined by Jackson,” when the animal may be clearly referred to its proper class. Since this period is distinctive for each class of animals and usually bears a sepa- rate name, it has been termed by the writer’ the protaspis stage of trilobites. It closely approximates the protonauplius form, or the theoretical, primitive, ancestral larval form of the crustacea. Like the homologous nauplins of modern higher erustacea, it is the characteristic larval type common to the class. The nauplius is therefore considered as a derived larva modified by adaptation. The post-embryonic stages of ontogeny have received the names nepionic, for the infantile or young; neanic, for the immature or adolescent; ephebic, for the mature or adult; and gerontic, for the senile or old. When especially applied to trilobites, the nepionic stages may include the animal when the cephalon and pygidium are distinct and the thorax incom- plete. There would thus be as many nepionie stages as the number of thoracic segments. The neanic stages would be represented by the animal with all parts complete, but with the average growth incomplete. Final progressive growth and development of the individual would fall under the ephebic stage. Lastly, general evidences of senility would be interpreted as belonging to the gerontie stage. . Beecher— Natural Classification of the Trilobites. 99 Application of Principles for Ordinal Divisions. In other classes of animals above the lower ccelenterates, the phylembryonic stage is the starting point from which correla- tions are made, and out of which all the higher groups are developed by a series of changes along certain lines. The protoconch represents this period in the cephalopods and gas- tropods; the prodissoconch, in the pelecypods; the proteg- ulum, in the brachiopods, and the protechinus, in the echinoids. In the trilobites, the protaspis, as already stated, has the value of the phylembryo, and in its geological history and the metamorphoses it undergoes to produce the perfect trilobite, accurate information can be gained as to what the primitive characters are, and the relative values of other features acquired during the long existence of the class. The simple characters possessed by the protaspis are the fol- lowing, as drawn from the study of this stage in all the prinei- pal groups of trilobites: Dorsal shield minute, not more than 4 to 1™™ in length; circular or ovate in form; axis distinct, more or less strongly annulated, limited by longitudinal grooves; head portion predominating; axis of cranidium with five annulations ; abdominal] portion usually less than one-third the length of the shield; axis with from one to several annu- lations; pleural portion smooth or grooved; eyes, when pres- ent, anterior, marginal, or submarginal; free cheeks, when visible, narrow and marginal. Examples, Plate ILI, figs. 1, 5. During this stage, several moults took place before the com- plete separation of the pygidium or the introduction of thoracic segments. These brought about various changes, as the stronger annulation of the axis, the appearance of the free cheeks on the dorsal side, and the growth of the pygidium by the introduction of new appendages and segments, as indicated by the additional grooves on the axis and limb. A full repre- sentation of the variety and succession of these early protaspis stages is presented in the writer’s paper on the “ Larval Stages of Trilobites.”°® Some of the conclusions and discussions in that paper are made use of here. In the earliest or Cambrian genera, the protaspis stage is by far the simplest expression of this period to be found. In the higher and later genera, the process of acceleration or earlier inheritance has pushed forward certain characters until they appear in the protaspis, thus making it more and more complex. Taking the early protaspis stages in Solenopleura, Liostra- cus, or Ptychoparia, it is found that they agree exactly with the foregoing diagnosis in its most elementary sense. Since they are the characters shared in common by all the larve at this stage, they are taken as primitive and accorded that value 100 LBeecher-——Natural Classification of the Trilobites. in dealing with adult forms possessing homologous features. Therefore, any trilobite with a large elongate cephalon, eyes rudimentary or absent, free cheeks ventral or marginal, and glabella long, cylindrical, and with five annulations, would naturally be placed near the beginning of any genetic series or as belonging to a very primitive stock. Next must be considered the progressive addition of char- acters during the geological history of the protaspis, and in the ontogeny of the individual during its growth from the larval to the mature condition. It was shown in the paper already referred to, that there was an exact correlation to be made between the geological and zodlogical succession of first. larval stages and adult forms, and therefore both may be reviewed together. The first important structures not especially noticeable in all stages of the protaspis are the free cheeks, which usually man- ifest themselves in the meta- or para-protaspis stages, though sometimes even later. Since they bear the visual areas of the eyes, when they are present, their appearance on the dorsal shield is practically simultaneous with these organs, and before the eyes have travelled over the margin, the free cheeks must be wholly ventral in position. When first discernible, they are very narrow, and in Ptychoparza and Sao include the genal angles. In Dalmanites and Cheirurus, however, the genal angles are borne on the fixed cheeks. If, as Bernard’ concludes, the crustacean head has been formed by the bending under, to the ventral side, of the anterior segments of an ancestral carnivorous annelid, this furnishes a means of further determining and also of satisfactorily correlating the prime significance and importance of the free cheeks. Since the free cheeks are ventral in the earliest larval stages of all but the highest trilobites, and as this is an adult feature among a number of genera which on other grounds are very primitive, this is taken as generally indicative of a very low rank. It seems to mean that the second segment remains where it was mechanically placed, and retains its full somitie nature, though from the necessities of such a condition, true ventral segments must soon disappear through modification into other structures or through disuse as segments. The genera Harpes, Agnostus, Trinucleus, and their allies agree in having well-developed, continuous, ventral free cheeks, and constitute a natural group. As they possess one expression or type of the genesis of an important common character, based upon facts of development, it should stand as an ordinal char- acter, and as such it is here taken. For this group, the name Hypoparva is proposed. It is fully defined, and its limitations established in the proper place in the classification. Beecher—Natural Classification of the Trilobites. 101 The remaining genera of trilobites present two distinct types of head structure, dependent upon the extent and character of the free cheeks. In both, the free cheeks make up an essential part of the dorsal crust of the cephalon, being continued on the ventral side only as a doublure or infolding of the edge, ‘similar to that of the free edge of the cranidium, the ends of the thoracic pleura, and the margin of the pygidium. They may be separated only by the cranidium, as in Ptychoparia, or by the cranidium and epistoma, as in /lle@nus and Homalono- tus, or they may be united and continuous in front, as in Ayglina and Dalmanites. One type of structure is distin- guished by having the free cheeks include the genal angles, thus cutting off more or less of the pleura of the occipital segment. The genera belonging to this group constitute the second order, the Opesthoparia. The third and last type of structure includes forms in which the pleura of the occipital segment extend the full width of the base of the cephalon, embracing the genal angles. The free cheeks are therefore separated from the cranidium by sutures cutting the lateral margins of the cephalon in front of the genal angles. Genera having this structure are here placed in the order Proparia. Several genera, as Calymene and Triarthrus, have been described as having the facial sutures beginning at or cutting the apex of the genal angle, thus making it indeterminate whether they should be classed with the Opisthoparia or Pro- paria. It will be found, however, that some species of these genera leave no doubt as to the anterior or posterior position of the suture. The small genal spines of Calymene calli- cephala Green are situated on the ends of the fixed cheeks, while similar but larger spines in Z7riarthrus spmmosus Billings are on the free cheeks, making the former belong to the Proparia and the latter to the Opisthoparia. Application of Principles for Arrangement of Families and Genera. The remaining characters to be noticed have chiefly family and generic values, and naturally follow the preceding discus- sions. ‘They are of great assistance, both in determining the place of a family in an order, and the rank and genetic position of a genus in a family. © There is very satisfactory evidence that the eyes have migrated from the ventral side, first forward to the margin and then backward over the cephalon to their adult position. The most primitive larvee should therefore present no evidence of eyes on the dorsal shield. Just such conditions are fulfilled 102 Beecher—Natural Classification of the Trilobites. in the youngest larva of Ptychoparia, Solenopleura, and Lios- tracus. ‘The eye line is present in the later larval and adoles- cent stages of these genera, and persists to the adult condition. In Sao, it has been pushed forward to the earliest protaspis, and is also found in the two known larval stages of Zriarthrus. Sao retains the eye line throughout life, but in 77rzarthrus the’ adult has no traces of it. A study of the genera of trilobites shows that this is a very archaic feature, chiefly characteristic of Cambrian genera, and only appearing in the primitive genera of higher and later groups or as an evidence of degen- eration. It first develops in the later larval stages of certain genera (Ptychoparia, ete.); next in the early larval stages (Sao); then disappears from the adult stages (Zriarthrus); and finally is pushed out of the ontogeny (Dalmanites). In Ptychoparia, Solenopleura, Liostracus, Sao, and Triar- thrus, the eyes are first visible on the margin of the dorsal shield after the protaspis stages have been passed through, and later than the appearance of the eye lines; but in Proétus, Acidaspis, Arges, and Dalmanites, through acceleration, they are present in all the protaspis stages, and persist to the mature or ephebie condition, moving in trom the margin to near the sides of the glabella. Progression in these characters may be expressed, and in so far taken for general application among adult forms to indicate rank, as follows: (1) absence of eyes; (2) eye lines; (8) eye lines and marginal eyes; (4) marginal eyes; (5) submarginal eyes; (6) eyes near the pleura of the neck segment. : The changes in the glabella are equally important and inter- esting. Throughout the larval stages, the axis of the cranid- ium shows distinctly by the annulations that it is composed of five fused segments, indicating the presence of as many paired appendages on the ventral side. In its simplest and most primitive state, it expands in front, joining and forming the anterior margin of the head (larval Pétychoparia and Sao). During later growth, it becomes rounded in front, and termi- nates within the margin. In higher genera, through accelera- tion, it is rounded and well-defined in front, even in the earliest larval stages, and often ends within the margin (larval Z77ar- thrus and Acidaspis). From these simple types of simple pentamerous glabellee, all the diverse forms among species ot various genera have been derived, through changes affecting any or all the lobes. The modifications usually consist in the progressive obsolescence of the anterior annulations, finally pro- ducing a smooth glabella, as in Zllenws and Niobe. . The neck segment is the most persistent of all, and is rarely ob- scured. The third, or mandibular, segment is frequently marked by two entirely separate lateral lobes, as in Aczdaspos, Beecher—Natural Classification of the Trilobites. 103 Conolichas, Chasmops, etc. Likewise, the fourth annulation carrying the first pair of maxillee is often similarly modified in the same genera, also in all the Proétide, and in Cheirurus, Crotalocephalus, Spherexochus, Ampyx, Harpes, etc. Here again, among adult forms, the stages of progressive differentia- tion may be taken as indicating the relative rank of the genera. The comparative areal growth of the free cheeks is expressed by the gradual moving of the facial suture toward the axis. As the free cheeks become larger, the fixed cheeks become smaller. In the most primitive protaspis stages, and in Agnos- tus, Llarpes, and Trinucleus, the dorsal surface of the cepha- lon is wholly occupied by the axis and the fixed cheeks, while in the higher genera, the area of the fixed cheeks becomes reduced until, as in Stygina and Phillipsia, they form a mere - border to the glabella. Therefore, the ratio between the fixed and free cheeks furnishes another means of assisting in the determination of rank. The pleura from the segments of the glabella are occasion- ally visible, as in the young of Qlenel/us, but usually the pleura of the neck segments are the first and only ones to be distinguished on the cephalon, the others being so completely coalesced as to lose all traces of their individuality. The pleura of the pygidinm appear soon after the earliest protaspis stage, and in some genera (Sao, Dalmanites) are even more strongly marked than in the adult state and much resemble separate segments. The growth of the pygidium is very con- siderable through the protaspis stage. At first, it is less than one-third the length of the dorsal shield, but by successive addition of segments, it soon becomes nearly one-half as long. In some genera, it is completed before the appearance of the free thoracic segments, all of which are added during the nepionic stages. An interpretation of these facts, to apply in valuing adult characters, would indicate that a very few seg- ments, both in thorax and pygidinm, may be evidence of arrested development or degeneration. On the other hand, the apparently unlimited multiplication of thoracic and espe- cially of abdominal segments in some genera is also to be con- sidered as a primitive character expressive of an annelidan style of growth. Genera like Asaphus, Phacops, etc., having a constant number of thoracic segments accompanied by other characters of a high order, undoubtedly represent the normal trilobite type. These analyses and correlations clearly show that there are characters appearing in the adults of later and higher genera, which successively make their appearance in the protaspis stage, sometimes to the exclusion or modification of structures present in the most primitive larve. Thus, the larvee of Dal- 4 104 Beecher—Natural Classification of the Trilobites. manites or Proétus, with their prominent eyes and glabella — distinctly terminated and rounded in front, have characters which do not appear in the larval stages of ancient genera, but which may come in their adult stages. Evidently such modifications have been acquired by the action of the law of earlier inheritance, or tachygenesis. In a classification of trilobites, for the purpose of illustrat- ing the principles here enunciated, the ontogenies of Sao and Dalmanites, Plate III, figs. 1-8, are selected. Sao belongs to the ancient family Olenide of the order Opisthoparia, and naturally may be expected to furnish very clear evidence as to the relations of many lower and older genera. Dalmanites, also, with its simple head structure, will give similar data regarding the Proparia. The early protaspis stage of Sao, Plate III, fig. 1, has no dorsal development of the free cheeks, and with the elongate form of the cephalic portion may be compared with the cephala of Agnostus and Jficrodiscus, and therefore correlates with the Hypoparia. The cephalon, at a later period of devel- opment, when the animal has two free thoracic segments, Plate III, fig, 2, shows the narrow marginal free cheeks and distinet eye lines. Here the resemblance to the cephala of Atops and Conocoryphe, Plate ILI, figs. 14, 15, is very marked, and indi- cates that the Conocoryphide is genetically the first family of the Opisthoparia. When Sao has eight thoracic segments, Plate ILI, fig. 3, the characters of the cephalon accord closely with Ptycho- paria and Olenus, showing that these genera should precede it in arranging the genera of the family Olenide. Evidence is thus furnished for the proper position of the first two families of the order. Now, if the relative values of the differentia- tion of the glabella, the position of the eyes, and the size of the free cheeks are considered in the light of the preceding analyses of these features, the remaining families of the order, as represented in their typical genera, naturally arrange them- selves as indicated in Plate III, figs. 18-23. There results (1) the Conocoryphide (represented by Atops and Conocoryphe, figs. 14, 15); (2) the Olenidee (Ptychoparia and Olenus, figs. 16, 17); (8) the Asaphidee (Asaphus and Jilenus, figs. 18, 19); (4) the Proétidee (Proétus, fig. 20); (5) the Bronteide (Lron- tews, fig. 21); (6) the Lichadide (Lichas, fig. 22); and (7) the Acidaspidee (Aczdaspis, fig. 23). For the Proparia, similar results are brought out by the study of the ontogeny of Dalmanites, and by comparisons with the characters governing the sequence of families in the Opisthoparia. The narrow marginal free cheeks place the Encrinuridee and Calymenide as primitive. The small or obsolete eyes and the larval form of the glabella in the former Beecher—WNatural Classification of the Tritobites. 105 further show that this family should be placed at the beginning. The nepionic Dalmanites, with seven thoracic segments, has a head structure very similar to the adult Cheirurus (Hecopto- cheile), fig. 28, thus making the Cheiruride precede the Pha- copide. The arrangement of families under the Proparia accordingly will be (1) the Encrinuride (Placoparia and Encrinurus, Plate Ill, figs. 24, 25); (2) the Calymenidee (Calymene and Dipleura, figs. 26, 27); (8) the Cheiruridee {Cheirurus (Eccoptochedle), fig. 28); and (4) the Phacopide (Dalmanites, Chasmops, Acaste, Phacops, figs. 29-88). The sequence of families in the most primitive order, Hypo- paria, may now be easily disposed of. The genera are so aber- rant and offer such conspicuous differences from ordinary trilobites, that it was considered better to delay their disposi- tion until the variations in structure governing the arrange- - ment of families in the higher orders were clearly shown. The degree of specialization of the glabella, of the form and character of the fixed cheeks, and the great range in the num- ber of segments in the thorax and pygidium are strong evi- dence that we are dealing with the terminal genera of the order which must have attained its-normal development in pre- Cambrian times. Agnostus and Aficrodiscus have so many protaspidian and larval characters that they must be considered more primitive than the other genera, although in some respects they show a high degree of specialization and even degeneration, as will be noticed under the family Agnostide. Moreover, Harpes, in its elongate cephalon, persistent ocelh, and many thoracic segments, is also quite primitive. Z7rinu- cleus, with ocelli only present in larval stages, a transverse cephalon, and genal spines belonging to the free cheeks, is con- siderably higher and properly comes last in the order, thus making the arrangement of families as follows: (1) Agnos- tide (Agnostus, Microdiscus, Plate ILI, figs. 9, 10); (2) Har- pedidee (//arpes, fig. 11); and (8) Trinucleidee (Zrinucleus, Ampyzx, figs. 12, 13). Diagnoses and Discussions. Subclass Tritopira. Marine crustacea, with a variable number of metameres ; body covered with a hard dorsal shell or crust, longitudinally trilobate from the defined axis and pleura; head, thorax, and abdomen distinct. Head covered with a cephalic shield com- posed of a primitively pentamerous middle piece, the cra- nidium, and two side pieces, or free cheeks, which may be Am. JOUR. Ee eck SERIES, VOL. III, No. 14 —FrEBRuARY, 1897 106 Beecher—Natural Classification of the Trilobites. separate or united in front, and carry the compound sessile eyes, when present; ‘cephalic appendages pediform, consisting of five pairs of limbs, all biramous, and functioning as ambu- latory and oral organs, except the simple antennules, which are purely sensory. Upper lip forming a well-developed hypos- toma; under lip present. Somites of the thorax movable upon one another, varying in number from two to twenty-nine. Abdominal segments variable in number, and fused to form a caudal shield. All segments, thoracic and abdominal, carry a pair of jointed biramous limbs. All limbs have their coxal elements forming gnathobases, which become organs of mandu- cation on the head. Respiration integumental and by branchial fringes on the exopodites. Development proceeding from a protonauplius form, by the progressive addition of segments at successive moults. , Heretofore it has been impossible to give an adequate diag- nosis of the Z7rzlobita, owing to the absence of information regarding certain important characters, and the obscurity of. the information relating to some other features. It is believed that enough is‘now known to frame a definition of the class, which, in accuracy and completeness, will compare favorably with any based upon living groups. Such a definition brings out the fact that the differences between the trilobites and other large groups are clearly recognizable, and do not consist of a statement of anomalous characters whose real significance is unknown. [To be continued. ] Barus—Trial of Interferential Induction balance. 107 Art. 1X.—Preliminary Trial of an Interferential Induc- tion balance; by O. Barus. 1. THE following device is capable, I believe, of a variety of applications in relation to alternating currents and to mag- netic induction. The idea underlying the apparatus is briefly this: Let the slender iron cores of two identical helices be placed at right angles to each other, in the same (horizontal) plane and at like distances from the point of convergence. Let the distant ends of the iron cores be rigidly fastened, while the other ends are free to move (expand and contract) in the direction of the axes. Then it is possible to adapt Michelson’s interferential refractor in such a way that the fringes are visible whenever the excursions of the free ends of the cores are either zero or vibrating in the same phase, ampli- tude and period to and from the point of convergence. The fringes vanish more or less fully for all other phases. Let the two helices be traversed by the same alternating current, and let the reduced length of the electrical conductor between them be negligible and the vibrations in the same phase initially. Then if the reduced length of the conductor be gradually increased, the two vibrations will be dephased and the fringes will gradually vanish. If the lengthening of the conductor be continued until the excitement of one helix is belated one complete period of the vibrating core with ref- erence to the other, the fringes will reappear caet. par. with their original clearness. A method of estimating the speed of signalling from one helix to the other is thus given in terms of the period of longitudinal vibration of the vibrating core. If the helices are not identical, similar deductions apply for the difference of time lag of current behind the electromotive force for the two helices respectively. Finally vanished fringes are open to observation by strobo- scopic methods. §2.-The form of apparatus used is shown diagrammatic- ally and largely in sectional plan in the figure. A and B are the two helices and @ and 6 their cores of soft Norway iron. ‘These with the helices are rigidly fixed at the further ends. The cores pass axially through the helices without touching them, and carry small light plane mirrors m and 7 at their free ends. With the thick plane-parallel plate J/, these mirrors make up Michelson’s refractometer. Light (thus far furnished by a sodium flame) is emitted by the distant burner £’ provided with Cushman’s sodium collar,* and condensed by *This efficient device, due to Holbrook Cushman, consists of a layer of filter paper impregnated with sodium carbonate and wrapped around the top of the burner. The sodium flame so obtained is. very dense and lasts for days. 108 Barus—Trial of Interferential Induction balance. the train of lenses e, f,c. As a rule the flame with a single lens would be sufficient and the train is here given to admit of the introduction of the stroboscopic disc G when an inter- mittent beam is wanted. Other small fixed screens are often advantageous. Falling on J/ (preferably covered with a trans- parent coat of silver), the light undergoes interferential refrac- tion in the now well-known way. The optical parts of the apparatus are fully described in Michelson’s* great memoir, and further reference here is needless. To obtain the inter- ference it is usually sufficient (good glass presupposed) to bring the images of the same small distant luminous point into coin- dence. Removing 7 and ¢, a small hole in G answers the pur- poses. If the helix A and its mirror is fixed, the helix & and its mirror must be movable in azimuth and altitude. I found the large round table of an old Kirchhoff spectrometer excel- lent for this purpose. The helices A and 6 were mounted in place of the telescopes, and the mirror J/ on a small adjustable tripod could then be moved on the flat table into position. If the glass is not very good, as is apt to be the case with thin mirrors, m and n, and ordinary plate glass, it is necessary to find the parts of the plates sufficiently plane, by trial. In general it is best to bring the edges of the mirrors to overlap, in which case the interference figure is of lune-shaped outline and relatively clear. The adjustments may be made with the naked eye. For observations, however, the small telescope 7, * Michelson: Valeur du Métre, Trav. Bureau Int., xi, 1894. Barus—Trial of Interferential Induction balance. 109 provided with cross-hairs, must be so mounted as to be easily focussed, moved about and inclined. Whenever a current passes through either of the helices A and & the cores a, 6 will in general expand, unless the field be very intense. Current for this purpose is furnished by one or more storage cells e, passes through a commutator OC and a switch board Y, by which either one or both the helices may be put into the circuit. The circuit also contains inductive resistance //, non-inductive resistance / and an interrupter Z for intermittent currents. A is a key. As the magnetic expansion due to complete saturation is only a few mil- lionths, the effects of thermal expansion must be excluded. Hence the frame of each helix is a thin shell-like water- jacket. Water arrives at w,-passes in a thin sheet between the core and the wire of helix A, then similarly through helix £& and leaves at w’. The same temperature is thas obtained for both helices and strong electrical currents passed through them are without thermal effect on the core. The two helices were wound identically and chiefly for con- tinuous currents. They each consisted of eight layers of No. 21 copper wire with 410 turns in each layer, and they were 37™ long. The diameter of the internal canal was about 0°8™, and the outer wall of the water-jacket on which the wire was wrapped about 15°". In view of the long helices the enclosed iron cores (about 28°" long, 0:6 in diameter) were in a fairly constant field.* Their rear ends were soldered into a rod of brass snugly fitting the box of the helix and rigidly fastened by screws. The front ends were soldered to very thin copper tubes which protruded beyond the helix and served as holders for the mirrors m, n. The cores were made as straight as possible, adjusted symmetrically and coaxially with the helix, leaving a clear space of about i™™ between core and nee The fastening of the cores is merely suggested in the ure. eWiien mounted the apparatus must necessarily be placed on a pier or other firm support ; but even thisis insufficient unless the vibrations are damped at different parts of the apparatus. Thus it is necessary to support the helices from two points near their ends, and independently to prevent transverse vibra- tions of the core. After many trials [ found that a short plug of vasilene in the tubes near the ends m, m was preferable to mechanical devices. Prof. Mayer’s important paper should be consulted on all these points. § 38. If both mirrors move towards the point of convergence symmetrically and equally, the same number of wave lengths * Mayer: Phil. Mag. [4], xlvi, p. 177, 1873. 110 Barus—Trial of Interferential Induction balance. are cut off by both and the interference fringes will not be dis- placed however fast the mirrors may vibrate. If an appreciable time is spent by the effective current in passing from helix A to helix 4, then with increasing retardation the fringes will gradually vanish and gradually reappear when the time of delay has reached the period of vibration of the mirrors. Let the motion of the mirrors m, 2 be given by a, sin wt and a, sin (wt+ 0). Then the motion of the fringes is proportional to : a’ sin(wt— 6"); where w =/a,?+a,?—2a,a, cos 6 and tan 6' = a, sin 6/ (4, +4, cos 6). If the amplitudes are equal a’ = 2asin 0/2 (1) If the self-inductions Z, and Z, of the two helices are not equal, if, for instance, the cores are chosen of different diame- ters, the currents in the helices will lag behind the correspond- ing electromotive forces by different amounts. The result will bring the interference fringes to vanish more and more. fully at first, and finally less and less so, as the difference of self- induction increases, till the dephasing approaches one com- plete period. If 0, and @, be the current lag in each helix and if their ohmic resistance # is the same, then Le) R= Lai een) carer . Pi iby Jb, or tan 6 = w Lis cs Ss As (2) ®, yy There are thus two equations, (1) and (2), for @. If LZ, be made very small tan@=o@Z,//. Similar results may be obtained for different amplitudes a, and a,, and different values for fr. Equation (1), however, may be supposed to include the retardation due to the external resistances as well as the lag in equation (2). From this wider point of view the fringes will be at vest if a’ =0 or if Gis an even multiple of 7. The motion of the interference fringes will be a maximum or they will vibrate to and fro with greatest intensity when @ is an odd multiple of 7. The criterion is, therefore, visibility and clearness of the fringes, and the method of observation is not continuous. §4. With the object of obtaining a continuous series of data the above stroboscopic disc G was introduced. It is made of very thin tin plate about 40™ in diameter, with a sufficient Barus—Trial of Interferential Induction balance. 111 number of equidistant small holes cut near the margin. Mounted on an axle and rotated by a small electromotor, it was supposed that the speed could be regulated nearly enough by inserting resistances. I did not have any success with this dise. Though I was able to increase the flashes to 3000 per second, they were never instantaneous enough for the work, even when the holes were reduced to an extreme of smallness. It ought not to be difficult, however, to reach the end in view by replacing the sodium flame /’ by a hydrogen Geissler tube. The light would have to be analyzed prismatically and the period of intermittance suitably chosen. Michelson* has shown that the lines of the hydrogen spectrum are not well adapted for refined refractometer work. In the present case the number of fringes in question is only a very few (2 or 3), so that I do not think there will be difficulty for this reason. 5. The apparatus being differential in type, is particularly adapted for observing the Joule expansion phenomenon, its variation with the field, with temperature and dimensions of the core, etc. The results of the table are random examples among many data obtained with continuous currents. The first column shows the field within the helix, the second and third columns the expansions in millionths of the length of core for each helix separately, and the fourth column the corre- sponding results when both helices were in circuit. When results differed the mean value was taken. Since the motion _-of one fringe across the cross-hairs of the telescope corresponds to about 30/10°™ (in view of reflection) and since the iron cores are about 30 long, the elongation is about 10~° per fringe. Now the motion of less than one-tenth fringe is discernable ; hence an elongation of one ten-millionth is easily recognized. Field _ Expansion x = aroeet dynes/em?. | Core A. Core B.| Both. Aes ‘0 0-0 0-0 ge | —-0-9)-7 070 11 3 —0°6| 0-2 |Incipient shifting. 14 6 0°6 (Voy 18 8 —0'8 Adis 20 8 —1:0 0°3 2? 1-2 —1:5} 0°4 Continual shifting. No rotation. 30 P23 —1°2 nah 44 oy °5| 0-2 ‘Much rotation. LOCO” BS eee 0°5 Rotation. 600 Sais SECON ysilhs 5. Rotation. * Michelson: J. ¢c., Annex, III. 112 Barus—Trial of Interferential Induction balance. Taken as a whole, the results of the table agree in character with bidwell’s* interesting investigations. They show that the expansion increases to a maximum, attaining it in a field between 20 and 30 dynes, at which the iron is probably satu- rated. After this the data are obscured. The present maxi- mum expansions are smaller than bidwell’s and occur at an earlier stage of magnetization. Thus an elongation of 1 to 1:5. millionths was rarely exceeded. Although my experiments were made in fields as large as 600 dynes, the results above 30 dynes are of no value. For im larger fields than this the fringes showed more and more rota- tion intermixed with the translation. This indicates that the suspended cores were flexed by the field, or that they contained inherent torsion which in a field gives rise to a Wiedemann effect, or. both. Probably flexure supervened. To give an example among many: In a field of 330 dynes the fringes for helix A shewed clockwise rotation, the fringes for helix B counter-clockwise rotation. The two cores together showed a definite translation of about one fringe, without rotation. This. would correspond to compensatory flexures,} and in large fields. the real phenomenon was masked by this discrepancy. If the results are studied with reference to weak fields, the curve is found to terminate abruptly in the abscissa. No. elongation was observable in fields less than about 7 dynes. For fields larger than this, an increment of field less than a single dyne produced an observable change of elongation. This. result is sufficiently remarkable to deserve detailed study. There is a final point to be mentioned. Im a field of about 20, in which the elongation of the above rods was most pro- nounced, the observed effect with a single helix in circuit was such as to correspond to a shifting of the frmges whenever the circuit was closed, with no return of fringes apparent on open-. ing. On closing, about 1°5 fringes visibly and invariably moved across the cross-hairs of the telescope; on opening, the return was either partial or (apparently) absent altogether. I was even able to make over 100 fringes move across the field by continually closing and opening the circuit. On inserting the other helix alone the result was the same but reversed in direc- tion. With both helices in circuit there was but little motion. I was at first inclined to refer this observation to an increase of temperature due to magnetization. The passage.of 4 fringe corresponds to an elongation of 5X107’. The equivalent rise * Bidwell: See Ewing’s Magnetic induction in iron, p. 238; Phil. Trans., 1888, p. 205; Proc. Roy. Soc., xlvii, p. 469, 1890. . + Unfortunately the soft Norway iron obtainable here is not quite straight ; and since all mechanical treatment hardens and injures the iron, I was obliged to- content myself with the best samples in the lot at my disposal. Barus—Trial of Interferential Induction balance. 1138 of temperature in iron would be 0:04° C., which in case of several hundred repetitions would be otherwise appreciable. Hence this observation, though distinctly marked, must be an illusion and due to the fact that the magnetization takes place much more ‘slowly than the demagnetization. It is a curious feature, however, that for fields above and below the sensitive value in question, continued shifting does not occur. 6. The results obtained are applicable to intermittent cur- rents of low frequency at once. From the dimensions given, the self-induction Z of the two helices is not above 0°5 henry, each. If the frequency be kept below 100 per second, the current would not be dephased more than about 0:2 period behind the electromotive force; and there would be no serious interference with visibility from this cause since the partial electromotive forces of each helix as well as the total electro- motive force are in advance of their respective currents by the same fraction of a period. Fora simple harmonic variation of electromotive force at the source, the impedance introduced at each coil for frequencies, 0, 10, 50, 100 would be about 11, 34, 158, 316, respectively. To produce appreciable expansion of the iron cores one would, therefore, merely have to increase the electromotive forces accordingly. In the following experiments, however, the current is actu- ally made and broken. Since the extra current is thus evoked and necessarily introduced at the maximum of electromotive force, its effect will be a maximum. From the given dimen- sions one should expect the current in successive intervals of 0-01 sec. each to be 0:20, :37, :50, -60, etc., of the maximum. It will thus be necessary, if the frequency is as high as 100, to use five times the battery power needed in case of continuous currents under the same circumstances. This is also the effec- tive quantity of current whether one or both helices are in circuit, seeing that resistance and self-induction in the two cases increase in like ratio. In the following table I give the results obtained for low frequencies, using a Foucault interrupter of the usual pattern. The first and third columns show the value of the fields had continuous currents been used; the second and fourth columns the corresponding appearance of the interference fringes (“clear ” denoting maximum visibility). 114. Barus—Trial of Interferential Induction balanee. One helix in circuit. Both helices in circuit. pele ot or |) aS aS —~ Fields (maxima). Fringes. ‘Fields (maxima). Fringes, 18 Hazy 16 all clear. 22 Just vanished 18 clear. 27 Vanished oT clear. 36 do 26 2 of face clear. 57 do 34 do. 70 do 40 do. 100 do 49 vanishing. Conformably with $5 the fringes vanish much sooner (i. e. for smaller fields) when only one helix is in circuit than when both are inserted. Since only about 30 per cent of the full current is effective, the minimum field producing elongation and, therefore, vibration of the cores is about of the same value as before (§ 5). It is difficult to give the vanishing point. The fringes do not disappear uniformly over the whole face, but parts of the field of view linger longer than others. This again points to rotation of fringes, parts nearer the axis vanishing last. Many other experiments of a similar kind were made, without, how- ever, adding essentially to the above. 7. The question finally occurs whether the above arrange- ment is sufficient to test the speed of transmission of an elec- tric signal or impulse from helix A to helix £. The equations for this case were originally worked out by Lord Kelvin* and since by others; but I do not know that much direct experi- mentation has been given to the subject, seeing that in deep sea cables the leakage obscures the law. As the underlying differential equation is of the same form as that which occurs in heat conduction, many of the electrical problems have vir- tually been solved in the analytic theory of heat. Thus if the potential at the initial point varies as sin 2@¢, the distribution of potential along the wire is given by Ro Vo sin(2wt—z/ wo ) where z2=avkc, « being tae distance of transmission from the initial point, c the electrostatic capacity per unit of length, % the electrostatic resistance per unit of length. Thus the speed of transmission of like phases is 2Vw/ke , increasing, there- fore, as the square root of the frequency of vibration, while * See Math. and Phys. papers of Sir William Thomson, p. 61 et seq., where the expression for capacity is also given. Barus—Trial of Interferential Induction balance. 115 the time of transmission (¢/2V @) varies as 2 already given and proportionally to the square root of the periodic time. Finally the magnitude of oscillation decreases in geometric progression as distance increases arithmetically. In case of a circuit made and broken as above, the quantity which may be called the retardation a (seconds) has the value e= 02882 Che 10" where C is the capacity in microfarads per knot, / the resist- ance in ohms per knot, w the length of cable in knots (185,526™). The equation would not be changed if all data were referred to the centimeter. The current reaches 50 per cent of its value after the lapse of 6a seconds, and its full value (nearly) after about 20a seconds, no matter what cable be chosen. Finally the capacity of the cable is C= -0486 Kau/log D/d, where D/d is the ratio of diameters of cable envelope and ‘eable core, A (microfarads) the specific inductive capacity of the substance of the envelope. To test the speed of transmission along a wire with resist- ance and capacity only, let the wire be wrapped bifilarly and the capacity of the bobbin increased by surrounding the insula- tion of the wire by a continuous conductor. J will suppose this to be done with silk-covered wire about 1™™ in diameter with sheets of tin foil around the successive layers to increase the capacity. Thus if D/d=11 and H=2, then C=2'5 microfarads. Hence the retardation per knot, if the resistance per knot be 40 ohms, is a = 2:'3/10°, and the current would attain its full value in about 50 millionths of a second, for the first knot of cable. For succeeding knots the conditions are more favorable. It follows, therefore, that if the above apparatus is to _ measure retardations along about 1°5 knot of the wire speci- fied, the period of vibration of the iron cores must be of the order of 0°0001 second. In this time the change of phase would be complete and one fringe would pass the cross-hairs. The point finally at issue is whether the above apparatus can respond to 10,000 vibrations per second. It ought not to be difficult to break the current as often as this, Gordon’s* rapid speed break reaching 6000 per second. If the iron cores of the above helices be clamped in the middle with the ends free to vibrate as an elastic solid, longitudinally, a length slightly over 30°" would vibrate with the given frequency. This is about the length of core used above. On the other hand, it is necessary to wind the helices for *Gordon: Electricity and Magnetism, p. 49. 116 Barus—Trial of Interferential Induction balance. smaller self-induction, retaining the condition of identity. Thus if only a single layer of wire is used on the above frame, ‘and if the core of half the above diameter (made tubular or cross-shaped in section) be inserted, the self-induction will be reduced to ‘002 henrys, and the impedance for 10,000 vibra- tions to 126 ohms. The field of this helix is 14 dynes per ampere, and since half this value is just sufficient to elongate the cores appreciably, about 90 volts would be needed to pro- duce the vibration, apart from the favorable condition of reso- nance. Finally, in view of the large factor &/Z = 750, after 0:0001 second the extra current has kept the current strength reduced to about 1 per cent of the maximum value. Hence the current must be 100 times stronger than when it acts con- tinuously, 4/L = 750, requiring 70 volts to maintain it. To summarize: Setting considerations of mechanism aside, an arrangement of the kind shown above supplied with about 100 volts ought to indicate the retardation along something over a single knot of wire of the high capacity stated, inserted between the helices. This retardation will be exhibited by the passage of one yellow interference fringe across the cross-hairs of the telescope. Brown University, Providence, R. I. Trowbridge and Richards—Multiple Spectra of Gases. 117 Art. X.—The Multiple Spectra of Gases; by JouN Trow- BRIDGE and THEODORE Wwm. RICHARDS. IN a recent paper upon the spectra of argon we have shown that the two different spectra of this gas are dependent primarily upon the electrical conditions which cause the gas to glow. The continuous discharge of a high tension accumula- tor through the gas -produces the red spectrum, while the dis- charge of a condenser, provided that its oscillations are not damped by the resistance of the tube, or other resistance or impedance, produces the blue spectrum. It becomes now a matter of great interest to determine if the same conclusions apply to other gases, a subject which has already been studied in detail by Willner and others. The chief difference between our work and the earlier investiga- tions are: first, the use of a high tension accumulator instead of an electrical machine or Ruhmkorff coil as the source of electricity ; and secondly, the introduction of varying ohmic resistance or impedance between the plates of the condenser in order to study the damping of the spark. It is the object of this paper to emphasize anew the importance of the electrical conditions of the circuit, and to call attention once more to the fact that the behavior of most elementary gases is in every respect similar to that of argon. With regard to the spectrum of nitrogen, it has been known for a long time that two spectra could be obtained by means of appropriate changes in the density of the gas, as well as by the introduction of the condenser; but not all investigators have put the same interpretation upon their results. So vary- ing are the views, that Angstro6m* and Thalent believed the familiar channelled spectrum to be due to impurities in the gas. Plicker and Hittorf,t Wiillner§ and Salet| have proved this view to be false, but they had not at hand the constant current of high tension at our disposal and their nitrogen was obtained from air containing argon, so that a revision of their work promised to be of great interest. With our Planté battery of ten thousand volts we have obtained the usual two different spectra of nitrogen by vary- ing suitably the electrical conditions of the discharge. By means of a continuous discharge with no spark gap or brush dis- * Poge. Annalen, exliv, 300. + Bull. soc. chim. (2), xxv, 183, t Roy. Soe., Proc., xiii, 153. Phil. Mag. (4), xxviii, 64. § Pogg. Ann., exxxv, 497, exxxvii, 337, cxlvii, 321, exlix, 103, cliv, 149. || Ann. Chem. Phys. (4), xxviii, 52. Hasselberg, Ames and others have also studied the nitrogen spectra. 118 Trowbridge and Richards—Multiple Spectra of Gases. charge in the circuit through nitrogen under varying pressure, we always obtained the channelled spectrum. The glow in the capillary tube, as well as the positive and negative light, was of a delicate pink color under these conditions; a color not unlike the red glow of argon. When an air-gap, over which the battery discharges in a brush, is introduced into this circuit, the glow becomes more violet in tinge, and the spec- troscope shows that the red bands almost if not wholly. disap- pear, while the blue and green ones retain their positions. Under these conditions the capillary tube is filled with a pure blue glow, less intense and vivid than that of the argon, how- ever. When the condenser is introduced, the whole appearance of the tube is utterly transformed. The blue color of the tube at once changes to a rich bluish green, and the channelled spec- trum gives place to bright lines, already well known and mapped. This line spectrum corresponds to the blue spectrum of argon. When the oscillations of the condenser-discharge are damped by means of a suitable resistance or self-induction interposed between the condenser plates, a channelled spec- trum reappears; but in this case the glow in the tube is of a bluish white color, the positive and negative lights being of a bright yellow. Whether or not this channelled spectrum is, as it seems, exactly like the one obtained by means of a con- tinuous discharge, photographic measurement willshow. This last appearance is probably the usual one obtained by means of the Ruhmkorff coil, for then the oscillations induced by the primary condenser are damped by the impedance of the secondary coil and the resistance of the tube. The spectrum of hydrogen is usually supposed to consist of four bright lines, Ha 6563-0 (C), H® 4861°5 (F), Hy 4340-7 (G*), Hé 4101-9 (h), as well as several in the extreme violet and ultra violet.* Other spectra have been observed also; but owing to the partial understanding of the conditions required to produce them, the voluminous literaturet upon the subject leaves a confused idea in the mind of the reader. The continuous discharge of a high tension accumulator through hydrogen gas at tensions varying from 0:05" to 3™" and more yields a beautiful white glow in the capillary of a Geissler tube, while the strata in the positive and negative light are often alter- nately pale pink and pale blue. When examined by a spec- troscope with a broad slit, the light from this discharge appears * Ames, Phil. Mag. (5), xxx, p. 48 (1890). + Angstrom, Vogel, Lockyer, Fievez, Wiedemann, Huggins, Wiillner, Hassel- berg, Balmer, Griinwald, Villari, Schuster, Salet, Smyth and others. For refer- ences see O. Dammer; Anorgan. Chem. i, 369. Trowbridge and Richards—Multiple Spectra of Gases. 119 to consist of bands similar to that of nitrogen, as well as of bright lines; but when the slit is narrowed every band is resolved into a multitude of sharp lines of varying intensities,* among which the four usual hydrogen lines, although present, are by no means especially prominent. A large capacity is required to change this spectrum into the familiar four-line spectrum which is comparable with the blue spectrum of argon. The change is-marked by a sharp alteration in the color of the glow from white to adeep red. In the process, the bluish green line (Hf) as well as the two in the violet, which retain their early position unaltered, become nebulous at their edges;+ while the red line Ha remains sharp and clear. The most marked change in the spectrum, however, is the com- plete obliteration of all the host of other lines covering the whole spectrum, and the obvious contrast between the oscilla- tory and non-oscillatory spectrum of this gas is quite as strik- ing as in the ease of nitrogen, although somewhat different in nature. This four-line deep red glow appears satisfactorily in a tension of gas of about a millimeter,—when the tension of the gas is much higher or lower the resistance is increased, the oscillations are damped, and other lines begin to appear. Curi- ously enough, however, the damping of the oscillatory dis- charge does not at first replace all the lines which were extin- guished by the introduction of the condenser. At first only a sharp line in the yellow and one in the green begin to appear, and gradually others are added as the impedance is increased. The relation of these conclusions to the varying spectra of hydrogen observed in stars leads to interesting speculations regarding the nature of the electrical and thermal conditions in the photospheres of these bodies.{ Each of the halogens gives two spectra, one with and one without the condenser. With iodine, if any of the solid ‘itself is present in the tube, the vapor tension is so soon altered by the heat of the discharge that the oscillatory discharge is damped and the non oscillatory substituted. Hence the former can be obtained only for a few moments. A tube of helium made by Professor Ramsay, the kind gift of the Hon. R. J. Strutt, gave a brilliant yellow glow under the infiuence of the continuous discharge, and a brilliant blue with the condenser discharge, but since the bright helium lines remained in each, and every other important line in the blue spectrum proved to be an argon line, it is evident that the oscillations produced no considerable effect upon the helium. * Smyth, Wied. Ann. Beiblatter (2), vii, p. 286. Willner observed this spec- trum but did not measure the lines. + E. Villari, Fievez, and Salet. ¢ HE. Ebert, Wied. Ann., liii, 1894. 120 Trowbridge and Richards—Multiple Spectra of Gases. As Crookes and others have already pointed ont, since many gases yield different spectra under the influence of varying electrical conditions, it is evident that the fact of the existence of two well-marked spectra of argon gives not the slightest presumption in favor of the hypothesis that the new gas isa mixture. In order to discover if argon possesses a dual nature, the gas must be split up in such a way that its components give different spectra under like electrical conditions; then alone would the evidence of the spectroscope be of weight in proving the dissimilarity of the several parts. The results of this work are thus far only those which were to have been expected from a high tension galvanic battery, reasoning from the work of other investigators with the Toep- ler-Holtz machine. The battery, however, gives a current so admirably constant and so easily regulated as to its tension, that we hope to be able to use it as a means of determining whether the oscillatory discharge produces its effect simply by increasing the temperature, or because of some inherent prop- erty in the manner of the discharge. It is our intention to extend the investigation by the sys- tematic photographic study of the action of the varying dis- charge upon all the elementary gases in the purest condition, as well as upon mixtures under widely varying conditions of temperature and pressure. Harvard University. T. Holm—Studies in the Cyperacec. 121 Art. XI. — Studies im the Cyperacee; by THro. Houm. Hl. Carex Fraseri Andrews, a morphological and anatom- ical study. (With Plate LV.) THis very rare and local plant has a peculiar and striking appearance which at once distinguishes it from most of the other species of Carec, and since we have, also, observed sev- eral peculiarities in its internal structure, we have thought it worth while to treat this species for itself. As regards the morphological characteristics of our plant, we have already in a previous paper® called attention to its monopodial ramification. The foliar organs are very singularly developed and represented by four or five scale-like, membra- nous leaves and one very broad and deep green leaf, being the only assimilating one of the shoot. The seale-like leaves have long and perfectly closed sheaths, while the single, proper leaf is not. sheathing, the base being merely convolute. This leaf is, then, destitute of any sheath and has no ligule, and the one margin or sometimes both are minutely folded, giving the leaf the appearance of being finely serrate along the margin. ‘These characters, the lack of sheath and of ligule, distinguish this species from most, if not all the other representatives of the genus Carex. It is true, however, that in some of our very broad-leaved species, e. g., C. plan- taginea, C. platyphylla, etc., the leaf-sheath is very short and breaks open at an early stage, but it is, nevertheless, easily dis- tinguishable while the leaf is still in bud, and the ligule is well developed and persists for a long time. Species with the Jeaf-margin undulate, as in Carex Fraseri, are not known in this genus except that a form of C. pallescens L. is occasionally met with, of which the bracts are more or less undulate, hence the distinction of the variety “undulata Kze.” The question is, now, to decide the proper situation of these two forms of leaves, whether tie large, green leaf belongs to the same axis as the’ scale-like ones. It appears, however, as if the floral bud is lateral, and that it is directly developed from the axil of the large, green leaf, this being the only leaf of the central, vegetative bud. The central bud is, therefore, of a short dura- tion, of about one year only. By examining specimens of this Species in the winter-time, the floral buds are observed to be situated exactly as in the other species with monopodial ramifi- eation, with the exception that each vegetative shoot has, in C. Fraseri constantly, not more than one single, assimilating leaf developed. * See this Journal, vol. i, May, 1896, p. 349. Am. Jour. Sc1.—Fourts Series, Vou. III, No. 14.—FrEsruary, 1897. 9 122 T. Holm—Studies in the Cyperacee. By considering the stem above ground, this is not terete or triangular as we are used to find it in the Cyperacew, but flattened, almost, in its entire length. The inflorescence bears a large number of staminate flowers at the apex, and a number of pistillate ones at the dase of the staminate, all of which are supported by broad, hyaline and silvery shin- ing bracts, almost entirely destitute of chlorophyll. The utriculus is large, membranous and much inflated. It encloses the pistil, which (Plate IV, fig. 2) is distinctly stipitate (Stp.) and with the rhacheola (Ith.) extended beyond the pistil. The development of this rhacheola is to be seen in our figure 1, where a very young pistillate flower of Carex Hraseri has been illustrated. It has been drawn from a flower at a very young stage, while the entire inflorescence was still inclosed in the bud for the winter. This figure shows the utriculus (Utr.) forming a low wreath and surrounding the three carpels (Carp.) ; on the front-side of the utriculus is a small, roundish body to be seen (Kh.), which represents the free apex of the rhacheola, being here extended beyond the flower. This fact, the exten- ‘sion of the rhacheola, has, already, been observed in this species by Baillon (1. ¢.); besides it is far from uncommon in other species of Carex, as recorded by the writer in a previous article upon this subject.* In Carex Fraseri this processus is most frequently without any rudiments of flowers, but sometimes a few imperfect have been observed. Another peculiarity of the pistillate flower in this species is the occasional development of four stigmata; besides that Boott (1. ¢.) has described and figured a flower with two utriculi, thus originating from two separate prophylla, a fact which seems to be exceedingly rare in the Cyperacec. But by considering these morphological characteristics of C. Frasert as a whole, there does not seem to be any character important enough for distinguishing it generically from the other Caricés, inasmuch as the peculiarities of the pistillate flower, the extended rhacheola for instance, is not sufficient for admitting any such distinction. Its most peculiar morpholog- ical character is, no doubt, the absence of a closed sheath and ligule in the assimilating leaf. We will, thereupon, examine the anatomical structure of our lant. . The leaves, the scale-like ones, are membranous and aimost destitute of chlorophyll. The epidermis forms, on both faces of the leaf a homogeneous tissue of rectangular cells with the walls slightly undulate. Stomata are either entirely absent or merely present in a very small number on the dorsal face. A * This Journal, vol. ii, September, 1896, p. 216. T. Holm—Studies in the Cyperacec. 123 transverse section of one of these scale-like leaves shows us a structure, which in most respects reminds us of a leaf-sheath, viz: the development of large lacunes between the mestome- bundles, the non-differentiation of the mesophyll and the par- tial absence of vessels in the mestome-bundles, while the leptome-elements are well-developed. The stereome is, also, characteristic, forming large groups on the dorsal face of the leaf underneath the leptome-side of the mestome-bundles ; besides it, also, forms isolated groups between the bundles. The corresponding groups of stereome on the ventral face are smaller, and the cell-walls are relatively thin, leaving a large lumen. The free part of the scale-like leaf shows a similar structure, and the free margins consist only of two strata of cells, corresponding to the dorsal and ventral epidermis. The large, green leaf has a similarly developed epidermis on both faces, but stomata are present in large number on both faces, but most numerous on the inferior, the dorsal face. These stomata show the same, characteristic form as recorded by Schwendener (I. c.), but they are constantly in niveau with the surrounding epidermis-cells. Very characteristic of the epidermis is the total absence of epidermal expansions such as hairs or thorns, which are so common in most of the other Cyperacece. There is, however, one kind of epidermal expansions which our plant has in. common with the other Cyperacee, viz: the peculiar silicious cones, which are noted to project from the bottom of the epidermis-cells and often reaching the superior wall. These cones may occur from one to two in each cell (Plate IV, fig. 3), but they are only present in those strata of epidermis which cover the groups of stereome on the dorsal - face of the leaf-blade. We have stated above that the epidermis forms an almost homogeneous tissue on both faces of the blade, excepting a few strata of somewhat narrower cells, covering the mestome- bundles, but this divergency in shape is too small to be of any importance. We should, however, expect to observe quite a considerable variation in some of the epidermal strata, at least on the ventral face of the blade. We know from the numer- ous and most important writings of Duval-Jouve upon the structure of the Cyperacew and the Gramineew, that certain strata of the epidermis have usually attained a special develop- ment, widely different from the other epidermis-cells, and of _which the function is to facilitate the folding or closing of the leaf-blade so as to prevent the surface from being exposed to the strong sunlight, thus protecting the leaf against a too rapid evaporation. These cells were by Duval]-Jouve named “ bulli- form-cells” (Duval-Jouve: Histotaxie |. ¢.) and they differ 124 T. Holm—Studies in the Cyperacec. very much from the surrounding cells in regard to their size and shape. ‘They are often developed as vesicles with a rather , thick and strongly cuticularized outer-wall, and show a great tendency to collapse, when exposed to excessive drought, while they rapidly swell up again, when brought in contact with moisture. These peculiar cells are, as a rule, located on the superior face of the leaf-blade and they are most often to be observed above the midrib, but also between some of the larger mestome-bundles in several genera, espe- cially in the broad-leaved species. It is easily understood, that being located where they are, these cells by collapsing naturally force the blade to become folded either as simply “ conduplicate ” or as “convolute.” In the genus Carex, of the species which have been examined, this form of cells has never been observed to be missing. We have figured the most common shape and size, which these bulliform cells attain in Carex, e. g. figure 5, which is from @. virescens Muhl., a species with leaves of ordinary width, and which grows in localities not exposed to excessive drought or moisture. In some species, which inhabit places exposed to heavy winds, and which are especially characteristic of the sand-dune vegetation, the bulliform-cells are especially devel- oped and are even supported by several strata of similar, but smaller cells, all constituting a typic closing-apparatus for the leaf-blade. Figure 6 shows such layers of bulliform cells from the leaf of C. trinervis Degl., a common sand-sedge from the west coast of Europe. The very narrow-leaved Carices have, on the other hand, not the bulliform cells so well differentiated, as for instance in C. exilis Dew. (fig. 4), besides similar marsh- sedges, as CU. diwca L., C. gynocrates Wormsk}., ete. But our C. /raseri is entirely destitute of any such bulli- form cells on either side of the blade, which is the more sur- prising when we consider the extraordinary width of the leaves; the plant seems, therefore, to be very poorly adapted for existing outside of the damp and shaded ravines where it usually occurs. In considering now the structure of the other tissues which compose the leaf, our plant does not differ in any essential respect from the other species of Carex. The mesophyll forms a homogeneous tissue all through the blade, and it con- sists of polygonal cells, which are closely packed on both faces of the leaf, while the central part of the leaf-blade is traversed by broad lacunes, which are only separated from each other by the mestome-bundles and the adjoining mesophyll. Very distinct from the mesophyll is the thin-walled paren- chyma-sheath, which partly surrounds the mestome-bundle, at least on the sides, extending from the stereome of the superior T. Holm—Studies in the Cyperacece. 125 face of the blade to that of the inferior. This parenchyma- sheath is not green in our plant, but colorless. The mestome-bundles are, furthermore, surrounded by a rather thick-walled mestome-sheath (M. S. in fig. 7) which forms an uninterrupted ring all around the leptome and hadrome. The group of leptome is very well differentiated and the hadrome contains a larger number of vessels, especially ring-vessels, than is usually observed in Carex. On both faces of the mestome-bundles are groups of stereome (St. in fig. 7) which show the same structure as generally known in this genus. This tissue, the stereome, forms also an isolated group in the leaf-margin itself. One feature seems, however, somewhat striking, when we consider a number of transverse sections from the entire width of the blade, viz: the uniformity in size and development of the mestome-bundles, at the same time as we observe the equality in thickness of the mesophyll between the mestome- bundles. There is, generally, in the broad-leaved species of Carex a distinct difference to be found in regard to the size of the mestome-bundles, besides that the mesophyll often becomes constricted in certain places so as to leave room for groups of bulliform-cells, which as stated above are absent in our plant. ‘The sections of the leaf, as described above, were all taken from the middle of the blade, and we will now examine the very base of the same leaf, which in the other Carzces forms a closed sheath, but which is free and merely convolute in our plant. Such sections, taken from the base of the leaf, show a structure much like that we have described for the broad part of the blade. We merely note that the groups of stereome are longer and much more narrow, besides tbat their cells have a considerably larger lumen. ‘The mestome-bundles themselves show a somewhat weaker development, and the lacunes are much reduced in size from what we have observed higher up in the leaf. The structure of the utriculus may be described in this place. Its thin, bladderlike texture (fig. 8) is due to the extraordinary thin cell-walls of not only the epidermis, but also of the meso- phyll underneath this. A few mestome-bundles are observable, all of which are very weakly developed, especially in regard to the hadrome-part, and the cells of the supporting stereome are relatively very thin-walled. Utriculus, so far, corresponds to Wilczek’s second type (1. ¢.), characterized by its thin structure and the non-differentiation of the mesophyll. The pericarp shows a similar thin texture and is composed of three tissues, viz: a very thin-walled epidermis, a stratum of three rows of relatively thick-walled sclereids, and finally a layer of horizontally stretched cells, which form the inner 126 T. Holm—Studies in the Cyperacee. coating of the pericarp (fig. 9). The stem above-ground, which at its apex bears the inflorescence, is in our plant flattened almost in its entire length, except at the base, where it is strictly cylindric. The anatomical structure is, however, the same whether we examine the cylindric or the flattened part, except in regard to the pith, which is broken in the upper part of the stem, rendering this hollow; this character, the hollow stem, is rather uncommon in the Carzees, although it has been observed in certain species. The epidermis of the stem is very uniform and thin-walled; it possesses the interior silicions cones, but is like the epidermis of the leaf, entirely destitute of exterior expansions. The bark-parenchyma is quite large and consists of very thin-walled roundish cells, which gradually pass over into a large tissue of polygonal cells, which occupy the greater inner-part of the stem and which represents a typical pith! Mestome-bundles are quite numerous and form two concen- tric rows, viz: an outer row of large bundles, supported by heavy layers of stereome, especially on the outer side and bor- dering immediately on the epidermis, while the inner row of mestome-bundles is only supported by a very small layer of stereome. The bundles of the inner row are smaller and are imbedded in the bark in alternation with the mestome- bundles of the outer row. Concerning the structure of these bundles, the larger ones are exactly built up as those of the leaf, except, of course, that no mestome-sheath is developed, this being always confined to the leaf-bundles. The rhizome of our plant is very short on account of its cespitose growth, and the ascending shoots do not push out in any considerable distance from the main rhizome. ‘The interior structure corresponds in most respects with that of the stem above ground, viz: the development of the bark, the arrange- ment of the mestome-bundles in two concentric, alternating rows, and finally by the central pith. We notice, however, some characters in the rhizome by which this differs from the stem above ground, viz: the presence of an endodermis, sur- rounding the bundles, thus separating them from the proper bark-parenchyma. The bark itself is in the rhizome composed of rather thick-walled cells, all of which contain deposits of starch ; the endodermis consists of roundish cells, which are thickened all around. As regards the mestome-bundles these are more or less surrounded by stereome, the cells of which are not very thick-walled. We notice here, as in the stem, that the two rows of bundles show a different development in regard to their relative size, besides that the larger are here perihadromatie, i. e. the leptome is central and surrounded by the elements of the hadrome. The central part of the rhizome is occupied by a thick-walled pith, the cells of which contain T. Holm—Studies in the Cyperacee. 127 starch. Considered as a whole, the structure of the rhizome of our plant agrees in most respects with that of a number of other species, which by Laux (I. ¢.) has been classified as repre- senting his eighth type. The roots of Carex Fraseri are numerous and cover the short rhizome with a dense mass of long and strong fibers. The interior structure of the root in the Cyperacee has long since been most ably discussed by Treub (1. ¢.) and Klinge (1. c.), whose comparative studies have given us the principal features by which the Cyperacee may be distinguished from the Graminee and other families of the Jonocotyledonee. We have learned, for instance, that the inner bark cells in the roots of the Grasses collapse radially, while tangentially in the Cyperacee. Another and very distinct character is that in the Caricee the Pace ranie borders immediately on the endo- dermis, thus interrupting the pericambium. The root of our plant shows now the following structure: an epidermis of usual form; a bark-parenchyma of about ten layers of roundish, thick-walled cells, some of which collapse tangentially and form thereby lacunes of very considerable width. The inner- most layer of the bark is differentiated as an endodermis, the cells of which are thickened so as to represent a typical U- endodermis. We observe, thereupon, the pericambium, which in our species forms a completely closed ring, without being interrupted by the protohadrome. This fact seems to form a very singular character of our plant, thus differing from all the other Caricee which have hitherto been examined. The central-cylinder is occupied by six relatively large groups of hadrome in alternation with groups of leptome, besides the distinct but small elements of the protohadrome, situated close to the pericambium. The greater part of the central cylinder is, however, occupied by a mass of conjunc- tive tissue,* which surrounds the vessels and forms in the center of the root a very compact tissue of rather large and thick-walled cells, the cell-walls being strongly lignified. If we will now consider these morphological and anatomical characters, which we have observed in Caree Fraseri, it may not be denied that this species is one of the most remarkable in the whole genus, at least of those which, so far, have been examined by the various authors. The monopodial ramification of its rhizome with its single assimilating leaf destitute of sheath, ligule, epidermal expan- sions and bulliform-cells in connection with its flat and hollow stem, besides the uninterrupted pericambium of the root, con- stitute a structure that seems almost unique in the family of the Cyperacec. - * Russow’s ‘‘Geleitzellen” (1.c.), Klinge’s “ Leitzellen” (1. c.) and Van Tieghem’s ‘‘Cellules conjunctives ” (1. c.). 128 T. Holm—Studies in the Cyperacee. Some of these characters, considered by themselves, may even prove sufficient for the distinguishing of this singular species. Washington, D. C., December, 1896. Bibliography. Baillon: Histoire des plantes. Monographie des Cypéracées. Paris, 1893. Boott: Illustrations of the genus Carex, vol. iv, London, 1867, Plate 484, p. 150. Duval-Jouve: Sur une forme de cellules épidermiques qui parais- sent propres aux Cypéracées, Mém. de l’Acad. d. sc. Mont- pellier, 1872, p. 227. Duval-Jouve: Histotaxie des feuilles de Graminées. Ann. d. se. nat. Bot. Série VI, tome i, p. 294. Klinge: Vergleichend histiologische Untersuchung der Gramineen- und Cyperaceen-wurzeln. Mém. de lAcad. imp. d. se. de St. Petersbourg, VII Ser., vol. xxvi, No. 12. St. Peters- burg, 1879. Laux: Ein Beitrag zur Kenntniss der Leitbiindel im Rhizom monocotyler Pflanzen. Inaug. diss. Berlin, 1887. Lemcke: Beitraige zur Kenntniss der Gattung Carex. Inang. diss. Keenigsberg. Pr. 1892. Mazel: Etudes d’Anatomie comparée sur les organes de végéta- tion dans le genre Carex. Inaug. diss. Genéve, 1891. Russow: Vergleichende Untersuchungen der Leithtindelkrypto- gamen. St. Petersbourg, 1872. Schwendener: Die Spaltceftfnungen der Gramineen und Cyperaceen. Sitzungsber. d. K. Pr. Akad. d. Wiss., vol. vi. Berlin, 1889. Treub: Le méristeme primitif de la racine dans les Monocotyleé- dones. Leyden, 1876. Van Tieghem: Recherches sur la symétrie de structure des plantes vasculaires. Ann. d. sc. nat. Bot., vol. xii. Paris, 1870. Wilczek: Beitrage zur Kenntniss des Baues der Frucht und des Samens der Cyperaceen. Bot. Centralbl., vol. 51, Nos. 5 and 6, 1892, p. 262. EXPLANATION OF PLATE IV. FIGuRE 1.—A young female flower of Carex Fraseri. Carp: the carpels; Rh: the rhacheola; Utr: the utriculus. FIGURE 2.— Mature caryopsis of C. Frasert. Rh: the rhacheola; Car: the cary- opsis; Stp: the stipe. FiguRE 3.—Cellsof the epidermis of the leaf of C. Praseri, showing two internal, silicious cones. St: stereome. Figure 4.—Bulliform-cells of the leaf of Carex emxilis. FiguRE 5.—Bulliform-cells of the leaf of C. virescens. FIGURE 6.—Bulliform-cells of the leaf of C. trinervis. Figure 7.—Mestome-bundle of the green leaf of C. Fraseri. Ep: epidermis; St: stereome; M.S: mestome-sheath; transverse section Figure 8.—Transverse section of the utriculus of C. Frasert. Hp: epidermis. Figure 9.—Transverse section of the pericarp of C. Frasert. Hp: the exterior, proper epidermis. T. A. Jaggar, Jr.—Simple Instrument for inclining, etc. 129 Art. XII.—A Simple Instrument for inclining a Prepara- tion in the Microscope; by T. A. JAGGAR, JR. THE instrument* here described was devised for use in petrography, especially in connection with the Michel-Lévyt and von Fedorowt optical methods of determining feldspars. The device may be of service to microscopists in other fields as well, where it is frequently necessary to examine in reflected light the surface of objects, whose relief above the object- glass makes tipping imperative if one desires to view the pre- paration on all sides. In the study of thin sections of rock in the polarizing microscope, the constituent minerals are sliced on planes bear- ing usually no definite relation to the optical symmetry of the individual crystals ; in order to obtain sections parallel or per- pendicular to directions of optical constancy in particular min- erals of such a rock section, it has been necessary, in the past, to seek ont individuals whose orientation chanced to give the results desired, or to average a number of approximately accurate sections, or to isolate the mineral and grind special sections oriented with reference to known cleavages or crystal faces. By using the so-called “ Universal Stage,” which per- its rotation of the object-glass about two axes at right angles to each other, von Fedorow has shown that a crystal section may be oriented at will with reference to the polarization plane of the nicols, by simply inclining it, first in one direction, then in a second at right angles to the first, until at length the desired optical effect is observed. For certain measurements of extreme precision in micro- scopical crystallography, the graduated circles of the universal stage are indispensable: but for the determination of the feld- spars by symmetrical extinctions, or for conoscopic work, in petrography, all that is required in practice is the means of tipping the slide with delicacy in any azimuth at will. An instrument for this purpose should be simple, inexpensive, easily manipulated, adaptable to any microscope and instantly removable; to meet these requirements the instrument here described has, been constructed. The accompanying figure (fig. 1) shows the construction of the instrument in vertical section and plan, natural scale. It is * The author’s model was made by C. Milton Chase, mechanician, 45 Cam- bridge st., Boston, Mass. The Bausch and Lomb Optical Co, of Rochester, N. Y., have completed a new model, adaptable to the stage of any microscope, when size and distance apart of clip-holes are specified. + A Michel-Lévy: Etude sur la détermination des feldspaths dans les plaques minces, au point de vue de la classification des roches. Paris, 1894. ¢ EK. von Fedorow: Universal-(Theodolith-) Method in der Minéralogie und Petrographie. II. Theil: Krystall-optische Untersuchungen, Zeitschrift fiir Krystallographie, etc., xxii, 229-268, 1893. 130 T. A. Jaggar, Jr.—Simple Instrument for simply a clip (¢) for the object-glass, supported 15™™ above the surface of the stage by a ball-and-socket joint (6); a long key (A), fitting a inane aperture (/) in the ae -arm (a) from Figure 1 (natural scale). either side, secures delicacy of manipulation and enables the operator to incline his preparation in any direction by a simple movement of the fingers. A thumb-screw (s) in the end of the barrel (¢) that serves for ball socket, permits adjustment of the tension on the ball by pressure on a brass plate (p) faced with cork, which fits the surface of the ball.* In the model figured, the foot-ring (7) fits the center of the Fuess mechan- ical stage, and the rectangular movements of the latter allow ample horizontal adjustment, for bringing different portions of the section into the visual field. The foot of the new model (fig. 2, see foot note p. 129) is made adjustable to the clip-holes of any stage, and the elevated clip attached to the ball-and-socket is of such-form that free horizontal movement of the object- glass by hand is possible in all directions ; furthermore, the foot- plate is so shaped that in special cases it may be simply slipped under the ordinary clips, and thus held in any desired position. * This arrangement was suggested to the writer by Professor Goldschmidt’s two-circle contact goniometer, where the crystal is centered by a ball-joint and key: v. this Journal, Oct., 1896, p. 285, On Crystal Measurement by means of Angular Codrdinates, and on the use of the Goniometer with two Circles; by Charles Palache. inclining a Preparation in the Microscope. 131 The chief uses of this instrument in petrography are the rapid determination of maximal and minimal extinction and absorption values, the orientation of feldspar twins for deter- g.2. New model of instrument, showing revised form of foot-plate, with sliding posts aida ble to clip-holes of any stage ; an extra pair of these posts shown at the left. Key isin position. The clips may be inserted in ail four corners of the elevated clip-plate, and thus turned backward or cross-wise, mination by Michel-Lévy’s method, and the inclination of the section to give well-defined interference figures in convergent light. This latter use has been tested by the writer with a section of calcite cut parallel to the face of the cleavage rhom- bohedron R (1011). The interference figure given by this section, when placed horizontally in the microscope in conver- gent polarized light, is a black bar, representing one arm of the basal cross, and a small are of two of three faintly defined chromatic rings on the extreme border of the field. If we now incline the section in the ball-and-socket clip, the complete cross and rings may be brought into the field and accurately centred, thus bringing the ¢ axis into coincidence with the axis of the microscope. This implies rotation of the section through an angle of over 44°, which is about the working limit of the instrument, and is of course far greater than would usually be required. For this work, where very high powers are used, a tube must be provided for raising the conoscopic condenser 15" to the level of the elevated object-glass. In addition it is desirable that this condenser be mounted in a steeply conical cell, in case any considerable tipping of the section about its apex is required; the high-power objectives offer no obstruction in this respect, as they are usually mounted in cells sufficiently tapering. or interference figures from mineral sections where such high magnification is unnecessary, a special long-focus conoscopic combination of objective and condenser is recom- mended, as these allow ample space for free inclination of the object- elass.* * R. Fuess in Berlin makes such a combination for use with the Fedorow Uni- versal Stage: vy. Erganzungen zu den Preis-Verzeichnissen 1891 und 1894, R. Fuess, 1895, p. 18, No. 14, last paragraph. 132. = Verrill— Coloration in Mammals, Birds, Fishes, ete. Art. XITI.— Nocturnal protective coloration in Mammaits, Birds, Fishes, Insects, etc., as developed by Natural Selec tron ; . by A. E. VERRILL. Mucsa has been written in respect to the imitative and pro- tective colors of these groups, as seen by daylight, and the bearing of these facts on natural selection is well known. Very little attention has been paid to their colors, as seen by twilight, moonlight, and starlight. Yet it is evident that pro- tection is more needed during the night than in the daytime, by a very large number of species. This is the case with those that move about in search of their food at night, as is the habit of numerous forms of small mammals, such as rodents (rats, mice, arvicole, etc.), insectivores (moles, shrews, etc.), many herbivores, various marsupials, and members of other orders. Many carnivorous species, which seek their prey at night, will also find advantages in such protective | colors, for thus they will more easily escape the notice of their prey. Hence many nocturnal carnivores are black or nearly so, as the mink, fishes, some bears, ete. The same principles will apply to birds, reptiles, fishes, and to insects, both in their larval and adult states, for many members of all these groups are very active at night and hide away in holes or beneath dense herbage by day. Moreover, large numbers of birds, fishes and insects, that are active by day, rest in exposed situations at night, and are thus liable to be destroyed by nocturnal ene- mies. Most small birds roost in trees, bushes, or reeds, and therefore need protection while sleeping. Most small fishes, that are quiet at night, rest among sea-weeds, grasses, and stones, or else directly upon the bottom, exposed to the attacks of many nocturnal carnivorous species. The struggle for existence is severe among such species. It is to be expected, therefore, that instances of nocturnal protective coloration will become numerous when looked for. The chief object of the present paper is to call the attention of more observers to this subject. In many eases the same colors are equally protective in day- light and at night. This is the case with the green colors, so often seen in the plumage of birds that live among foliage, and with the various shades of brown and gray,—common colors of birds, and mammals that live on the ground, among rocks or dead leaves, and of tlrose that live on or among tree trunks. The same applies to the white colors of mammals and birds in winter and in the arctic regions. But there are * Abstract of a paper read before the Morphological Society, Dec. 30, 1896. Verrill—Ooloration in Mammals, Birds, Fishes, etc. 1388 many colors that are not in the least protective by day, yet are eminently so by night. In general, the black and very dark colors, common in mammals, birds, and insects, are protective at night and not by day. One of the most obvious effects of moonlight is to give very strong or black shadows, in which black or dark animals become invisible, or nearly so. This invisibility is often increased by sharply contrasted stripes or patches of white or light yellow, which look like patches of moonlight falling across a dark shadow, and thus serve to break up the outlines of bird or beast that might otherwise be recognized. Transverse black or dark brown bands on fishes that rest among eel grass or sea weeds, tend to render the out- lines of the fish indistinct, because they look like the shadows and shaded surfaces of the weeds. Black fins and tails have a similar effect, in concealing or destroying the outline of fishes. The striped colors of the tiger have the same effect when it lives among the stalks of reeds, ete., and is probably much more effective in twilight or moonlight than by day. The same is true of the spotted: pattern of the leopard, pan- ther, and jaguar. : A great number of small nocturnal mammals, belonging to diverse groups, have dark gray and grayish-brown colors (mouse-colors) which are highly protective at night, but are usually not at all so in the daytime, for such colors are con- spicuous among the green herbage which they frequent, and on which most of them feed. Moreover, nearly all such mam- mals hide away in holes in the daytime. I have noticed that our common meadow mouse (Arvicola) which is very dark gray, is scarcely to be seen even in a moonlight night, in localities where it is very abundant among grass, and when large numbers are so near that the sound made by their teeth in feeding is very evident. Among insects there are multi- tudes of instances of colors that are evidently nocturnally pro- tective and which can be explained only on the basis of natural selection, favoring the variations in color that are in this way most useful. Such colors may or may not be more or less pro- tective in the daytime. Frequently they appear to’be just the opposite of protective in the daytime. Thus many butterflies have bright colors that are very conspicuous by daylight and which do not in any way match their customary surroundings. This applies to those species that are black or dark blue, striped or blotched with white, yellow, or orange, and to many species that are spotted or striped with red, orange, and black on the upper surface of the wings, and often also beneath, so that they are conspicuous whether flying or at rest. Their active habits and acute senses probably give them fair protection by day. At night, when resting with the wings folded, the colors 1384 Verrill—Coloration in Mammals, Birds, Fishes, ete. of the under side of the wings usually blend very perfectly . with that of the flowers on which they roost. Many of our species of Argynnis and allied genera are marked with red, orange, and brown, while there are bright silvery patches on the under side of the wings, which are exposed when at rest. I have observed that these butterflies become very inconspicu- ous in the moonlight, when sleeping on the goldenrod and other favorite flowers, and that their silvery spots imitate very closely the dew-drops that surround them. Numerous nocturnal insects that live on the ground are black or dark brown, which are colors that are protective only at night. This is true of most ground beetles, many crickets, cockroaches, ants, etc. Many of these insects hide away in the daytime, so that no protective colors are then needed. But many insects that are exposed both during the day and at night have acquired green or yellowish colors that are protective at all times, when living among foliage. Green grasshoppers, katydids, ete., are examples. In general, patches, stripes, or spots of strongly contrasted dark and light colors are more likely to be of use by moon- light than by daylight, whether on birds or insects. Reptiles are to a large extent diurnal in their habits and many kinds hide in holes and crevices when at rest, so that our native species of this group appears to afford few good instances of evident nocturnal protective colors, though many may occur when the habits of tropical species become better known. Among nocturnal amphibians protective colors are common, and in many cases they appear to be exclusively for nocturnal protection. Our native nearly black species of salamanders (Amblystoma punctatum and A. opacum) have conspicuous spots or blotches of white or light yellow. It is evident that these colors have been acquired by natural selection in conse- quence of the nocturnal protection that they afford. | Verrill— Changes in the colors of certain fishes, etc. 135 Art. XIV.—WNocturnal and diurnal changes in the colors of certain jishes and of the squid (Loligo), with notes on therr sleeping habits ;* by A. E. VERRILL. _WSILE investigating the nocturnal habits and colors of some of our native marine fishes, in 1885 to 1887, at Wood’s Holl, Mass., in the laboratory of the U.S. Fish Commission, of which I had charge at that time, I made the unexpected discovery that a number of species had the peculiar habit of assuming, while sleeping, a style of coloration quite unlike that seen in the daytime. Numerous other duties prevented me from making as many observations of this kind as I wished, at that time, nor have I since had opportunities to continue them. Therefore I have decided to publish these incomplete observations, with the hope of inducing other naturalists to continue such studies in some of the various zoological stations that are now established. Most of my observations were made late at night, between midnight and 2 o’clock A. M., when everybody else had retired. The gas jets near the aquaria were turned down so low as to give barely light enough to distinguish the forms and colors of the fishes. Under these conditions, by using great care not to cause any jar of the floor, nor sudden movements of any kind, I succeeded in observing many species asleep. Most fishes sleep very lightly and are aroused by-almost impercepti- ble vibrations of the air or water. Some of these fishes took unexpected attitudes while asleep. In many eases the change of color from that seen while awake, or in the daytime, consisted in a simple increase in the depth or intensity of the colors, the pattern of colors remain- ing the same. This was the case with several species of flounders. Those that are spotted or mottled with dark pig- ment showed their markings much more strongly, or in greater contrast with the ground-color, than by day. Several species of minnows (/undulus) which are marked either with longitudinal or transverse dark bands, have these markings more decidedly black and better defined than by day. The same is true of the king-fish (Wenticerrus nebulosus), in which there are obliquely transverse dark stripes that come out more strongly at night than by day. The black sea-bass (Serranus furvus) and the sea-robins (Prionotus palmipes and P. evolans) presented the same phe- nomena. Several species of trout (Salvelinus fontinalis, etc.) were observed to become much darker at night than in the daytime, but I was not sure that any of those observed were asleep at the time. * Abstract of a paper read before the American Morphological Society, Dec. 30, 1896. These observations were also communicated to the Connecticut Academy of Sciences, in 1888, but were not published. 136 Verrall—Changes in the colors of certain fishes, ete. It is well known that trout, flounders, and some other fishes are able to change their colors, even in the daytime, according to the color of their surroundings. Therefore a darkening of the colors at night is to be expected, even if not asleep. But in all the cases mentioned above the nocturnal change of color is of a protective character, as explained in the preceding article. Other fishes, however, show much more remarkable changes. Among these the seup or porgy (Stenotomus chrysops) is one of the best examples. This fish, when active in the daytime, usually has a bright silvery color with iridescent tints. But at night, when asleep, it has a dull bronzy ground-color and the body is crossed by about six transverse black bands. When one of these fishes, with this coloration, was awakened by sud- denly turning up the gas, it immediately assumed the bright silvery colors belonging to its daytime dress. This experiment was repeated many times, on different individuals, with the same result. As this fish naturally rests among eel-grass and sea- weeds, the protective character of its nocturnal colors is obvious. A common file-fish (Jfonacanthus, sp.) was observed that presents a very decided change in color pattern. This species, in the daytime, is mottled with brown and dark olive-green, and the fins and tail are a little darker than the body, but when asleep, at night, its body becomes pallid gray or nearly white, while the fins and tail become decidedly black. These colors are decidedly protective at night, or in a feeble light, among rocks and weeds, where it lives. This and other species of file-fishes, when sleeping, would usually rest on the bottom with the back leaning against the glass of the aquarium or against a stone at a considerable angle. ~The common tautog or black fish (Zautoga onitis) has the curious habit of resting upon one side, half buried among gravel, or partly under stones, and is often curved in strange positions. It is easy to imagine that the flounders originated from some symmetrical ancestral form that acquired, like the tautog, the habit of resting upon one side, at first only when sleeping, but afterwards continually, owing to the greater pro- tection that this habit and its imitative coloration afforded. The one-sided coloration and the changes in the position of the eyes, ete., would gradually follow in accordance with well known laws of evolution. The common squid (Loligo Pealec) was observed sleeping on several occasions. At such times it rests in an inclined position, on the tip of its tail and on the basal parts of the arms, which are bunched together and extended forward, so that the head and anterior part of the body are raised from the bottom, so as to give room for breathing. The siphon tube is then turned to one side. Under these circumstances the color is darker and the spots more distinct than when it is active, owing to the expansion of the brown and purple chro- matophores. b td ( < 0. C. Marsh—The Stylinodontia. 137 Arr. XV. — The StyLrnoponTia, a Suborder of LHocene Edentates; by O. C. Marsu. In the autumn of 1873, the writer obtained in the Eocene deposits of Wyoming the remains of an extinct mammal of great interest. The most striking feature was the lower molar teeth, all essentially alike, and inserted in deep sockets. They were nearly cylindrical in form, and all grew from persistent pulps. The outer and inner faces were covered with a thin layer of enamel. This type specimen was described by the writer, in this Journal, vol. vii, p. 532, May, 1874, under the name Stylinodon mirus, as representing a new genus and species. The affinities of this new form, so far as then deter- mined, were recorded as follows: “These specimens resemble in some respects the correspond- ing partsof the genus Zovodon Owen, from the Quaternary of South America; but may, perhaps, have some more affinities with the Edentates.” The writer subsequently made this new form the type of a distinct family, the Stylznodontide, and placed it under the order Zi2llodontia (this Journal, vol. ix, p. 221, March, 1875), and this reference, instead of the original suggestion as to its affinities with the Edentates, has been generally followed. Fragmentary remains of the genus Stylinodon were subse- quently obtained by the writer trom time to time, in essentially the same horizon, but none of them threw much additional light on the affinities of this peculiar form. A fortunate dis- covery, in the spring of 1882, at a new locality, was a consid- erable part of the skull and skeleton of a second specimen apparently of the same species, and this material seemed sufficient to determine definitely the systematic position of Stylinodon, as soon as the speeimen could be fully prepared for investigation. Owing to a pressure of other work, it was not until ten years later that this specimen was ready for the artist, and careful drawings made, when the Edentate aflinities of the animal became more strongly apparent. The problem, however, was not a simple one, and the relation of the genus to other allied forms required careful consideration. In an interesting paper recently published, Dr. J. L. Wort- man discusses the affinities of this family, and presents an argu- ment in favor of their being true Edentates.* This announce- ment makes it more important that the type specimen of the genus Stylinodon be figured, and that the second more perfect specimen be also illustrated and described, and this is the main object of the present communication. * Bulletin American Museum of Natural History, vol. viii, pp. 259-262, 1896. Am. Jour. Sci.—Fourts Surtzs, Vou. III, No. 14.—FrEprvuary, 1897. 10 138 O. C. Marsh—The Stylinodontia. The Skull and Teeth. Figures 1 and 2, below, represent respectively a portion of a large front incistform tooth of the lower jaw of Stylino- don, and several of the adjacent molar series, all natural size, and pertaining to the original type specimen of Stylinodon mirus. The peculiar sculpture of the enamel of the anterior tooth, showing both the longitudinal grooves and the trans- verse lines of growth, is a characteristic feature of these teeth. In the molars, the two bands of enamel, external and internal, show markings similar, but less distinct. The large front tooth is apparently “from the left side, the enamel shown being thus on the outer face. The other specimen containing the ‘molar teeth is part of the right lower jaw, with the inner face removed, showing the base of the teeth. The sockets of six of these are represented in figure 2, and behind them one more may be seen in the inner part of the jaw, making together seven in this series, all of similar form and size. The lower jaw containing these eight teeth was short, deep, and massive, with a strong coronoid process, the base of which was in advance of the posterior teeth. In the second specimen of Stylinodon already mentioned, the lower jaws agree in all respects with the type. The seven molar teeth have the same position and proportions as in that specimen. The roots of the large incisiform teeth extended backward as far as the base of the penultimate molars. The condyle of the lower jaw is massive and transverse, the articu- lation looking upward. Its motion was not limited by a post- glenoid process. The posterior margin of the jaw above the angle is thickened into a distinct process, which is somewhat ineurved. The lower jaw is especially deep below the last molars, and the entire ramusis robust. The teeth of this genus and the great depth of the jaw below the last molars will dis- tinguish it from Dryptodon, described by the writer from a lower horizon. The skull of Stylinodon is short and massive. The tem- poral fossze are especially large, and are separated above by a high ridge. The brain cavity was small. The occipital plane is narrow, and the sides converge above and meet at the junction with the sagittal crest. The occipital condyles are small, and there are no distinct paroccipital processes. The Vertebree. The cervical vertebree of Stylinodon are well shown in figure 3, below, which represents the series in the natural position essentially as found. The centra are very short, with the articular faces nearly flat. The axis has a long neural spine directed backward, but the succeeding cervicals have only rudimentary spines, as indicated in the figure. O. C. Marsh—The Stylinodontia. 139 The first dorsal vertebra has a very high, strong spine, as shown in figures 3 and 4. The other anterior dorsals have also elevated neural spines, and nearly flat articular faces on the centra. The Scapular Arch, The scapula of Stylinodon is narrow, with the acromion projecting but slightly below the glenoid fossa, as shown in figure 5. The posterior border is not expanded. The anterior portion is somewhat wider than the posterior, and there is no coracoid process. There is a well-developed clavicle. This is of moderate size, with the shaft somewhat flattened. It articu- lated above with the lower end of the acromion, and below with the sternum, but is not represented in the figures. fle. & } ] a Ui | \ ) \ vit ih h A All i FIGURE 1.-—Left incisiform tooth of Stylinodon mirus, Marsh; outer view. FiguRE 2.—Molar series of right lower jaw of same individual; inner view. I, socket of first premolar; VI, socket of penultimate molar. Both figures are natural size. The Fore Leg. The fore leg of Stylenodon mirus, as represented in the second specimen above mentioned, is shown in figure 5, one- fourth natural size, with the scapula (s), the whole nearly in the position in which it was found. The humerus (A) is seen in this figure from the outside, and its connections above and below are clearly indicated. As this bone is especially charac- teristic, both of the genus and to a eertain extent of the order, it is important to present here all its typical features. 4 . 140 O. C. Marsh—The Stylinodontia. FIGURE 3.—Cervical vertebree of Stylinodon mirus, with first dorsal vertebra, rib, and sternum, in position; seen from the left. One-half natural size. a, atlas; az, axis; r, rib; s, spine of first dorsal vertebra; st, sternum. These are seen to good advantage in figures 6 and 7, where the bone is represented one-half natural size. These figures render a detailed description unnecessary. This bone, like all those of the skeleton, is solid, there being no medullary cavity. 0. C. Marsh—The Stylinodontia. 141 5./ FIGURE 4.—First dorsal vertebra of Stylinodon mirus, with ribs and sternum in position; front view. One-half natural size. m, mneural canal; 7, rib; s, spime; st, sternum;. 2, anterior zygapophysis. FIGURE 0.—Left fore leg of same individual; outsideview. One-fourth natural size. h, humerus; J, lunar; m, magnum; jp, place for pyramidal; 7, radius; s, scapula; uw, ulna; wn, unciform; III, third metacarpal; IV, fourth metacarpal. ; 149 O. C. Marsh—The Stylinodontia. The inner structure of the shaft is shown in figure 7,6. The pecu- liar head of this humerus, with its strong tuberosity, the promi- nent deltoid ridge, and the supinator crest below, together with the supracondylar foramen and distal articulation, are all char- acteristic features, and taken together clearly indicate the Edentate nature of the animal to which this bone belonged. Figure 6.—Left humerus of Stylinodon mirus; front view. cf, supracondylar foramen; d7, deltoid ridge: h, head; ¢, external tuberosity ; rs, supimator ridge. FIGURE 7.—Ends and section of same bone. a, proximal end; 0, transverse section; c, distal end. All the figures are one-half natural size. The radius and ulna are shown in position in figure 5. The radius (7) is much the smaller bone, and is placed nearly in front of the ulna (zw). The latter is quite robust, and has a strong, powerful olecranon process, as shown in the figure. These bones also are Edentate in type. O. C. Marsh—The Stylinodontia. 143 The bones of the carpus and manus of this individual are only in part preserved, but those represented in figure 5 will serve to indicate the general nature of the fore foot. There were apparently five digits in this foot, although the fifth was small or rudimentary. The metacarpals were quite short, as indicated by the third and fourth. The phalanges were also short, and the median ones, at least, possessed claws. The above description and figures of Stylinodon will in themselves be conclusive evidence to most anatomists that this genus has close affinities with the Edentates, if it is not a typical member of that group. Its relation to other allied genera will be discussed in a later communication. 8. 8a. OF ea A ( Zz Figure 8.—Left femur of Morotherium gigas, Marsh; front view. FIGURE 8a.— Proximal end of same bone. FIGURE 9.—The same bone; outer view. All the figures are one-sixth natural size. The reference in the original description of Stylinodon to its resemblance with Zoxodon is also worthy of some consideration, as the latter genus is now believed to be nearly related to various forms long considered Edentates, but at present regarded by many as aberrant Ungulates, since, notwithstanding the appar- ent resemblance of the feet to those of Edentates, the teeth indicate affinities with the perissodactyles. This peculiar group, the Chalicotheria, of which Chulico- therium, Kaup, is the type genus, is now known to have its 144 O. C. Marsh—The Stylinodontia. representatives in America, Europe, and Asia. One genus, Ancylotherium, from the Miocene of Greece, was described by Gaudry as an Edentate. The term Ancylopoda derived from this genus has been recently used for the whole group, but, as the writer has already suggested, should be replaced by the name Chalicotheria.* To this group, the genera Moropus and Morotherium, . described by the writer as Edentates,t have been recently referred by some authors not familiar with the specimens on which these genera were based. While it is possible that the nature of some of the remains attributed to the former genus may be fairly in doubt, there can be no question that the two known species of J/orotherium are both true Edentates. The type specimen of the latter genus is well represented in figures 8 and 9, above. It was mainly the remains of these two genera that suggested to the writersome important conclusions as to the early history of Edentate mammals, which are in part quoted below, since they seem to have been overlooked by recent writers. Origin of the Hdentates. The affinities of the Stylinodontia as now determined have brought up again a most interesting question as to the origin and former geographical distribution of the Edentates, and it may not be out of place to repeat here what the writer said on this subject nearly twenty years ago, in an address before the American Association for the Advancement of Science, at the Nashville-‘meeting, August, 1877.{ The main passages relat- ing to the Edentates are as follows: “The Edentate mammals have long been a puzzle to zodlo- gists, and up to the present time no clew to their affinities with other groups seems to have been detected. A comparison of the peculiar Eocene mammals which I have called the Z%lo- dontia, with the least specialized Edentates, brings to light many curious resemblances in the skull, teeth, skeleton, and feet. These suggest relationship, at least, and possibly we may yet find here the key to the Edentate genealogy. At - present, the Tillodonts are all from the lower and middle Eocene, while J/oropus, the oldest Edentate genus, is found in the middle Miocene, and one species in the lower Plio- comegrt te *The term Ancylopoda is preoccupied, having been used by Gray, in 1848, for a group of Brachiopoda. + This Journal, vol. vii, p. 531, May. 1874; and vol. xiv, p. 249, September, 1877. + Introduction and Succession of Vertebrate Life in America. This Journal, vol. xiv, pp. 338-378, November, 1877. 0. C. Marsh—The Stylinodontia. 145 ‘©The Edentate Mammals are evidently an American type, and on this continent attained a great development in numbers and size. No Eocene Edentates have been found here, and, although their discovery in this formation has been announced, the identification proves to have been erroneous. In the Miocene of the Pacific coast, a few fossils have been discov- ered which belong to animals of this group, and to the genus Moropus. There are two species, one about as large as a tapir, and the other nearly twice that size. This genus is the type of a distinct family, the Jd/oropodide. In the lower Pliocene above, well-preserved remains of Edentates of very large size have been found at several widely-separated local- ities in Idaho and California. These belong to the genus Morotheriwm, of which two species are known. East of the Rocky Mountains, in the lower Pliocene of Nebraska, a large species apparently of the genus A/oropus has been discovered. The horizon of these later fossils corresponds nearly with beds in Europe that have been called Miocene. In the Post-Plio- cene of North America, gigantic Edentates were very numerous and widely distributed, but all disappeared with the close of that period. These forms were essentially huge sloths, and the more important were Megathervum, Mylodon, and Megal- GRY NE “Tt is frequently asserted, and very generally believed, that the large number of huge Hdentata which lived in North America during the Post-Pliocene were the results of an extensive migration from South America soon after the eleva- tion of the Isthmus of Panama, near the close of the Tertiary. No conclusive proof of such migration has been offered, and the evidence, it seems to me, so faras we now have it, is directly opposed to this view. No undoubted Tertiary Eden- tates have yet been discovered in South America, while we have at least two species in our Miocene, and, during the depo- sition of our lower Pliocene, large individuals of this group were not uncommon as far north as the forty-third parallel of latitude, on both sides of the Rocky Mountains. In view of these facts, and others which I shall lay before you, it seems more natural to conclude, from our present knowledge, that the migration, which no doubt took place, was from north to south. The Edentates, finding thus in South America a con- genial home, flourished greatly for a time, and, although the larger forms are now all extinct, diminutive representatives of the group still inhabit the same region.” * * * “The Edentates, in their southern migration, were probably accompanied by the horse, tapir, and rhinoceros, although no remains of the last have yet been found south of Mexico. 146 O. C. Marsh—The Stylinodontia. The mastodon, elephant, llama, deer, peccary, and other mam- mals, followed the same path. Why the mastodon, elephant, rhinoceros, and especially the horse, should have been selected with the huge Edentates for extinction, and the other Ungulates left, is at present a mystery, which their somewhat larger size hardly explains.” * * * : ‘‘T have already given you some reasons for believing that the Edentates had their first home in North America, and migrated thence to the southern portion of the continent. This movement could not have taken place in the Miocene period, as the Isthmus of Darien was then submerged ; but, near the close of the Tertiary, the elevation of this region left a much broader strip of land than now exists there, and over this the Edentates and other mammals made their way, per- haps urged on by the increasing cold of the glacial winters. The evidence to-day is strongly in favor of such a southern migration. This, however, leaves the Old World Edentates, fossil and recent, nnaccounted for; but I believe the solution of this problem is essentially the same, namely, a migration from North America. The Miocene representatives of this group, which I have recently obtained in Oregon, are older than any known in Europe, and, strangely enough, are more like the latter and the existing African types than like any of our living species. If, now, we bear in mind that an elevation of only 180 feet would close Behring’s Straits and give a road thirty miles wide from America to Asia, we can easily see how this migration might have taken place. That such a Tertiary bridge did exist, we have much independent testimony, and the known facts all point to extensive migrations of animals over it.” The discoveries made within the last two decades, or since the above was written, have added much to a knowledge of the subject here discussed, but have not modified materially the conclusions given in the foregoing quotations. In regard to the origin and distribution of the Edentates, present evidence tends to confirm the opinion there recorded, that this great group of peculiar mammals originated in North America, and migrated to other parts of the earth, where their remains have since been found, or their living representatives still exist. Yale University, New Haven, Conn., January 19th, 1897. Chemistry and Physics. 147 SOR ON Dern. SON TB LT G HNC ER: I. CHEMISTRY AND PHYsICcSs. 1. On the Diffusion of Metals——In his Bakerian lecture Ros- ERTS—AUSTEN has given the results of some elaborate experiments showing the tendency of two or more metals to mix spontane- ously and thus to form a homogeneous mass. Already in 1883, he had observed that ‘“‘ while molten copper and antimony inter- penetrate but slowly, the mobility of gold and silver in molten lead is comparatively rapid.” In his later experiments molten Jead and bismuth were selected as the fluids into which the diffu- sion of the other metals took place. The tubes containing this molten metal, which were about 200™™ long and 10™™ internal diameter, were arranged in an air bath with double walls, which could be readily maintained at fixed temperatures, determinable accurately by means of thermoelectric junctions. To avoid con- vection currents the tubes were kept hotter at top than at bot- tom. After a suitable time, varying from six hours to seven days, the diffusion tubes were removed, cooled, carefully meas- ured and cut into transverse sections; the contents of each sec- tion being weighed and analyzed. Kelvin, as a deduction from Fourier’s theory of heat-conduction, states the law of diffusion as follows: “The rate of augmentation of the ‘ quality’ per unit of time is equal to the diffusivity multiplied into the rate of aug- mentation per unit of space of the rate of augmentation per unit of space of the ‘quality.”” By “quality ” is here meant the con- centration of the diffused matter. Hence the law may be repre- sented by the differential equation dv [dt = k(d’v /dx’) where ~ is distance in the direction of diffusion, v is the degree of concentration of the diffusing metal, ¢ the time, and & the diffu- sion constant; ¢.¢., the number which expresses the quantity of the metal in grams which diffuses through unit area (one sq. cm.) in unit time (one day) when unit difference of concentration (in grams per cubic cm.) is maintained between the two sides of a layer one centimeter thick. Since the unit of diffusivity has the dimensions [L’I"] the diffusion constants may be expressed in square centimeters per day. Each pair of metals has a perfectly definite constant of diffusion at a given temperature. On tabu- lating the results obtained it appears that gold diffuses more rap- idly in bismuth and in tin than in the heavier metal, lead, as also does platinum; the diffusion of both being about equally in- creased when bismuth is replaced by lead. Platinum diffuses more slowly in lead than gold does, while rhodium diffuses almost as rapidly ; suggesting that the platinum metals are molecularly more complex than gold or silver. 148 Scientific Intelligence. The second part of the research was devoted to the question whether gold would still permeate lead if the temperature were maintained far below the fusing point of the latter metal. In the first experiments, thin plates of gold were fused to the lower ends of cylindrical rods of lead 14™™ in diameter and 70™™ long, and these cylinders were maintained for thirty-one days in a little iron chamber lined with asbestos and kept at 250° C, 75° below the melting point of lead. ‘The cylinders were then measured, cut into sections and assayed. Gold was found through the entire length of the cylinder, the diffusivity being in one case 0:023 and in another 0:03°4 °™ per diem. At 200°, the experiment lasting only ten days, the values were 0°:007 and-0:008. Since the eutectic alloy of gold and lead fuses at 200°, experiments were made to see if gold would diffuse into solid lead at 165°, a temperature below this point. The diffusivity was found to be 0:005 and 0:00454 “™ per diem. Even as low as 100°, the diffusivity was found to be 0:00002. Experiments were also made on the diffusion of gold into solid silver at 800°; a temperature 160° below the fusing point of silver and 50° below that of the eutectic alloy of these metals. The results gave a diffusivity of the same order as that of gold in lead at 200°.— Phil. Trans., elxxxvii, A, 383-415, 1896. G. Bige. 2. On Optical Rotation in the Crystalline and the Liquid States.—The specific rotation of several uniaxial crystals which polarize circularly, has been examined by Travss, in connection with the rotatory power possessed by them in the melted or dis- solved states. Thus the hexagonal-trapezohedral-tetartohedral crystals of patchouli camphor have a rotation for the D line of —1°325° per mm. in the optic axial direction; while in the fused state the specific rotation [a]»>=-—118° and in alcoholic solution [a],—=—124°5. Since these values correspond to rotations of —1:240° and —1°308° per mm. respectively, it follows that this camphor has practically the same specific rotation in the crys- talline and amorphous states. The same is true of ordinary camphor. Matico camphor on the other hand, although crystal- lizing in the same system, gives a rotation of —1:877° for the D line, in plates 1™™ thick; while the melted substance has the specific rotation [a]»>=—29°17°. Hence the rotatory power is about six times as great in the crystalline as in the melted state. Rubidium tartrate crystallizes in the hexagonal-tetartohedral system, the crystals of the dextrotartrate being levorotatory and those of the levotartrate being dextrorotatory. Plates 1™™ thick rotate 10°12° to 10°24° for the sodium line; while in aqueous solu- tion and for a thickness of 1™™ the rotation is only 0°69° and of the opposite sign to that of the crystals. Czesium dextrotartrate is isomorphous with the rubidium salt, the crystals giving a rotation of —14° to —19° per mm. for the D line; this rotation being oppo- site in sign to that of the aqueous solution.— Ber. Ak. Berl. x, 195-205, 1895; J. Chem. .Soc., 1xx, 11,509, September 1896. G. F. B. Chemistry and Physics. 149 3. On the Electrolysis of Water.—The electromotive force required for the electrolysis of water was shown by v. Helmholtz to depend upon the density of the oxygen and the hydrogen at the electrodes, being lower in proportion as this density is smaller ; so that if all gas be removed from the liquid, its value must be zero. From an equation connecting the electromotive force of polarization A, for any given pressures of the oxygen and hydro- gen p, and Pu with the electromotive force when the pressure is the pressure of the atmosphere p,, he has determined the value. A, to be 1°783 volts. SoKko.torr has sought to obtain a direct proof that water can be decomposed with any electromo- tive force however small. For this purpose he used a voltameter containing two platinum plates, an insulated platinum point being placed near each plate. On passing a current, the electrodes became polarized by gas layers of definite density. Since these gases are electrically neutral and therefore free, they diffuse through the liquid, reach the platinum points, and polarize them also. By using a sensitive electrometer, the author has shown that any electromotive force, however small, can effect the electro- lysis. Moreover, he finds that an electromotive force of one volt suffices to produce gas having a measureable pressure. In one experiment, which lasted 16 months, a calomel cell (1:072 volt) produced gas of 2°53™™ pressure; and this even seems not to be the limiting value, as the pressure continued to increase. Similar difficulties were met with here to those encountered in ordinary electrolysis. Forces are active on the surfaces of the electrodes which hinder the free diffusion of the gases and bring about the absorption of these gases by the platinum and other metals. Hence the accurate determination of A, is not easy. Measure- ments at low pressures gave the author a value of 0°745 volt for A,; a value much smaller than that given by v. Helmholtz.— Wied. Ann., II, lviii, 209-248, June, 1896. GapiegB: 4, On the Electrolytic Production of Hypochlorites and Chlo- rates.—The electrolysis of solutions of potassium chloride has been investigated by Oxrtret. The current from four storage cells was passed through (1) a copper voltameter for measuring its strength, (2) an electrolytic gas voltameter, (3) the cells for the experiments, (4) an ampere meter, and (5) a resistance box. The electrodes of the gas voltameter were of nickel rolled into two concentric cylinders, and immersed in solution of caustic soda. The experiment cell had a capacity of about 115°° and was closed tight by means of a rubber cover. Through this cover passed the wires to the electrodes, both being of platinum, a capillary delivery tube and another glass tube. reaching to the bottom of the cell, by means of which it could be ‘filled or emptied. The current strength employed was from 1 to 1:2 ampere, and it was continued for two hours. Using neutral solutions, the author finds that the main product is hypochlorite, 83 per cent of the active chlorine existing in this form at the end of the experiment, and 17 per cent as chlorate. Addition of 150 Scientific Intelligence. alkali, since it favors the decomposition of the water, increases the amount of chlorate formed. NRaising the temperature acts similarly. Diminishing the density of the current at the kathode favors the reduction of the hypochlorite, the effect being greatest in a strong solution either neutral or slightly alkaline. At the anode, however, such a diminution increases the amount of water decomposed, the effect being less marked in a strongly alkaline solution. Since in an alkaline solution the reduction is a mini- mum, no diaphragm is necessary.—/J/. Chem. Soc., 1xx, ii, 517, September, 1896. G. F. B. 5. On the Action of Nitrous acid in a Grove cell.—It has been observed by Inte that if the nitric acid in a Grove cell be grad- ually diluted with water, the electromotive force remains nearly constant until the acid contains 38 per cent of HNO,. On fur- ther dilution, the electromotive force falls from 1°8 to 0°7 volt, having the lower value with 28 per cent nitric acid. If, how- ever, potassium nitrite be added to the 28 per cent nitric acid, the electromotive force rises to 1°8 volt again, but falls to 0°7 volt when the nitrous acid present is destroyed by potassium per- manganate, hydrogen peroxide, carbamide, etc. Conversely the electromotive force of an acid stronger than 38 per cent is low- ered by the addition of permanganate or of carbamide. It is evi- dent, therefore, that nitrous acid is the real depolarization agent in a Grove cell.— J. Chem. Soc., |xx, ii, 554, October, 1896. (See the author’s paper in Zeitschr. physikal. Chem., xix, 577, May, 13896.) Gs FB 6. On the Spectra of Fused Salts of the Alkali Metals.—Because of the comparative simplicity of the spectra of the alkali metals, DerGramonr finds the salts of these metals to offer special advan- tages for the study of the line spectra of the non-metals by the action of a highly condensed spark on the fused salt. The spec- trum thus obtained differs considerably from that obtained with the metal itself or that given with the fused salt and a non-con- densed spark. Chlorides, bromides and iodides decompose readily under these conditions; while fluorides show but little tendency to dissociate, and carbonates, though dissociated with difficulty, give the spectra of the metals in their simplest form, no lines of carbon being observed. Salts of sodium show three intense double lines 6160-6154, 5895-5889, 5867-5862, the other lines being weak, though 5675, 5669, 5155, 5152, and a broad nebulous band 4983-4978, are discernable. Salts of potassium show 7698, 7665, 6939, 6911, 6308, 6245°5, 6117°5, 5832, 5811, 5801, 5783, 5360, 5344, 5340, 5323, 5113, 5099, 4828, 4389, 4309, 4264, 4223, 4185 and 4045, Lithium salts give 6706, 6103, 4972, 4603, 4273, 4132; the small number of lines making the salts of this metal particu- larly well suited for studying the spectra of the non-metals. When fused phosphates are subjected to the action of the con- densed spark in this way, a line spectrum of phosphorus is obtained superior to that seen in a Pliicker tube. Using the potassium or sodium salt, the following lines appear: 6506 (dif- Chemistry and Physics. 151 fuse) 6458, 6088, 6042, 6034°5, 6025, 5498°5, 5462 (feeble) 5453 (feeble) 5423-5, 5409, 5385, 5340, 5311, 5292, 5250, 4968 (diffuse) 4941, 4603, 4588-5. Thetripletin the red 6042-6034°5-6025, and the doublet in the blue 4603—4588°5 are the most distinctly recognized -lines.— C. R., exxii, 1441, 1534, June, 1896. Grey Ey 7. On the Preparation of Lithium and Beryllium.—The follow- ing mode of preparing lithium and beryllium has been described by Borcuers. In the case of lithium the solution of the chlo- rides of the alkalis and alkali earths is.made slightly alkaline, evaporated in an iron vessel, fused with ammonium chloride to render it neutral, and electrolyzed with a current of 1,000 amperes per square meter of cathode surface, the electromotive force being 5 volts. The upper rim cof the iron crucible is kept cool by a circulation of cold water, so that a thin crust of solid material is formed on the surface which prevents the metallic lithium from coming to the air. The metallic globules are placed in a paraffin bath at 130°-200° when the pure metal rises to the surface. Solutions of beryllium chloride are evaporated down with an alkali chloride and ammonium chloride and electrolyzed in the same way as magnesium chloride. Calcium and magne- sium chlorides must not be present. If the temperature be not kept as low as possible, the beryllium forms an alloy with the iron of the crucible—J. Chem. Soc., 1xx, 11, 520, September, 1896. Eon 5, 8. Light of the glow beetle—H. Muraoxa has studied the light given by a large collection of the “ Johanniskifer” and has found some interesting relations between this light and the effects of the radiation from uranine salts observed by H. Becquerel. The chafer or beetle used by the author constitutes one of the sights of Kyoto, Japan. About the middle of June one sees thousands of these beetles lighting up the environs of the towns. The experiments were conducted with over 300 of these chafers confined over the experimental sensitive plates by means of a net. It was discovered that the natural light of the chafer resembles ordinary light; but on filtration through cardboard or through copper plates the radiations exhibit phenomena similar to those observed by Becquerel, and also phenomena analogous to the Rontgen rays. The filtered rays manifest an extraordinary phe- nomenon (das Saugphanomen) which is analogous to the behavior of the magnetic force lines toward iron. The peculiarities of the filtered glow-beetle rays appear to depend upon the physical character of the substance through which they are filtered—per- haps its density. The peculiar phenomena observed are obtained by filtration. In an analogous way can X-rays be obtained by filtration—and this process suggests a means of rendering such rays homogeneous. The filtered glow-beetle rays show clearly reflection. It is difficult to show refraction, interference and polarization, yet the author believes that these phenomena are present. The filtered rays of this insect appear to occupy, like the fluorescence rays of Becquerel, a mean position between the 152 | Screntific Intelligence. ultra-violet rays and the Roéntgen rays.—Ann. der Physik und Chemie, No. 12, pp. 773-781. 3 aie 9. Réntgen Rays..—In some recent experiments conducted in the physical laboratory of the University of Glasgow by Lord KELVIN, assisted by J. C. Beattie and M. Smoluchowski de Smo- lan (Nature, Dec. 31, 1896), it was shown that air drawn through an experimental tube in which it was exposed to the radiation from a Crookes tube became electrified, sometimes positively, and sometimes negatively. Professor Richard Threlfall and James A. Pollock (Phil. Mag., Dec. 1896) conclude from careful experiments that no sensible amount of matter is projected from a Crookes tube, and that the hypothesis that the action of the X-rays is due to a projection of matter is untenable. They also conclude that the phenomenon is not due to the projection of ether streams, and that no disturbance of the ether is caused which is sufficient to affect electromagnetic radiations. Je Pe 10, Hlectriclightin Capillary tubes.—O. Scuortt has observed the extraordinary briiliancy of electric discharges through very fine capillary tubes. The discharges were excited by an induction coil of 25°" spark length. If one makes the assumption that the — duration of the spark discharge is no longer than 77,495 of a second, a capillary of from one to two square millimeters in sec- tion radiates as much light as 1—2000 Hefner flames emit. Since the ordinary arc light has a much larger radiating surface, the capillary light is far more powerful than is created by any other source.— Ann. der Physik and Chemie, No. 12, p. 768-772. Tihs 11. Zemperature of the sun. W. E. Witson and P. L. Gray have established (Phil. Trans. Acad., vol. clxxxv, 1894, p. 361), that the radiation of platinum up to the point 1600° C. obeys the law g = a (T’—T,’), in which T is the absolute temperature of platinum, T’, the absolute temperature of the surrounding medium. They have now deduced another formula, g = 0 (T’—T,’) +a(T —T*,’), which, with suitable choice of the constants a and 6, repre- sents the radiation of polished platinum or blackened platinum. The authors find by extrapolation in their formula that the tem- perature of the hottest portion of the positive carbon attains a temperature of 3300° C. The radiation of the hottest portion of the positive carbon is nearly three times as great as that from the hottest portion of the negative carbon. Likening the sun to a black body, in respect to radiating power, the authors find that the temperature of the sun should be in the neighborhood of 8000° C.— Proc. Roy. Soc., London, vol. lvit, p. 24, 1895. J. 7. 12. Argon and helium.—LockxyY&R has examined the spectra of the gas obtained from uraninite or cleveite, broggerite, and divers other minerals, and cites various cases in which lines are obtained in certain gases which do not appear in other gases, and con- cludes that argon and helium are mixtures of which the separa- tion will be very difficult. He has also compared the wave- lengths of the lines of argon with the wave-lengths of lines Geology and Natural History. 153 observed in the chromosphere, and in the nebula of Orion. A great number of these lines are identical, and this fact appears to throw a new light upon the numerous lines in the sun and the fixed stars the origin of which has been obscure. Argon and helium apparently establish a close connection between the ele- ments of our planet and those of the other celestial bodies. When the intensity of the electric discharge through the gas from uraninite is augmented, carbon lines are rendered more feeble and others stronger. The author believes that this phenomenon indi- cates that the gas is a mixture. If one passes the discharge for a long time through helium, the yellow luminescence disappears. This results from the products of the combination of platinum which are decomposed when one afterwards heats the discharge tube.—Proc. Roy. Soc., vol. lviii, 1895, p. 67, 118, 116, 192, 193. aj. 15 IJ. Gronocy AND NATURAL HISTORY. 1. U. S. Geological Survey.—The seventeenth annual report of the director, C. D. Walcott, gives a full account of the opera- tions of the survey for 1895 and 1896 (pp. 200). There were 32 geological, 6 paleontological, and a still larger number of topo- graphic parties in the field distributed through the United States from Alaska to Florida. Shaler continued work upon the Narragansett coal field. Emerson was engaged in mapping central Massachusetts; Dale. the adjoining portions of Vermont and New Hampshire ; and Hobbs, the Cornwall district of Massachusetts. Wolff and Clark in New Jersey; Keith in Maryland; Taff, Darton and Campbell in- the Virginias; and Hayes in the southern Appalachians, con- tinued geological mapping and completed a number of sheets. Eldridge studied the phosphates of Florida, and the gilsonite of Utah; David White continued his work on the coal-bed floras and his results are proving very valuable in identifying coal hori- zons. Wan Hise, aided by Bayley, Smyth and others continued areal work in the Lake Superior region. Gilbert was in Colorado and Kansas; R. T. Hill in Texas; Weed in Montana. Emmons, assisted by Spurr and Tower, surveyed the Aspen mining district and made a reconnaissance into Montana and Idaho; Cross _ mapped the Telluride district of Colorado; R. C. Hills continued the survey of the coal and iron areas of south central Colorado. Turner and Lindgren were in the gold belt of California; Branner and Lawson in the San Francisco region; Diller in northwest Oregon; and Willis in northwestern Washington. Becker and Dall studied the gold and coal of the coastal regions of Alaska. Chamberlin, assisted by Salisbury, Leverett and others, continued his glacial investigations. Stanton studied the paleontology of the Cretaceous of ‘Texas; Ward, the paleobotany of the Cretaceous and Jurassic of California: Knowlton of the Denver basin, while Am. Jour. Scil.—FourtH Series, Vou. III, No. 14.—FEpruary, 1897. shat 154 Scientific Intelligence. the director himself did field work in Massachusetts, North Dakota, Montana, Idaho and Washington. Separates of four accompanying papers in the seventeenth annual report have already been issued,* ‘‘The underground waters of the Arkansas Valley in eastern Colorado” (pp. 51), by G. K. Gilbert, ‘‘ Geological reconnaissance in northwestern Ore- gon” (pp. 80), by J. S. Diller, ‘“‘The Uintaite (Gilsonite) De- posits of Utah” (pp. 41) by G. H. Eldridge, and “‘The Water Resources of Illinois” (pp. 147) by Frank Leverett. Monograph XXV, ‘The Glacial Lake Agassiz,” (pp. 658) by Warren Upham, is a culminating publication of many papers. It contains, besides numerous maps, a full account of the facts and of Mr. Upham’s views concerning them, and also a brief statement (pp. 244-251) of some alternative views by Prof. Chamberlin. Six bulletins have appeared since July}: “The Catalogue and Index of contributions to North American Geology,” 1732-1891; Bulletin No. 127 (pp. 1054), by N. H. Darton; and the “ Bibli- ography and Index of North American Geology, Paleontology, Petrology and Mineralogy” for 1892-3; Bulletin No. 130 (pp. 210), for 1884; Bulletin No. 135 (pp. 141), and for 1895, Bul- letin No. 146 (pp. 130), all three by F. B. Weeks, are invaluable aids to the working geologist, and their prompt publication by the survey is especially commendable. Florence Bascom, in Bulletin No. 136 (pp. 124), on “The Ancient Volcanic Rocks of the South Mountain, Pennsylvania,” described an acid and basic series of pre-Cambrian lavas. M. E. Wadsworth and G. H. Williams have advocated the volcanic origin of similar rocks at a number of points in the Atlantic states from Maine to Georgia, but this is the first full presentation of the facts from a single region (see further, p. 160). F. H. Newell, in Bulletin No. 140 (pp. 356), gives the report of progress in the division of hydrography for 1895, and records stream measurements in many states and territories. A brief contribution to the geology and paleontology of north- western Louisiana, Bulletin 142 (pp. 63), by T. Wayland Vaughan, considers the Cretaceous, Tertiary and Pleistocene of that region and describes twelve new species of mollusks. C. D. Perrine, in Bulletin No. 147 (pp. 22), notes the earth- quakes in California in 1895. About forty shocks are recorded by the seismographs of the Lick Observatory. Bulletins Nos. 68, 96, 112, 114, and 129 contain records for previous years. The following folios have been issued : ; No. 26, Pocahontas, Va., lat. 37° to 37° 30’, long. 81° to 81” 30’, by M. R. Campbell. No. 27, Morristown, Tenn., lat. 36° to 36° 30’, long. 83° to 83° 30’, by Arthur Keith. No. 29, Nevada City, Cal., by Waldemar Lindgren. The maps in folio No. 26 show the distribution of twelve for- * For notices of two of these papers, see pp. 155, 156. +See this Journal, vol. ii, pp. 84, 303, 306, 395 and 456, 1896. ee Geology and Natural History. 155 mations in the Carboniferous, nine in the Devonian and Silurian, and two in the Cambrian. Those in folio No. 27 show two in the Carboniferous, two in the Devonian, ten in the Silurian, and six in the Cambrian. Folio No. 29 contains the Grass Valley, Nevada City and Banner Hill special maps, all upon a scale of ;,4;;>- This important placer and quartz mining district, which is de- scribed in six folio pages of text, has already produced about 120,000,000 of dollars. The maps show Carbonifer ous, Jura-trias, Eocene and Pleistocene formations, besides eleven eruptives, as well as many quartz veins and auriferous gravel deposits. J. S. D. 2. A geological reconnaissance in Northwestern Oregon; by J. S. Ditter. (From the Seventeenth Annual Report of the U.S. Geol. Survey. Part I.)—The author has given brief but clear accounts of the topography, the features ot the coast range and of the Oregon coast, and notes of the work reported by previous authors. He describes the geological formations met with, con- sisting of slight traces of pre-Cretaceous rocks. He concludes that, contrary to previous opinions, ‘‘it is quite improbable that pre- Cretaceous sedimentary rocks form any considerable portion of the Coast Range in Oregon north of Coquille.” Much the same statement may be made regarding the Cretaceous. “ No certain Cretaceous rocks are known in the [Coast] range, and yet it is probable that they do occur where it joins the Klamath moun- tains. South of Roseburg a short distance along Myrtle Creek, are Cretaceous conglomerate sandstone and shales..... indi- a by the fore of Aucella and other fossils they contain” (p. 16 A large tract of Eocene rocks forms the great mass of the coast range from near the Columbia to the Coquille. These rocks are of both igneous and sedimentary origin. The tufas in the former contain fossils of Eocene age. The shales generally contain marine fossils and in several places coal has been discovered in them. ‘They are named the Arago beds from: Cape Arago, near which the coal-bearing strata have been known for several years through the observations of Prof. Thomas Condon. The upper part of the series are sandstones and have yielded Kocene fossils as determined by Dr. Dall. Oligo- cene, Miocene, Pliocene and Pleistocene deposits are distinctly recognized; and the author has sbown the geological history of the region as indicated by evidences of erosion, elevation and depression and the relation of the several beds to each other. Special tribute is given to the contributions made by Professor Condon to the knowledge of the geology of Oregon. In the second part, on economic geology, account is given of the coal fields and a few other mineral products of the region. The age of the coal is in some cases known to be that of the Eocene formation (Pebble Creek, Cape Arago, Callahans, and north Fork of the Umpqua). In other places the coal occurs in beds known to be older than some Miocene beds. At Coos Bay 156 Scventific Intelligence. the age is somewhat uncertain, but the evidence seems to point to probable Eocene rather than Miocene age. H. 8. W. 3. The underground water of the Arkansas Valley in Eastern Colorado ; by G. K. Girpert. (From the Seventeenth Annual Report of the U. 8. Geol. Survey. Part II.)—This is an excel- lent piece of descriptive geology, adapted to the needs of those who will reap most benetit from it, i.e., the people living in the region described. | The author has used fossils as true ‘‘ Leitfossilien,” marks ot the various formations in which they occur; and among them he has included a beautiful plate of ‘“ nodules of marcasite,” charac- teristic of the lower part of the Fimpas limestone. The ordinary well driller should be able, with the use of this essay, to locate any one of the formations passed through. The Dakota sand- stone is said to be the only valuable source for artesian water in the region discussed, and much care is taken to demonstrate the depth undergound and slope of this sandstone, by the aid of diagrammatic sections across the region in various directions, and by clear definitions, in simple but thoroughly scientific language. H. S. W. 4. The Geological Society of America.—The ninth annual meeting of the Geological Society was held in Washington, D. C., December 29th, 30th and 31st. The following officers were elected for the ensuing year :—President, Edward Orton, Colum- bus, O.; 1st Vice-President, J. J. Stevenson, New York City; 2d Vice-President, B. K. Emerson, Amherst, Mass.; Secretary, H. L. Fairchild, Rochester, N. Y.; Treasurer, I. C. White, Mor- gantown, W. Va.; Editor, J. Stanley-Brown, Washington, D. C.; Councillors, J. S. Diller, Washington, D. C., W. B. Scott, Princeton, N. Y. The sessions were held in the Hall of the National Museum and were presided over by the President, Joseph Le Conte, Berkeley, Cal. Memorials of deceased fellows were read as follows :—of Robert Hay, by R. T. Hill: of Charles Wachsmuth, by Samuel Calvin; of N. J. Giroux, by R. W. Ells. The following is a list of the papers presented for reading: JOSEPH LE Conte: The different kinds of earth-crust movements and their causes. (President’s address.) J. 8. Dim~LeR: Crater Jake. J. F. Kemp: The Leucite hills, Wyoming. N. H. Darton: Physiographic development of the District of Columbia region. N. H. Darton: Dikes in Appalachian Virginia. FRANK LEVERETT: On the changes of drainage in the Ohio river basin. C. WittiAM Hayes: The solution of quartz under atmospheric conditions. MakiIus R. CAMPBELL: Erosion at base level. Marius R. CAMPBELL: The origin of certain topographic forms. J. B. WoopwortH: Homology of joints and artificial fractures. ARTHUR KEITH: Notes on the’ structure of the Cranberry district in North Carolina C. H. Hitcucock: Notes on the stratigraphy of certain homogeneous rocks. J. B WocpwortH: Unconformities in Martha’s Vineyard and Block Island. RosBert Bei: Evidences of uortheasterly differential rising of the land along Bell river, (Read by title.) Geology and Natural History. 157 GeoRGE EH. Lapp: Surface tension of water as a cause of geological phe- nomena. ERASMUS HaworTH: Cementing materials of the Tertiary sands and gravels of western Kansas. ; H. W. TurNER: The work of the U. S. Geological Survey in the Sierra Nevada. J. W. SPENCER: Geomorphy of Jamaica as evidence of changes of level. (Read by title.) RapH S. Tarr: The Cornell glacier, Greenland. H. L. Farrcuiup: Shorelines of lake Warren and of a lower water level in western-central New York. G. K. GiuBert: Old tracks of Erian drainage in western New York. ANGELO H&ILPRIN: The assumed glaciation of the Atlas mountains of Africa. FRANK LEVERETT: The relation of an abandoned river channel in eastern Iowa to the western edge of the Iilinois icelobe. Grorce H. Barton: Glacial observations in the Umanak district, Greenland. F. B. Taytor: The Nipissing-Mattawa river, the outlet of the Nipissing great lakes. F. B. TAytor: Moraines of recession and their significance in glacial theory. HARRY FIELDING RIED: Mechanics of glaciers—-moraines and stratification. Harry FIELDING RIED: Variations of glaciers. BAILEY WILLIS: Preliminary note on the Pleistocene history of Puget sound. WARREN UPHAM: Modified drift in St. Paul, Minnesota. I. C. Russet: Note on plasticity of glacial ice. (By title.) CHARLES R. Keyes: Physical basis for general geological correlation. (By title.) F. D. Apams and A. HK. Bartow: Origin and relations of the Grenville-Hast- ing series in the Canadian Laurentian (with observations by R. W. ELLs.) J. F. Kemp: The pre-Cambrian topography of the eastern Adirondacks. J. E. Wourr and A. H. Brooks: The age of the white limestone of Sussex county, New Jersey. JOSEPH F. JAMES: Notes on the Potsdam and Lower Magnesian formations of Wisconsin and Minnesota. Henry 8. WituiAMs: On the Southern Devonian formations. I. C. Waite: A complete oil well record in the McDonald fieid between the Pittsburgh coal and the Oil Sand. (Read by title ) David Waite: The age of the lower coals of Henry County, Missouri. Henry B. KUMMEL: Structure of the Newark formation of western New Jersey. WinuiAM B. CLARK: The Upper Cretaceous formations of the northern Atlan- tic coastal plain. T. W. Stanton and F. H. KNowtton: Notes on the stratigraphy and paleon- tology of the Laramie and related formations in Wyoming. I. C. RUSSELL: Geology of northeastern Washington. EK. H. Barsour: A study of nature, structure,and phylogeny of Dzemonelix. (TEORGE P. MERRILL: Notes on rock-weathering. Hexey B. KuMMEL: New evidence on the origin of some trap sheets of New Jersey. C. WILLARD HAYES and ALFRED H. Brooks: The crystalline and metamor- phic rocks of northwest Georgia. ALFRED C. LANE: The grain of rocks. G. PERRY GRIMSLEY: The origin and age of the gypsum deposits of Kansas. (Read by title.) 5. Pre-Cambian rocks and fossils—In a paper read in the geological section of the British Association, Liverpool Meeting, September, 1896, by Str W. Dawson, an abstract of which is published in the Canadian Record of Science, July, 1896, an account of the present state of knowledge regarding the pre- Cambrian stratigraphy is given. The base of the Cambrian is 158 Scientific Intelligence. fixed at the lower limit of the Olenellus fauna. With this it is pro- posed to include the Protolenus horizon of Matthew, which terminates below in a barren sandstone in both southern New Brunswick and Newfoundland. Beneath this Cambrian system lies the Aicheminian system of Matthew, composed of red and greenish slates and a basal con- glomerate, and containing no trilobites but fossils referred to Ostracods, Mollusks, Worms, Brachiopods, Cystideans, and Proto- zoa. The following formations are recognized as belonging to the Echeminian system: viz. the Signal Hill series and Random Sound series of Newfoundland, the Keweenian or Keweenawan series of Lake Superior, and the Chuar and Grand Canyon for-. mations of Arizona. ‘Che author regards Algonkian as a term “unhappy in form and cause, and perhaps should be dropped.” The Echeminian is regarded as the earliest system of Paleozoic time. Below the Paleozoic rocks are two systems of the Eozoic. The upper member, the Huronian system, is separated above and below by unconformities from the contiguous rocks. ‘‘ Laminated. bodies comparable with Eozoan, burrows of worms, spicules of sponges and indeterminate fragments referable to Algez or to Zoo- phytes,” are reported from the rocks of this system. Rocks of the system are recognized in New Brunswick, Newfoundland, Lake Superior and Lake Huron, and also apparently in Colorado. The lower member of the Kozoic is named the Grenvillian sys- tem (the upper part of Logan’s Lower Laurentian). The rocks of this system are found in the St. Lawrence and Ottawa val- leys, in New Brunswick, the Adirondacks and eastern slope of the Appalachians. Among the rocks are found belts of limestone “associated with whut seem to be altered sedimentary beds, and in places rich in graphite and in apatite.” The fossils recognized in this system are said to be “Protozoan alone, represented by peculiar and gigantic forms, as Eozoan and Archzozoan, and some smaller types (Archzeospherine).” H. S. W. 6. Antiquity of man in Britain.—The number of Watural Science tor January, 1897, contains a note of recent discoveries by Mr. W.. J. Lewis Axzort of what are believed to be evidences of the existence of man in Britain at a much earlier period than that which has been previously assigned. The specimens in question are a series of. flints which at least bear a striking resemblance to the work of man, and which were obtained from the Cromer Forest Bed at Runton. They were found there sticking in the iron pan. This Forest Bed is now usually regarded as forming the top of the Pliocene series, and contains forms of the cave-bear, rhinoceros, elephant, deer, and other mammals living and extinct. A detailed account of the speci- mens is promised for the February number. 7. On the Age of the Lower Coals of Missouri; by Davin. Wuire. (Abstract of a paper read before the Geological Society of America, Dec. 31, 1896.)—As the result of the study of the Geology and Natural History. 159 composition, vertical range, and geographical distribution of the plants from the lower coals of Henry County, Mo. (Zhe Age of the Lower Coals of Henry County, Missouri), David White con- cludes that the two approximate low coals, which occasionally rest on the eroded surface of the Mississippian series in that region, are slightly younger than the Brookville, Clarion, or Mazon Creek horizons of the northern bituminous fields, though they are perhaps not so young as the middle Kittaning coal. Thus the period of the erosion of the Missisippian appears to include the time represented by the lower portion of the Lower Coal measures and the Pottsville series, a succession of sediments attaining a thickness of twelve hundred feet in the anthracite regions, or over twenty-five hundred feet in the Virginian portion of the Appalachian trough. The plants from Missouri are found by Mr. White to be probably nearly contemporaneous with those of the D (“ Marcy ’’) vein of the northern anthracite fields, but are possibly slightly younger. Forty-two of the forty-three plant genera and nearly one-half of the species occurring in this county are also present in the European basins. A critical com- parative examination of the American and of the European floras leads the author, who regards the species as generally synchron- ous and indicating continental conditions involving greater facili- ties for inter-migration than geologists generally admit, to consider the Henry County coals as partially contemporaneous with the Transition Series between the Upper and the Middle Coal Meas- ures of Great Britain, and the Third or Upper Zone of the Valen- ciennes series in the Franco-Belgian basin, or as referable to the Geislautern beds near the top of the Westphalian (Saarbr ticker Schichten) of the Rhenish district. 8. The relation of the fauna of the Ithaca group to the faunas of the Portage and Chemung ; by Epwarp M. Kinpie. Bull. Am. Paleontology, vol. 2, No. 6; pp. 1-56, with two plates. Dec. 1896.—This is an admirable example of what can be done by an exhaustive study of the fossil faunas of a single restricted area. The author has taken as his thesis the determination of the disputed question whether the fauna of the Ithaca group should be ranked with the Portage or Chemung. He has given a con- cise review of previous discussions over the general and particular points involved. He has collected the fossils from over 80 different stations in the locality representing the faunas under discussion, and has iden- tified all the species, compared the fossils, tabulated their range upward and downward in the general section, thus strongly sup- porting by a compact scientific argument his conclusion that “ the Ithaca fauna should be classed in the Portage epoch.” H.. S. W. 9. The Phosphate-Deposits of Arkansas ; by Joun C. BRAN- NER. (Read before the Am. Inst. Min. Eng. at Colorado meet- ing, Sept. 1896.) The author, late State Geologist of Arkansas, has presented in this paper a concise statement of the present 160 Scientific Intelligence. known facts regarding the phosphate deposits in Arkansas. They appear to be restricted in this region to the interval between Lower Paleozoic rocks and the Carboniferous, occupied also in part by the black shale, which has been considered to be of Devonian age. The author concludes that “we are reduced to the neces- sity of believing that this interval, with its phosphate-deposits, represents the slow accumulation of organic matter over a com- paratively deep sea (not abysmal, however) during the upper Silurian and Devonian periods.” Mr. C. W. Hayes has already discussed the similar phosphate eae occurring in the same interval in the central Tennessee rocks.* It may be observed that the presence of rolled and rounded pieces of fish bones, fragments of osseous plates that were $ to in. thick, which occur among these phosphate nodules both in Arkansas and in Eastern Kentucky, as known to the writer of this note, fix the age of the deposit as not earlier than the Devonian era. H. S. W. 10. Die Leitfossilien, ein Handbuch fiir den Unterricht und fir das Bestimmen von Versteinerungen ; von ERNEST KoOKEN, pp. | 1-848, with nearly 900 figures. Leipzig, 1896 (Tauchnitz).—This elaborate treatise must prove of great value to students of paleontology in Germany in facilitating the determination of fossils in the laboratory. But the fact that the characteristic fossils are European species, and from European faunas will pre- vent American students from getting from it the help they might otherwise gain. Still, the analyses made of the characters pre- sented by the fossils of each grand division of invertebrates dis- cussed and the orderly listing and description of the characteris- tics of the fossil faunas of Europe, will render the work of value - in the American laboratory of paleontology whenever compara- tive geology is studied. The volume is composed of two parts: the first “ Palaontolo- gische Uebersicht,” gives general descriptions of the characters of fossils of the chief invertebrate types, followed by analytical tables of families and genera. The second part is ‘‘ Die Leittossilien,” in which the genera are given under each class for each era, and the characteristic species of each genus selected are distinguished from each other by analytical descriptions and tables, thus bring- ing out with great distinctness the prominent observed features ot the characteristic fossils of each horizon. H. 8. W. 11. Ueber die neue geologische Uebersichtskarte der Schweiz. 1: 500,000; von C. Scumipt. Extract from Compte-rendus in Congres géologique international 6th Session, 1894, Zurich, pp. 352-360).—This brief report gives detailed account of the classi fication adopted for the series of deposits from the Alluvium to the Devonian and of the crystalline rocks. 12. The ancient volcanic rocks of South Mt., Penn.; by FLORENCE Bascom. U. 8. Geol. Surv., Bull. No. 136, Washington, 1896. * 16th Ann. Rep. U..S. Geol. Surv., Part IV, pp. 620-623. Geology and Natural History. 161 8°, 124 pp., 28 pl.—The perception of the importance of the fact that considerable areas of previously unrecognized ancient rhyo- litic lavas occur along the Atlantic coast and in the Appalachian region is due to the late Pror. G. H. Wrix.rams, and to his keen- ness and enthusiasm we owe primarily the appearance of this memoir, and the one noticed in the following section. In the present work Miss Bascom has selected a small and typical area and has studied it thoroughly in the field and the material col- lected with equal thoroughness in the laboratory. Especially in the study of the characteristic structures of these ancient acid lavas, the ways in which they have been altered and modified and the means and criteria by which they may be recognized, is the work a valuable one, which must serve as a model for investi- gators of such rocks. It is illustrated with a large number of excellent plates which add greatly to its value. It is impossible, in the brief limits of this notice, to do more than to call the attention of the petrographer (and geologist as well) to its impor- tance. L. V. P. 13. The geology of the Fox Islands, Maine; by Gro. Oris Smita. (Inaug. Diss. Johns Hopkins Univ., 1896.)—What has been said in the foregoing applies well to this, which is chiefly a study of a series of ancient lavas occurring off the coast of Maine. The geology of the islands has been mapped and the occurrence of a small, interesting area of the Niagara is described, in which 80 species of fossils have been found by Prof. Beecher. A variety of interesting acid volcanics have been carefully studied and the results given. The work contains one plate and an excel- lent map on a scale of a mile to the inch. Wee: 14. The Cell in Development and Inheritance ; by Epmunp B. WIitson, pp. 1-371. New York and London, 1896. (The Macmillan Co.)—The author has presented in brief compass and in a clear and lucid manner the principal facts of our present knowledge of the cell; its morphology, chemistry, physiology and development. Without entering into so full an exposition of the theory of the cell as is given by Hertwig, the chief points in its history are given. No endeavor is made to give an exhaustive account of the cell, but rather the attempt is made to consider, within moderate limits, those features of the cell that seem more import- ant and suggestive to the student of development. The omis- sions are particularly noticeable on the botanical side of the sub- ject. ‘The book is fully illustrated by some original figures, and by a large number of reproductions of the classic figures which have been produced by the many workers in this branch of biology. Each chapter is provided with an ample list of the liter- ature in which the subjects discussed are elaborated. In the last chapter the chief theories of inheritance and devel- opment are defined and discussed. In reply to the question, “What is the nature of the germ- plasm and how has it been acquired ?” Prof. Wilson takes an agnostic position. He says, “The truth is that an explanation of 162 Scientific Intelligence. development is at present beyond our reach,” and again: “ But when all these admissions are made, and when the conserving action of natural selection is in the fullest degree recognized, we cannot close our eyes to two facts: first, that we are utterly ignorant of the manner in which the ideoplasm of the germ-cell can so respond to the play of physical forces upon it as to call forth an adaptive variation; and second, that the study of the cell has on.the whole seemed to widen rather than to narrow the enormous gap that separates even the lowest forms of life from the inorganic world ” (p. 330). Hy GW 15. Tables for the Determination of Minerals by Physical Properties, ascertainable with the aid of a few Field Instruments. Based on the System of Prof. Dr. Albin Weisbach, by Prerstror FRAZER, 163 pp. 1897 (J. P. Lippincott Co.).—The fourth edition of the Weisbach tables prepared by Prof. Frazer has recently been issued. A large amount of new material has been added and the tables have been adapted in other respects so as to make them even more useful than before to the student. These tables are now too well and favorably known to need more than this brief notice. | 16. Der Lichtsinn augenloser Tiere von Witipatp A. NAGEL. 8vo. 120 pp. Jena, 1896 (Gustav Fischer).—This memoir dis- cusses at length the function of sight in eyeless animals. It begins with a lecture on the apparent paradox of “seeing without eyes.” This is followed by an account of the author’s interesting experiments showing that many animals are still sensitive to light and distinguish the direction from which it comes after the re- moval of the eyes. The concluding part discusses the theories of vision and perception of light in the lower animals. 17. Additional information concerning the giant Cephalopod of Florida ; by A. E. Verritt.—Atter the publication of the notice in the January number of this Journal, I received additional facts concerning this huge creature from Dr. Webb. He has also sent me photographs,* taken two days after it came ashore, giving four different views of it. These photographs show that it is an eight-armed cephalopod, and probably a true Octopus, of colossal size. Its body is pear- shaped, largest near the broadly rounded posterior end. The head is scarcely recognizable, owing to mutilation and decay. Dr. Webb writes that a few days after the photographs were taken (Dec. 7th), excavations were made in the sand and the stump of an arm was found, still attached, 36 feet long and 10 inches in diameter where it was broken off distally. This probably represents less than half of their original length, as the arms of Octopus generally taper very gradually and are often five or six times longer than the body. What looks like the remains of the stumps of arms is shown in the front view.f * These were taken by Mr. Edgar Van Horn and Mr. Ernest Howatt, to whom my thanks are due for the proofs. + I have had drawings made from the photographs of the front view and side- Geology and Natural History. 163 The length, given as 18 feet, includes the mutilated head region. The photographs show that the “breadth,” given in the first account as 10 feet, applies to the more or less divergent stumps of the arms (?) and the body taken together, as they lie on the sand. The body, itself, is almost 7 feet wide, and rises at its thickest part 34 feet above the sand in which it is partially imbedded. The body is not greatly flattened and probably had a diameter of at least 5 feet when living. The parts cast ashore probably weighed at least 6 or 7 tons, and this is doubtless less than half of its total mass when living. This species is probably one of those upon which the sperm whale feeds regularly on the whaling grounds off our southern coast. Whalers have told me, years ago, that sperm whales killed in that region often vomit great masses of cephalopod flesh, includ- ing sections of huge arms. One reliable whaling captain used to say that he had seen very large suckers “as large as a dinner plate’ on such fragmerits of arms. The suckers of this Florida Octopus would have been as large as that, if they had the pro- portions to the arms and body usual in small species of Octopus. This species is evidently distinct from all known forms, and I therefore propose to name it Octopus gigaunteus. It is possible that it may be related to Cirroteuthis, and in that case the two posterior stumps, looking like arms, may be the remains of the lateral fins, for they seem to be too far back for the arms, unless pulled out of position. On the other hand, they seem to be too far forward for fins. So that they are probably arms twisted out of their true position. This is, at any rate, the first gigantic Octopod that has been described or figured from actual specimens. Note. Since the above was in type, I have learned that Dr. Webb had caused the sand around the monster to be removed, and by means of six horses and powerful tackle he has moved it higher up the beach. He says that the true length of the body is 21 feet. The head is mostly or entirely gone. The outer integument has dried to a firm mass several inches thick. view, which will be published in the American Naturalist. The photographs themselves are not strong enough for reproduction, having been over-exposed. A notice of this Octopus, written by me, was published in the New York Herald, Jan. 3d, but my signature was omitted without my consent. A figure, furnished and described by me as a restoration, was inserted without any expla- nation: it is needless to say that it does not closely resemble the mutilated remains. 164 Scientific Intelligence. OBITUARY. GENERAL Francis A. Waker, President of the Massachu- setts Institute of Technology, and for 9 years (from 1872 to 1881) Professor of Political Economy and History in the Sheffield Scien- tific School, died suddenly in Boston on the 5th of January at the age of fifty-six. So large a part of his active and varied life was devoted to the public service, and the leading facts of his career have been given so fully in other publications, that no detailed account of his life will be expected in this Journal. But as he was the first economist to be elected to the National Academy of Sciences, the first President of the American Economic Associa- tion, and the author of numerous widely-read and important books on economic subjects, some reference must be made to the scientific side of his career. His economic work lay in three distinct lines. He was at once the Jeader of an economic movement, a theoretical economist, and a statistician. His name is chiefly associated in the popular mind with the movement for the establishment of international bimetal- lism by an agreement among the leading states. His attitude on this subject has frequently been misunderstood. He had no sympathy with the national bimetallic movement of the campaign of 1896 which aimed at the introduction of the free coinage. of silver by the United States alone, but he followed in his views very closely those of Cernuschi and other European economists. As a theoretician his most important contributions to the science are his Law of Wages and his Law of Profits. The former was first suggested in his treatise on - Wages, published in 1876, and more fully developed in his text book of Political Economy published in 1882. This theory stood in direct opposition to the wage-fund theory of an earlier period, and has had an important influence upon economic thought. His theory of profits established a close parallel between profits and rent, and held that, just as rent is the remuneration for special advantages in the way of land, so profits are remuneration for special advantages in the way of business ability. As a statis- tician Gen. Walker’s most important work lay in the management of the ninth and tenth censuses, which he developed from a mere enumeration of the population into a great statistical investiga- tion, reinforced by numerous special studies of the principal resources of the United States. He combined in a rare degree the logical mind of the scholar, the vivid style of the popular writer, and the organizing power of the administrator. His lit- erary activity lay in many fields, and many departments of economic science will feel his loss. H. W. F. A) yee P Fale ¥ yeh PE RECON DAR ATS Ta Ls SER Dye ee eee ee ai " » a _ A ” _ > hae” Oe PP re Sy Pek ae eee >. Ey Bs ee Be CR AT en COD a . 4 ‘nd ee re r oy yf ce < Ee =e STILL FINER CALCITES! MOST WONDERFUL STRIKE IN YEARS. Gorgeous! GRAND!! MAGNIFICENT!!! Where are words strong enough to describe the superb CALCITES which, during the past month, have been coming exclusively to us from a new /mine near Joplin, Missouri!? Never before at any locality have such beautiful yellow Calcites been found. The crystals range from delicate lemon-yellow, through rich golden to deepest amber, occasionally with beautiful purple cen- = ters. Their faces are very brilliant and a con- siderable part of the crystals are transparent. Sizes run from 2 inches up to 12 inches, and weigh from 5 ounces to 450 lbs. Prices, 10c. to - $5.00; extra large museum crystals, $5.00 and upwards. YELLOW CALAMINES, a new find at Joplin. Bright, beautiful and well crystallized specimens, a few of them in most excellent pseudomorphs after Calcite ; 10c. to $3.30. JOPLIN GALENAS. Several hundred of the attractive specimens for which this locality has become famous, have just arrived; 10c. to $2.00. HERDERITE. A new locality in Maine has yielded us a choice - little lot of good-sized, bright crystals in small groups, $1.00 to $3.50. CRYSTALLIZED CINNABAR. Just received direct from the mine in California, a fine lot of specimens; 50c. to $3.50. RHODOCHROSITE. Half a dozen groups of very large crystals from the new Colorado locality and several good groups from Alicante ; $1.00 to $5.00. AMETHYSTS. 25 good groups from Schemnitz ; 50c. to $2.50. COVELLITE from Montana, a new find of this exceedingly rare mineral. Showy specimens, 2dc. to $2.50. SULPHOBORITE, a new Mineral, in small, bright, loose crystals, d0c. each. BORACITE, good, clear, little crystals, 10c. to 50c. each. ZIRCON twins and groups, Canada, $2.00 to $5.00. APATITE, excellent crystals and fine museum groups. RUTILES FROM GRAVES MT. Choice, brilliant crystals, twins, and 8-lings, loose and on the matrix, 50c. to $5.00. LAZULITES FROM GRAVES MT. A fine lot of good loose crys- tals, both single and twinned, 25c. to $1.00; a few good matrix groups, $1.00 to $3.50. : UTAH MINERALS ai half former prices, including extra fine crystallized ORPIMENTS ; UTAHITE in remarkably good crystals ; a splendid new lot of Brochantites, Olivenites, Mixites, Anglesites, etc. ; also excellent Jarosite, Conichalcite, Tyrolite, Martite, Hematite Pseudomorphs after Pyrite, etc. 124 pp» ILLUSTRATED CATALOGUE, 25c. in paper, 50c. in cloth. 44 pp. ILLUSTRATED PRICE-LISTS, 4c.; Bulletins and Circulars free. GEO. L. ENGLISH & CO., Mineralogists, 64 East 12th St., New York City. be CONTENTS. s. Art, VIII.—Outline of a Natural Classification of the Trilo- base bites; by C. E. Bercuger. (With Plate IIL.) ihe IX.—Preliminary Trial of an Interferential Induction Bal- : ance *\.byC, Bakuse. to 20) 0 Se Beas. <. ‘X.—The Maltiple Spectra of Gases; by J.. TrowpripeGE and pees T. W. Ricuarps Roa XI.—Studies in the Cyperacer; by T. Hom. (eae IV.) Renee XII.—Simple Instrument for inclining a Preparation in the eee : Microscope; by T. A. Jaaear, Jr. i | “XIII.—Nocturnal protective anloracan in Mammals, Birds, bah | Fishes, Insects, ete., as developed by Natural Selection ; Pe leg ~ by A. ‘E. Vertu. XIV.—Nocturnal and diurnal changes in the colors Be certain fishes and of the squid (Loligo), with notes on their sleeping habits; by A. E. VERRILL XV.—The Stylinodontia, a Suborder of Eocene Edentates; by O. C. Marsu SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Diffusion of Metals, RopeRts-AUSTEN, 147.—Optical A Rotation in the '‘rystalline and the Liquid States, TRAUBs#, 148.—Electrolysis of bee Water, SokoLOFF: Electrolytic Production of Hypochlorites and Chlorates, Ont- TEL. 149.—Action of Nitrous acid in a Grove cell, [HLE:; Spectra of Fused Salts of the Alkali Metals, DeGRamont, 150.—Preparatiou of Lithium and Beryllium, Borcvers: Light of the glow beetle, H Muraoxka, 151.—Rontgen Rays, KEL- vin: Electric hght in Capillary tubes. O. ScHortt: Temperature of the sun, W. E. Wiuson and P. L. GRay: Argon and helium, LocKYER, 152. Geology and Natural History—U. S Geological Survey, 153.—Geological reconnaissance in Northwestern Oregon, J. S. DILLER, 155.—Underground water of the Arkansas Valley in Kastern \’olorado, G. K. GILBERT: Geological Society of America, 156.—Pre-Cambrian rocks and fossils, 157.—Antiquity of _ man in Britain W. J, L. Appotr: Age of the Lower Coals of Missouri, D. | sy WuitkE, 158 —Relation of the fauna of the Ithaca group to the faunas of the ie Portage and Chemung, E M. KinpLE. Phosphate-Deposits of Arkansas, J. C. BRANNER. 159,—Die Leitfo-silien. ein Handbuch fiir den Unterricht und fur das Bestimmen von Versteinerungen, hk. Koken: Ueber die neue geologische Uebersichtskarle der Schweiz. U. ScHmMipT: Ancient volcanic rocks of South Mt, Penn., F. Bascom. 160.—Geology of the Fox Islands, Me., G. O. SMITH: Cell in Development and Inheritance, E. B. Winson, 161.—Tables for the Deter- mination of Minerals by Physical Properties. ascertainable with the aid of a few Field Instruments. P. FRazerR: ler Lichtsinn augenloser Tiere, W. A.. NaGeL: .Additional information concerning the giant Cephalopod of Florida, A. EK. VERRILL, 162. Obituary—GEN. FRANCIS A. WALKER, 164, Chas. D. Walcott, U. S. Geol. survey. THE mow eG ALN oion EDWARD S: DANA. ASSOCIATE EDITORS Proressors GEO. L. GOODALE, JOHN TROWBRIDGE, H, P. BOWDITCH ann W. G. FARLOW, oFCAMBRIDGE, | Prorzssors 0. C. MARSH, A. ErVERRILL anv H. S. : WILLIAMS; or New Haven, PRorussor GEORGE F. BARKER, or Puamapetpuia, Proressor H. A. ROWLAND, or Batrimorz, Mr. J. 8. DILLER, or Wasuineron. FOURTH SERIES, VOL. II—[WHOLE NUMBER, CLIIL] N WITH PLATE Y. NEW HAVEN, CONNECTICUT. 1897. TUTTLE, MOREHOUSE & TAYLOR, PRINTERS, 125 TEMPLE STREET. : : eribers of countries in the Postal Union. Remittances should be made either’ by ey hte registered letters, or Sear checks. Published monthly. Six dollars per year (postage prepaid). $6.40 to foreign sub- consignment just received. Our collector made special trips to localities and _ ‘* sprinkled ” over tufts of the Calcite. A sample lot received last summer Gantatied a few good specimens whic. were quickly sold to the best European and American collections. og gave — promise of better things, and the hope it aroused then is fully realized in an the best things collected were at once shipped to Philadelphia. ae ee CHABAZITE VAR. PHACOLITE in wonderfully beautiful enecwnenae, ¥ exhibiting a variety of twins and complex forms. White or colorless crys tals of great brilliancy, often a half-inch in diameter, are scattered over dark basalt, making strikingly handsome examples of this variety. gee iH PHILLIPSITE in sharp crystals of the same high luster ; apparently he ph simple form; also the familiar cross twins and anew and rare pps co a habit Pouoranitiale Cumengéite trilling. ji FIBROUS AND ACICULAR CALCITE in wheat chia and lt shapes common to Stilbite and Aragonite and exactly resembling the former in its yellow color. Mo as Re Phillipsite and Phacolite in charmihg association with the above, daintily SPHERICAL CALCITE, in odd specimens. i GMELINITE. A few perfect and symmetrical crystals grouped vith “i Analcite and Natrolite. Rare! The choicest of thése are rapidly selling here and abroad at $2. 00, $8. 00 and $4. 00 each. Equally good but smaller at 50c. to $1.50. Choice micro- scopic mounts 25¢. to $1.00. eeabe If you want them, order at once. HERKIMER QUARTZ, es We have just purchased an old collection of these popular crystals, which ~ - contained a Startling Novelty in the way of an inclusion. One of a group of three crystals shows a cavity filled with fluid in which moyesa minute spider-like form, the body an amber-colored bead, what corresponds to the re legs being acicular crystals of a lustrous black hydro-carbon (?) Other crystals of the ‘‘first water’ at le. to $5. 00 each, according to size. Prices lower than formerly. ; : ie FOREIGN MINERALS just in. Selected and partioulale ane speci- mens of the following: Wiserine, Freieslebenite, Eisenrose, Mesotype, Sal Gemma (Halite) serbian a tetrahexahedral Crystals, Geikielite, Magnetite, etc., etc. setae, ie CABINET SPECIMENS AND SCHOOL | MATERIAL. ee Free. Dr. A. E. FOOTE, | 4 LS 7 Anes STREET, PHILADELPHIA, PA., U. S. A. ESTABLISHED 1876. a. THE \ AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES.] + Oo Art. XVI.—Crater Lake, Oregon; by J. 8. DILLER. With Plate V. | THE Crater lakes, Bolsena and Bracciana in Italy, Paven in France and Laach in Germany, besides numerous other exam- ples in various parts of Europe, South and Central America and Asia, have long been known to science, but the one in the United States which, all things considered, is the most impos- ing of the series, has scarcely been mentioned in scientific publications and its very existence even appears to be generally unknown to persons interested in such features. The Crater Lake of Southern Oregon is deeply set in the summit of the Cascade Range and is remarkable, not alone for its geological history, of which it contains some especially interesting chapters, but also on account of its geographic position and depth, its beautiful blue transparent waters and the grandeur of its completely encircling cliffs, affording no outlet. The summit of the range at this point is broad, with gentle, canyoned slopes surmounted by numerous voleanic cones. The rim of the lake, which is nearly circular, with an aver- age diameter of six miles, rises a thousand feet above the general level of the range. Its outer slope is gentle and rather regular from 10° to 15°, but within, the descent to the lake is precipitous. In general the rim may be described as the hollow base of a very large but deeply truncated cone. Here and there, two to four miles from the crest upon the outer side, are cinder and lava cones, adnate to the great cen- Am. Jour. Sci.—FourtH Srriss, Vou. Ill, No. 15.—Marou, 1897. 2 > a ' 166 J. 8. Diller—Crater Lake, Oregon. tral voleano. The crest of the rim varies in height from 6759 to 8228 feet above the sea, i.e. 520 to 1989 above the lake. Its prominences stand at the head of spurs radiating from the lake. Some of these ridges were formed by single streams of lava, but others result from erosion and are separated by deep canyons, such as those of Sand Creek and Sun Creek, which pass directly through the rim. The rim is composed wholly of lava streams and beds of voleanic conglomerate, dipping away from the lake as shown by the accompanying figure. This is the normal composition and structure of the basal portion of a large volcano. Mt Mazama SS Crater Lake Fic. 1.—Section of Crater Lake and its Rim with the probable outline of Mt. Mazama, structural details generalized—vertical and horizontal scales the same. The southern and western parts of the rim are made up of many beds; those of lava generally predominate, both in size and number. The northeastern portion of the rim at the Pali- sades is made up almost wholly of one great flow. The same is true to aless extent at Llao Rock, which is formed of a short but broad stream over 1200 feet in thickness. The exposures upon the inner slope of the rim show sections of the lava streams radiating from the lake, and that of Llao Rock furnishes a good example. Its greatest thickness is in the middle, where it fills an ancient valley, down the northwestern slope of the rim and tapers upon both sides to a thin edge as seen in Plate V. This massive flow rests locally on an irregular layer of pumice and is overlain by the same sort of material. Andesites predominate, especially among the earlier lava flows so well exposed in section upon the inner slope of the rim ; but rhyolites are common among the later ones, and are usually associated with pumice. Basalts were not observed in the rim. However, they occur upon its outer slope several miles away from the crest. They are connected with promi- nent cinder cones adnate to the once greater central peak and are the newest lavas of the region, excepting that of Wizard Island within the lake. This succession of lavas has been J. S. Diller—Crater Lake, Oregon. 167 observed in many large volcanoes and clearly points to Crater Lake as the site of such a mountain. The rim is intersected by a number of vertical dikes, some of which stand out prominently upon itsinner slope. The largest of these, locally known as the Devil’s Backbone, varies from five to twenty-feet in thickness, and as seen in the plate, cuts the rim from water to crest one and a half miles southwest of Llao Rock. Nearly a dozen other dikes appear in various parts of the rim, and all radiate more or less directly from the lake. Some of them cut through the older lavas only. Others reach to the top of the rim, but none was seen to penetrate the late flows of rhyolite. The dikes are of andesite, and their radial arrangement, as well as the succession of lavas, point to the middle portion of the lake as the center from which they emanated. The name Crater Lake suggests that the lake occupies a crater, and it would naturally De supposed that the crater was originally as large as the rim of the lake. That this is not the ease, however, is indicated by the following consideration. No lava came out through the flanks of the rim excepting the basalts which are associated with cinder cones low on its outer slope. The lava streams generally reach up to the crest and radiate from the rim. If the crater were as large as the rim, the lavas must have escaped by overflow from its lowest point, and the inner slope of the rim would bear traces of the rise and fall of the lava within, instead of exposing sections of all the coulees and sheets of conglomerate of which it is made up. The inner slope of the rim is one of fracture and not of flow, and shows that the lava streams once extended farther towards the center of the lake than now and were more elevated in that direction. They issued from the crater or sides of a huge voleanic peak which once stood upon the present site of the lake. The rim of the lake has been extensively glaciated and affords ample evidence concerning a change in the topography of the region since the glacial period. Moraines are widely spread over a large part of the rim, and extend in some cases far beyond it down the principal lines of drainage upon both sides of the range. The glacia! debris is occasionally accumu- lated in well-defined ridges transverse to the direction of glacial motion, but more commonly is spread over the region in an irregular veneering composed of bowlders, gravel and sand. From the main highway to the lake is a wagon road which ascends over a steep slope littered with many bowlders and irregular piles of glacial debris. On the east of the road is the crest of the range. Against the rim of the lake this crest ends in a prominent moraine ridge marking the line of \ tt 168 J. S. Diller—Crater Lake, Oregon. separation between the glacial lobes which descended Anna and Castle Creeks on opposite sides of the range. At many points on the very crest of the rim, glacial debris is well exposed, resting on striated rocks. Occasionally the debris is over 50 feet in thickness, but is rarely composed of large fragments. The largest bowlders, about 10 feet in diam- eter, were seen some distance from the crest upon the outer side. Glaciers descended all the valleys upon the outer slope of the rim for from two to five miles. Below, the moraines termi- nate in plains through which the present streams have cut narrow deep canyons with columned slopes or cliffs, rendering them inaccessible, Their sculptured walls gave name to Castle Creek, but an equally fine display occurs along Anna Creek. Glacial strize are well marked at many points, on the very crest of the rim, radiating down the outer slope in some cases for a distance of five miles. The slopes were so generally covered with moving ice that the outlines of the glaciers were not well marked. Probably the largest mass was that of the Divide glacier, which had a width of over four miles upon what is now the rim of the lake. The main portion descended Castle Creek and its branches toward Rogue River, but a large lobe extended down Anna Creek on the eastern slope of the range. i large mass of ice descended the valley next south of Llao Rock, reaching far down over the broad stream of basalt from Red Gone. The lavas are deeply planed off and striated, but the effect of glacial erosion upon the general topography is not so marked as upon the sonthern side of the lake, where deep: U-shaped canyons have been cut in the older lavas. That the voleano was active at intervals during the glacial period is well shown by the glaciated flow of Round Top, upon the northeastern edge of the rim. This flow is overlain by two layers of pumice separated by a sheet of rhyolite, all of which were erupted after the glaciation of the surface upon which they rest. The eruption must have been accompanied by great floods. from the snow-capped mountain. Such floods would account for the fact that all the valleys radiating from Crater Lake: have been extensively filled with sediments. Reference has already been made to the occurrence of gla- cial strize on the very crest of the rim. They may be seen in many places along the crest northwest of Victor Rock ; also a few miles beyond, in the sag of the rim next southwest of Llao Rock, as well as near its summit. They occur also in Round Top, in Kerr Notch and over Eagle Crags, completing J. S. Diller—Crater Lake, Oregon. 169 the circuit of the lake from which they radiate. The topmost rocks of the crest are planed off and well striated upon the outer slope, but not upon the steep, broken surfaces which they present to the lake. Glaciation is a feature of the outer ‘slope only, and it is evident that the ice armed with stones to scratch these rocks must have come from above, that is, from a peak which once stood upon the site of the lake. In the deep, U-shaped canyons of Sun and Sand Creeks there is evidence to the same effect. They extend directly through the rim to the brow of the cliff, over 500 feet above the lake’s surface, and belong toa system of drainage which has been completely decapitated. It is impossible to account for the glacial phenomena and drainage features of that region on the supposition that its topography is unchanged. The succession of lavas and the system of dikes point to a voleanic center within the circle of the rim; but the structure of the rim, the condition of its inner slope and its glacial phe- nomena go a step further and suggest that during the glacial period the lake did not exist, and that its site was then occu- pied by a huge volcano which furnished the coulees and sheets of fragmental material in the rim, as well as the masses of ice and snow for its glaciation. In figure 1, the probable outline e the peak is indicated. Judging from ‘the character of the lava, the slopes of the rim, its size and the extent of its glaciation, it is reasonable to sup- pose that Mt. Mazama* was once a rival of Shasta and Rainier for the supremacy of the range. The greatest feature of the region is the enormous pit or caldera containing the lake. It is 4,000 feet deep, extending from the crest of the Cascade Range down half-way to the ‘sea-level. More than a square mile of its bottom is below the level of Klamath Lake, at the eastern base of the range. ‘The pit is half concealed by the lake which so greatly beauti- fies the scene. The volume of the pit is nearly a dozen cubic miles, and, if we include the lost peak, it would possibly be half again’as large. The problem presented for solution by the removal of such an enormous taass and the development of so great a completely enclosed pit in the process, is one which has not yet been completely solved. There are, however, some phe- nomena about the lake that throw considerable light upon it, especially when taken in connection with those of similar features in other parts of the globe. The composition and structure of the mass removed con- nects its transfer at once with volcanism, and the form of the * The ‘‘ Mazamas,” a society of mountain-climbers of Portland, Oregon, met at Crater Lake last summer, and christened the mountain, whose remnant still ncircles the lake, ‘* Mt. Mazama.” 170 J. S. Diller—Crater Lake, Oregon. remnants renders it necessary to suppose that it was either blown out by a tremendous volcanic explosion or swallowed up by an equally great engulfment. The occurrence of a distinct rim at once looks favorable to. the explosion method, for rims more or less complete, made up of the ejected materials, are known about many pits formed in that way.* As we have already seen, however, the rim is neither made up wholly or in any part of fragments blown out of the pit, but, on the contrary, is composed throughout of layers of solid lava, alternating with those of volcanic con- glomerate and tuff, all of which were erupted from Mt. Mazama before the pit originated. There are, indeed, great quantities of pumice in the region, but that is, at least in part, clearly associated with the last eruptions of Mt. Mazama, for it underlies some of the latest flows, and differs widely from the material (andesite) of which the basal portion of that great voleano was made up. The entire lack, about the pit, of material corresponding in kind, form or quantity to that which would necessarily arise, if the pit were produced by a great explosion, compels us to look in the other direction for the solution of the problem. While it is true, as shown by Lyell,t Scrope,t Judd,$ Geikie| and others that the majority of crater lakes occupy basins produced by volcanic explosions alone, there are others, which the same authors recognize as occupying areas a portion of whose depression is attributed to subsidence. Lake Lonar in India has a low rim of fragmental material, but its volume is much less than that of the pit which it surrounds. For this reason the pit is ascribed chiefly to subsidence. Major Dutton4 has pointed out a number of similar sunken areas or pits, but without lakes, in the Hawaiian Islands, and Dana has described** the sinking of the lava column (molten material), thus deepen- ing the pit, in the yet active voleano of Kilauea. The subsi- dence of the molten material in the throat of the voleano and * Volcanoes: The character of their phenomena, etc., by G. Poulett. Scrope,. p. 215. + Elements of Geology, sixth edition, pp. 679-82. +t Volcanoes, The character of their phenomena, etc. G. Poulett, Scrope, pp. 222-225. § Volcanoes, by J. W. Judd, pp. 170-171. || Text-Book of Geology, third edition, p. 240. {Major Dutton who made'a study of Crater Lake, after having visited the great volcanoes of the Hawaiian Islands, came to the conclusion that this depres— sion was ‘formed in the same manner as the great calderas of the Hawaii Islands” (U. S. Geol. Survey, Eighth Annual Report, p. 158), that is, ‘‘ by the dropping of a block of the mountain crust which once covered a reservoir of lava, this reservoir being tapped and drained by eruptions at much lower levels.” (U. 8. Geol. Survey, Fourth Annual Report, p. 105.) ** Characteristics of Voleanoes (1891) p. 127, and Manual of Geology, fourth. edition, pp. 284-5. J. S. Diller—Crater Lake, Oregon. 171 consequent collapse of the floor of the pit with portions of the surrounding cliffs, are shown to result from the escape of the lava through fissures at much lower levels upon the slope of the mountain. In connection with the subsidence of the lava in Kilauea in 1840 a mass was erupted 27 miles from Kilauea and about 2700 feet below the summit of its cliff. In some cases, however, the sinking of the lava in Kilauea is not known to have been accompanied by an eruption of lava upon the surface. The view that the caldera of Crater Lake originated by sub- sidence appears especially applicable for the reason that it occurs in the summit of a prominent range. The western slope of the range, although not steep, is considerably more inclined than that of Kilauea; and within 15 miles of the lake, the level of its deepest bed appears upon the surface. The engulfment of Mt. Mazama and the production of such an enormous pit in the process may well be expected to have given rise to eruptions upon the lower slopes of the range at no very great distance from the lake. That Mt. Mazama actually disappeared by subsidence is plainly suggested by the behavior of the last erupted lava. It is a rhyolite which escaped from the north slope of Mt. Mazama. Its broad stream follows a shallow valley that now appears in the rim at the head of Cleetwood Cove. Upon the outer slope of the rim, opposite the cove, there is a long depres- sion down the surface of the flow where cliffs, columns and angular blocks occur in great confusion. They have resulted from the caving in of the crust of the lava tunnel in the val- ley filled by the thickest portion of the flow. But on both sides the lava surface is smooth and easily traversed. Upon the inner slope of the rim is an equally remarkable ~ and exceptional feature. Descending from the Rugged Crest, at the upper end of the broken tunnel, to the lake is a flow several hundred yards in width. Itis part of the large stream of rhyolite which spread upon the outer slope of the rim, and has flowed inwards towards the lake over the broken ends of the older coulees of andesite. The fluidal structure, as well as the parting planes produced by it, are well marked approxi- mately parallel to the inner slope of the rim. It is the only inward-flowing lava found anywhere on the rim, and appears to indicate that before the thickest portion of the final flow of rhyolite had completely solidified, Mt. Mazama was engulfed, and the yet viscous lava followed it towards the abyss. It might be supposed that a study of the features upon the bottom of the great pit would disclose the character of the change by which it was produced; but this is not the case. Its original surface is in large part, if not wholly, covered by 172 J. S. Diller— Crater Lake, Oregon. the products of later voleanic eruptions. We have an excel- lent opportunity to study this portion of the pit on Wizard Island, where the bottom rises above the surface of the water. The island, near the center of the view in plate V, is formed of a cinder cone and lava field, which are practically unchanged since their recent eruption. The steep-sloped cinder cone, 845 feet in height, is sur- mounted by a perfect crater 80 feet in depth. The encircling lava field, made up of angular blocks of dark lava broken up at the time of the er uption, is rough in the extreme. Judging from Wizard Island, one might expect that there were other piles of recent cinders and lava upon the bottom of the pit, and so it seems; for according to the soundings of the lake made by Captain, now Major, Dutton in 1886,* there are two such prominences rising from great depths, but failing by over four hundred feet to reach the surface of the water. The eruptions of lava and fragmental material upon the bottom of the pit have partially filled it up, but how much has been lost in depth by these final outbursts cannot be estimated. To epitomise: The history of Crater Lake and its rim began in the upbuilding, by normal volcanic processes, of a large vol- cano, Mt. Mazama, comparable in the nature of its lavas, struc- ture and size with the greater peaks of the Cascade Range. Crater Lake did not then exist. Its site was occupied by Mt. Mazama, which was an active voleano in the glacial period. Glaciers descended from its higher slopes, scratching the rocks and depositing moraines about its base. The later eruptions of Mt. Mazama occurred in the glacial period and doubtless produced extensive floods which filled with debris the valleys of all the streams radiating from the mountain. In approximate connection with its final eruption, the molten material of the interior withdrawing, the summit of Mt. Mazama caved in and sank away, giving rise to a caldera nearly six miles in diameter and 4,000 feet deep. Thus origi- nated the great pit in which Crater Lake is contained, encircled by a glaciated rim, the hollow base of the engulfed Mt. Mazama. Upon the bottom of the caldera, volcanic activity continued. There were new eruptions building up cinder cones and lava fields and partially refilling the great pit. __Precipitation is greater than evaporation in that region. Volcanic activity ceasing, the conditions were favorable for water accumulation, and Crater Lake was formed in the pit. Washington, D. C. * Highth Annual Report, Part I, pp. 157-8. Adams, Barlow and Elis—Canadian Laurentian. 173 Arr. XVII.—On the Origin and Relations of the Grenville and Hastings Series in the Canadian Laurentian ; by Frank D. ApAMS and ALFRED E. BARLow, with remarks by RK. W. Euts.* As the exploration of the more remote portions of the great Canadian protaxis of the North American continent progresses, accompanied by the detailed mapping of its more accessi- ble parts, the true character, structure and origin of the Lau- rentian System is being gradually unfolded. The work of Logan during the early years of the Canadian Geological Sur- vey, though excellent in the main, is being supplemented and, in certain directions, corrected ; and as the work is now being pushed rapidly forward, it is believed that the time is not far distant when, difficult as the study is, we shall possess as com- plete a knowledge of these ancient rocks as we now do of Many more recent formations. In a paper which appeared in 1893,+ it was demonstrated that Logan’s “ Upper Laurentian” does not exist as an independent geological series, the anortho- sites, which were considered as constituting its main feature, being in reality great intrusive or batholitic masses; while in a subsequent paper,{ it-was shown that in the remaining por- tion of the Laurentian, two distinct classes of rocks could be distinguished, the first being beyond all doubt igneous rocks, and the second consisting of highly altered rocks of aqueous origin. In addition to these two classes of rocks of which the origin could be recognized, there was yet a third class, concern- ing the genesis of which there remained some doubt. Since the appearance of these papers, the present writers have been working together in mapping a large area (about 4800 square iniles) of the Laurentian in central Ontario, com- prising map-sheet No. 113, and a portion of 119, of the Ontario series of geological maps, the district lying to the north of Lake Ontario, along the margin of the Protaxis, and being especially well suited for purposes of study. Portions of three summers have already been spent in the district, and as two years more must probably elapse before the work can be completed, it is desired here to present a general outline of the results so far obtained, indicating certain conclusions which seem likely to be reached concerning the origin of the rocks in question. 3 The Fundamental Gneiss, as shown by the work of the Cana- dian Geological Survey, occupies by far the larger portion of the protaxis as a whole; while the Grenville Series has prob- * Published by permission of the Director of the Geological Survey of Canada. + Adams F, D.—Ueber das Norian oder Ober-Laurentian von Canada, Neues Jahrbuch fiir Mineralogie, Beilage Band viti, 1893. ¢ Adams, F. D.—A Further Contribution to our Knowledge of the Laurentian, this Journal, July, 1895. 174 Adams, Barlow and EHlls—Canadian Laurentian. ably its principal development along the southeastern margin, although as the exploration of this vast areais continued, new and possibly more extensive areas of these rocks may yet be found. Strata belonging to this series are already known to occur on the upper Manicuagan River, the lower Hamilton River, on the Manouan Branch of the Peribonka and on the lower part of the Ungava River, in the Labrador peninsula; while similar rocks, which would seem to belong to this series, but which have not as yet been thoroughly examined, have been met with about southern Baffin’s Land, and possibly about Baker Lake near the head of Chesterfield Inlet, as well as on the west coast of Hudson Bay and also at Cross Lake on the Nelson River. The Fundamental Gneiss consists of various igneous rocks closely allied in petrographical character to granites, diorites and gabbros, and which almost invariably have a more or less. distinct foliation. Where this foliation is scarcely perceptible it becomes very difficult to decide whether the rock is an intru- sive granite or diorite, or a very massive form of the gneiss in question. ‘The different varieties of gneissic rock alternate with or succeed one another across the strike, or sometimes cut one another off, suggesting a complicated intrusion of one mass through the other, but there is usually a general direction of strike to which, in any particular district, the foliation of all the varieties conform. The associated basic rocks are very dark or black in color and are usually foliated, but sometimes this foliation is absent and the rock occurs in masses of all sizes and shapes scattered through the acid gneisses, and in the great majority of cases so intimately associated with the latter that it is impossible to separate the two in mapping. The smaller of these masses can be distinctly seen to have been torn from the larger, which latter are often of enormous size. This process can be observed in all its stages. The granitic gneiss invades the great basic masses, sending off wedge-like arms into them, which tear them apart and anastomose through them in the most complicated manner. These smaller masses can then be observed to be separated into still smaller frag- ments, which either from the fact that they split most readily in the direction of their foliation or owing to subsequent movements, when the rock was in a more or less plastic condi- tion, often assume long ribbon-like forms. That great move- ments have taken place in the whole series during or after this invasion is shown by the complicated twisting of these darker bands and masses into all manner of curious and intricate forms, as well as in the frequent rolling out of great blocks of the amphibolite, after having been penetrated in all directions. by small pegmatite veins, resulting in masses of a dark basic gneissoid rock, filled with strings, bunches, separated frag- ments or grains of quartz or feldspar, giving to the mass a pseudo-conglomeratic appearance. , Adams, Barlow and Ells—Canadian Laurentian. 175 There can be but little doubt that the various gneissic rocks, constituting the more acid part of the series, are of truly igneous origin; and there is no evidence whatever of their having ever formed part of a sedimentary series. The true character of the more basic members is more uncer- tain, but they are probably closely related to the pyroxene granulites of Saxony, and doubtless represent either differentia- tion-products of the original magma, or basic intrusions whose structural relations and characters have been largely masked by the great movements which have taken place in the whole series at a later date. The Grenville Series differs from the Fundamental Gneiss in that it contains certain rocks whose composition marks them as highly altered sediments. These rocks are chiefly limestones, with which are associated certain peculiar gneisses, rich in silli- manite and garnet, having a composition approaching ordinary shale or slate, or else very rich in quartz and passing into quartzite, having thus the composition of sandstone. These rocks, as has been shown in one of the papers before referred to, usually occur in close association with one another, and are quite different in composition from any igneous rocks hitherto described. They are considered as constituting the essential part of the Grenville series. They usually, however, form but a very small proportion of the rocky complex in the areas in which they oceur, and which, owing to their presence, is refer- red to the Grenville series. They are associated with and often. enclosed by much greater volumes of gneissic rocks, identical in character with the Fundamental gneiss. The lime- stones are also almost invariably penetrated by masses of coarse pegmatite, and occasionally large masses of the limestone are found embedded in what would otherwise be supposed to be the Fundamental gneiss. The whole thus presents a series of sedi- mentary rocks, chiefly limestones, invaded by great masses of the so-called Fundamental Gneiss, and in which, possibly, some varieties of the gneissic rocks present may owe their origin to the partial commingling of the sedimentary material with the igneous rocks by actual fusion. There is, however, no reason to believe, from the evidence at present available, that any considerable proportion of the series has originated in the last mentioned manner. It will be readily seen that an exact delimitation of areas of the Grenville series is thus sometimes a matter of great diffi- culty, as they often appear to shade away into the Fundamental gneiss, aud it has hitherto been difficult in the case of the Grenville series to account for the existence of such a compar- atively small proportion of sedimentary strata, intimately asso- ciated with such great volumes of igneous gneisses. The relations of the two series, as determined by the investi- gations of the last two seasons, throws new light upon the sub- ject, and indicates the probable explanation of the difficulty. 176 Adams, Barlow and Hils—Canadian Laurentian. The northwestern half of the more restricted area at present under consideration is underlain by Fundamental Gueiss, pre- senting the characters described above. A smaller area of the same gneiss occurs at the southwestern corner of the area, in the townships of Lutterworth, Snowdon and Glamorgan, while in the southern and southeastern portions of the area there are other occurrences, which, however, present a-more normally granitic character. The southeastern portion of the area is underlain by rocks of the so-called Hastings Series, consisting chiefly of thinly-bedded limestones, dolomites, etc., cut through by great intrusions of gabbro-diorite and granite. These limestones and dolomites are usually fine-grained and bluish or greyish in color, with thin interstratified layers, holding sheaf-like bundles of hornblende erystals. As compared with the limestones of the Grenville series they are comparatively unaltered. They form beyond all doubt a true sedimentary series, and in the southeastern corner of the area are associated with conglomerates or breccias of undoubtedly clastic origin. Between the great area of Funda- mental Gneiss in the northwest, and the Hastings series in the southeast of the sheet, there lies an irregular-shaped belt of rocks, presenting the characters of the typical Grenville series as above described, the limestones having in all cases the form of coarsely crystalline, white or pinkish marbles, although more or less impure. The strike of the foliation of the Gren- ville series follows in a general way the boundaries of the Fundamental Gneiss, and is seen in an especially distinct man- ner to wrap itself around the long and narrow development of the gneiss exposed in the southwest corner of the area. Iso- lated masses of the limestone and gneiss characteristic of the Grenville series are also found in the form of outlying patches about its margin, as for instance in the townships of Lutter- worth and Stanhope. The relations of the Grenville series to the Fundamental gneiss are such as to suggest that in the for- mer we have a sedimentary series later in date than the Funda- mental Gneiss, which has sunk down into and been invaded by ~ intrusions of the latter series when this was in a semi-molten or plastic condition. The limestones, while themselves rendered more or less plastic by the same heat which softened the lower gneisses, do not show any distinct evidence of absorption or solution by the invading rocks, unless some of the highly gar- netiferous gneisses usually associated with the limestones are formed by a commingling of the two rocks. Masses of the highly crystalline limestone or marble in some cases lie quite isolated in what are, to all appearances, the lower gneisses, as if they had been separated from the parent mass, and had passed outward or downward into the gneissic magma. The contact of the Fundamental Gneiss and the Grenville series would appear therefore to be a contact of intrusion, in very many cases at least. Adams, Barlow and Ells—Canadian Laurentian. 177 The question of the relations of the Grenville series to the Hastings series then presents itself. Although repeated tray- erses have been made from one series into the other, no sharp line of division has been found. Towards the southeast the limestones of the Grenville series in many places, though still highly crystalline, seem to be less highly altered, and finally, as the Hastings series is approached, present in places the bluish color of the limestones of the latter series; so that it is often impossible to determine to which series they should be referred. The limestones of both series also have the numer- ous small interstratified gneissic inclusions or bands so fre- quently referred to in the descriptions of the limestones of the Grenville series, making the resemblance still more complete. In fact, although the true relations of the two series are obscured by the presence of numerous great intrusions of gran- itic and basic pyroxenic rocks, and can only be determined with absolute certainty by the completion of the mapping, the Investigations so far indicate that in the region in question the Hastings series would seem to represent the Grenville series in a less altered form. In other words, the Hastings series, when invaded, disintegrated, fretted away and intensely metamor- phosed by and mixed up with the underlying magma of the Fundamental Gneiss, constitutes what has elsewhere been termed the Grenville series. The Grenville series may, how- ever, represent only a portion of the Hastings series, and the work so far done in this district has not been suflicient to determine the stratigraphical position of this portion. Concerning the age of the Hastings series but little is known as yet. To the southeast of the area under consideration, how- ever, its clastic character is well marked, breccias and conglom- . erates, often greatly deformed by pressure, being present as well as certain fine-grained and comparatively unaltered lime- stones, in which a very careful search may yet be rewarded by the discovery of fossils. Both lithologically and stratigraphi- cally the rocks bear a striking resemblance to rocks mapped as Huronian in the region to the north and northeast of Lake Huron, and it seems very likely that the identity of the two series may eventually be established. The two areas, however, are rather widely separated geographically, so that the greatest care will have to be exercised in attempting such a correlation. Like the Grenville series, the rocks of the Hastings series are unconformably overlain by and disappear beneath the flat- lying Cambro-Silurian rocks of the plains, which limit the pro- taxis on the south and are separated from it in time by an immense erosion interval. Further investigation in this area, as well as in that adjoining to the east, now being mapped by Dr. R. W. Ells, will, however, it is hoped, before long throw additional light on the age of this very interesting and impor- tant series of rocks. If further investigation proves that the relations of the several series have been correctly diagnosed, 178 Adams, Barlow and ELlls—Canadian Laurentian. and that the explanation of these relations as given above is correct, the Laurentian system of Logan will resolve itself into an enormous area of the Fundamental Gneiss, which is essen- tially of igneous origin and which there is every reason to believe forms part of the downward extension of the original crust of our planet, perhaps many times remelted and certainly in many places penetrated by enormous intrusions of later date ; into which Fundamental Gneiss, when in a softened condition, there have sunk portions of an overlying series, consisting chiefly of limestones. Farther east, in that portion of the province of Quebec where the Grenville series was first studied by Logan, the rocks of the Hastings series proper have not been recognized. The Lower Paleozoic strata rest directly upon the Grenville series and would cover up the Hastings series to the south should it extend as far east as this. The limestones of the Grenville series, moreover, here extend much farther back from the edge of the protaxisin bands and streaks conforming to the strike of the underlying gneissic rocks, so that the origin of the series and its relations to the Fundamental Gneiss is not so clearly indicated. When, however, its relations here are interpreted in the light of the Ontario occurrences, there seems to be no reason why the same explanation might not be offered to account for. its origin also. The bands of limestone, which often vary in thickness from place to place, and are frequently interrupted in their course or abruptly cut off, might be con- sidered as having taken their form from long folds in the series from which they were derived as it settled down into the magma beneath, or as having been separated by great lateral intrusions of the gneissic magma. Their original shape and character has, however, without doubt been greatly altered by the enormous movements to which both series of rocks have been subsequently subjected. If again this proves to be the true explanation of the rela- tions of these series, the Grenville series will cease to be an anomaly among our Archean formations and will, so far as its mode of occurrence is concerned, bear the same relation to the Fundamental Gneiss as the Huronian does farther west in the Lake Superior and Huron district, as shown by Lawson and Barlow; the similarity in position, however, not imply- ing identity in age. The recognition of the Grenville series as consisting of a series of sedimentary rocks, largely limestones, invaded by igneous material which now makes up by far the greater por- tion of the series and consists largely of extravasations of the Fundamental Gneiss, is now pretty certainly established by the field evidence. Its recognition as a portion of the Hastings series which has been intensely metamorphosed, will probably be more clearly established as the field work progresses. Since subordinate areas of the Grenville series also occur to the Adams, Barlow and Ells—Canadian Laurentian. 179 south of the St. Lawrence in the Adirondack region, and are now being mapped, it will be of great interest to ascertain whether the same relations do not also exist in that area, and whether a continuation of the Hastings series to the south can- not be recognized in the “ Huronian Schist” of St. Lawrence and Jefferson counties, shown upon the Geological Map of the State of New York, which has just been issued by the Geolog- ical Survey of this State. It is perhaps unnecessary to draw attention to the fact that the recent investigations of Messrs. Wolff, Brooks, Nason, Kemp, Westgate and others on the crystalline limestones of New Jersey have a certain bearing on this subject. Remarks by R. W. Ells: | In connection with the statements advanced in the preceding paper by Dr. Adams and Mr. Barlow, it is but right that the conclusions arrived at from the study of the similar rocks in their eastern and northern extension should be stated. The investigations in this quarter have now been carried on for six years, and have extended over a very large area to the north of. the Ottawa, in which is included the typical Grenville series of Sir W. E. Logan, and extending far up the Gatineau River; while to the westward, the work has been carried on till the vicinity of the area, described in the accompanying paper, has been reached. It may be said therefore that the detailed examination of the rocks which make up the Gren- ville and Hastings series has extended over an area about 250 miles in length by 75 miles in breadth. In the early days of the study of these rocks much difficulty was experienced. Firstly there was a great and almost inac- cessible wilderness, the only available means of travel over the greater portion being by canoes ; and in the second place there was an almost entire lack of trained observers to carry on the work. Add to this the entire absence of microscopical deter- minations, and one can readily comprehend the difficulty expe- rienced in the attempt to solve this most difficult of the problems in Canadian geology. Foliation and stratitication were considered conclusive evi- dence of sedimentation, and as most of the rocks of the great Laurentian complex gave evidence of these forms of structure, the inference naturally followed that the greater portion of the gneissic, granitic and anorthositic rocks were of sedimen- tary origin. So far was this sedimentary theory carried out that, in the earlier reports of the Geological Survey, even the masses of binary granite and many of the pyroxenic rocks were included in the same category. This was at the time a very natural conclusion, since many of these masses have a regular bedded structure and conform, over very considerable areas, to the regular stratification of the rocks, either gneiss or crystal- line limestone. As the country became more accessible the 180 Adams, Barlow and FLlls—Canadian Laurentian. field investigations showed very clearly the intrusive nature and later age of many of these masses, while the aid of the: microscope fully established the non-clastic and igneous char- acter of the great bulk of the gneisses. The more recent and probably sedimentary origin of the limestones and associated gneisses of the Grenville series, as distinct from the great. mass of the underlying Laurentian Fundamental Gneiss, was. pointed out some years ago in a paper by the author, read before the Geological Society of America. The subsequent investigations on these rocks, to the west and southwest, showed that the conclusions then presented were correct, but that as the work extended westward to the south side of the Ottawa the character of the various groups of rocks gradually changed. The areas of limestone became much more extensive, and there was. a large development of hornblende and other dark-colored rocks, rarely seen to the north of the Ottawa. The limestones also were very often highly dolomitic, and in certain areas were blue and slaty, with but little of the aspect of the Grenville limestones, except where they were in close contact with masses of intrusive granite or diorite. There is also in the rocks of this group to the south of the Ottawa, where they have been styled the Hastings series, from the fact that they were first studied in the county of Hastings, a very consider- able proportion of schists, micaceous, chloritic and hornblendic,. with certain regularly slaty beds, and others of true conglom-. erate containing quartz pebbles. In certain portions the litho- logical resemblances between the Grenville and Hastings rocks are very close, and they may, for all practical purposes, be regarded as one and the same series. From a number of sections made in the counties of Renfrew on the south of the Ottawa, and in Pontiac, to the north of that river, it would appear that the original Grenville limestones and associated grey and rusty gneiss form the lower part of the series, since it is only on their development westward towards the typical Hastings locality that the characteristic Hastings schists and associated strata are met with. In character and general aspect these rocks of the Hastings series are almost identical with many of these which in the Eastern Townships and in New Brunswick have been regarded as probably Huronian for many years; and so marked is the resemblance that the author, in presenting his summary report for 1894, referred the rocks seen near the Bristol iron mines: to that division. It now appears very conclusively established that both in the eastern and western areas we have a well developed series of rocks, including limestones, gneiss and schists, which are of undoubted sedimentary origin, but which: have been enormously acted upon by great intrusive masses as- well as by other dynamic agencies, so that in many parts their original characters have almost entirely disappeared. Beecher— Natural Classification of the Trilobites. 181 Art. XVIIIl.—Outline of a Natural Classification of the Trilobites ; by CHARLES E. BEEcHER. (With Plate III.) [Continued from page 106. ] Arrungement of the Hamilies of Trilobites. Susctass TRILOBITA. Order A. Hypoprarta. Family 1. Agnostidee. Family 8. Trinucleide. Family 2. Harpedide. Order B. OPpisTHOPARIA. Family 4. Conocoryphide. Family 8. Brouteide. Family 5. Olenide. Family 9. Lichadidee. Family 6. Asaphide. Family 10. Acidaspide. Family 7. Proétide. Order C. PROPARIA. Family 11. Encrinuride. Family 13. Cheiruride. Family 12. Calymenide. Family 14. Phacopide. The order Opisthoparia, with nearly one hundred and fifty enera, has a much greater geological distribution than either of the others, and was by far the dominant group during the Cambrian and Ordovician, being represented by about eighty- five genera in the former age and forty-five in the latter. Nineteen genera of this order are found in the Silurian and ten in the Devonian, most of them having continued on from older ages. our genera represent the order in the Car- boniferous and one in the Permian, thus marking the extinc- tion of the subclass as well as the last genera of the Opisthoparia. The comparative abundance and duration of the three orders are expressed in the table on page 182, from which it appears that the Hypoparia probably culminated in pre-Cambrian times, the Opisthoparia during the Cambrian, and the Proparia during the Ordovician. In the following classification, the families adopted by Salter** and Barrande’ are in the main adhered to, and the number cor- responds very closely with that in Zittel’s “ Handbuch der Palzontologie” ** and also in the “ Grundziige”’ ** of the same author. The order of arrangement, however, is very different. A great number of family divisions have been proposed, and undoubtedly many others will yet be made, but it is not within the province of this paper to determine the precise value and Am. Jour. Sci.—Fourts Series, Vou. III, No. 15.—Marcu, 1897. 13 182 Beecher—Natural Classification of the Trilobites. limitations of the families. This would require discussions of priority and synonymy, and otherwise detract from the direct purpose of the writer; viz., to establish a basis for a natural classification, and in this way to apply what is currently known and accepted regarding the trilobites. Nevertheless, some notice must be taken of several families and genera which for various reasons do not appear here. The family Aglaspide, including the genus Aglaspis Hall, proves to belong to the Merostomata, and is therefore omitted. The family Bohemil- lide has been shown by the writer’ to have no foundation, because the type of the genus Lohemilla Barrande was based upon a mutilated specimen of Aglina. Hypoparia Opisthoparia. Proparia. TRILOBITA. Permian. Carboniferous. Devonian. Silurian. Ordovician. Cambrian. Pre-Cambrian. Table of Geological Distribution of Trilobites. Several genera still commonly adopted are not here recog- nized in the lists under the families, since from the minute size of the individuals described and their immature charac- ters, they must be considered as the young of larger forms. Such are Conephrys Callaway, Cyphoniscus Salter, Holometo- pus Angelin, /socolus Angelin, and Shumardia Billings. TZrzo- pus Barrande has been shown to be a chiton. Much could be said against some of the recognized genera, but, as with the families, the writer has preferred in almost every case to adopt, for the present, what has been commonly accepted, and thus to avoid the entanglements of dates and synonyms which would be out of place in any general discus- sions. ‘The type species of every genus is here made the cen- tral idea. It is taken as representing the genus more closely than any fortuitous assemblage of diverse species, which the Beecher—WNatural Classification of the Trilobites. 188 next investigator may show belong to another or to several genera. Our ideas of a genus are naturally based mainly upon the species with which we are most familiar. Until forced to make authoritative comparative statements, it does not occur to one that the type of the genus under consideration may be quite different. An American student’s conception of Homa- lonotus will probably be formed largely upon the species com- monly known as H. delphinocephalus Green, from the Niagara, and 77. DeKayi Green, from the Hamilton. The first time the type of the genus, H. EKnightt Murchison, is seen he will be puzzled to place it. Similar examples could be multiplied indefinitely, and only show that the type must be taken as the ultimate unit of comparison. Diagnoses and Discussions of Orders and Families. Order A. HYPOPARIA, n. ord. (urd under, and wapea cheek piece.) Free cheeks forming a continuous marginal ventral plate of the cephalon, and in some forms also extending over the dorsal side at the genal angles. Suture ventral, marginal, or submar- ginal. Compound paired eyes absent; simple eyes may occur on each fixed cheek, singly or in pairs. Including the families Agnostide, Harpedide, and Trinu- cleide. This order includes the groups O and D, or the Ampycini and Agnostini of Salter, and also the family Harpedide of that author, which he included in the Asaphinz. The special recognition of characters, however, between Salter’s groups and the order here proposed is different. The presence of a part homologous with the free cheeks of other trilobites has generally been more or less overlooked in the families of this order. In Zrinucleus, Dionide, and Hurpes, the sutures have been correctly determined by Bar- rande.® Likewise, Angelin’ gave the right structure in Ampyz, but in Agnostus, this feature has escaped notice. The exami- nation of extensive series of Agnostus, in the National Museum and i the Museum of Comparative Zodlogy,* has proved that under favorable conditions of preservation this genus shows a distinct plate, separated from the cranidium by a suture, and it can be compared only with the free cheeks in other trilobites, * In the former, through the courtesy of C. D. Walcott and OC. Schuchert, and in the latter, of A. Agassiz and R. T. Jackson. | 184 Beecher-——Natural Classification of the Trilobites. especially where they are continuous around the front of the cephalon, as in Zrinucleus and Ampyx. The presence of a hypostoma in Agnostus was also determined. Even in the higher genera of this order, the suture is frequently unnoticed in descriptions, but it can be seen in all well-preserved speci- mens. In Zrinucleus” and Harpes, it follows the edge of the cephalon, and separates the dorsal from the ventral plate of the pitted limb. Since eye spots occur on the fixed cheeks in the young Zrinucleus and adult LHarpes, it is probable that this character is a primitive one in this order, and has been lost in Agnostus, Microdiscus, Ampyx, and Dionide. | The ontogeny of Sao, Ptychoparia, Triarthrus, Dalmanites, etc., shows that the true eyes and free cheeks are first devel- oped ventrally, appearing later at the margin, and then on the dorsal side of the cephalon. Therefore, the Agnostide, Trinu- cleide, and Harpedide have a very primitive head structure, characteristic of the early larval forms of higher families. Other secondary features show that this order, though the most primitive in many respects, is more specialized that either of the others, except in their highest genera. The characters referred to are the glabella and pygidium. Very few species show the primitive segmentation of the glabella, it being usually smooth and inflated, and resembling in its specializa- tion such higher genera as Proétus, Asaphus, and Lichas. The pygidium often fails to indicate its true number of seg- ments. Some Agnostus and Microdiscus show no segments either on the axis or imb of the pygidium. Z?rinucleus and others may have a many annulated axis and fewer grooves on the pleural portions. The number of appendages corre- sponds to the axial divisions, as determined by the writer.* The multiplication of segments in the pygidium and their con- sequent crowding makes them quite rudimentary. Family I. AgGnostrip2£ Dalman. Small forms, having the cephalon and pygidium elongate, nearly equal, and similar in form and markings. Free cheeks ventral, continuous; suture marginal or ventral. Eyes want- ing. Thorax composed of from two to four segments, with grooved pleura. Cambrian and Ordovician. Including the genera Agnostus Brongniart, and Microdiseus Emmons. é The genera in this family are primitive in their form and structure, as shown by their ventral free. cheeks, marginal or ventral suture, elongate cephalon, and large pygidium. Some Beecher—Natural Classification of the Trilobites. 185 species have spines at the genal angles, corresponding to the interocular spines of Holmza, and young Hlliptocephala, and not to the spiniform projections of the free cheeks. From their abbreviated thorax, and progressive loss of annulations on the glabella and axis of the pygidium, they must also be considered as degraded. J/icrodiscus, the earlier genus, has three or four free segments, and in some species (J/. speciosus Ford) preserves the normal pentamerous glabella and annulated pygidial axis, while the later genus, Agnostus, has but two free segments, and has lost the annulations of both glabella and pygidinm. Matthew” has described the protaspis stage of Microdiscus, which agrees with the similar stage of Ptychopa- ria and Sao. Fully a dozen generic names have been proposed for forms of the general type of Agnostus, but none of them has ever come into current use. Nine were first published by Corda,” but as Barrande* subsequently showed that one was based on an Orbicula, another on a poor specimen of #glina, and three others on a single species, this grouping soon fell into disuse. Moreover, Barrande was inclined to give no generic value to the form and lobation of the glabella, and therefore all the species were placed by him in the single genus Agnostus. At the present time, more weight is given to the characters of the glabella and pygidium, as indizating generic differences in dor- sal and ventral structure, so that further study may show the desirability of restoring such of Corda’s names as were founded upon natural groups of this family. Family I]. Harprpipa Barrande. Cephalon large, margined by a broad expansion or limb; glabella short and prominent. Free cheeks ventral, continu- ous; suture marginal, following the outer edge of the limb. Paired simple eye spots, or ocelli, single or double, at the dis- tal ends of well-marked eye lines on the fixed cheeks, extend- ing outward from the glabella. Thorax of from twenty-five to twenty-nine segments, with long grooved pleura. Pygidinm (in Harpes) very small, composed of but three or four seg- ments. Cambrian to Devonian. Including Harpes Goldfuss, Harpina Novak, and Harpides ? Beyrich. : The genus Harpes presents considerable variation in the lobes of the glabella. ZZ. wngula Sternberg shows the full num- ber of five lobes, but in some species, as H. @ Orbignyianum ey UR Ow 186 Beecher—Natural Classification of the Trilobites. Barrande, the structure is like Cyphaspis, with separate basal lobes. Arvraphus Angelin was apparently based upon a specimen of Harpes denuded of the pitted border. Harpides Beyrich is imperfectly known but seems to belong here. The ocular ridges and tubercles on the fixed cheeks, the broad limb, the glabella, and the narrow weak thoracic. segments are all in accord with Harpes, though in other features it has affinities. with the Conocoryphide. In many respects, Harpes is one of the most interesting genera of trilobites since it is so unlike other forms. The broad hippocrepian pitted limb of the cephalon has its counter- part in Zrenucleus and Dionide, although not so well devel- oped in these genera. The head is also comparatively longer and larger, both features being decidedly larval. It is the only family known in which functional visual spots, or ocelli, are situated on the fixed cheeks. The young Z7rinucleus has similar eye spots, or ocelli. The great number of free seg- ments in the Harpedide is another primitive character, although the cephalon (in Harpes) still remains larger than the thorax and pygidium. Family Ii] TrinucLtem Barrande. Cephalon larger than the thorax or pygidium; genal angles. produced into spines. Free cheeks continuous, almost wholly ventral, carrying the genal spines; suture marginal or submar- ginal. Paired simple eyes or ocelli generally absent in adult forms; compound eyes wanting. Segments of thorax five or six in number, with grooved pleura. Pygidium triangular; margin entire; axis with a number of annulations; limb grooved. Ordovician and Silurian. Including the genera and subgenera Z7rinucleus Lhwyd, Ampyx Dalman, Dionide Barrande, Andymionia ¢ Billings, Lonchodomus Angelin, Raphiophorus Angelin, and Salterza ? W. Thompson. The leading genera of this family form a tolerably homoge- neous group, although each has sometimes been recognized as characterizing a separate family. Zrinucleus and Dionide have a broad pitted border, but this hardly seems of sufficient importance to remove them far from Ampyz, since the three genera agree in nearly all important structural details, as the extent and character of the free cheeks, the glabella, the num- ber of free segments, and the character of the pygidium. Lon- chodomus and Leaphiophorus of Angelin are commonly ad- mitted as subgenera of Ampy«. Beecher—Natural Classification of the Trilobites. 187 Both Salteria W. Thompson and Andymionza Billings have been described as subgenera of Dzonide Barrande, though there is little positive evidence for this disposition of them. Until more perfect material representing these forms has been described, it will not be possible to decide satisfactorily upon their relationships or place in a classification. Therefore, they are left with doubt in the present family. Order B. OPISTHOPARIA, n. ord. (émucGe behind, and zapea cheek piece.) Free cheeks generally separate, always bearing the genal angles. Facial sutures extending forwards from the posterior part of the cephalon within the genal angles, and cutting the anterior margin separately, or rarely uniting in front of the glabella. Compound paired holochroal eyes on free cheeks, and well developed in all but the most primitive family. Including the families Conocoryphide, Olenide, Asaphide, Proétidz, Bronteidee, Lichadidze, and Acidaspide. This order is nearly equivalent to group B, or the Asaphine of Salter, which included. also the families Calymenidz and Harpedide, which belong elsewhere. The families which are here placed under this order lend themselves quite readily to an arrangement based upon the characters successively appearing in the ontogeny of any of the higher forms. Thus Sao, Ptychoparia, and other genera of the Olenide have first a protaspis stage only comparable in the structure of the cephalon with the genera of the preceding order, the Hypoparia. Therefore this stage does not enter into consideration in an arrangement of the families of the Opisthoparia. In the later stages, however, there is a direct agreement of structure with the lower genera of this order. The nepionic Sao, with two thoracic segments (Plate ITI, fig. 2), has a head structure agreeing in essential features with that in Atops or Conocoryphe (Plate III, figs. 14,15). A later nepionic stage, with eight thoracic segments (Plate III, fig. 3), agrees closely with adult Ptychoparva or Olenus (figs. 16, 17). These facts clearly indicate that the family Conocoryphide should be put at the base of this extensive order. Then, as Ptychoparia and Olenus are more primitive and simple genera than Sao, they, as typifying the family Olenidze, should govern its posi- tion, which accordingly would be next after the Conocoryphide. In each case, a family is considered as represented by its typi- cal and most characteristic forms. It would be impossible to 188 Beecher—Natural Classification of the Trilobites. consider the advanced specialized genera of some families as representing their normal facies, for each one has undergone | an independent evolution, and some characters have reached - as great a degree of differentiation as will be found in much higher families. It has been recognized that variations in the positions of the eyes, the relative size of the free and fixed cheeks, and the degree of specialization of the glabella have a definite order in the ontogeny of any trilobite, and also that these characters have a greater taxonomic value than many others. Applying these principles in arranging the families which come under the Opisthoparia, we have the sequence as indicated above, beginning with the Conocoryphide and followed by the Olenidse, Asaphidee, Proétidee, Bronteidee, Lichadide, and Acidaspidee, in regular progression. See Plate ILI, figs. 14-23. Family IV. Conocorypuipa Angelin. Free cheeks very narrow, forming the lateral margins of the cephalon, and bearing the genal spines. I ixed cheeks large, usually traversed by an eye line extending from near the ante- rior end of the glabella. Facial sutures running from just within the genal angles, curving forward, and cutting the ante- rior lateral margins of the: cephalon. Eyes rudimentary or absent. ‘Thorax with from fourteen to seventeen segments. Pygidium small and of few segments. Cambrian. Including the genera and subgenera Conocoryphe Corda (= Conocephalites Barrande), Aneucanthus Angelin, Atops Emmons, Avalonia Walcott, Barliella Matthew (=Salterca Walcott and Hrinnys Salter), Bathynotus Hall, Carausia Hicks, Carmon Barrande, Ctenocephalus Corda, Dictyocephal- ites Bergeron, Hryx Angelin, Larttza Walcott, and Towotrs Wallerius. The genera coming under this family present a number of very primitive characters such as are shown only in the larval stages of higher forms. ‘The free cheeks are narrow and mar- ginal, and can be compared with those in the nepionic stages of Sao and Ptychoparia. The eyes have not been detected, but the presence of an eye line suggests their possible existence. The variations of the glabella are very marked, and are as great as those which in higher forms attain some importance as family characteristics. In Zoxotes, Carausia, and Aneucanthus, the glabella expands in front, joining and forming part of the anterior margin, as in the glabella of the larval stages of Solen- Beecher— Natural Classification of the Trilobites. 189 opleura, Liostracus, Piychoparia, and Sao. Ctenocephalus and Hryzx are slightly more advanced, as the glabella no longer marks the edge of the cephalon. In is the diffusivity thus stated AX = 86400 « and is OG) A IN Of course the tabulated values of 6 answer to a period for *The values of g were computed from the seven-place table of the integral given by Lord Kelvin, Phys and Math. Papers, vol. iii, 1890, p. 434. The inter- polations were made by a known formula of the same order of accuracy as the formula for second differences, and the results were tested by substitution in the formula for second differences. Seven-place logarithms were used. The abbrevia- tion to five places introduces an apparent inaccuracy, inasmuch as the log 2g, as tabu- lated, sometimes varies in the last place from the five-place log. of the tabulated value of 2g. The tabulated logarithms are, however, the nearest in five places to the values of 2q expressed in seven places. 284 G. F. Becker—Computing Diffusion. which A= 1/k, or in the ease of salt a period of 100,000 days, while the values for y correspond for salt to a lapse of time of 100,000 years. In connection with this table it may be convenient to set down a few diffusivities interesting to the geologist, either immediately, or for comparison with the diffusivity of rock magmas. The data for the diffusion of heat and motion are taken from Lord Kelvin’s memoir on Heat. The diffusivities of solutions were all originally published in terms of days, not seconds. In the list below they are stated both in this way under A and in seconds under «x. In choosing illustrations from the many experiments of Scheffer and of Schuhmeister* I have when practicable selected two nearly at the same temperature with different concentrations, and a third at a different tem- perature but with concentration as nearly as may be equal to that of one or other of the first two. There seems no doubt that « in the case of liquids varies both with concentration and temperature. The variations of « however, are far from com- plete elucidation. Mr. Schuhmeister remarks that “the rapidity of diffusion runs almost exactly parallel with the larger or smaller values of the coefficient of friction,” but makes no further comment on the relations of viscosity to diffusivity. In an investigation undertaken by Mr. A. Sprungt in Prof. Wiedemann’s labora- tory, for the purpose of determining the viscosity of salt solu- tions over wide ranges of temperature and concentration, diagrams and tables are given which enable one to fix the viscosities of seven salts of which Mr. Schuhmeister has deter- mined the diffusivities for the same temperature and for nearly the same concentration. Sprung’s concentrations are given in terms of the weight of anhydrous salt per 100 parts by weight of solution. Schuhmeister’s data are for weight of anhydrous salt per unit volume of solution. The concentrations beng 1/10 and the temperature 10° C., « the diffusivity, and pw the coeflicient of viscosity, the results of the two observers are expressed in the first three lines of the following little table. Salt KBr KI KCl NaCl g (aCe? Neeso! mogoo™ kX 10° 13°08 12°96 12°73 972 (etl 7°64 3°24 pe 0°126 0°124 O28 O1s6 Ores 07179: 0247 Ke ehO sO, 207 On99 0°208 O0°2386 0°224 0°244 0°398 *Mr. J. Schuhmeister’s investigation, undertaken at Prof. J. Stefan’s advice, appeared in Wieu. Sitz. Ber. Ak Wiss., vol. xxix, 1879, Part II, p 603. At the close of the paper mean values of A are given for 10°C. Stefan recomputed Graham’s diffusion experiments. His paper is in the same volume as Schuh- meister’s, p.161. Mr. J. D. R. Scheffer’s memoir on diffusion is published m Verh. Kon. Ak. van Wet., xxvi, Part, 1888, with separate pagination. + Pogg. Ann, vol. clix, 1876, page 1. ca a wi la he at. CRI Ig lc cE te G. F. Becker— Computing Diffusion. 285 In the last line I have shown the product «p’, which evi- dently varies slowly and irregularly. Expressed in two figures the average of the two culpbates gives 0°22; the average of the three chlorides gives 0°22; the average of the five alkaline salts gives 0°22; the average for the seven compounds is 0°22. The average for the three potassium salts on the other hand is only 0-20, and the average for the two sodium salts is 0°24. The variation of «u* from a single value is perhaps not greater than might be expected from the data on diffusivity (which, as Mr. Schuhmeister himself points out, accord only approximately) if xp” were really a constant. It is clear that the hypothesis—diffusivity is inversely proportional to the square of the viscosity for uniform temperature and concentra- tion, expresses approximately the facts for these seven com- pounds. The number of compounds is small, but it embraces salts of four bases and of four acids showing rather a wide range of diffusivities and of viscosities. The probability that the accord is a mere coincidence seems to me extremely slight, and I infer that the hypothesis just stated probably expresses, at least roughly, the relations of diffusivity to viscosity for an important class of compounds. Should it prove that solutions such as rock magmas are to be included in this class, it would greatly facilitate the discussion of rock differentiation. Some approach to a quantitative determination of the viscosity of. lavas might be attained. Direct investigation of their dif- fusivity would seem impossible. Liquids diffusing into water. Solution of Temp. Concentration. A K Authority. Chlorhydric acid, 11° 1HC1/7:17H?O 2671 -000 030 91 ) os 11°, 1HCI/108-4H?0 1837 -000 021 26 | a e Nes 1HC1/6 86H?O 27080 -000 024 07 !& Sulphuric acid, 75° 1H°S0!/685-7H20 1:042 -000 012 06 re He See tE2SO2/36 270) 1-008 -000 01L 6i | a : 13s 1H°S04/35°4H?0 1244 -000 014 40 J Sodium chloride, HOTA al A Sy per em? 810 -000 009 38 ) e Sse =F 322638 ‘798 000 009 24 : aie SOTO Y "9015 ‘000 010 43 | ‘© (Mean), 10°° 10 ‘840 -000 009 72 Sodium carbonate, Spe = 0998 86 Es ‘319 -000 003 69 | - So 1347103 ae 375 =°000 004 34 3 20755 2 182442 : °613 -000 007 10 = ‘* (Mean), 10-2 aES 390 000 004 51 | = Sodium sulphate, Disarm 04 O19 i 614 000 007 11 [2 LOT OT3 24 ss “678 000 007 85 } & io(Mean); 107° - -10 660 -000 007 63 | Copper sulphate, eel Oly ae (7) 0°3859 0°3864 0°0005 + ae * (8) 0°3859 0°3860 0°0001 + - Ss (9) 0°0772 O754 0°0018— 300 5 (10) 0°0772 0°0757 0°0015 — & “s (11) 0°1543 0°1532 0:;001LI— ¢ f (12) 071544 0°1524 0°0020— s i (13) 0°0772 0°0744 0°0028 — 500 by) (14) 0°0772 0°0737 0°0035 — oh os (15) 071544 071521 0°0023— a a aye esO1544 - 0-1512 ~~ 0:0032— ee ce (17) 0°3859 03827 0:00382 — og (18) 0°3859 0°3831 0°0028 — - - (19) 0:0772 0:0744 0:0028— 500 10 (20) 0°0772 0°0757 00015 — A va (21) 0°3859 0°3828 0°0031— rH os (22) 0°3859 0°3827 0°0032— . os Blank tests made upon a solution obtained by mixing the maximum amount of the iodate with 5™ of dilute sulphuric acid (1:3), neutralizing as usual with potassium bicarbonate, adding the iodide from the trap and 5° of starch emulsion, showed that a single drop of iodine was invariably sufficient to bring out the starch blue. Occasionally it was found that the mixture, particularly when chlorides or bromides were present, of itself developed a trace of color, but by no means a read- ing tint. 0-5 ey. (7) 00772 0-°0802 0:0080-- Si 0-2 (8) 00773 0:0853 0:0080+ is 0-2 (9) 00772 00873 95 O-0lOlee ps 0°5 (10) 0:07 7271) 10-0861 ) 0:0089E ea 05 4 (11) 071544 01646 0:0102+4 i 0°5 (12) 01548 0:1626 0:0083+ a 0°5 The influence of sodium chloride and potassium bromide in increasing the amount of iodine liberated is plain. The increase comes without doubt from the iodate, and is doubtless due to the formation of iodine chloride or bromide, during the Gooch and Walker—Analysis of Lodides. 299 acidifying, by the interaction of the free iodine, the iodie acid, and the hydrochloric or hydrobromic acid, according to the reactions previously discussed. It is plain, therefore, that the valne of the process in the determination of iodine in an iodide is restricted of necessity to those cases in which it is known that chlorides or bromides are not present to any con- siderable extent. For determining) the standard of a solution of nearly pure potassium iodide, employed in so many labora- tory processes, it should find useful application. “ TaBLe IV. Analysis of Pure Potassium Iodide. KI taken. KI found Error. grm. grm. grim. (1) 0°0814 0°0816 0°0002 + (2) Osta. 0:0813 00001 — (3) 0°0814 0:0805 0 0009 — (4) 00815 00809 00006 — (5) 0°0814 0°0808 0°0006 — (6) 0:0814 0°0806 0:0008 — (7) 00814 0°0812 0°0002 — (8) 0:1628 = 021624 0°0004 — (9) 0°1628 01617 0:0011— (10) 0°1628 01621 0:0007 — (11) 0°1628 0°1619 0:0009 — (12) 0:1628 0°1624 0:0004 — (13) 0°1628 0°1621 0:0007 — (14) 0°1628 0°1626 0:0002 — (15) 0°2442 0:2451 0°0009 + (16) 0°2442 0°2442 0°0000 (17) 0°2442 0°2439 0:0003 — (18) 0°3256 0°3258 0°0062 + (19) 0°3256 0°3256 0:0000 (20) 0°3256 0°3258 0:0002 + (21) 0°3256 0°3272 00016 + (22) 0°3256 0°3256 00000 (23) 0°4071 0°4076 0°0005 + (24) 0°4071 0°4080 0°0009 + (25) 0°4071 0°4073 0°0002 + e \ In Table IV are comprised a number of experiments made exactly like those which seemed to give the best results in the series of Table I]. The iodide and an excess of iodate (5° of Vf é NG the 0°5 per cent solution to every 20°™ of Fi iodide) were made to interact in a volume of about 150™°, 5° of sulphurie acid (1 : 3), were used to bring about the reaction, 10° of potassium bicarbonate were added after the neutralization of the sulphu- 300 Gooch and Walker—Analysis of Lodides. ric acid was complete, and the free iodine was estimated by titrating decinormal arsenious acid, the manipulation being like that previously described in detail. The average result of a series of several determinations in which a great excess (0°1 grm.) of potassium iodate was used, proved to be practically identical with that of a similar series in which only a small excess of the iodate was employed, so that it appears to be unnecessary in any practical work to restrict the amount of iodate below the amount necessary to decompose the maximum quantity of potassium iodide which we have handled, namely, 0°4 grm. It appears that for the estimation of iodine in a soluble iodide free from notable amounts of chlorides or bromides, this method, depending as it does upon a single standard solu- tion, is simple, fairly accurate, and rapid. eta! W. Lindgren—Granitic Rocks of California. 301 Art. XXVIII.—The Granitie Rocks of the Pyramid Peak District, Sierra Nevada, California ; by WALDEMAR LIND- GREN. [Published by permission of the Director of the U. 8S. Geological Survey. ] It has long been known that the summit region of the Sierra Nevada is occupied by an enormous mass of granitic rocks, and that a large part of it consists of granodiorite, a rock intermediate between a granite and a diorite, but no detailed maps have thus far been made of the granitic areas. An opportunity was offered for the study of these granitic rocks during the survey of Pyramid Peak atlas sheet, which was undertaken in the summer of 1894 by the writer, assisted by Mr. H. C. Hoover. The results are shown in the accompany- ing map (p. 302), compiled from the Pyramid Peak folio now in press. The region embraces half a degree of latitude by half a degree of longitude, and contains 927 square miles. The southern end of Lake Tahoe falls within the northern corner, and the main divide of the mountain range runs near the east- ern boundary of the sheet. While the western part is oceu- pied by an approximate plateau deeply trenched by canyons and gulches, the crest of the Sierra Nevada rises in the eastern part to lofty snow-capped mountains. The drainage of the western part is toward the Sacramento and San Joaquin rivers, while the drainage toward Lake Tahoe eventually finds its way to the deserts of the Great Basin. The older bed-rock series consists of slates, schists and gra- nitic rocks. These are extensively covered by Tertiary erup- tives, andesite, rhyolite and basalt, which have not been indi- eated in the map accompanying this paper. To the west the slates, schists and accompanying basic eruptive rocks continue down to the foothills of the range and contain several small masses of granitic rocks. ‘Toward the east the latter continue over to the eastern escarpment of the Sierra Nevada. The range in this vicinity contains two summit-ridges. -The west- erly, dividing the Pacific from the Great Basin, is found on this sheet, while the easterly summit divides the drainage flow- ing into Lake Tahoe from that running into the Carson River. The Sedimentary Areas. No fossils have been found in any of the sedimentary rocks of the bed-rock series within the limit of this sheet, and the age assigned is in all cases tentative only. The determinations are, however, based upon a coniparison with formations of Am. Jour. Scr.—Fourtu Series, Vou. III, No. 16.—Aprin, 1897. 302 W. Lindgren—Granitic Rocks of California. approximately known age in adjoining areas, and they there- fore possess a strong degree of probability. vi Ty ~yt ~ =)h) B ZoeaS: 4 [aves (SN N77 = tae \7 -<7yS!ippex Ny se AYE \ 7 \33 AW ‘ x . y : 1" l \ \y x 5 : t1 a“ ; ~ x EAS =/.NN ‘ Vl 7847 2, x Sond ee Ee i Y4 Bee) ayy N Aq A Rog Ss \ t WEY Ay Zo UM &e CoO sy E A B Cc D A, Sedimentary slates and schists. B, Augite porphyrite. C, Granitite. D, Granodiorite. EH, Diorite and Gabbro. € Under the name of the Calaveras formation the beds of Paleozoic age have been comprised which cannot at present be further subdivided. The larger part of them probably belong to the lower or middle Carboniferous. To this formation all of the sedimentary rocks along the western boundary of the W. Lindgren—Granitic Locks of California. 303 sheet belong. The sedimentary rocks are separated from the granite by an extremely irregular contact line. The bays of granitic rocks reach far into the schists and slates, and all along this contact the sedimentary rocks have been subjected to an intense contact metamorphism, but of fusion or absorption there is absolutely no evidence. The whole of the Calaveras formation on the eastern part of the Placerville sheet and on this sheet has a pronounced sili- ceous character ; it consists of altered sandstones, grading into quartzite, and clay slates grading into micaceous schists. The cause of the metamorphism is partly of a regional char- acter and caused by dynamic movements affecting a large part of the Sierra Nevada, chiefly prior to the great granitic intru- sions, partly of a local character, and caused by the heat and the emanations from the intrusion of enormous massses of gra- nitic magmas. While the latter metamorphism is superim- posed upon the former, and the phenomena resulting from each not always easy to discriminate, it is clearly seen that the extremely altered sediments are found only at the contacts with the granitic rocks, and that the degree of metamorphism gradually decreases away from it. The contact zones are here very wide, typical contact metamorphic rocks often being found two miles from the contacts, or even more in case of projecting masses of sedimentary rocks surrounded on all sides by granite. It does not appear probable that any of these rocks are old, pre-Carboniferous or Archean schists. Less altered rocks, the clastic character of which is clearly apparent, occur at a few places near the western border of the sheet. They are principally dark clay-slates and quartzitic rocks, which under the microscope show their fragmental ori- in. Thus on Silver Creek, near the western boundary of the sheet, on Sly Park Creek, at Fort Grizzly and southeast of Tar’s Sawmill. But the larger part of the Calaveras formation in this sheet is occupied by the contact metamorphic schists. In places especially exposed to the action of the granitic magma, the rock is converted to normal, medium-grained gneiss or mica- schist, and at these places the contacts with the granite, usually sharp, are liable to become indistinct. Somewhat farther away the schists are finer-grained, generally of a brownish color, from the biotite contained, or of a silvery lustre caused by scales of muscovite on the planes of schistosity. The surface is fre- quently knotty, changing to normal “ Knotenschiefer.” They often carry andalusite, characteristic for contact rocks, in well- developed crystals, and such rocks may be found more than one mile distant from the contact. Excellent exposures are found in the deep canyons of Silver Creek, Camp Creek and the north fork of the Cosumnes, 304: W. Lindgren—Granitic Rocks of California. but they are accessible only with difficulty. The schistosity is indicated on the outcrop by lines, straight on the whole, but delicately wavy in detail ; heavy benches alternate with streaks, in which the lamination is very fine. Nodules and nests of apparently segregated quartz are common. On the ridges and slopes satisfactory outcrops are rarely seen, as the rock here weathers to a dark-red soil. A part of the area in the southwestern corner, is also intensely altered; micaceous schists and a striped green ~ and white schist, consisting of pyroxene, quartz, feldspar and wollastonite, evidently a product of contact metamorphic action on limestone, appear in this vicinity. The stratification can only rarely be observed beyond doubt, as for instance where quartzite and black clay-slate alternate, but it is probable that in most cases the stratification approxi- mately coincides with the superimposed schistosity. In the northern area the strike of the schistosity is generally due north and the dip either about vertical or westward at a steep angle, this being contrary to the general rule further down the ~ slope. South of the south fork of the American River the strike is more irregular, but generally east-west, while the dip is always within 20° of the perpendicular and usually to the north. Comparing with the 8... part of Placerville sheet and the N.E. part ‘of Jackson sheet, it will be seen that the series in these regions also has an abnormal east-west strike; the cause may possibly be sought in the mechanics of the intru- sion, the slates in this vicinity being especially torn up by deeply incised bays of granitic rock. Horizontal and inclined joints also traverse the schists, separating them into rhomboidal fragments. The contact of the schists with the granitic rock is usually best defined where that line runs parallel to the schistosity. Wherever the contact cuts across the strike a stronger metamorphism, accompanied by a feathering out of the schists and by an injection of granitic magma, is often noted. The cleavage of the schists has not been produced by the pressure of the intruding magma; it existed before the granitic irruption. A few isolated areas of schists, quartzites and highly-altered tuffs referred to the Jura-trias are scattered on both sides of the crest in the northern part of the area of the Pyramid Peak sheet. One of the principal reasons for assigning them, with doubt, however, to the Jura-trias is their position in the contin- uation of strata known to be of that. age in the area of the Truckee sheet adjoining northward ; another is that the prin- cipal mass, near Mount Tallae, is intimately connected with large masses of dark-green diabase-porphyrite and porphyrite tuff, which is characteristic of the Jura-trias at Sailor Canyon (Colfax sheet) and northward. The color of the outcrops of W. Lindgren—Granitic Rocks of California. 305 these schist areas is usually reddish brown, contrasting strongly with the light-gray granodiorite. The two small areas at the northern boundary consist of quartzite and black slate, the latter altered near the contacts to gneissoid micaceous schist; the contacts are usually sharp, extremely so where the road crosses the western area. At other places, as for instance, on the west side of Loon Lake, the contact is very unsatisfactory, the reddish granitic outcrops being everywhere mixed with schistose fragments. The long and narrow area west of Tells Peak is strongly metamorphosed and composed of gneissoid schists, quartzites and mica-chlorite-andalusite schists. The largest area of supposed Jura-trias lies in Rockbound Valley between Mount Tallac and the Pyramid Peak Range ; it has a roughly triangular form and is distinguished by out- crops of dull gray or brown color. The rocks consist of a series of clearly stratified black slates and white quartzitic rocks; beautifully banded hard rocks, dark-gray and white, also ocecur in Rockbound Valley. The normal strike appears to be N.N.W., with a dip of about 45° to the east; the schistosity is not prominent. In the western point of the area the rocks are disturbed and dip in different directions. In the vicinity of Suzy Lake white quartzitic rocks crop, less clearly stratified, often indeed appearing massive. The microscope shows that the banded rocks from Rockbound Valley and in the Suzy Lake region are porphyrite tuffs, probably depos- ited contemporaneously with the eruption of the large por- phyrite mass of Mount Tallac. Dikes of typical diabase porphyrite were noted on the west shore of Suzy Lake; on the western slope of Rockbound Valley uralite-porphyrites appear which would seem to lie conformably in the sedimentary series and are made somewhat schistose by pressure. The Granitic Rocks. The granitic rocks exhibit a rather unexpected variety in composition and structure. They include granites, aplites, grano- diorites, diorites, and gabbro, by far the largest areas being, how- ever, occupied by the granodiorite. The structure is always massive, a well defined schistosity being nowhere observed. Joints are frequent, however, and near the summit the rocks are intersected by extensive fissure systems. Granite.—A normal biotite-granite or granitite occupies sev- eral large areas along the Pyramid Peak Range at Echo Lake, at Mokelumne Peak, and about the headwaters of the Cosum- nes River. Its outcrops are generally distinguished by a light- yellowish or reddish color, due to the sesquioxide of iron con- tained in the orthoclase. it is harder and of a firmer texture 306 W. Lindgren—Granitic Rocks of California. than the granodiorite, and its areas include the highest and roughest ridges in the region. For the same reason bowlders and cobbles of granitite are much more abundant than those of granodiorite. While it varies somewhat in appearance and constitution, yet it is a typical granite. The rock is coarse- grained and has often a decided tendency towards a rough porphyritic structure. The orthoclase appears as large grains and imperfect prisms of reddish gray color up to two or even three centimeters long; the quartz is very prominent in dark- gray rounded grains up to one centimeter in diameter, while the black mica and smaller quartz and feldspar grains le between these larger constituents. Hornblende occurs only rarely; when it appears plagioclase usually also enters into the composition, and transition-forms to granodiorite result. TABLE OF ANALYSES OF GRANITIC ROCKS. Analyst: Mr. Geo. Steiger. I II OGL IV V VI VIL Villy pa SiO ees hee "7°68 72°95 6745 68°13 67°14 68:32) 65°83 “bia oe TiO. Be eR mice 58 Ads Oxy See EB 15°51 Fe.03 eS 2 1:76 HeOe Te a “51 2 21. Mm@eetis ae trace CaO 22 Neier © ihe 1:16 SOO Meir oll 4°07 o°2Len 41S Ope 5°32 MG. O Mee eer 18 1°10 GO) Pere es 5:00) 1b 430) 3566.95 37508 2°70 Bion 2°88 “1 L084 Nai@ abit POY | sierpahie 6 30207 BIS) 3°09 2:0) 2°41 2°86 3:06 Water 100— 04 ee Water 100+ Nt 63 1250) SN, = eA Day "10 2109) 100°13 100°25 I. 164 Pyramid Peak collection; granitite: Placerville Ditch, 4 mile northof Ditch-camp 7. Lat. 38° 45.5’; Long. 120° 36-1’. II. 20 Pyramid Peak collection; granitite; south side Pyra- mid Peak, 1,200 feet below summit. Lat. 38° 50-2’; Long. 120° 9:5’. III. 103 Pyramid Peak collection; granodiorite ; road, + mile west of Silver Lake House. Lat. 38° 39:8’; Long. P20 Mite IV. 69 Pyramid Peak collection; granodiorite; trail Emer- ald Bay to Rubicon Point, 14 mile south of latter. Lat. 8° 48°6’; Long. 120° 6°23’. V. 86 Pyramid Peak collection: granodiorite; 1 mile E.S.E. of Rockbound Lake. Lat. 38° 58:8’; Long. 120° Noite VI. 120 Pyramid Peak collection; granodiorite; Big Mud Lake bears N. 30° W. and is 14 mile distant. Lat. 38, 35 64s, Lone. 20 mor W. Lindgren—Granitic Rocks of California. 307 VII. 177 Pyramid Peak collection; granodiorite; Meek’s Creek, 2 miles up from mouth at Lake Tahoe. Lat. 39° 0-8’; Long. 120° 9’. Vill 24 Pyramid Peak collection; diorite; Pyramid Peak bears S. 50° E, and is 3 miles distant. Head of Blakeley Creek. Lat. 38° 52:3’; Long. 120° 12’. IX. 93 Pyramid Peak collection; Glen Alpine Spring bear N. 40° W. and is 2 mile distant. Lat. 38° 51:9’; Long. 120° 5:2’. Under I in the table of analyses is given one of these gran- ites of an unusually fresh type. It is a light-gray, hard, gran- ular rock with an approximation to a por phyri itic habit. Macro- scopically are noted large quartz grains up to five millimeters in diameter, large grains and imperfect prisms of feldspar up to ten milimeters in length. Between these larger grains lie the remaining feldspar-quartz mass with somewhat finer grain. Biotite in foils up to three millimeters in diameter are scattered through the rock. Microscopically, the structure is almost allotriomorphic, the rock being principally made up of large, _ irregular and very inter locking ¢ grains of microcline, microper- thite, and quartz. In the microline lie imbedded smaller grains and prisms of an acid plagioclase, as well as some quartz. Between the larger grains lie in places aggregates of quartz, microcline, albite, and oligoclase, also with “interlocking struc- ture. Biotite is sparingly present. The analysis may be cal- culated as follows : Per Cent. Ope wen e Ss. iyi }e):* 2 18-23 nO PEACE Sh oa: 5°18 AAG hr 2} eeere "iat 4°76 KA 81,0 28°17 810, PG: OF SF ok aT 17°24 PLOY Eee e Oe NS of 89 IN ORT a ewe ST 2-96 Na Al 8i,O 25°09 SE iki aedaiglge tah Bes 1°05 Al.O, See Ree = ele Sete e! 93 CO ee OF i oe : 49 Ca Al Si,O 2°4.7 PO = SE ee Cee ra 19 SG de a ve eae ee 13 Ga ou te ce ia i 02 Apatite "25 308 W. Lindgren—Granitic Rocks of California. LiOnges) ascee eae 14 SO} rn igh tee eee os sili CaQh e238 os ie 10 PRS, eee oes 9 NEL a 35 Bighitec: 5 ee eee xk eink 3°10 Majometiten! fates mc. ofa. it)t saree 61 CURIEUZ) 28 a 5 Seen hen eae 39°80 IW etter: ge ap is a a aay pil 100°15 From the total amount of potash 0°24 was tentatively sub- tracted for the biotite. The total lime, after subtraction of amount needed for titanite, was counted to CaAI],Si,O,; the total soda as NaAI$i,O,. On basis of remaining AI,O, a portion of the silica was referred to the biotite and half the. amount of iron oxide and sesquioxide subtracted as magnetite. There remains then for the biotite: SiGe Bh Maes 1:25 40°32 ANE SURAT OUP. Ait 81 26°13 BeiO Rh VRP 22 ate : Fi obiaad. eee 62 20:00 MoO ion a0). Sole 18 5°81 Her yay! goat eh A te 24 774 3°10 100°00 This corresponds fairly well with a normal biotite. A specimen from the south side of Pyramid Peak was sub- jected to a partial analysis (II), which shows it to be of practically the same composition as (1), though a little more of the anor- thite molecule is present. The granite is on the whole very constant in mineral composition and it is believed that these analyses well indicate its average composition. Granodiorite.*— As mentioned before, the granodiorite is the prevailing rock, occupying a broad belt extending across the whole area from north to south. It is of a crumbling nature, falling an easy prey to the destructive forces of weathering and * The name of granodiorite was first proposed by Messrs. G. F. Becker, H. W. Turner and W. Lindgren in 1892. References to the rock may be found in the following places: Geologic Atlas of the U.S., Folios 3, 5, 11, 18, 29. W. Lindgren: The Auriferous Veins of Meadow Lake, this Journal, III, vol. xlvi, p. 201, 1893; U.S. Geol. Survey, 14th Ann. Rep., pt. 2, p. 243; U.S. Geol. Survey, 17th Ann. Rept., pt. 1, p. 35. H. W. Turner: U. 8S. Geol. Survey, 14th Ann. Rept., p. 478; U. 8. Geol. Survey, 17th Ann. Rept., pt. 1, p. 724. W. Lindgren—Granitic Rocks of California. 309 erosion. The outcrops are of rounded form, often weathering into huge detached boulders; the color is very light gray. The granodiorite is a medium to coarse-grained rock, the average diameter of the grain being 2 to 3 millimeters. The grayish quartz and white feldspar grains are of about equal size. The quartz is decidedly less prominent than in the granite, and the feldspar does not reach the dimensions attained in the latter rock. Black mica and hornblende are usually present in about equal quantities. The foils of the former reach 2 or 3 milli- meters in diameter, while the hornblende is roughly prismatic, the crystals sometimes attaining | centimeter in length. By reason of this development of the hornblende, a somewhat por- phyritic aspect may occasionally be obtained. ‘Titanite is nearly always present in small isolated brownish grains. A little mag- netite is also a constant accessory mineral. The appearance and composition of the rock is very constant over large areas, with only small variations in grain and in the quantity of hornblende and biotite. Ina few places the quantity of hornblende dimin- ishes and the rock then assumes a habit more similar to that of granite; thus, for instance, at Buck Island Lake, between Rubi- con Peak and Rubicon Point and in the area east of Fallen Leaf Lake. Microscopical and chemical investigation shows the rock at this point to be a granodiorite, though rather rich in orthoclase. . Analysis III shows the composition of a typical granodiorite from the northwestern shore of Silver Lake. It is a light-gray granular rock composed of white or yellowish feldspar in grains and imperfect prisms, grayish quartz, biotite foils up to 1 mil- limeter in diameter, and a rather abundant dark-green horn- blende in well-defined stout prisms up to 8 millimeters in leneth. Grains of brown titanite are also present. The microscrope shows the structure typical for granodior- ites: Very plentiful, roughly idiomorphic prismatic crystals of an acid plagioclase, sharply outlined foils of yellowish brown biotite partly decomposed into chlorite, and grains of imperfect prisms of ordinary brownish green hornblende, accompanied by a few grains of magnetite. These constituents are cemented by anhedral quartz, orthoclase, and a little microline. Titanite occurs in small grains enclosed in biotite. On the borders of the plagioclase and orthoclase a little micropegmatite often occurs. Some slight evidences of crushing are present in this rock, though such phenomena are in general very rare in the granodiorites. As neither the hornblende nor the biotite has been analyzed, it is clear that no exact calculation of this analysis can be made. It is, however, possible to arrive at the approximate composi- tion by means of the following calculation : 310 W. Lindgren—Granitic Rocks of California. Per Cent. SiO ly shi) eee 11°48 AVON DSL ean Be £0 SAE Es 3° KAl$Si,O, LIS SiO. 20. encom Fes yee 20:20 Al OM eiSiposen ce gine 5:74 UN aS Oe sree 8 oe ene eae 3°47 NaAlsi,O, 29°41 SiO, HARRY Mi Taya os as ritene 0°14 PURO pacman sa tdi Pans ha Ley CAO aan! a: E 2°40 CaAl,Si,0, ite) DIOL Hea eee Nees 49 iO Ret Gens Maken be 58 CaO id sade ie leg 40 Titanite 1°40 BAO) Meco eaace eee 12 CaOicas Sires oreo 16 (iF oe satis la bc MS UR 02 (NDAD G Saas ee rear 30 Maenetite: 2. ta name "84 Hornblende and biotite. -_-- 12°79 Quartz Se Vyas Se eran 25m 100°11 SiOs ee sees aha ah elas (4°50) YAO lig Bie te Na 2°1 Be Oy ors reyes le eae ee 1°13 He axe2is sik Sched geh Sun apes 2°00 CaO Se eng 8s 1) en ae eee 64 Me (icaseis ty ip sates eee 1:10 LO ae OE SS eae 66 HO (SOR Oy See 63 12°79 The estimation is made in the following way: A small amount of potash and soda being first subtracted for the bio- W. Lindgren—Granitic Rocks of California. 311— tite and hornblende, the remainder was calculated as orthoclase and albite. Further, 1:20 per cent lime was tentatively sub- tracted from the total as belonging to the hornblende, and 0°56 per cent for titanite and apatite, the remainder being calculated asanorthite. The amount of magnetite is estimated. From the remaining silica 4°50 per cent was subtracted to approximately correspond with the AJ,O,, and MgO available for biotite and hornblende. The analysis is entirely typical for granodiorite and is ex- tremely similar to the analyses of the granodiorite from Nevada City and Grass Valley, Nevada County.* Under No. IV a partial analysis is recorded of a rock near the shore of Lake Tahoe, not far from the northern boundary line of the sheet. The rock is coarse granular, consisting chiefly of slightly reddish feldspar with much quartz. Horn- blende and biotite are present in about equal quantities, but the hornblende occurs in small grains and prisms only. It was thought that this rock presented a certain similarity to the granitite and it was therefore analyzed, but the figures ob- tained indieate it to be a normal granodiorite. The microscope shows a few large carlsbad twins of microcline and microper- thite; an abundance of imperfect prisms of plagioclase imbed- ded in anhedral quartz, orthoclase, microcline and microper- thite, the latter two often as Carlsbad twins; well-defined foils of biotite, a little hornblende and a few grains of titanite. , Analysis V shows the composition of a granodiorite from the valley of the Rubicon, 2 miles south of the northern boundary line of the sheet. Macroscopically the rock is very similar to IV. Under the microscope the same normal granodiorite structure is apparent. Asin IV much of the potassium feld- spar cementing the plagioclase prisms is microcline. The analy- sis indicates a normal grancdiorite. Analysis VI shows the composition of the granodiorite 6 miles south of Silver Lake in Bear Riverecanyon. The rock is normal except for the fact that the quartz is rather prominent in grains up to 6 millimeters in diameter and that the biotite predominates over the hornblende. Under the microscope it is evident that the cementing ortho- clase and microcline are rather abundant, but the granodiorite structure is well-defined. Titanite is abundant. The analysis shows a closer approach to the granitite than in any other of the rocks here examined ; this is expressed in the high percent- age of silica and potash, as well as in the relatively low per- centage of lime. Still there isa wide gap between this rock and a normal granitite. Analysis VII shows the composition of the granodiorite of * 17th Ann. Rept. U. 8. Geol. Survey, pp. 38 and 42. 312 W. Lindgren—G@ramtic Rocks of California. Meeks Creek, Truckee sheet, a few miles north of the northern boundary of the Pyramid Peak sheet. Both in appearance and composition the rock is a normal granodiorite. Comparing the analytical and microscopical results with the field notes, it is clear that the granodiorite, as it appears in the High Sierra, is a rock of well-defined and fairly constant com- position, structure, and appearance. It is neither a normal diorite, nor is it a granite; it is clearly an intermediate type, occupying a place “between normal quartz-mica-diorite and quartz-monzonite (Brogger).* All transitions toward diorite and, more rarely, toward granitite, may be found, but they are local and do not cover large areas, while the normal grano- diorite is the prevailing rock of the Sierras. Comparing the type here described with the granodiorites of the many smaller areas enclosed in the slates on the western flank of the range, it can be stated that the latter as a rule approach more closely to the quartz-diorites, the percentage of lime being higher and the percentage of potash more often smaller than equal to that of soda. A few of these smaller granitic areas could, in fact, almost as well be indicated as quartz-mica-diorites. In the general habit, however, in the percentage of quartz, horn- blende, and biotite, and in the constant presence of titanite, they are entirely similar to the granodiorites of the High Sierra. Microcline, not common in the granodiorites of the foothill region, occurs abundantly in those of the High Sierra. Diorite and gabbro.—When the amount of hornblende and biotite in a granodiorite increases, it is usual to find the quartz and orthoclase relatively diminished in quantity and rock types more closely allied to normal diorites result. At the same time pyroxene frequently appears, and transitions into gabbro are formed. These more basic rocks in places form smaller areas enclosed in granodiorite or granite; more frequently they he between the two rocks or on the contacts between granite or granodiorite and the schists. The rock in these areas is of a very variable structure and composition, ranging from a quartz- diorite to a gabbro, almost approaching a peridotite. The lat- ter type is, however, rare. The normal diorite, such as occurs in the canyon of the south fork of the American River, is medium to coarse-grained, composed nearly entirely of horn- blende and plagioclase. A little quarz, however, very fre- quently enters into the composition. Typical coarse-grained gabbros with large reddish gray basic feldspars and dark green * Though it is often difficult in practice to separate the normal quartz-mica- diorite from the granodiorite, it would seem suitable to restrict granodiorite to the following limits: S102 59—69 per cent, Al,O3; 14—17 per cent, “Re,Os 14—24 per cent, FeO 14—44 per cent, CaO 3— 64 per cent, MgO 1—24 per cent, K,0 1—32 Na,O 24— 44, W. Lindgren—Granitie Rocks of California. 318 uralitized pyroxene occur near Round Top and to the west of Slippery Ford, and are connected with the diorites by abun- dant transitions. Normal syenites have not been recognized, but intermediate rocks of the composition of monzonites may occur in places ; it does not appear practicable, however, to separate them from the diorites. A normal diorite from the western slope of the Pyramid Peak range was partially analyzed (VIII). The rock is gran- ular, dark grayish green, the grains, averaging two millimeters in diameter, consist of feldspar, dark green hornblende and a little biotite. The plagioclase, which does not exceed andesine in basicity, occurs in imperfect short prisms, occasionally cemented by a little orthoclase. There is no quartz. The mica is in well-defined yellowish brown foils, often ineluding small feldspar prisms. The hornblende occurs in irregular grains, but is sometime roughly idiomorphic. Small grains of titanite are present. The structure is typically hypidiomor- phic. The composition is that of a normal diorite, indicated by the high percentage of lime and soda and small amount of potash. The hornblende must contain much lime. Analysis IX shows the composition of one of the interme- diate rocks, occurring in a diorite area a couple of miles south ,of Tallac Peak. It js a coarse-grained, dark rock made up chiefly of hornblende, a little biotite and feldspar. It carries a considerable amount of orthoclase, and its composition corre- sponds nearly exactly to the monzonite from Mulatto, analyzed by Lemberg.* The relation and succession of the granitic rocks.—The contacts of the granodiorite with the granite aré sometimes sharp, but more commonly much pegmatite, diorite and gran- ite-porphyry occur on them, making them indistinct. In other places transition forms may be observed, such as at the north- ern end of Pyramid Peak granite area. Especially interesting are the exposures along the Pyramid Peak range. Wherever branches or bays of granodiorite reach into the granite a great variety of lighter or darker dioritic rocks make their appear- ance, in places bordering sharply against the granodiorite, at other times forming extremely graded transitions into it. Near the contacts of the schist areas it is quite common to find the granodiorite gradually growing darker and changing to diorites. The contacts between the granite and the diorite are usually sharper, and south of the south fork of the American fe abundant well-defined dikes of granite occur in the lorite. * W.C. Brogger: Die Eruptionsfolge der triadischen Eruptivgesteine bei Pre- dazzo in Siid Tyrol, p. 62. 314 W. Lindgren—Granitic Locks of California. There is no doubt that all of the granitic rocks are later than the altered sedimentary rocks and the augite-por- phyrite, but it must be confessed that in spite of good expo- sures the evidence as to the relative age of the granite, grano- diorite, diorite and gabbro is not decisive, and even in some . respects contradictory. There is some evidence, based on the general form of the Pyramid Peak granite area and the man- ner in which it includes the slate fragments, as well as on the occurrence in it of dikes of a rock allied to granodiorite, tend- ing to show that the granite was intruded earlier than the granodiorite. On the other hand, it is unquestionably true that the granite of the southwestern corner sends out numer- ous dikes into the diorite of the south fork of the American River ; this diorite again shows numerous local transitions to apparently normal granodiorite, so that if it be conceded that this diorite area is of approximately the same age as the main granodiorite mass, it would follow that the granite would be later than the granodiorite. The probability is that the intru- sion both of the granite and of the granodiorite was accom- panied by minor intrusions of acid and basic magmas, and that there are diorites, pegmatites and aplites of the age of the granodiorite and of that of the granite, the latter being the older rock ; only on this supposition can the contradictory tes- timony be explained ; the diorites of the canyon of the South Fork and in the smaller areas along the western boundary would then belong to the period of granitic intrusions, while those of Round Top and the Pyramid Peak range would belong to the granodiorite. Washington, D. C., Jan., 1897. R. S. Tarr—Climate of Davis and Baffin’s Bay. 315 Art. XX1IX.— Difference in the Climate of the Greenland and American sides of Davis’ and Baffin’s Bay ;* by RALPH S. Tarr. Route followed.—During the summer of 1896, I made the journey along the Labrador and Baffin Land coasts, northward in the latter part of July and southward in early and middle September.t The period between these times was spent on the Greenland coast, the northernmost point visited being in lat. 74° 15’. Some features concerning the marked difference in climate on the two sides of the sea separating Greenland from the American land, apparently well known to Arctic voyagers, but new to me, have seemed to warrant brief state- ment. Climate of American and Greenland sides.—On the north- ward trip a landing was made at the island of Turnavik, about in lat. 55°, on July 20. There extensive snow banks were found to be still lingering in the sheltered valleys, at points no more than one or two hundred feet above the sea level. Indeed several hundred miles to the south, on the west coast of Newfoundland, snow was seen. Later a landing was made in southern Baffin Land, on Hudson Straits, and here snow banks were also found. Nearly the entire distance traversed by the ship, from southern Labrador to the mouth of Cumberland Sound, one or two degrees south of the Arctic Circle, where we left the American coast, we met with floe ice, often so heavy that progress through it was slow and difficult. This was an unusual season, and hence the conditions were some- what more severe than common. The ice is rarely so abund- ant at this season of the year. The presence of great floes, moving southward in the Labrador current, so chilled the air that the thermometer often stood in the thirties, and frequently descended to the freezing point. On July 30 snow fell in some quantity upon the ship, which then lay off Frobisher Bay. On July 30 an unsuccessful attempt was made to enter Cumberland Sound, which was then, as often in this season of the year, completely shut in by the heavy ice floe. Not being *Tt is a rather remarkable fact that the great bay which extends north to the Arctic circle, from the main Atlantic to Davis’ Straits, should have been given no name. Baffin’s Bay extends from Cape York to Davis’ Straits at the Arctic circle; but no name is applied to the bay south of this. Because of the difficulty experi- enced in the attempt to write this article, on account of the absence of such a name, I would propose for the bay between Labrador and Greenland and south of Davis’ Straits, the name Davis’ Bay after the first navigator of these waters, who, in the year 1587, made the perilous voyage in a sailing vessel as far north as Upernavik. + As a member of the Peary Greenland Expedition. 816 &. 8. Larr—Chimate of Davis and Bafin’s Bay. able to penetrate the ice, the ship left this bleak coast and crossed Davis’ Straits to Disco island on the Greenland coast, going north four or five degrees. There we founda decided change in climatic conditions. The air was balmy, and although the highland portions of the mainland and island were ice-capped, and snow banks were seen in the protected valleys, the season was distinctly more advanced and more pleasant than on the American side two or three hundred miles southward. The flora was richer, insects were abundant, and everything betokened summer. Passing still farther northward to the Upper Nugsuak peninsula, in lat. 74° 10’, although the climate was somewhat more severe, it was distinctly farther advanced than on the American side which had been left a few days before, and which lay five or six hundred miles to the southward. From August 7 to September 7, although by the latter date the night had begun, with the sun setting at about 7.30 in the evening, we were able to live comfortably in lat. 74° 10’ with no other protection than that of tents. Dur- ing this time the lowest recorded temperature was 28°, which came at the coldest part of the night. The storms brought rain and not snow. Returning, we left Disco on September 11 after a beautiful, warm day. Snow had recently fallen upon the uplands, but many flowers were still in blossom near sea level. Coming southwards to the mouth of Cumberland Sound, our first view of the American land showed a snow- covered surface, and from this point to the north end of New- foundland, freshly fallen snow was seen on the land. Although we later found that the floe ice had disappeared from the Labrador coast, it still blocked a portion of the entrance to Cumberland Sound, and for sixty hours the ship was held:a prisoner in this ice before we could pass into the Sound. After entering Cumberland Sound we encountered violent snow storms, and the fall of snow was sufficient to cover the surface of the land. Hence in this latitude, in the summer of 1896, snow fell on July 30 and on some day pre- vious to September 13, the latter fall being sufficient to whiten the hills as seen from the sea. Influence of Ocean Currents.—The eauses for this difference between the climate of the two sides of a sea separated by only afew hundred miles in the broadest portions, are perhaps several, though the chief cause is to be found in the ocean cur- rents. The icebergs and floes of the Arctic pass southward on the American side, and the cold ice-laden waters influence the climate of the neighboring land very perceptibly. On the Greenland side there is a current in an opposite direction, hence carrying warm water northward. It is difficult to say how far the north-moving current LR. S. Tarr—Climate of Davis and Baffin’s Bay. 317 extends, and also how far from the land its influence is felt. I should say that the current does not probably reach far to sea; because if it did, the bergs and winter sea ice would not readily escape, and the Greenland coast would be ice-bound in summer, rather than ice-free as we found it.* There are some reasons for thinking that the north-moving current extends as far as the northern end of Melville Bay. The fact that great quanti- ties of ice accumulate and remain there, rendering this bay notoriously difficult to navigate, seems to show that something interferes with the southward movement of the ice. Pointing toward the same conclusion that the warm current reaches into Melville Bay, are the observations which were made from our camp at Wilcox Head, the point of the Upper Nugsuak peninsula which projects into Baffin’s Bay near the southern end of Melville Bay. During two or three weeks encampment at this exposed point, it was noticed that the ice- bergs floated almost uniformly toward thenorth. Itis true that this was a period during which south winds prevailed, but even when these did not blow, the bergs still moved northward. Moreover with but one hundred feet, or less, above the water, it seems improbable that the great ice masses which floated past our camp could have been driven so rapidly by mere wind action, particularly when the wind was not strong. Infiuence of Winds.—There are two other reasons for cli- matic moderation on the Greenland coast; in the first place the winds from all directions, excepting the east, reach the shore from water, the temperature of which is well above the freezing point during thesummer. The entire coast of this part of Greenland is then free from floe ice, and is encumbered only with scattered bergs and berg fragments. Hence on this coast the water is warmer than that on the opposite side of Baftin’s Bay. On the one shore there is a cold ice-laden current, on the other a slowly moving drift of water toward the north. One might expect that the wind blowing from the ice cap of Greenland, the prevailing wind while we were there, would be cold; but the reverse is true. This wind, descending from great altitudes and passing over the surface of snow and ice, is really warmer than that from the sea, and at times it was * Incidentally I would call attention to the fact that this clearing of the various forms of ice from the Greenland coast is usually made possible by the winds that blow from the ice cap It is an interesting coincidence that the very cause for most of the ice gives rise to conditions which permit this to be carried away. From the cold highland area of inland Greenland, the dense air settles and blows outward, producing off-shore winds, which keep the fjords free of ice encum- brances ; and at times, extending out from thé coast, the wind drives this ice wel] to sea, where it comes under the influence of the south-moving currents, which I believe must exist not far from the Greenland shore. Am. Jour. Scil.—FourtH SERIES, VoL. III, No. 16.—Aprin, 1897. 22 318 RK. S. Tarr—Climate of Davis and Baffin’s Bay. noticeably warm, rapidly descending air. The explanation is no doubt the same as the explanation of the chinook and foehn winds. Cause of difference in Glaciation.—On the Greenland side the land is mainly submerged beneath the great ice cap, with branches extending to the sea through the valleys. On the American side there are only a few isolated glaciers on Bafiin Land, and none are known to exist in Labrador. On the American side the southernmost glacier is located on the southern side of Frobisher Bay in lat. 62°. It is evident that this difference in ice-covering is not due to the climate near the coast line, fora much greater development of glaciers is found on the side where the climate is more moderate. Pos- sibly one of the causes for the difference is the greater humidity of the air that comes to Greenland after crossing over the waters of Baftin’s and Davis’ Bays; but the main cause for the difference is evidently the greater elevation of the land on the eastern side. Former Glaciation on the American side.*—Labrador and Baffin Land have been recently glaciated. So recent was this time of glaciation that at Turnavik Island, on the Labrador coast, glacial striz still remain distinctly on the exposed rock-faces. The violent frost action on Baffin Land has in most cases removed the striz ; but the form of the hills, and the presence of erratics on the surface, show recent glaciation. Comparing the conditions seen here with those of New England, it seems certain that the glacier has left this northern region more recently than New England. Changes in Level of Baffin Land.—Before the ice covered this land it was much higher than now. The evidence of this is as striking as that on the coast of Maine. The land valleys are in all stages of submergence; in Hudson Straits there are entirely submerged strike valleys, others into which the sea. enters, and still others entirely above the sea level. The fjords, the nearly land-locked bays and sounds, and the land valleys extending beneath the sea, make this coast one of the most irregu- larly indented shores in the world. It is possible to navigate for fifty miles behind the land on the southern side of @umber- land Sound, being all the time behind hills which reach five hundred or a thousand feet above the sea. It is practically certain that the glacial conditions came when this land was higher. When the ice disappeared the ancient highland was reduced in elevation, and the level of the hills was lowered three or four hundred feet below the surface of the sea, for beaches are found at this elevation, in various parts of < Tarr, Am iGeol 189ipxx a sileand IO: R. S. Tarr—Climate of Davis and Baffin’s Bay. 319 Baffin Land.* Perhaps in this depression is found a potent cause for the disappearance of the ice sheets. Now the American land is engaged in the reverse movement of uplift. It stood three hundred feet lower at a time so recent that the boulder beaches are distinctly visible, and the individual boulders scarcely injured by the action of the weather in this region of extremely violent frost work. Their surfaces bear lichens, but they are still rounded, and they he directly on the surface, with scarcely any soil. accumulation since they became a part of the dry land. Very recently this land was moving upward, for there are beaches directly above the present ones, and yet so closely connected with them that at first it seemed that they must be forming even now. This is in harmony with the evidence obtained by Dr. Bell on the shores of Hudson’s Bay. Relation of Changes in Level to Glaciation.-—Elevation is a potent cause for glaciation. Baffin Land and Labrador have been glaciated in a recent period, during a time when the land stood a thousand feet or more above the present level. The ice of this period disappeared when the land was lowered, and possibly because it was lowered. Now the land is rising: at present the climate is so rigorous that even with the present elevation the conditions are nearly severe enough for glaciers to develop, and in some places they do actually exist. Even near the sea level the snow banks do not disappear before the first of August, and snow commences to fall in the autumn as early as the middle of September. So far as I can estimate from my short visit, it seems that there must be places not far above the sea level where even now the snow stays throughout the summer. A slight change in climate is all that is needed to increase the number of these, and to add to their area and depth until glaciers begin. The elevation needed for this increased rigor cannot be many hundred feet in the higher regions. Such elevation, if widespread (and the recent uplifts have extended over a broad area), would remove another of the causes for moderation in this northern climate, namely the disappearance of some of the neighboring water: the recent uplift in Baffin Land has added a large amount of land to the former area, even though the elevation has been only about three hundred feet. This uplift has so decreased the depth of some of the bays and straits that an additional elevation of three hundred feet would very perceptibly reduce the amount of water both north and west of Labrador. This added land * The question of the elevated beaches of Baffin Land is discussed by Mr. T. L. Watson in a paper published in the Journ. of Geol., 1897, v, 17. 320 RR. S. Tarr—Climate of Davis and Baffin’s Bay. and decreased water area would increase the rigor of the climate. , . The point which I wish chiefly to make is that the climatic conditions of Baffin Land and Labrador are wonderfully near those which produce glaciation. I would not predict that these lands are about to enter into a glaciated condition again, but it is safe to say that if the elevation now in progress con- tinues, the time is not far distant when valley glaciers will again come in the Labrador peninsula and when those of Baffin Land will increase in extent, provided of course that there are no general changes of climate in progress, the nature of which we do not understand. From this condition of local glaciation to a general ice sheet, spreading over the land, the step is not great. _ While upon this topic, and in conelusion, it may be pointed out that in addition to the moderateness of the summer climate of the Greenland coast, there is also a submergence of the land at present in progress. ‘Topographical evidence of this is plain in the places which I visited, and the Danes have proved the point for at least some portions of the coast. Accompanying this subsidence, the Greenland glacier has recently withdrawn, at least from the land of the Upper Nugsuak peninsula, and the amount of withdrawal has been great indeed. Even now the ice-front at this point is moving backwards. Is Greenland now passing through the stage of ice-with- drawal from which the American, Labrador and Baffin Lands have so recently escaped ? and is there any relation between the down-sinking of Greenland and the uprising of Labrador and Baffin Land? Is the ice-withdrawal directly due to the land movement, and is the load of ice really the cause for the sinking which allows its withdrawal? that is, does the ice increase in area and extent with no other result than its own destruction by depressing the land, and hence removing the cause of supply? These questions very naturally arise and others even more speculative come to mind; but as their answer is not definitely at hand they may well be left as mere queries. Ithaca, N. Y. E. C. Case—foramina perforating, ete. 321 Art. XXX.— On the Foramina perforating the Cranial Region of a Permian Reptile and on a Cast of tts Brain Cavity ; by E. C. Case. Durineé the spring of 1896 the author collected from the Middle Permian of Texas the nearly complete skeleton of a reptile (Dimetrodon incisivus Cope) belonging to the Pelyco- saurian group of Cope’s order Zheromora. It has recently been shown* that the order Zheromora has no existence and that the Pelycosauria are merely specialized Phyncocephalians closely allied to Palwohatteria. The cranial region in the specimen is especially well pre- served and permits a close study of the different foramina. The bones are all in their natural relations and nearly free from distortion, so that the brain cavity when freed from its enclosed matrix showed its natural form. The occipital region closely resembles that of Sphwnodon. The condyle is formed by the exoccipitals and basioccipital. The exoccipitals meet in the median line above, excluding the supraoccipital from any part in ‘the foramen magnum. - Laterally they join the expanded proximal ends of the paroccipitals. The supraoccei- pital is a triangular plate inclined forward as it ascends and joining by the base of the triangle the parietals above. Lat- erally it joins the paroccipitals and inferiorly the exoccipitals. The paroccipitals are expanded proximally, joining the supra- occipital and exoccipitals. Distally they are elongated out- wards, backwards and downwards and join the greatly flattened quadrates. The lower edge of the proximal end is marked by a notch which, in union with similar notches in the basioccipital and petrosal form the fenestra ovalis. The paroc- cipitals remained free during life or until advanced age. This feature is found only in turtles and the young Sphenodon. It has been noticed in young lizards before leaving the egg.+ The basioccipital forms the lower portion of the condyle ‘and hes between the exoccipitals and paroccipitals. The lower surface is trough-like for its posterior half and supported a pos- terior extension of the basisphenoid. Laterally a slight notch forms the inner wall of the fenestra ovalis. Anterior to the horizontal, trough-like portion the inferior surface rises . sharply ; the angle thus formed is marked by a large foramen through which the hypophysis passes into the interior of the * Baur and Case: On the Morphology of the Skull of the Pelycosauria, and the Origin of the Mammalia, Anat. Anz., xiii, Nr. 4 and 5, 1897. + Siebenrock, F.: Das Skelet der Lacerta Simonyi Steind. und der Lacertiden familie uberhaupt; Sitzunberichten der kaiserl. Akademie der Wissenschaften in Wien. Mathm. Naturwiss. Classe., ciii, Abth. 1, April, 1894. 322 £. C. Case—Foramina perforating the basioccipital, fig. 3, Hy./. The petrosals join the parocci- pitals, exoccipitals and the basioccipital, but the sutures are not distinguishable. The lower part of the anterior edges were continued forward as long processes, the anterior inferior processes of Siebenrock.* These are partially destroyed in the specimen. A deep notch in the anterior edge of the petrosals just above the origin of these processes, the incisura otosphenoidea Sieb., marks the point of exit from the brain cavity of the fifth pair of nerves (trigeminus), fig. 3, 5. The superior end of the anterior edge is separated from the supraoc- cipital by a notch which is continued on the sides of the bone as a shallow, short groove. The posterior edge contributes the last portion to the walls of the fenestra ovalis. The basisphenoid remained free. The posterior edge is greatly thickened vertically and its lower edge stood well away from the basioccipital. The otic region and the posterior edge of the basisphenoid were covered with a large mass of earti- lage. The lower surface of the basisphenoid is excavated by a deep pit, fig. 2, 47, which opens on the posterior as well as the inferior surface of the bone and divides the posterior into two parts. The upper edge of the posterior surface, forming the base of the pit, was continued backward as a spout-like process articulating with the lower surface of basioccipital. The anterior edge is extended forward as a parasphenoid ros- trum originating between the short and stout pterygoid processes. The foramina penetrating these bones are remarkably similar in position to those penetrating the same bones in Sphenodon. The condylar foramen transmitting the twelfth pair (hypo- glossus) penetrates the exoccipital just anterior to the edge of foramen magnum. Its outer end opens in a notch (the ¢ncisura vene gugularis Sieb.) in the side of the exoccipital. A little below and further forward a second and much smaller foramen opens in the same notch; this may transmit either the ninth or tenth pair of nerves or a minor blood vessel. Passing for- ward the notch deepens and is very soon converted into a foramen by the adjacent portion of the paroccipital. This is the foramen vene jugularis of Siebenrock and transmits the jugular vein and either the ninth or tenth nerves or both of them. In Sphenodon the foramen transmits not only these but the twelfth pair as well, the nerves being separated from the vein by very thin walls of bone and may be separated from each other or have a common canal. The opening of the twelfth pair into the notch which forms the beginning of the * Siebenrock, F.: Zur Osteologie des Hatteria-Kopfes ibid. Bd. cii, Abth. 1, June, 1893. Cranial Region of a Permian Reptile. 323 jugular foramen is then very similar to the condition found in Sphenodon. Fig. 1. Side view of a cast of the brain cavity. Fig 2. Lower view of the basisphenoid. Fig. 3. Lower view of the cranial region. Fic. 4. Lower view of the cast of the brain cavity. 5. The trigeminus nerve. 7. The facial nerve. i2. The hypoglossus nerve. Ju. The jugular foramen. Ty. Cast of tympanic cavity. Hy. Hypophysis. Hy. F. Foramen penetrating base of basioccipital. # O. Fenestra ovalis. J. C. Foramina for internal carotids. Hw. Opening of eustachian tubes. Cb. Cere- bellum. 324 ft. C. Case—foramina perforating the The fenestra ovalis, fig. 3, /” O., is a single opening leading by a very short canal directly into the brain cavity, a character found in fishes and the amphibian J/enopoma and existing imperfectly in some recent reptilia, as the turtles. The same thing is described by Cope as existing in another Permian reptile, from the same horizon as the present specimen, but belonging to a separate family, the Diadectida, and his order Cotylosauria™. The foramina for the seventh (facial) pair of nerves appear on the outer surface of the petrosal just anterior to the fenestra ovalis, fig. 3,7. They are located relatively a little further back than in Sphenodon. On the inner face of the same bone the foramina appear at the side of the base of the brain cavity a little anterior to their external opening. They are located just anterior to a slight ridge which defines the limits of the tympanic cavity. In Sphenodon this is about the point of location of a foramen common to the seventh and eighth nerves, which, however, almost immediately divides, the pos- terior branch penetrating the inner wall of the tympanic cavity and leading the auditory nerve to the inner ear. The foramen for the fifth (trigeminus) nerve is completed from the incisura otosphenoidea by the membranous wall of the anterior portion of the brain case, as in Sphenodon and many lizards. Fig. 3, 5. ’ The deep pit excavating the lower surface of the basisphe- noid'is in all probability the lower opening of the eustachian tubes. In most reptilian forms the tubes pass into the pharynx in the neighborhood of the basioccipital-basisphenoid suture and anterior to the fenestra ovalis. In the crocodilia and the aglossal batrachians they have a common opening into the mouth. In the present form the tubes probably penetrated the large mass of cartilage covering the otic region and the posterior end of the basisphenoid and found a common open- ing in the deep pit described. It is difficult to imagine the use of such an extensive cavity in the basisphenoid, but in the Teleosauria an equally large cavity is found roofed over with bone. Anterior to this pit two foramina, fig. 2, 7. C, penetrate the lower surface of the basisphenoid bone and on its upper surface a large foramen appears just posterior to the origin of the presphenoid rostrum. Through the pair on the lower sur- face the internal carotid arteries enter the bone and through the upper it gains access to the brain cavity by way of the pituatary fossa. On either side of the single foramen a pair of small foramina carry branches of the internal carotid. All of * Cope, EH. D.: On the Structure of the Brain and Auditory Apparatus of a Theromorphous Reptile of the Permian Epoch, Proc. Am. Phil. Soc., vol. xxiii, 1885, Cranial Region of a Permian Reptile. 325 these foramina are very similar in position to the same ones in Sphenodon. The cast of the brain cavity shows fairly well all parts pos- terior to the fifth pair of nerves, and the hypophysis anterior to them. As is well known, the brain in the reptilia does not fill the brain cavity but is supported by a mass of connective tissue carrying lymph and fat masses, so a cast of the brain cavity does not give an exact copy of the brain; however many points can be brought out by such a east. If the cast be held with the short terminal portion of the medulla horizontal, the lower surface pitches downward at a sharp angle to a point anterior to the tympanic region and then ascends as sharply to the point of origin of the hypo- physis. Thesuperior surface is horizontal and arched from side to side to a point over the tympanic cavity and there turns upward at an angle of 45°. The angle thus produced is marked by a low, narrow ridge running across the cast and marking the position on the brain of a narrow and elevated cerebellum, figs. 1 and 4 Cdb., such as occurs in Sphenodon. This region was probably the seat of a large amount of con- nective tissue and it is probable that the upper surface of the medulla descended at as sharp an angle as the lower. This would thake still more marked the resemblance to Sphenodon and to the cast figured by Cope.* This sharp bend of the medulla downward is not found in other for ms, though in the brain of Chelonia and some lacertilia a bend is apparent. The sides of the medulla show most posteriorly the begin- ning of the twelfth nerves, tigs. 1 and 4 (12), anterior to these the cast of the eet foramen, figs. 1 and 4, J/w., and finally the large casts of the tympanic cavity, figs. 1 and 4, Ty. The nature of the matrix and the cavities prevented the tympanic cavities being cleaned so that the semicircular canals could be determined, but it is probable that they were very similar to those described by Cope. Anterior to the tympanic casts a sharp constriction marks the ridge defining the limits of the tympanic cavity and thena sharp outswelling the point of exit of the trigeminus nerve, figs. 1 and 4 (5). Near where these leave the body of the cast a small stub on each side marks the origin of the seventh pair, figs. 1 and 4 (7). The hypophysis is the most interesting feature of the brain. Descending between the anterior inferior process of the petrosal and turning posteriorly, it oceupies a small notch in the posterior edge of the upper surface of the basisphenoid and then passes directly into the body of the basioccipital through the foramen mentioned. In the Crocodilia a somewhat similar condition exists. The basisphenoid is excavated for a con- * Baur and Case, loc. cit. 326 EF. C. Case Foramina perforating, etc. siderable extent to accommodate the hypophysis. This makes — it probable that the excavation of the bone is merely a second- ary character to make room for the hypophysis, for in the Crocodilia the basisphenoid takes a large part in the floor of the brain-cast, and in the present form it is pusbed so far downward that it is excluded and the hypophysis encounters the basioccipital as soon as it turns toward the rear. Marsh (* and ¢) has described in the family Atlantosauride@ of his suborder Sawropoda of the Dinosauria a condition in which the pituatary cavity becomes a canal perforating the basisphenoid and opening into the pharyngeal cavity, consider- ing it an embryonic character snch as exists in the chick at the fifth day of incubation.* If the hypophysis occupied the entire cavity in the basioc- cipital it extended back nearly as far as the tympanic region and much further back than in most reptilian forms. In Sphenodon, the Crocodilia and some amphibians it reaches well back, but not so far as in the present form. Compared with Sphenodon, the specimen shows the follow- ing points of resemblance. The foramina for the blood ves- sels and nerves are almost identical in position and nature. The contour of the medulla and cerebellum was similar and the hypophysis extended far back. The only point of differ- ence is the excavation of the basioccipital to receive the distal end of the hypophysis. The free communication of the tym- panic cavity is a character which is found in many existing primitive forms and is of secondary importance. The points here brought out confirm the close relationship of the Pelycosauria to the primitive Rhyncocephalia already asserted by Baur and Case.{ * Marsh, O. C.: Principal Characters of American Dinosaurs, this Journal, vol. xxvi, August, 1843. + Marsh. O. C.: American Dinosaurs, Sixteenth Annual Report U. 8. Geol. Survey, 1896. t+ Baur and Case: On the Morphology of the Skull of the Pelycosauria, and the Origin of the Mammalia, Anat. Anz., xiii, Nr. 4 and 5, 1897. Paleontological Lab., Univ. of Chicago, January, 1897. Trowbridge and Richards— Temperature, ete. 327 Art. XXXI.—The Temperature and Ohmic Resistance of Gases during the Osciilatory Electric Discharge ; by JOHN TROWBRIDGE and THEODORE Ww. RicHarps. With Plate VE IN our payor open the spectra of argon and the multiple spectra of gases,* we have emphasized the importance of con- sidering the electrical condition of the circuit in which is placed the Pliicker tube containing the gas under examination. We have pointed anew to the fact that in general the continu- ous discharge of an accumulator produces one spectrum while the oscillator y discharge of a condenser produces another. In considering this question one is immediately struck by the fact that, although the gas acts as if it presented a resistance of several hundred thousand or even several million ohms to the eurrent while under the influence of the continuous discharge, nevertheless this same tube allows oscillations which are wholly damped by a few hundred ohms to pass through it under the influence of a condenser. These considerations led us to measure the resistance of such a tube to the oscillatory dis- charge, and we found by means of a novel method that in fact a mass of gas at low tension contuined in a capillary tube may act as though it opposed a resistance of only jive or six ohms to the spark of a large condenser. In order the more clearly to grasp the situation, the poten- tial differences between the ends of the tube during a con- tinuous discharge may well be considered first. A number of measurements of such potential differences have been made by Hittorft+ and others, but it may be well to give two of the many series of measurements which we have made, in order to facilitate comparison with the discharge of the ‘condenser through the same tubes. The tubes employed throughout this research were of the ordinary type devised by Plicker, con- sisting of two cylindrical bulbs separated by a capillary 132 in diameter and 7™ long. The electrodes were of aluminum. Unless otherwise “sated. all the experiments here recorded were made with tubes of exactly this shape and size; and most of the experiments were made with a single tube. The voltmeter used for measuring the potential differences was a Thomson electrostatic electrometer, and the current used was not much over a milliampere. As the voltmeter was only graduated to 1800 volts, the readings above that amount are merely approximations. * This Journal, vol. iii, pp. 15 and 117, 1897. + Wied. Ann., xx, 705. 328 Trowbridge and Richards— Temperature and Potential differences between electrodes of spectrum tube. Hydrogen. Nitrogen. Pressure in Voltage. Pressure in Voltage. 7°0 2600 (?) 8°5 very high. 6:0 2100 (?) 5: very high. 4°0 1900 4: 2600 (?) 35 1500 3° 2100 (?) 2°0 1340 25a 1750 1°5 1260 iby 1600 1°25 1220 1°4 1410 1:15 1150 V2 1340 1°00 1100 Le 1180 7) 1140 O°7 1140 "50 1220 0°6 1080 13 very high 0°5 1040 0'3 980 0°25 1030 0°13 1700 0°06 2800 + (?) Each gas evidently has its minimum of potential difference, that of shydrogen lying at about 1 millimeter of pressure, and that of nitrogen at about 03 millimeter. These minima, as well as the total potential differences, are undoubtedly modified by the strength of the current; but the results given above are comparable with one another because they were all made under the same conditions. Hittorf found a minimum at about 0°35 millimeter for nitrogen, and he pointed out by means of his extra electrodes that the fall of potential was very irregular, the greater part of it residing at the cathode. His results have been confirmed by others, and Wood* has shown that the heat evolved at different parts of the tube follows the same irregularities as these potential differences. Neglecting the factors of the potential difference which reside at the electrodes, the sum of which increase with the exhaustion of the tube, we find that according to Hittorf’s results the resistance of the gas itself steadily diminishes as the exhaustion proceeds. For example, with a current of two milliamperes he found a fall of potential of about 120 volts between two parts of the middle of the tube eight centimeters apart, the tension of the nitrogen being 0°35 millimeter. When the current was about one milliampere and the tension of the gas was only about 0:001™™ the voltage sank to fifteen. These two figures correspond to resistance of 60,000 ohms and 15,000 ohms respectively, the resistance of the gas diminishing as the pressure is decreased. Of course we have no certainty as to how much of this opposition to the current is due to true * Wied. Ann:, lix, 238. Resistance of Gases during an Oscillatory Discharge. 329 resistance and how much to a kind of polarization, but it is convenient for present purposes to count it all as resistance. In any ease this opposition, if maintained, is far too great to permit the passage of oscillations, even under the most favor- able conditions. In order to prove that the opposition is not maintained, but is in fact broken down by the spark, it was only necessary to photograph the discharge with the help of a rapidly revolving mirror, after the method of Feddersen. Unfortunately the light in the tube itself is too faint for direct instantaneous photography : but the light of the spark between two cadmium electrodes in the same cireuit is quite bright enough for the purpose, and of course any oscillations which crossed the spark gap must also go through the tube. Our next step was, therefore, to make a series of such photographs of a spark discharged through hydrogen, at first when the gas glowed with a white hight and showed its magy-line spectrum, and afterwards, when it exhibited the characteristic red tint and a spectrum ‘of only four lines in the ee portion of the spectrum. BGS TY. - = Battery of 5,000 to 10,000 storage cells. = Condenser of 1,000 to 18,000 electrostatic units. — Small resistance to damp oscillations. S = Spark gap between cadmium terminals. T = Plicker tube containing gas. W = Chief water resistance of 5 to 50 megohms. In order to obtain the white light in the hydrogen tube, it was necessary to increase either the impedance or the resistance in the circuit containing the tube. With a definite very small amount of impedance we increased the resistance until the 330 Trowbridge and Richards—Temperature and red glow disappeared in the tube, and discovered, on develop- ing the photographs which were obtained by means of the revolving mirror, that the discharge was non-oscillatory. When, however, the resistance in the condenser circuit was dimin- ished, the red glow began to appear, and the photographs taken when all the resistance except the tube itself was removed showed that the discharge was oscillatory. This also was evident from the peculiar crackle of the spark, which Hertz remarked was essential in performing his experiments on electric waves. The apparatus used in this and subsequent experiments is sketched in the accompanying diagram (fig. 1). An examination of our photographs showed the interesting fact that there were in general not more than two or three complete oscillations; the remaining ones which could have been obtained from the given capacity aud self-induction hav- ing been damped by the resistance of the gas. The question immediately arose: what is the resistance of the gas at the instant of the discharge? For if an idea of this can be obtained we can get an estimate of the amount of heat devel- oped in the gas during each oscillation. A Thomson electro- static voltmeter connected to the ends of the hydrogen tube indicated a difference of potential of over 1,800 volts, and this difference of potential could only be obtained by substituting for the Geissler tube a resistance of many thousand ohms. The indications, however, of this instrument in this case are of no value; for we discovered that a resistance of from ten to twenty ohms was sufficient to produce the same amount of damp- ing which the gas exerted. The resistance of the gas, therefore, could not be greater than these amounts.* It is evident, there- fore, why the voltmeter gives erroneons readings. On account of the inertia of the moving parts and the very short time of the discharge, it does not indicate the fall of potential through the small resistance of the tube during the instant when the discharge passes, but maintains an indication of a high differ- ence of potential. In order to apply systematically thisnew method of measur- ing resistance, our next step was to prepare a series of stand- ards,—photographs of the oscillatory sparks of condensers of different sizes, damped by known resistances, which were sub- stituted for the Geissler tube in the condenser circuit. In all these experiments, of coarse, the small resistance on the left hand side of the sketch was cut out by a suitable key. Three large Leyden jars, each 30 centimeters in diameter and 50 centimeters high, having a capacity of 6,000 electrostatic units apiece, were used either singly or together to act as the con- denser; the waves generated by these large capacities were * Damping of electrical oscillations, Proc. Am. Acad., 1891. Resistance of Gases during an Oscillatory Discharge. 331 much too long to interfere with one another upon so short a circuit. The resistances were wires of manganin 0:2™™ in diameter, stretched on both sides of long strips of thin vulean- ite plate, the idea of this arrangement being to eliminate selt- induction and yet to prevent the short-circuiting of the high potential. The spark gap usually consisted of cadmium ter- minals arranged in the focus of a revolving mirror driven very rapidly by means of a small electric motor. In a few eases zine terminals were used, with no appreciable difference in the results (Rhigi*). The terminals were re-pointed from time to time and were always kept at a distance of 1°3 millimeters apart. With this apparatus the photographs of perhaps five hundred sparks were taken; and the results are recorded in the following table. Asa general rule the spark containing the highest number of oscillations upon any plate was taken as the representative one. The first column below records the resistance through which the discharge had to pass before reaching the spark gap, while the second, third and fourth record the number of half oscilla- tions observed upon the photographs. . These figures correspond in general tendency with the less precise determinations made by Feddersen ;{ they show, as his determinations did, that the larger the capacity, the fewer the number of oscillations. This tendency is especially noticeable between two and ten ohms, the part of each curve which is most capable of accurate determination. While not perfectly regular, these curves manifestly furnish the means of measur- ing approximately any small resistance through which a spark, followed by as much as one-half of an oscillation, is able to pass. Having now our scale of measurement, we substituted for our known resistances a Pliicker tube attached to an admir- able automatic Toepler air-pump (of Kiss, Budapest), as well as to receivers containing pure hydrogen and nitrogen. These gases could be delivered individually into the tube at any desired pressure. The bulbs of the pump, aggregating over a liter in volume, were always in communication with the Plicker tube, while the circuit was closed, so that the discharge took place under essentially constant pressure. The hydrogen was made electrolytically and purified by passing through a solution of potash, and over fused potash and phosphoric anhydride; the nitrogen was made by passing ammonia over an excess of heated cupric oxide, through much water, and over the same two driers as the hydrogen. The length of the spark gap remained always the same, excepting for the very lowest and the very highest pressures of gas, through which * Nuovo Cimento, If, xvi, 97 ; + Pogg. Ann., exiii, p. 437. 3382 Trowbridge and Richards—Temperature and Resistance Standards. Capacity=6,000. Capacity=12,000. Capacity 18,000. Resistance No. of half oscilla- No. of half oscilla- No. of half oscilla- Ohms. tions f= fonss— Lions = 9) 37. 32 1 Zl. ea 2 16. 14, 13 3 12. ta 10 4 oD 8.5 A 5 8. fe 6 6 io 6. 5) 7 6.5 oD. 4 10 3) 4 3 15 3 3 20 Zz: 2 ie 30 ] ] if aes beeeeGeesteeeenisce 2 4 6 8 10 12 14 Fig. 2. Ohms are plotted horizontally. Half oscillations are plotted vertically. SEESSEESESGEE Boe! SSGeees—en=>- ~~ == gueuneeeoncee SESEEEESESGL SL” OL oe! a Resistance of Gases during an Oscillatory Discharge. 333 the electricity refused to pass unless the spark gap was nar- rowed. In the first column of each table below is recorded the pres- sure of the gas, in the second, third and fourth are recorded the numbers of half oscillations obtained with the three differ- ent capacities respectively, while in the fifth, sixth, and seventh of these columns are to be found the resistances corresponding to these oscillations, each value being taken from its proper eurve in fig. 2. In order to give a better idea of the compari- son and the way in which the oscillations are damped, repro- ductions from two photographs are given in Plate VI. The Resistance of Hydrogen. Number of half oscillations. | Resistance in ohms. Pressure of | | | ag ‘Capacity = Capacity = |Capacity = Capacity = |Capacity= Capacity = 6,000. 12,000. | 18,000. | 6,000. | 12,000. 18,000. | millimeters | | ohms ohms over ohms 13°5 | 0 | 100 | 10-70. | - 2 50 ? | Feces i 5°0 | val 2 2 | 30 | ZONE, 15 a0... || 2 | | 20 Semel 3 3 | etree steered elias 1) DEO. aliss © 2 3 Slarey Wat Ley Peet ee ew 3 | eee: | [7 ogee 3 + (Lees