An tte On he’ JN he cs js sh aig Ps lth =n ee ee hn ten On Vig mst ttn am fas oo abc thatres Raye tretn tn he © Perera er ee beeen ee a « Foes * ome ote Sen ale ealieioka ert Onin tate Taal SPN i Fh tha ey Hn ye ihe as AA ibs hn Merb. Dieter ant ae ~ Serr rere ee ee cent rn an tina MnO ein tn Ho en AA Ae tate Realy MH Be BO A se all ll Mth tne any Ae Ne lpnten BTM BE MD = Mew ee a Rea ae ee e errerears a rer i tn a ag ry mt A At “ale Deny eh Raat AP Ate ie cee ee ee ae ee Wee ees ae ntberait-war eer are aes rc aie Atte fe fae”, he fen oti iA aint ote thatiabhn oD Aastatie RaW saa MME AR Sm acta Ae ah Ee. ake Sm thre te MB AR fu A Be te Re Oy 1 wn Ale tial a Sarre 0 een eer ere ee ‘i Wate geet de = te Ne Mn Fetch ot oT aintnning etna tena mottelle PLN, ann t= At Nanette aha fe A th s'Aoe NaS, ~w- a ast A lata BL Iie A a aPC =} een Sag item ey at ea Nae ee pee pe eae eR rein pease mre en ed repgenaren tl © ater wr GPa RAO hehe ete mete Pan Pas Sahn hee a AT he ek th et A de ee, eee ta MM we rath Rete ant Basanti A MAMI SIAL NS ae ee eee ae a ne anna nt et tA AR ptm telnet ite aes a iw fg a them Vapi Me hy Tenth B= ie tla ket Bet eo a OR en lah alte Mel dint“ sn r= fhe Rha Milt He MO Sn SRT i ahaenserseainen Pe ete chet Ae Ll te th etelipicN ip patie te Rie AY AN Ren ee No Se ene i tt nn nt hahaa at tne STR ah cnet ye ae Lee ey So ee amie thee ihe eds ram nate tia Talat Mn RIM Resta AON! een Ta ee Re Melee N. La let PIA MS PDP Llp Mond as Bae OA RO tine eae Nee Mae RM tert Ne ter Ante Dati CULE A tate ERE AN arta tt he — ers SS RN aed alata te an thn Pace Asi Sg Mathie Te iw et. at tt A eee IA Me UM, peat kh the. A AT sien et eR pthes tac ee Sen ta Anetta eM Me on heg SNe Lega Bac: Mei Rm iva orn t= Reh actA RATS On Se Sega st apqytrecar te ease Ma Retar Mey ttn tertin Mer Me Bete myo gE ee arte ee Pita Pe te tam hype Ten teen Mh —— Pere A Rem cD Ste eee Det EM Rie Bi Pe AN ats ee net Na NN Nga Alisa RA Reman bk it ig inet te ascaca armen angen he ett ~ 2 Ue bite tate Aen tM tert ae Wem Hee te hin in hn sie Hi a an 1 tel eG ye et Oy LAM ere Fe AA RE dd = Pata east he hE en As hs em eens, meth Me le HO Ce ad Pan Ae Mt Ae hg Ain eA en At tA ee re a ant ee ee NT rere no ee eT See ieee ae rae Re aah RAL eh oN Sain Mgnt dE, Resa Bn om Pit Pn ha ols Nalin ey en Ri init ee eee vee tL Reh Me Poneto Sa ere tend Rl Pa aan dA Mall Nt utah Ao Leia Ne aT mR SECC Na ono er te CIE a1 Ne Mae Me an Mos EO stn Pha the she Tne Mee Mat De tg in He Me etn ote PP tad sah Pn te ithe Io A iat tee ee ee SM. he ie Pace Pe ti Mie, ne Pit | eat Arsen. te Mtn Reb ah me Pe cteniv he ee ba crt TR te nth ok ted te A ancl a9 ead Aenea AANA RAE ae Rin asic tg the Cag Arlen ana ath, Cate Taye ten eRe’ tg % tate te Ma Meet Nei to ons ar ee a ee cere eS amarteetin ee elt Ma ete A oe Mi Na de ete Rl MBS!» + Sf hm No le A cht A Me ee Tig tt Nas it Man, Melttip> Se MA, edie ett MO Nh nh Wath ee es nS een anh nn Dt MDC NR eS ha RN re Se eta Nhe? hm he Ph es Mn Mea We Pen ee Roti sae ade Pin Ne ge ne Cer Bn New By A= Ta a in Nan A Mt lan he Ne TA, NA AI Bae sa A A Tin te = es Mente ee ee Poe any AD ne Mc NAN ene PTMOA CM an heen And SF Tn aa tte the Mig that |t Na re are BATES SET ie tripe ae wear on FP Rae aw ns on = im then MN ac tas Va: nants “ter tithe te Saws = Bn tw te Re the i eh th Tae A ee eee Ate Me ein ie Cs Pits aR a Bn haem ‘tes ms eS Pe a ee Raden PRL My he Be erent IAP Nae tt ki tee aM rere ese aie * ite ta AN, fe tones an arith tate een whet ey" 2 al, a the wai On Aes Bae Se NOD ender svar Srey weeny Sn ee EE emia ora Bo tae Nh hdc ochags Nhe te ee RE AMAT tm er teernarne ete eeheendi Ge RM, ER ee ait. tn eA IA ALSINM Mn Np nt Rael Me Boone Matta att aM etn Gol Ah th Ba A RAMEE ee aa tthe Anti, 5h dee a the sen Ung ny "ROT the ae pts Pe hn Mo at SN hen ets STW ERO aT. lt a i te ee i Bhat Tete mn ttle Poo OM sa gst De DA Pichi Mao Aa Min a I ee ee “he thee malonate MI cnt Me hn the ll a ile Se eh tn etal Es id a cece ta een ara as ce Aa haat Re tag or awe stay tae Me Aes ata a Se a he ty Pe I Ra Lhe TM EE ee a arr oar SD Ka. ath na ots te OA Mane ne phan AA sc en ee — Arne ack Mga pes ha Ranh, Saleh Nee Ma IR paarteiaparer tirsen te arte fe rere “aan tel A fhe Monto ew satan ee onsen Pe le Ha he OR ae ntl een aed teh ‘it = Do ke OMe ee altri ute Me tn Mw Nae Md on te SAS sh Bima th» Hy, ag a Pet nn ee eee eam thy Stee" a Rr ee rete Fe en ett es BAB AL ee ee are be RE Pe en a eo eer eepererare ha om tothe f= etme ein Miho ny ag tg tential = Su ita te Rael ng th tenn Meer tT inna 6 per er Peep re ee rae Tae thie Mu Be actA MTN! Oe Ms An BE Pigeon ee Be UA, Tate Se IR rena ent re Ie Si na ee an Ett et nT PRS atta ing alt eo he kateb Net tt ial Seale te ela Anetin: tathe te = Mall shes MBs! Wf Mb Sie Moi Mane Mat tn eH Antti Rn tia Misia sy te Rend Ws Khe SN ee ad eR Nats ett te Be ee NA titty SE lt Pmt tens 4 ® Pent ®.4. 2, he NN eta dh alg Rit Ae BPA SSD OS, eS Se The * he He Nn * Tn metinn'hn' ee PRM Malt Sattalites Leg He Atte le te et ae Map: Bat RNa tts eM HA en nw ni ee eh TL a hat le Rt Re Hr eth: MS Palad A&P ee Ur Pan A RE oe we ee ee Te & nt rear eT nr sree ae ee eee ra Fins Baas Baa Mal 5 arta athe ete Sw me ee Be ee nn ics Ranta rte tite Pa eRe ne Oe ee Ret Peat Pata SA eRe tan OB Be Ka Reh A A Oe, a Pe Tet te Meh eet be ae Paha ss aN est hs AB BN et A ALR AEM Re on ORT a ten hen ek. lh ee car eee nee rae aR Pa a4 nhs Tay Po Mey Sean Be ating ie Ho Ha fh fn, a ae Mali A FR i aH He ogee oh He Ba Mata 12. ee ee te eee meta My, Te A ee aOKE I KN 268.2 Toes ie etaeeimaeel en aan Timlin Abert eta PHY, Te tate Thy? Ree The ansehen tht ee eee ne ae he ince An Antee cer afin eA SB oi hs (2 ate Be S ba ie = re ad Se i aleed en th Coen 9 let hy Pith el ee Hee ae 8 ee ae ee Sete ie Ae Me Vine a eA FR alte ei ® ee ea ee ee ee WAAR ote Fe te i dl Pay Oe eee a De a me ec anor nated ella Ry a hat Re at cn Set Riser 950 ae Re da ten Meher etek BL ee Ree the a te the Meet eR et tne Mei Me PM A ee Ke ot SR en cnthn He teed Ae Meth, eee ae lei NN — Haile Pe ee ee ak ee et ae AER Pee es ae eA & wh Se ah MANN IN ttn etre, Pi Mire Be TB Me ~ i Page" an ea da Bn > Me Manlio ee eRe =! ee ee Fin ea nate a OR tae Tn Mee F lo An oe Kol. Me & cinders Se Pe Re a tt Reet ee OP ee ee IN Sy OF Rega hs te Pinte Me, ect Poth, Mon etn nem eh - ag the AR TA thee Tey Ma em te ee A hie Oe . SMa gt Ripe te ait aad Ce ek ee a eae i te eee ees * AO Oe ee te a ety Me Le Veils Te i tee. al eo Retha eds. em I Mt ath Net oe . - ow ——~-4 v 1 eet het sa Re i lee ie he ee A SO et ee a ee we ee 4 r wt Mn tng) eta Pea ie ee Me ——% 8 F s n " ae ew Te aiieratiy ar wee re ee Be Ate ks Se Aint nme ihn Ne en Ne ig gm & Rutan ta ee hy Bade ee nite pt AB Rae a! a Me te 9+ 9 ata Refine taste The Finan = m Dette AMM = 1's Penh ain eet a Re att Hatten a Me Maha he eS, x crn an cet i NR te ee Ba saat. ’ ee ee an ee ge Te ~~ o™ a setae Hl einem Mea ties Lela eet tee thettatine he ee Pitas Matin toy = hehe ‘ gM Has Rely. G1 Peete Rates ? ares ee ere , Marae P os pl t “ yea Oye Ain, ee ~ or fama oh ete Sa 4 eras LP MCP Re ae? ~ en - a ee SE hes Pi re Fe thet the “se + en ote” ee ” oh ve ae ee ee Uae en ee 2 see ee el Bet ate Beit otefe © eo RA Te tte tel ee OVD mee 2 Werk a A ek A ee Bride => ye Me, ote le mer oe a %e"*~! ee el Me oA”. oh a 4 0S ert ee Ln a en sae Maem, te axtecnthot SLO A ae Se oe en ee ERM Me pare oe Wren a wees ee Se tata Ne Re Ph ae ee a a Te ante we er et 7 tA t-te ee Rag ode a Pe ret A a ee eee a ee aa ects ae at a” ae eh th et “ ee Ae Atte, etl Sta MO Re Rit NP tle at Tr pte Sw Mien be eek ete ES a ae a ee ee i) wohl * ae the Enna Den Hts in Bethe) Se ee Pe tote b ~" & ee ee 2 a ie ®. herent ome Re A Bh et te B4q ofa owe % a Mp Min & Sette Care Rowe* ee Pa nals Me $A An RRA BLO. OH NS 8 mt em « ae “an Re the ntl, OA 7 Nag Des tee ete ny Me ratgt ois Pn? Ae Pertinent ee Onan ter a OR eT Oe ig ee ee we oe Tee CAEN EM - ee ee eMamga Gata Sm ne Fhe Ne hee Ret MMi Ne te ge sania Mel ae RA ee Se eee am bY —- et Fe agtd ~~ i LT Me! ee ee vem e CEN ps KP Rs = a Ak Ee et A pe Ae ee Pree RM ate = ee ve ae ard th & 7 te, a AeA Oe ee Aah Ae Oe ne) we ee, ee Be ete alee de het 4. ART * Tithe. wteMallet. ot + — Cee A ee A ee See eraw at Ratt a ta hs Se tetek Re te oS NO te te te eta ee A. Ane ~ ths ath a Thr tat Ve i a as » ee aS ™ - de eS a ua" no FS = - Nn te a Rm Awa’. &™% ah, Pat ee IA ated ht Mae ee th TS et tan Ste athe Mae Metin ee ee es + Nee Pe 4 24 qt 4 ge RL ~~ ine A ha ft alll, hes LAK o + taren —* a kun ~ ~ i" ‘ Sheth ate! ee ete, A ee iy © io ” A ae ‘ - ‘ +o met te eA ete Ree Ante 0 er Ane HSS 4D I -~o - we EES U2 tet bed . 2 oe th A Keen AS oo . . ib eyes Whe one tae ee Sr) ‘ > > - Be Tie ye tee TAL Ae’ oe 6 ew ~~ wy n ae ~~ a. eee ee ee - ——t eee as ‘ or ~~ & . "te - . ~- = RAAB ee ee eee An ee ee me, il +" 4445 nS “ a Ss yan ee - . a ™ a we * .3* a igne oh on aa EAE =a. = CAAA BAe ~ ~~ os =n * ~ a he e™ - Pa ee x -—~ ™ “a yp Ra -. ea - o% BA eit it, das SAA ~ 7s . — > : 7 Vos * ~*% . ~ V~ me a ee on am we ie ee ee a “ - se Ne = . me phat cite =m - a A. ~ » i tne ~ —e ~ . a ar ~* eee =e A a Es nt Oy be - a +o “ x Pe ~ ee ae = ws {> eae * : ee - . —" he - ~. An “7 “ ~o hed > » - Marte ® weet De 9 . ' = - om a ~ ~ = ~- we wana ka i ah Oe . . - wae ma Peery » om awe => a4 8 le a Tn te a 9 Smee &! See Ce oe = ~ “ - . os a me - oor 4 tee eM . - ne . ~ ' te = ~ * a ie a 7 . » ax . , 7 7 «* - ‘ rot ae z~% me + Js ~ - x a a. nae = a» * an = 7 eo we ite : CR a ee ee P :: = te oh Nm a + a x ta ete ~ 7 " oe > - ate “ ~ * Pirin sae . ‘. x . * “ i Ye <> +> ad - - . . . aA~ ‘. rn ‘ + x - _* a | a 3 —_ »~ “* - , mw ~ ~ we - ‘ ion = - “ - ~* ” ~ ~ = . ~ C= - 7 4 ——e sues a * as ‘ 7 ” “ -” ~~», . ' ” ~ a 7 ~ ~ rn = “ ~ - * ~ ™% . 4 a* vA . ‘ > *. is ' ~ - w ~ « - # AMERICAN JOURNAL OF SCIENCE. Epiror: EDWARD 8S. DANA. ASSOCIATE EDITORS Proressorss GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or CampBrwpce, Proressors ADDISON E. VERRILL, HORACE L. WELLS, L. V. PIRSSON anp H. E. GREGORY, or New Haven, Proressor GEORGE F. BARKER, or PHILADELPHIA, Proressor HENRY §S. WILLIAMS, or Irwaca, Proressor JOSEPH S. AMES, or Battimore, Mr. J. S. DILLER, or Wasuinerton. FOURTH SERIES VOL. XX—[WHOLE NUMBER, CLXX.] WitH 15 PLATES. NEW HAVEN, CONNECTICUT. 1905 9S EOE 4 : ; : THE TUTTLE, MORE HOUSE CERT pe eS ips VV VWUseUN TAVOLAN): ii, See MOTT 41 1: CONTENTS TO VOLUME XxX. Number 115. Page Art. I.—Iodine Tiv.. %» Voltameter; by D. A. Kreiper. 1 I1.—Handling of Preci, tes for Solution and Reprecipita- peetetnoy ll An GoOGw ose nt eee OS es ee iT IlJ.—Estimation of Sulphites by Iodine; by R. H. Asurey 18 IV.—Revision of the New York Helderbergian Crinoids ; fale Pareor ( (With Plates IDV). 22 24s ees: S Let V.—Petrographic Province of Central Montana; by L. V. oy TESS DIN Gs SRE a Be ag Mia Nee perma ees Ae UG Me 35 eee roomia paucitiora ; by I’. Horm ...2 .--2222..252-2. 50 VIl.—Relative Proportion of Radium and Uranium in Radio-active Minerals; by E. Ruruerrorp and B. B. Sep ATO Coe ge ee he eho oe eae Rs i BB VIII.—Side Discharge of Electricity ; by J. TRowBRIDGE -. 57 IX.—Effect of High Temperatures on the Rate of Decay of the Active Deposit from Radium; by H. L. Bronson-_. 60 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Amounts of Neon and Helium in the Air, W. Ram- say: Radio-activity of Thorium, O. Sackur, 65.—Use of Quartz Apparatus for Laboratory Purposes, Mytius and Mreusser: Permeability of Quartz Vessels to Gases, BERTHELOT: Outlines of Inorganic Chemistry, A. Goocu and C. F. Waker, 66.—Spectroscopic Analysis of Gas Mixtures, J. E. LILIENFELD : FitzGerald-Lorentz Effect, MorLEY and MILLER, 67.—Normal Klement: Influence of Character of Excitation upon Structure of Spectral lines: Radio-active Minerals, R. J. Srrutr: Absence of excited Radio- activity due to temporary Exposure to y-Rays, 68.—Handbuch der Spectro- scople, H. Kaysmr, 69. Geology and Natural History—United States Geological Survey, 69.—Pre- liminary Report on the Geology and Underground Water Resources of the Central Great Plains, N. H. Darton, 70.—Origin of the Channels surround- ing Manhattan Island, New York, W. H. Hoxsss, 71.—Isomorphism and Thermal Properties of the Feldspars, A. L. Day and E. T. ALLEN, 72.— Tin Deposits of the Carolinas, J. H. Prarr and D. B. Strerrett: Tubico-* lous Annelids of the Tribes Sabellides and Serpulides from the Pacific Ocean, K. J, Busu, 75.—Student’s Text-Book of Zoology, A. SEDGWICK : Preliminary Report on the Protozoa of the Fresh Waters of Connecticut, H. W. Conn, 76.—Etudes sur l’Instinct et les Moeurs des Insects, J.-H. Fapre: Rocky Mountain Goat, M. Grant: Catalogue of North American Diptera (or two-winged Flies), J. M. ALDRIcH: Fauna and Geography of the Maldive and Laccadive Archipelagoes, J. S. GaRDINER: American Museum Journal, 77.—Cold Spring Harbor Monographs, 78. Miscellaneous Scientific Intelligence—Vom Kilimandscharo zum Meru, C. Uutie, 78.—Glacial Studies in the Canadian Rockies and Selkirks, W. H. SCHERZER : Solar Observatory of the Carnegie Institution of Washington, G. E. Hate: United States Naval Observatory: Publications of West Hendon House Observatory, Sunderland, 80. 1V CONTENTS. Number 116. Page Art. X.—Mechanical Equivalent of the Heat Vaporization of Water; by R. H. Hover . 2.) >... + ee XI. —Phosphor escence of Zinc Sulphide through ‘the Influence of Condensed Gases obtained by Heating Rare-Harth Minerals ; by C. BaskERvitue and L. B. Lockwart.... 93 XII.—Action of Radium Emanations on Minerals and Gems; by C. Baskervitie and L. B. Lockmart __..-___-_-_- 95 XIII.—Behavior of Typical Hydrous Bromides when Heated in an Atmosphere of Hydrogen Bromide; by J. L. KRBIDER: .. so et us a Le 97 XIV.—Glacial (Dwyka) Conglomerate of South Africa ; by He 'T. MELLoR 250 eso ee XV.—Formation of Natural Bridges; by H. F. Cuezanp._ 119 XVI.—Quartz from San Diego County, California; by G. A. WARING SoS tosup oc Sel 10 Se 125 XV II.—Radio-active Properties of the Waters of the Springs on the Hot Springs Reservation, Hot Springs, Ark.;-by B. By Bottwoop.. i.e a 128 XVIII.—Genesis of Riebeckite Rocks; by G. M. Muregocr . 133 X1IX.—Purpurite, a new Mineral; by L. C. Graton and W: Ty. SCHALLER = 22o0uc.. epee 2 146 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Studies with the Liquid Hydrogen and Air Calori- meters. I. Specific Heats, J. DEwar, 152.—Thermo-electric Junction as a Means of determining the Lowest Temperatures, J. Dewar, 153. Geology—Geology of the Vicinity of Little Falls, Herkimer County, H. P. CusHING, 156.—Geology of the Watkins and Elmira Quadrangles, accom- panied by a geologic map, J. M. Cuarxe and D. D. Lutuer, 157.—Geologic map of the Tully Quadrangle, J. M. CuarkEe and D. D. LuruEer, 158.— Contribution to the Paleontology of the Martinez Group, C. EK. WEAVER: Faune cambrienne du Haut-Alemtejo (Portugal), J. F. N. DELeGapo, 159. —Paraphorhynchus, a new genus of Kinderhook Brachiopoda, S. WELLER : Sympterura Minveri, n. g. et sp.; a Devonian Ophiurid from Cornwall, F. A. BatHEeR: Ancestral origin of the North American Unionidez, or fresh- water Mussels, C. A. Wurre, 160.—Thalattosauria, a group of marine rep- tiles from the Triassic of California, J. C. MerRR1am: Geology of Littleton, New Hampshire, C. H. Hircucock: Vorschule der Geologie, J. WALTHER, 161.—Die Moore der Schweiz mit Beriicksichtigung der gesammten Moor- frage, J. Frtw and C. Scurorer, 162.—Study of Recent Earthquakes, C. Davison: Introduction to the Geology of Cape Colony, A. W. ROGERS, 168.—Ice Erosion Theory, a Fallacy, H. L. Farrconixp, 164.—Hanging Val- leys, I. C. Russewx, 165.—Glaciation of the Green Mountains, C. H. Hircn- cock: Ice or Water, H. H. Howorrts, 166. Miscellaneous Scientific Intelligence—United States National Museum, R. RatHBUN: Forestry ; Tenth Annual Report of the Chief Fire Warden of Minnesota, C. C. ANDREWS: Les Prix Nobel en 1902: Negritos of Zam- bales, W. A. ReEp: Magnetic Survey of Japan reduced to the Epoch 1895°0 and the Sea-level, A. TANAKADATE, 167.—Beitrage zur chemischen Physiologie, F. Hormerster ; Du Laboratoire & V Usine, L. HOULLEVIGUE : Traité Complet de la Fabrication des Bieéres, G. Moreau and L. Livy, 168. CONTENTS. ar Number 117. Page Art. XX.—Development of Fenestella; by E. R. Cumines. Oia lates V5 2 Vil ands EE) coi ee Se 169 XXI.—Age of the Monument Creek Formation; by N.H. ORPRESEEN 2 teed ee as Sse SN St A a a en Se te) 178 X XII.—Iodometric Determination of Aluminium in Alumin- ium Chloride and Aluminium Sulphate; by 8. E. Moopy 181 XXIITI.—Secondary Origin of Certain Granites; by R. A. 2 GT Sa eae pee riieaeae EN leh etna ata Sage tric aerate LU nots 185 XXIV.—Tychite, a New Mineral from Borax Lake, Califor- nia, and on its Artificial Production and its Relations to Northupite ; by 8S. L. PENFIELD and G. 8. JamMIEsoN 217 XXV.—Modification of Victor Meyer’s Apparatus for the Determination of Vapor-Densities; by B. J. Har- EMME CMNE Seo 8 Poe cs a LN se cael tre ces Geren ODDS XXVI.—New Lower Tertiary Fauna from Chappaquiddick Island, Martha’s Vineyard; by T. C. Brown. (With ere aed Sore Ce ee ee Se O99 XXVII.—Production of Radium from Uranium; by B. B. em OO pee Be eee ee aks Se ae 2 239 SCIENTIFIC INTELLIGENCE. Geology—Explorations in Turkestan with an account of The Basin of Eastern Persia and Sistan. Expedition of 1903, under the direction of RAPHAEL PUMPELLY, 245. al ' CONTENTS. Number ois: Page Arr. XX VIII.—Ultimate Disintegration Products of the Radio-active Klements ; by B. B. Bottwoop .__-_.--- 253 XXIX.—Use of the Rotating Cathode for the Estimation of Cadmium taken as the Sulphate ; by C. P. Frora ._ 268 XXX.—Crystallization of Luzonite; and other Crystallo- graphic Studies ; by A. J. Mosus__..-__.__.- Base aa XX XI.—Determining of the Optical Character of Birefract- ing Minerals; by F. EH. Wrigur..-. 2.0.75) 0 ees XX XII.—Groups of Efficient Nuclei in Dust-Free Air; by C. Barus eet 0 ls oes a rrr XX XIII.—Studies in the Cyperaceer ; by T. Hotm_____._. 301 XXXITV.—Preliminary Note on some Overthrust Faults in Central New York; by P. EF. Scunumur._ 725s - 808 XXX V.—Petrography of the Tucson Mountains, Pima Co., Arizona; by EH. N. Guinp. > (With Plate 1X¢) 7297s 313 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Gases produced by Actinium, Drprerne: New Heavy Solution, Dusorn, 319.—Hydrolysis of very Concentrated Ferric Sulphate Solutions, RecouRa: Separation of Gold from the Metals of the Platinum Group, JANNASCH and von Moyer: Determination of Sugar with Fehling’s Solution, Lavauur, 320.—Slow Transformation Products of Radium, E. RuTHERFORD, 3821. Geology and Mineralogy—Indiana, Department of Geology and Natural Resources, Twenty-ninth Annual Report, W. S. BLhatcHLEy, 322.—Geo- logical Survey of Louisiana: Geological Survey of New Jersey: Brief descriptions of some recently described Minerals, 323. CONTENTS. Vil Number 119. : Art. XXXVI.—A New Niobrara Toxochelys; by G. R. imine (With Plate X:)\ 2 le eo et 325 XXXVII.—Contributions to the Geology of New Hamp- shire. I. Geology of the Belknap Mountains; by L. V. Pirsson and H. 8. Wasuineton. (With Plate XI.) 344 XXXVIII.—The Fauna of the Chazy Limestone; by P. E. EPPA EOND esc ra oc aa ye at ey a eee a 353 XXXIX.—The Mechanical Properties of Catgut Musical Seiten Dye divs ISUNTON: 225222080 35 2208 oo Le 888 XL.—Use of the Rotating Cathode for the Estimation of Cadmium taken as the Chloride; by C. P. Fiora_-_- 392 SCIENTIFIC INTELLIGENCE. Chemistry and Physics.—Formation of Ozone by Ultra-violet Light, FiscHER and BRAEMER: New Reagent for Nickel, TscuHuGarrFr, 397.—Electrolytic Dissociation Theory with some of its Applications: Soils and Fertilizers : Engineering Chemistry, 398.—Text-book of Chemical Arithmetic: Text- book of Physiological Chemistry for Students of Medicine: Formation of Helium from the Radium EHEmanation, 399.—Blondlot’s ‘‘ Emission pesante”: Diffusion of Nascent Hydrogen through Tron, A. WINKLEMANN, 400.—Landolt-Boérnstein Physikalisch-chemische Tabellen, 401. Geology and Mineralogy.—United States Geological Survey, 402.—Osteology of Baptanodon, C. W. GitmMorez, 403.—Cambrian Fauna of India, C. D Watcort, 404.—Catalogue of Type-Specimens of Fossil Invertebrates in the Department of Geology, U.S. National Museum, 405.—Graptolites of New York; Part I, Graptolites of the Lower Beds, R. RUEDEMANN : Mesozoic Plants from Korea, H. YABrE: Paleontologia Universalis: Ninth Annual Report of the Geological Commission, Dept. of Agriculture, Cape of Good Hope, for 1904: Rock Cleavage, C. K. Lerry, 406.— Experiments on Schistosity and Slaty Cleavage, G. F. Becker: Die Alpen im Liszeit- alter, 407.—Structural and Field Geology, J. GrrKie: Clays and Clay Industry of Connecticut, G. F. Loucutin, 408.—Geology of Western Ore Deposits: Delavan Lobe of the Lake Michigan Glacier, etc., 409.— Platinum in Black Sands from Placer Mines, D. T. Day: Cassiterite, W. EK. Hippen, 410. Miscellaneous Scientific Intelligence.—Harvard College Observatory : Pub- lications of the Cincinnati Observatory: Report of Director of the Yerkes Observatory, Univ. of Chicago: Carnegie Institution of Washington, 411. —Annual Report Board of Regents Smithsonian Institution: Catalogue of Collection of Birds’ Eggs in the British Museum of Natural History: Bibliotheca Zoologica II, 412. Obituary.—Baron FERDINAND VON RICHTHOFEN, Professor LEO HERRERA, Mr. G. B. Buckxton, M. Eviste Recuvs, 412. Vill CONTENTS. Number 120. Page Art. XLI.—Two New Ceratopsia from the Laramie of Converse County, Wyoming; by J. B. Harcuer. (With Plates’ XII, XYIT)... 0202 a XLII.—Restoration of the Horned Dinosaur Diceratops ; by Ricuarp 8. Lunt. (With Plate XIV.) .._.._ 073 420 XLITI.—Triassic System in New Mexico; by Cuaruezs R. KEYES 0025952500200 423 XLIV.—Structure of the Upper Cretaceous Turtles of New Jersey; Agomphus; by G. R. Wiktanp. > 927 ae XLV.—The Cambro-Ordovician Limestones of the Middle Portion of the Valley of Virginia; by H. D. Campprtn 445 XLVI.—Relations of Ions and Nuclei in Dust-free Air; by CARL BaRus 2.040002 202 448 XLVIL—Additional Notes upon the Estimation of Oad- mium by Means of the Rotating Cathode, and Summary; by:Cuartes P) Frora 220s 454 XLVIII.—The Estimation of Cadmium as the Oxide; by CHARLES “Pe rors 2 ee22 Cees jee ee eee XLIX.—The Mounted Skeleton of Tecan oe prorsus in the U.S. National Museum; ie C. Scnucuerr. (With Plate XN.) (220222 22.25 22 re SCIENTIFIC INTELLIGENCE. Chemistry and Physics.—A New Formation of Diamond, Str W. CrooKkzs: A New Compound of Iron, Orro Hauser: Nitrosyl Fluoride, Rurr and SrAusER, 460.—The Atomic Weight of Strontium, T. W. RicHARDS: Qualitative Analysis, E. H. S. Barney and H. P. Capy: Charging Effect of Réntgen Rays, Kari Haun, 461.—Emission of Negative Corpuscles by the Alkali Metals, J. J. THomson : A New Method of showing the Presence of Neon, Krypton, and Xenon, S. VALENTINER and R. ScumiptT: The Mechanical Properties of Catgut Musical Strings, J. R. Benton, 462. Geology and Mineralogy.—lowa Geological Survey, Volume XV, 463.—Sum- mary Report of the Geological Survey Department of Canada, R. BELL: Glaciation of Southwestern New Zealand, E..C. AnpREws, 464.—Mastodon- Reste aus dem interandinen Hochland von Bolivia, J. F. PomprcKs: Description of New Rodents and Discussion of the Origin of Daemonelix, — O. A. Peterson, 465.—Economic Geology of the Bingham Mining District, Utah, 466.—Economic Geology, 467.—Minerals in Rock Sections, L. Mcl. LuqueEr, 468. Miscellaneous Scientific Intelligence.—National Academy of Sciences, 468.— The Geological Society of America: A Laboratory Guide in Bacteriology, Pauu G. HernemMann: British Tunicata, ALDER and Hancock, 469.—Cata- logus Mammalium tam viventium quam fossilium, EH. L. TROUESSART : Carnegie Institution of Washington: A Handbook of the Trees of Cali- fornia, A. Eastwoop, 470. Obituary.—Professor DEwitt BristoL Brace, Professor RALPH COPELAND, Professor D. W. von BEzoup, 470. InDEX TO Vou. XX, 471. U.S. Nat. Museum: MMMM a pe = | JULY, 1905. Established by BENJAMIN SILLIMAN in 1818. gue i0) AMEHRICAN JOURNAL OF SCIENCE. |. Epitorn: EDWARD S. DANA. ASSOCIATE EDITORS PROFESSORS GEORGE L. GOODALE, JOHN TROWBRIDGE, | W.. G. FARLOW AND WM. M. DAVIS, or CamsBrwpce, a Bee scors ADDISON E. VERRILL, HORACE L. WELLS, L. V. PIRSSON anv H. E. GREGORY, or New Haven, Proresson GEORGE F. BARKER, or Pumapenput, Proressorn HENRY S. WILLIAMS, or Itwaca, Proressor JOSEPH S. AMES, or Batrimorg, © Me. J. S. DILLER, or Wasuinerton. FOURTH SERIES VOL. XX—[WHOLE NUMBER, OLXX.] Noxtie JULY 1905. WITH PLATES I-IV. 1905 \ THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPDE ee Published monthly. Six dollars per year, in advance. $6.40 to countries in the Postal Union. Remittances should be made either by money orders, registered see letters, or bank checks (preferably on New York banks). . - Ve CRYSTALLIZED STANNITE — A small consignment direct from the silver mines in Bolivia affords new and unique examples of the following. They have been identified by high mineralogical authority. . Stannite, Splendent crystals, grouped in cavities of the massive — mineral. The crystals are most perfectly developed and very rich in planes. The species has been known to science for over a century, but until recently only in the massive form, thus leaving its crystallographic form in question. Hence the present find is peculiarly welcome. Andorite. A silver and lead sulph-antimonide. Massive and in small but characteristic crystals of highly modified orthorhombic habit. Some- times associated with the crystallized Stannite. RARE MINERALS Specimens are supplied by us to students and chemists for purposes of ~ comparison and investigation ; in commercial quantities for industrial uses, SYSTEMATIC COLLECTIONS OF TYPICAL SPECIMENS In sets of twenty-five up to fifteen hundred specimens. Prices $0.00 upwards per set, the average price for students’ specimens being about twenty cents. We have supplied the leading institutions for thirty years, having lately completed a single order for over 60,000 specimens. Our material is the accepted standard both as to correct labeling and high quality. Free Collection Catalog, containing lists and illustrations of General Mineral Collections, Series of Ores fore Prospectors, Sets of Crystals, Series _ illustrating Hardness and other Physical Characters of Minerals, with Price List of a Material and Individual Specimens. FOOTE MINERAL GO, Established 1876, by Dr. A. E. Foote. W. M. FOOTE, Manager. ea DEALERS IN MINERAL SPECIMENS AND COMMERCIAL RARE MINERALS, 1317 Arch Street, Philadelphia: EEE AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES.] Arr. 1—An Lodine Titration Voltameter; by D. ArBert KREIDER. THE rapidity, accuracy and sharpness of the end reaction of several of the methods of volumetric analysis, particularly of the iodometric methods, have suggested the possibility of applying them to the voltameter. A rather extensive investi- gation of quantitative electrolytic oxidation and reduction methods, with this aim in view, has resulted in the evolution of a titration voltameter the accuracy of which may be depended upon to about one part in ten thousand. The advantages in point of manipulation and time required, as well as its applicability to a greater range of current density, will, in the writer’s opinion, make it of service in many investi- gations. The basis of the method is the electrolysis of potassium iodide and the titration of the liberated iodine by sodium thio- sulphate. Herroun* first’ suggested the use of iodine in this connection. He electrolyzed zinc iodide between a platinum anode and zine cathode in a beaker; but gives the results of only one determination and leaves the method in an imprac- _ticable form. * Danneelt reports four comparative tests of Herroun’s zine iodide voltameter in series witha silver voltameter. The results show a difference of between +0°27 per cent and —1 per cent. His burette readings were made to only the nearest 0°1°°, and the total quantity of thiosulphate was small’; varying from 7 to BA. The work of Danneel, as well as that of Kistiakowskyt on a silver titration voltameter, seems never to have found its way * Phil. Mag. [5], xl, 91, 1895. + Zeitschr. fiir Elect. Chem., iv, 154, 1897. t Zeitschr. Phys. Chem., vi, 97, 1890. Am. Jour. Sct.—Fourts Serizs, Vou. XX. No. 115.—JuLy, 1905. 1 g D, A. Kreider—-lodine Titration Voltameter. into the general literature. The work of both men, on voltam- — eters, was incidental to another research and the titles of their articles fail to afford any clue to this particular contents. My own preliminary search of possible reactions had been com- pleted and I had settled upon the form of the iodide voltameter before I learned of the work just mentioned. Herroun rather erroneously points to the high electro- chemical equivalent of iodine as its chief advantage; this is obviously immaterial in a volumetric method. As a basis of a titration method, however, iodine has peculiar advantages in the rapid action of sodium thiosulphate upon it and in the extreme sharpness of the end reactions, intensified when desir- able by the addition of starch solution. Potassium iodide is, moreover, much to be preferred to the zine salt. It 1s obtain- able commercially quite free from iodates and its solutions may, therefore, be acidified without the liberation of iodine, except for the very slow action of atmospheric, or dissolved, oxygen. It is, of course, impracticable to electrolyze the potassium iodide directly, when the liberated iodine is to afford a measure of the current; this is because of the recombination of the electrode products by diffusion. Acidifymg prevents this recombination, or even after it has taken place, in an acid solution the original amount of iodine is liberated, providing ‘the iodine has not been allowed to diffuse to the cathode during the electrolysis, where it would tend to combine with hydrogen and would be irrecoverable. A most satisfactory action is obtained when the anode is submerged in a strong aqueous solution of the potassium iodide under dilute hydrochloric acid in which the cathode is suspended. With the electrodes thus placed vertically, the anode below, the liberated iodine, because of its great density and solubility in potassium iodide, diffuses very slowly. After a run of hours, and even on subsequent standing for hours more, the solution about the cathode remains Boe color- less and free of iodine. The Potassium Iodide Cell. Asa cell, I have employed a side-necked test tube, fig. 1, drawn out at the bottom to a long capillary. Into the junction of this capillary a glass rod was ground. The test tube was closed by a rubber stopper with three perforations, through the middle one of which the ground glass rod passes air tight, though not so tightly as to prevent easy motion when moist. The other perforations receive the glass tubes through which the wires connect with the platinum electrodes ; anode at the D. A. Kreider—TIodine Titration Voltameter. 3 bottom of the tube, cathode above. The size of the electrodes was, in the small cell, 1-6 X 2-7"; in the large cell, 2°5 & 6°. In the latter case the cathode was somewhat smaller and corru- gated. These electrodes were bent into cylindrical form and arranged coaxially with the tube. The smaller cell had a diameter of 2°, length, 12™, length of capillary, 7. Its capacity was about 30°; about 7° being required to cover the anode. The larger cell was made in the same way, of a tube 3 in diameter and about 15™ in length, with a capil- lary about 18 long. With the distance of 5™ between the nearest edges of the electrodes, the total volume required to cover the cathode was about 60°, about 20° sufficing to cover the anode. The apparatus was filled by raising the ground glass rod and by diminished pressure, effected through the side neck, drawing up through the capillary successively the required amount of hydrochloric acid and then the strong solution of potassium iodide in water. In this way the anode is completely submerged in the concentrated solu- tion of iodide without the use of excessive quan- tities. The hydrochloric acid serves as an elec- trolyte, permitting the separation of the electrodes sufficiently to preclude the possibility of any interaction of the electrode products. By this method of filling the cell the iodide is sufficiently acidified, and if not too rapidly drawn in, a sharp line will mark the junction of the two solutions of different density. The electrolysis results in a quantitative liberation of iodine unless the current density is too great. In the latter case oxygen is evolved along with the iodine. With the per- missible current densities indicated in Table I, so long as the potassium iodide is not impoverished about the anode, not a trace of oxygen appears and the action is entirely satisfactory. As the iodine is liberated it sinks along the electrode and by its convection effect renews the iodide at the surface of the electrode without the shghtest disturbance of the supernatant acid. In all cases where the current density has not exceeded the indicated maximum, or where the duration of the current has not been such as to exhaust the iodide (so that no oxygen is evolved) the supernatant liquid remains perfectly colorless and free of iodine, and continues to show a sharp line of demarcation between the iodine solution and the acid. Table I shows the possible current densities, potential change, and permissible time of run, under the given condi- tions. In the first column, the first figure represents the num- 4 D. A. Kreider—TIodine Titration Voltameter. ber of grams of potassium iodide used and the second figure the volume of its solution in water. In the fifth column, where two figures are given, the first represents the fall of potential through the cell at the beginning of the experiment and the second figure that at the end of the run. The interval is found in the sixth column. An asterisk following the figures of the sixth column indicates that the current was con- tinued until, and stopped at the moment of, the first appear- ance of gas on the anode. It is evident that the liberation of iodine is no longer quantitative for some time before this gas appears, the safety limit depending upon the current density. Within this safety limit the rise of potential is also not as great as that indicated in the fifth column. The fall of potential through the cell increases continuously and very regularly, after the first few minutes, in which it rises rather more rapidly. Just before the appearance of the gas the potential naturally rises quite rapidly. The gradual change of potential is doubt- less due to the increased resistance of the potassium chloride solution which is formed during the electrolysis at the bound- ary of the two solutions. This part of the cell, especially the small one, is always considerably heated by the current, and a conspicuous line of increased density of the solution, but entirely colorless, gradually creeps up the tube to a distance of several centimeters in an hours run. In the 7th experi- ment the solution became quite warm and small bubbles of gas appeared on the sides of the glass, long before there was any evidence of gas being evolved at the anode. To test the correctness of the supposition that this was merely dissolved gas, the 8th experiment was made with the cell in a water bath maintained at 9°. Under these conditions no gas appeared until it was evolved at the anode. This experiment also shows that the warming of the solution is, if anything, an advantage ; due, doubtless, to the more rapid diffusion of the iodide or to the greater solubility of the iodine. In experiments 38 and 10 the current was varied; the time of run for each value of the current is given in the sixth column. Three determinations with two of the small cells in series, each containing 2 grams of potassium iodide in 75°, with a current density of 0-015 amp./em’* showed the following satis- factory agreement in the amount of thiosulphate required : ( 31°55 © 28.60% | 16°62 (4) 4 31-58« (>) 98.66 « (*) 16-60 « A comparison between the values obtained by two of the iodine voltameters in series with each other and with one or more copper voltameters is shown in Table II. The titrations lodine Titration Voltameter. 5 D. A. Kreider in these cases were not as accurately made, nor the solutions standardized with the same eare that peed the later determinations, as will be seen from the sequel. TABLE I. Diam= of eel —— 2°": | Length of cell=12°™. ee : Electrodes 1:6 x 2°7™. | Dist. between Beaeeee edges of pibereades! pe | Current Time KI HCl Current! density | pp ot \grms. in cc. of (r:4) |(approx.)| (approx.) Wales oyu run, solution cc. amp. amp fie mins. | ql cm | Esa =e not s 1% | 9.9% 99% 1 lino ? measured 0 13 | 0 015 | 9) Ae Pee 5 a 0-13 | 0-015 2°00 (35* | (| | 0-25 | 0-029 15 ane 18 4 ry — 0°40 0°046 | | 2 more ( eres) | 2402058. } | 43 more* Ale gets: 7+5 ; | 0-5 0°058 |7-1to 7:8,15* BS Ne aw a 20 | 0-5 0-058 [3° “ 4°7)13* ‘Solid iodine at anode See o ee BEER aeeta HiaAS { Solid iodine at anode tee Sia Sd RL aa al | Solution alkaline ee a Se ee Solid iodine in large te ry. 9 . = IDeA 66 - % = 5875 | 2 | 05 | 0-058 paw aes | Peqnantity at anode { Diam. of cell = 3. 3, | Length of cell = 15. (0) Large ceil Electrode (anode) 2°5 x 6, Distance nearest edges, d°™. Sees 20 40 Peso ty Ob aI2+ = 65 Shia ‘Solid iodine at anode | 0-5 0-017 |1:5 ** 2°5/91 | 10)10 ** 20 40 1-0 0-038 |4:1 ‘* 4:7 2 more | | | 1°5 0-050 5°7 _ 1 more* Solid iodine at anode * Gas evolved at anode. TaBLe IT. . Iodine voltameter. Cu- voltameter, | Distance | Pe neeere | | ati eae | between a ewes toute S | Copper | | 5 elec- sre =| rent ee Pets Hales o bi: equiv- Copper | ae trodes, ae Praia ices |e bo ae alent, grins. [Ag an | eee nea Mie | grms. ia —— | | ieee alee 725) 73°90| 0°2391/| | ; ; >) "237 013 1 | (b) 14 75 | 0°5 0°058 25 73-72, 0-2385 ° 379 0-0 af) (2) 514 751) onl ogeolsn § | 182°87| 0:4202|| 0°4199] 13 a) (5) 5/4 * 75| 6 O°) 0845) | 139-71] o-4198) 0-4208 7” | | | | | 3 515 “© 7-5) 0510-05845 | 134°21| 0°4244 ae af 0-013 424 6 D, A. Kreider—TLodine Titration Voltameter. Standard Solutions. Sodium thiosulphate is the most convenient medium for the titration of iodine, but it is not a very satisfactory standard where great accuracy is desired. The salt is readily obtained quite pure, but there is always some uncertainty as to the amount of extraneous water the crystals may contain. Further, on long standing, especially when exposed to the hght, it un- dergoes a slight decomposition, with deposition of sulphur. However, this decomposition is so slow that, if kept in the dark, the solution will remain quite constant for months. In all of my titrations, except those that were merely relative, I have employed an approximately decinormal solution of sodium thiosulphate, standardized against arsenious oxide, by means of an iodine solution. The purest arsenious oxide obtainable was resublimed three times and weighed from a weighing bot- tle. This solution also was only approximately decinormal. The weight of the arsenious oxide was carefully determined and reduced to its weight in vacuo, and the solution accu- rately made up to one liter at 20° in a calibrated flask. The direct employment of the arsenic solution for the titrations is undesirable because it necessitates an excess of bicarbonate and the danger of loss of iodine in the neutralization requires a complicated series of traps. Moreover, the end reaction, when starch is employed, is much slower with arsenic than with the thiosulphate. With the latter it is practically instan- taneous. The Titrations. The titrations were performed in Ehrlenmeyer beakers. In filling the cell, care was always taken to draw up the last of the iodide solution as far as the ground glass joint, but with- out admitting air which would stir up the solutions. To avoid a possible loss of iodine through a considerable leakage of the joint and vaporization from the concentrated solution, the Ehrlenmeyer beaker containing about 150° of water was placed under the cell during the electrolysis. The capillary extension of the cell (fig. 1) reached to the bottom of the beaker. Whenever the leakage of the ground glass joint was sufficient to allow the iodine solution to descend the full length of the capillary during an experiment, the joint was reground with the finest emery. Under these conditions a loss of iodine is impossible. After the current was cut off, the ground glass joint was opened slightly and the cell allowed to empty slowly. The great density of the iodine solution keeps it continuously covered with a layer of water of considerable depth in the D. A. Kreider—lLodine Titration Voltameter. he beaker, and the supernatant acid of the cell washed out the iodine completely. A burette of 50° capacity was employed. This burette was carefully calibrated and was found to be surprisingly accurate. Its error indicates an extremely minute and regular taper of the tube and the graduations are such as to permit of accurate readings to 0°02°. In fact I have felt considerable confidence in reading it to 0°01°%. This, with the strength of solution employed, was equivalent to about 0:-1™€ of -silver. In the earlier determinations, when more than 50° of the thiosulphate were required, the burette was in some cases refilled, which, of course, multiplied the error of reading. In other cases, where the amount of thiosulphate was approximately known, a sufficient quantity was added to the beakers from calibrated pipettes of various size, so that the additional amount required should be less than 50°. This is, of course, less exact and impractical as well, unless the amount required is approximately known. In the later determinations a bulb burette, fig. 2, was employed. This contained 6 bulbs, each of approximately 25° capacity, connected by small tubes of about 2 to 3™™ internal diameter. At about the middle of these tubes marks were etched in such a way as to permit readings without error of parallax. The smallness of the tubes prevented filling the ‘burette from the top and to avoid the uncer- tainty of rubber connections a small side tube, also about 2 to 3", terminating in a funnel, b, was sealed on and supported by a section of cork as shown. A finger placed on @ as the liquid is poured into } regulates the flow, so that air bubbles are not carried along and the burette is filled quietly and accurately. Inclining the discharge tube, as shown, is of great advantage in preventing any of the grease from the cock soiling the interior of the burette, a very troublesome feature of the usual burette. The readings of this burette are naturally extremely accurate. It was employed in bleaching the larger part of the iodine. Experience enabled me to judge from the color about when the remaining iodine was less than that bleached by the con- tents of one bulb. If there was any difficulty in judging this, a comparison beaker of iodine solution could be employed. At any rate the 50° burette, with which the titration was com- » 8 D. A. Kreider—TIodine Titration Voltameter. pleted, was equal to two of the bulbs, and no difficulty was experienced in keeping within that limit. In case of acci- dentally overstepping the amount of thiosulphate required, a measured volume of iodine solution may be added and the titration be repeated, subsequent deduction being made for the amount of thiosulphate necessary to bleach the added iodine. The end reaction was in all cases taken as the bleaching of the iodine color, without starch. This in itself is quite deli- eate, and I have invariably been able to read it to a small frac- tion of a drop, when the beaker stood on white paper in a good indirect light. As a confirmation of the reading, 5° of starch solution was then added and produced a faint purple color. The delicacy of this end reaction was more thoroughly appreciated when, after a number of titrations had been made successively, as in the experiments of Table III, the addition of the starch produced almost precisely the same shade of color in all, despite the fact that a very small fraction of a drop of the thiosulphate produced a distinctly perceptible change in the color. TABLE III. oo KI a) ie - | Difference, Electrodes, 2s q| grms. ® S a S Se Fev Se om. |gSS| “in Se eee ee | 42 ce. ae a 5 vad ee fe —EE 2 ae 16X2-7| 2 | 6 in 75)... 80-4 (1) eee 5 Ub ae ol ae 0-058 | 18067] 5 ae pe eh 1128-49] | 9) . ° = . . (2) |} 16X27 5 | 5 BS nes 0-058 | 198-55) 0 Oops \§ 16x27] 2 | 5“ F5) (0-058. | 197-26) nema (3) eens 5 110 “901 ¢ °°) | 0017 | 1e746 7 ae eee Ae oa es) (0058 |170°67| (4)/4 16x27) 5 | 5 « 7-5/5 05 | 170°67) 0:23) 0-135 (25x60) 5 {10 “20 \ / 0-017 | 170-90 Table III is a record of a number of determinations of the constancy of this voltameter. Two or more of the cells were connected in series, and the conditions in each varied as shown. The time of run for 1 to 3 inclusive was about 45 mins., for the 4th, one hour. The amount of hydrochloric acid (1:4) was not measured. Enough of the acid was drawn in to insure the covering of the cathode when the iodide solution was drawn in, and to keep the solution acid throughout the experi- D. A. Kreider—Todine Titration Voltameter. 9 ment. The iodide was roughly weighed. No correction was made for the blank determinations, nor was special care taken to maintain exact constancy of temperature. Table IV shows the results of the only two determinations that were made by three of the iodine voltameters in series with each other and with one normal silver gravimetric voita- meter. The original readings of the-burettes for the required thiosulphate is given in the 6th column. In the 7th column is given the value corrected for the blank determinations. A number of blank determinations for the small cell, when 5 grams of lodide were used, gave, with the blank shown in (2), an average value of 0-07 of thiosulphate, which is the cor- rection applied to all of the small cells. The large cell, with 10 grams of potassium iodide, showed an average value for a blank determination of 0°21° of thiosulphate. This is three times instead of twice the value of the small cell, as would be expected were the result due to traces of iodate in the iodide. The uncertainty as to the amount of iodine liberated by iodate, or by dissolved oxygen, or by possible oxidizing impurities of the acid, make it rather more desirable to employ known weights of the iodide and known volumes of the acid and then to correct for the blank determination, than the alternative of securing absolute freedom from these extra sources of libera- tion of iodine. In the silver voltameter employed, the cathode was a plati- num bowl about 8™ in diameter and 3°5™ in depth. The anode was a silver disc, 5-8" in diameter, 0°8™™ thick, and sup- ported by three platinum wires bent over its edges. ‘This was wrapped in filter paper. The solution was made up of 20 grams of pure silver nitrate, dissolved in 106° of distilled water. The deposited silver was washed with water and allowed to stand under water over-night. Then washed again with water, finally with absolute alcohol and heated for 4 hrs. in an oven at 160°. Then allowed to cool for an hour in a desiccator before weighing. TaBLeE IV. le cs KI | Mee as ee Silver Difference. hehe mp ee) ae aS c Silver (in vac.) | poems eRe g) BS | pyrex) aa | NS pease seven ie oter,| Cems. of (Pry : Aas CG. Sree Zz Ss ; grms. silver. io | | (16x27 2 din 75 ) 0:058 152-07; 152°00| 1°63386 | 0-00150 |0:092 +16x27 5 | 5 * 85 -0°5 |0°058) 152-06) 151-99] 1°63375 || 1°68286 | 0°00189 |0-085 | 12-5 x 6-0 Se | ee j 0°017, 152-19) 151-98) 1°63364 | | 000128 0-078 | | | | | | | | / (1 6x2:°7) 2.15 7-5 0-0 10:00 0°06 | pete sce 7 |) bee XS 125 os § 0-058 156-40) 156-33) 1°68042 | | 167934 | 0°00106 |0°063 25x60! 5 110 * 2015 1 10°017' 156°55! 156°34! 1°68053 / 0°00119 !0°071 10° D. A. Kreider—Lodine Titration Voltameter. The titrations in the experiments recorded in this table were made with due regard to all possible sources of error. The solutions were brought precisely to the temperature of 20°, at which temperature the room was maintained. The burettes had been thoroughly cleaned with chromate solution, and of course, ample time was allowed for the burettes to drain to a constant reading. The results show that the iodine voltam- eter, even after the correction for the blank determinations, run uniformly higher by from 0:06 per cent to -09 per cent ; but that they agree among themselves to an order of accuracy of about 1 part in 10,000. Sloane Physical Laboratory, Yale University, June 5, 1900. = pty Gooch—Precipitates for Solution and Reprecipitation. 11 Arr. Il—TZhe Handling of Precipitates for Solution and Reprecipitation ; by F. A. Goocu. [Contributions from the Kent Chemical Laboratory of Yale Univ.—cxxxyv. ] In many processes of analytical chemistry, the preparation of substances in pure condition is brought about by precipita- tion, solution, and reprecipitation ; and sometimes this cycle of operations must be repeated. When a precipitate, gathered upon a filter, is easily acted upon by the appropriate solvent, the process of dissolving the precipitate from the filter is simple; but when the precipitate is refractory toward solvents or difficult to attack on account of its physical condition, as is the case with many gelatinous precipitates, the proper hand- ling of the precipitate involves some inconvenience and delay. In meeting such difficulties, I have found it advantageous to place within the ordinary paper filter, before filtering, a movable lining of platinum gauze upon which the precipitate rests for the most part and with which it may be removed. The simplest form of this device is easily made by cutting platinum gauze to the shape shown in the accompanying figure. In ordinary use, this piece of gauze, folded to make a cone of angle a little less than 60°, and held by pincers at the point of overlapping, is placed within this filter and allowed to fit itself closely by the natural spring of the gauze when released. Upon filters so prepared a precipitate may be collected and washed as usual; and, at the end of the operation, the cone with nearly all the precipitate may be transferred, by means of ivory-pointed pincers, to dish or beaker for suitable treatment. The small amounts of the pre- cipitate which have passed through the gauze, being somewhat protected by the gauze against the compacting action of filtra- tion and washing, are generally removable with ease from the filter by a jet of the washing-liquid. After washing, the gauze may be replaced within the same filter and serve for a second col- lection of the precipitate to be subsequently dissolved, in case double precipitation and solution are desirable. The final col- lection of the precipitate is, of course, made upon paper with- out the gauze lining, when precipitate and filter are to be ignited. This device has proved very serviceable in the handling of such precipitates as ferric hydroxide, aluminium bydroxide, and basic acetate precipitations. 12 Gooch—Precipitates for Solution and Reprecipitation. I have used also in the manipulation of such precipitates a regularly made cone of 60°, fitted with eyelets for handling ; but the simple folded cone is, on the whole, more convenient. Precipitates collected upon asbestos in the perforated crucible are frequently removable without difficulty by allowing a suitable solvent to percolate precipitate and felt; but in case the precipitate is pasty or compacted, solution in this manner may be unpleasantly slow. In such eases, it is convenient to remove the greater part of the precipitate, collected and washed in the usual manner, upon a dise of platinum foil, perforated, fitted with a wire handle, as shown in the figure and placed upon the asbestos felt before the transfer of the preci- pitate to the crucible. To make such a dise, shown in figure 2, is the work of a few moments only ; and by its use pasty precipitates, such as cuprous sulphoey- anide or the sulphides of the metals, are easily handled for solution. These simple devices so facilitate the manipulation of preci- pitates in many processes of analysis that they have seemed to be worthy of description. Ashley—Estimation of Sulphites by Iodine. 13 Art. II].— Zhe Estimation of Sulphites by Iodine ; by R. Harman ASHiEy. [Contributions from the Kent Chemical Laboratory of Yale Univ.—cxxxvi. | Votuarp’s method for the determination of sulphur dioxide and sulphites is accurate and reliable, but involves the incon- venience of making up every solution to be examined accurately to a standard volume of which portions are to be drawn from a burette and made to react with definite amounts of a stand- ardized solution of iodine. The method consists in running the unknown sulphite or sulphurous acid solution into a known amount of a standardized solution of iodine, acidified with hydrochlorie acid, to the disappearance of the iodine reaction with starch. This procedure rests upon the facts that the oxidation of sulphite is brought about in the acidified solution and that, as Bunsen showed, no more than a small proportion of hydriodic acid should be present at the point at which the bodies are made to react. The reaction for the oxidation of sulphur dioxide proceeds normally in dilute solutions according to the equation 2S0,+21,+4H,O = 4H1I+2H,80,,. In solutions too concentrated, however, the secondary reaction SO, +4HI = 2I,+2H,O+8 takes place as Volhard has shown%, and vitiates the indications. To avoid the inconvenience of the Volhard method it has been proposed by Ruppt to bring about the oxidation of sul- phites- by treatment with an excess of standardized iodine in a solution made alkaline by acid sodium carbonate, and then, after fifteen minutes, to titrate the excess of iodine by sodium thiosulphate. This procedure, however, is, as has been shown by Ruff and Jaroch{ and by the present writer,$ faulty in pr rinciple and pr actice, aud gives correct results only by a chance balancing of opposing errors. Theoretically it might be possi- ble to overcome the difficulties by treating with acid the alka- line mixture of iodine and sulphite and acid sodium carbonate * Ann, Chem. 242, 98. + Ber. Dtch. Chem. Ges. xxxv, 5694. t Ber. Dtch. Chem. Ges. xxxviii, 409. § This Journal, vol. xiv, p. 237. 14 A shley—Estimation of Sulphites by Lodine. before attempting to titrate by sodium thiosulphate the excess of iodine. In the experiments recorded in the table the following proce- dure was followed: the sulphite was treated with 1 grm. of acid sodium carbonate and an excess of standardized iodine solution. ‘The solution was then acidulated with a safe amount of hydrochlorie acid, it having been found by experiment that the presence of 10°* of 1: 4 hydrochloric acid in 125°™* of water was without effect upon the determination of iodine by sodium thiosulphate. The excess of iodine after acidification was titrated by standardized sodium thiosulphate. It will be noticed that in the experiments recorded under A of the table, - the excess of iodine used was small and in these experiments large negative errors are obtained; while in the experiments recorded under B, in which a large excess of iodine was employed, the results are better. They are best when at least twice as much iodine is added as is theoretically required to oxidize the sulphur dioxide. The length of time during which the iodine may act does not affect the results to any very marked degree. A Iodine Todine value Error. Excess Vol. at value of Iodine of Na2S.03 In terms Interms of HC! titra- SO, taken. taken. used. of Jodine. of SOz. 1:4. tion. erm. erm. erm, erm. erm. cm®, “ems 0°2197 0°3143 0°0978 —0°0032 —0:0008 (33) ea Re a 0°0965 —0°0019 —0:‘0005 Me ye es es 0°0970 —0°:0024 —0°0006 ig a 0°1535 0°1913 0:'0464 —0°0086 —0°0022 “ i ced 2 0°0467 —0'0089 -—-0°0022 a of ae ve 0:0472 —0°0094 —0°0024 he ee ie os 0'0465 —0°'0087 —0°0022 me of 0°2366 0'3194 0°0903 —0°0075 —0°0019 a ee 0°2906 0°3825 Oris? —0°0213 —0°0054 Y re 0°3825 0°4463 0:0750 —0'0112 —0.0028 ¢ 2 B 0°1143 0°3143 0°1990 +0°0010 +0°'0008 5°0>) Sie ef es 0°1982 +0:0018 +0°0004 ue My 29 eS 0°1992 +0:0008 +0:°0002 sé & ha me 0°1986 +0°0014 +0:°0003 Re Gs 0°1482 0°3187 0°1708 —0'0003 —0'0001 75 ee 0°15 76 0°3187 0°1586 +0°0025 +0°0006 us ie Rey of 0°1643 —0:0032 —0°'0008 se Be oe £6 0°1598 +0°00138 +0°'0003 “ ie ee 4 0°1606 +0°0005 +0°0001 + cs A 0°1602 +0:0009 +0°0002 ge or Ashley—Estimation of Sulphites by Iodine. 15 (B) Iodine Iodine value Error. Excess Vol at value of Iodine of Na.S.O; In terms In terms of HCl titra- SO, taken. taken. used. of Iodine. of SOx. 14: tion. erm. erm. erm. erm. erm. emea) > yeme. 0°1576 0°3187 0°1622 —0O:0011 —0°00038 FHS) oct led 219) 0°1560 0°3195 0°1660 —0 0025 —0'0006 cs ee 0°1992 _ 0°4460 0°2482 —9°0014 —0°'0003 ee ee O'1915 0°3825 0°1919 +0°:0009 —0 0002 ee a 0°2056 O37 71 0*1701 +0°0014 -+0°0003 ee sf Ss rfp 0°1697 +0°0018 +0°0004 ge - st ee 0°1707 +0°0008 +0°0002 ie rs nee’ Ss 0°1709 +0:°0006 +0°0002 $8 36 [0°2131 0°4470 0°2412 —0°0073 —0:0018] oh s 0°2354 0°3825 0°1490 —0°0019 —0:0005 %S ge 0°2597 0°4463 0°'1869 —0°'0003 —0'0001 aé rs 0°2638 0°4463 0°1847 —0°0022 —0°0005 es ee 0°2908 0°6375 0°3505 —0°'00388 —0:0009 ee $e 0°3187 0°4463 0°1326 +0°0050 +0 0012 &¢ ge 0°3395 0°6275 0°2842 +0°0038 +0:°0009 Be ce $2 oc 0°2852 +0:0028 +0°0007 ci oie a sé 0°2844 +0°00386 +0°0009 “¢ ne es ce 0'2855 +0°0025 +0:°0006 es sé Ruff and Jaroch* take the ground that in the favorable results occasionally obtained by Rupp’s process, an error due to the over-oxidation of the tetrathionate normally formed in the action of: sodium thiosulphate upon the residual iodine is apparently balanced by some oxidation of sulphur dioxide by dissolved air, the iodine in solution acting catalytically as well as directly. "The theory, however, is quite at variance with the evidence supplied in the table: for, if it were true, under no conditions could iodine in the presence of air act as a correct measure of sulphur dioxide, as it apparently does when used in a sufficiently large excess; nor does the theory of the cataly- tic action of iodine explain the fact that when a greater mass of iodine is used, under conditions otherwise similar, we get a larger oxidation of sulphur dioxide. The most obvious explanation is that at a low concentration of iodine an intermediate oxidation product may be formed and that the formation of this product may be prevented by sufti- cient concentration of the iodine. It is not unreasonable to suppose that the formation of a small amount of dithionate instead of sulphate is the occasion of the deficient expenditure of iodine noted when the concentration of this element is low, and that the dithionate is not formed appreciably when the * Loe. cit. 16 Ashley— Estimation of Sulphites by Lodine. iodine concentration is high. The dithionate once formed is but slowly attacked by iodine, and that is apparently the rea- son why long standing of the mixtures containing a small pro- portion of iodine does not result in complete oxidation of the sulphite to sulphate. From these considerations it will be seen that the secondary error of Rupp’s process may very probably be due to the formation of some dithionate from ae sulphite where the concentration of the iodine is low. The practical estimation of sulphurous acid or a soluble sulphite may, then, be accomplished with a reasonable degree of accuracy by adding to the solution of the substance, not exceeding 100° in volume and containing a gram of ‘acid sodium carbonate, at least twice as much iodine as is theoreti- cally necessary to effect oxidation, acidifying cautiously with hydrochloric acid, and determining with standard sodium thiosulphate the excess of iodine remaining in the acidified solution. The author takes this occasion to thank Prof. F. A. Gooch for much kind assistance. Talbot—New York Helderbergian Crinoids. fy Art. IV. — Revision of the New York Helderbergian Crinoids ;* by Mianon Tarror. (With Plates I-IV.) Tuts paper treats of the Crinoidea of the Helderbergian rocks of New York, and is a continuation of Dr. George H. Girty’s thesis, ““ A Revision of the Sponges and Coelenterates of the Lower Helderberg Group of New York.” In Dr. Girty’s paper, the term “ Lower Helderberg ” included the Tentaculite, or Manlius, limestone; but here “ Helderbergian,” as proposed by Clarke and Schuchert,}+ is used to include only the Coey- mans, or Lower Pentamerus; the New Scotland, or Delthyris Shaly ; and the Becraft, or Upper Pentamerus. With the exception of the work done by Wachsmuth and Springer, who probably used specimens that Hall had studied, the crinoids of the Helderbergian rocks of New York have not received much attention since Hall’s descriptions were published, in 1859. Very little subsequent collecting has been done, and for the most part the forms secured have been speci- mens of LZomocrinus scoparius and Ldriocrinus pocilliformis or simply stem fragments, the work of gathering being done in the New Scotland. _ A reopening of the old locality at Jerusalem Hill was made, however, in 1901, by Professors Beecher and Schuchert ; and a new locality was discovered at North Litchfield, both of these being in the Coeymans limestone. The majority of fossils found were crinoids, but there were also cystids in appreciable numbers and five ophiuroids representing two genera. In the fall of 1903, these collections were increased by more material collected at the same locality by Mr. C. J. Sarle; so that in the Yale University Museum there are now three collections—one from Jerusalem Hill and two from North Litchfield. The first of these consists mainly of Homocrinus scoparius, though it contains uncompressed forms of Cordylocrinus plumosus and several good specimens of Jelocrinus pachydac- tylus. In the region of Litchfield, the Coeymans limestone attains a thickness of one hundred and fifty feet and Momo- * This paper is part of a thesis presented to the Graduate Faculty of Yale University for the degree of Doctor of Philosophy, in June, 1904. The larger part of the work was done under the supervision of the late Professor Charles Emerson Beecher, for whose help and inspiration the writer wishes to make the most grateful acknowledgment. Type specimens have been studied in the Yale University Museum, the New York State Museum and the American Museum of Natural History: and the thanks of the writer are here expressed to Professor R. P. Whitfield, Dr. J. M. Clarke, Dr. E. O. Hovey and Mr. H. H. Hindshaw, for courtesies in connection with the study, and to Professor Charles Schuchert, who took up the direction of the work after Professor Beecher’s death. +Science, New Series, vo]. x, p. 876, 1899. Am. Jour. Sci.—FourtH Series, Vou. XX, No. 115.—Juty, 1905. 2 18 Talbot—New York Helderbergian Crinoids. crinus scoparius is said to range from the Manlius almost to the top of the Coeymans. Most of the specimens in the Yale collection were found about forty-six feet from the top of the section in a twelve-inech layer containing slabs rich in Homo- crinus scoparius and also specimens of Melocrinus pachydac- tylus, Anomalocystites cornutus, Lepocrinites gebharde and the ophiuroids. Cordylocrinus plumosus is abundant in the lower bed mentioned later. The collection from North Litchfield is chiefly from two horizons and is extremely rich. One of these beds is a lime- stone four inches thick in which are specimens of M/elocrinus — nobulissimus with very large crowns and very stout, long stems and a large form of Cordylocrinus plumosus in comparative abundance, the majority of the individuals showing many long cirri crowding around the calyx. The material from this zone has one specimen of Lepocrinites gebhardi and several of Hlomocrinus scoparvus. Although all the fossils in this bed are ot large size, especially is this true of Melocrinus nobilis- simus, whose columns are very thick and, though only frag- ments, measure from fifty to seventy centimeters in length. This is long for Paleozoic crinoids. Wachsmuth and Springer state that no columns over three feet in length have been seen from the Paleozoic and that generally they are not over one foot long.* Here there are numbers over two feet in length. The other horizon, a few inches higher in the section, has furnished slabs covering a floor space of some sixty-five square feet, slabs that are literally covered with ecrinoid stems and crowns. Here, too, as in the lower bed, are stems over two feet long. The forms represented are Mariacrinus beccheri, Melocrinus nobilissemus, M. pachydactylus, Thysanocrinus arborescens and Cordylocrinus plumosus. 'To show the rela- tive abundance of these species, an enumeration of the indivi- duals on the slabs was taken and by actual count there were found, of Mariacrinus beecherz thirty-one specimens, of JZelo- crinus nobilissimus six, of M. pachydactylus one, of Thysano- crinus arborescens ten, and of Cordylocrinus plumosus eight hundred and seventy-three, making a total of nine hundred and twenty-one specimens. In addition to these are numerous crinoid columns, several gastropods and brachiopods and one cephalopod. Ona small surface of six square feet there are three hundred and twenty crinoids. The cover of this bed is also in the collection and it is esti- mated that two-thirds as many more crinoids are on its lower surface. This enumeration was made before anything was done toward developing the slabs and such preparation may * North American Crinoidea Camerata, vol. i, p. 39; Mem. Mus. Comp. Zool., Harvard College, vol. xx, Cambridge, Mass., May, 1897. Talbot—New York Helderbergian Crinoids. 19 double the number now visible; hence in this one collection, there are undoubtedly more crinoids than in all other collec- tions from New York combined. The following species, listed by Hall from the Coeymans limestone at North Litchtield, have not been recognized in the Yale material: Mariacrinus paucidactylus (probably Melo- crinus pachydactylus), M. ramosus, M. plumosus, Platycrinus parvus (probably Cordylocrinus plumosus), P. ramulosus (seems to be restricted to the Cobleskill zone of the Manlius) and P. tentaculatus. This is not to be wondered at, however, as a slight change of position, horizontally or vertically, often reveals a different fauna; and as Hall’s collections represented gatherings not only from the quarries but also from the stone walls about the town of Litchfield, the fossils undoubtedly came from different horizons and localities. In the classification, nomenclature and terminology of the erinoids, Wachsmuth and Springer have been followed and the reader is referred to their works, ‘‘ The North American Crinoidea Camerata”* and “The Revision of the Paleoeri- noidea.”’ + Order, INapuNAtTA Wachsmuth and Springer. Suborder, Fistutata, Wachsmuth and Springer. Family, Cyathocrinide Roemer. Genus, Homocrinus Hall. Homocrinus scoparius Hall. Plate III, figure 3. Homocrinus scoparius Hall, Nat. Hist. N. Y., Pal., vol. iii, 1859, p. 102, pl. 1, figs. 1-9.—Wachsmuth and Springer, Rev. Paleocr., Pt. I, 1879, p. 79; Proc. Phila. Acad. Nat. Sci., vol. xxxi, 1880, p. 302.—Bather, Kongl. Svenska Vet. Akad., Handl. xxv, 1893, p. 105. In the collection of crinoids from Jerusalem Hill, N. Y., now in the Yale University Museum, there is a considerable number of slabs showing Homocrinus scoparvus in abundance. These slabs vary in size from a few centimeters to over half a meter in length and the surfaces are virtually covered with these beautiful fossils. One slab, thirty centimeters long and twenty-three wide, has eighteen specimens, three of which are complete, that is, have the crown and the whole length of the column, including the distal end. Aside from these, there are four other stems and two (possibly three) specimens of Anoma- locystites cornutus on the same slab. On other slabs from the same horizon are Melocrinus pachydactylus, Anomalocystites cornutus, Protaster forbesi, and Dalmanites sp. (7). Many * Memoirs of the Museum of Comparative Zoology at Harvard College, vols. xx and xxi, with Atlas, Cambridge, Massachusetts, May, 1897. 1 Proceedings of the Philadelphia Academy of Natural Sciences, vols. xxxi, XXxXiil, xxxvii and xxxviii. 20 Talbot—New York Helderbergian Crinoids. of the specimens of /Zomocrinus are in almost perfect condi- tion, and where the fine cirri are visible on the stem the grace and delicacy of this species are well shown (pl. III, fig. 3). The following additions are made to Hall’s description :— Ventral sac strong, elongated, sometimes three-fourths as long as the arms, the upper part composed of vertical rows of small hexagonal plates. The upper end of the sac probably has five large plates, which are drawn out into spines, some- thing like those in Scaphiocrinus unicus. Three of these spines and traces of a fourth can be seen in one specimen, and their position shows that a fifth was probably present origi- nally. These spines are not scattered irregularly over the upper surface, as is indicated in Hall’s figure. Column long and slender, consisting of irregularly alternating larger and. smaller joints, round below and becoming obtusely angular and enlarged above. Canal small and round, Shortest column observed 4™ in length; longest, which is still incomplete, 15™ long. Very delicate cirri are preserved, but in no specimen are they found above the middle of the stem. Wherever the distal end of the column is present, there is a coil or loop, as if the stem twined around some support (pl. HI, fig. 3). No indications of the clustering of columns mentioned by Hall were seen in the Yale collection. Horizon and tecality.—Comimon in the thinly laminated or shaly layers of the Coeymans or Lower Pentamerus, at Scho- harie, Jerusalem Hill and North Litchfield. Hall reports the species from the Manlius, or Tentaculite, limestone,* but no such specimens have come under the wr iter’s observation. Cotypes (used by Wachsmuth and Springer for the revised genus) in the American Museum of Natural History, from Litchfield, N. Y. Family, Edriocrinide n. fam. In the specimens of Adriocrinus under observation, there are differences that at first seemed to have specific, 1f not generic value. There are two quite common forms—one (No. 1 and No, 2)+ the small hemispherical cups, so well known to. collectors in the Helderberg Mountains; and another (No. 3) like the preceding only that the cup has a prominent band or ring around the upper margin. There are other forms that are not so common, however; and they can be divided into two groups, or even three. One specimen (No. 4) about twice as high as the common ones has the hemispherical cup, above which and fused to which is a solid band; and above this still another band of six fused plates, twice as high as the lower * Nat; Hist. N:Y., Pals vols ain, dda, 1S8a0: + Numbers refer to those on pl. IV, figs. 1-6. Talbot— New York Helderbergian Crinoids. 21 band. Another individual (No. 5) does not show the first band, and the second is broken up by weathering into five comparatively broad plates and one narrow one. The next specimen to attract attention (No. 6) resembles the one just described only that on one side the plates succeeding the cup have the appearance of a row of three short plates, instead of one high one. It was not until these forms, seemingly so different, had been most carefully compared that any conclusion concerning them could be reached. The difficulty was due, mainly, to the fact that in most cases the suture lines are wholly obliterated ; but, with a trace of a suture here and another there, there was something on which to base an interpretation. The following solution is offered : The genus Agassizocrinus is said to be dicyclic because young specimens have infrabasals, although the latter are obliterated before maturity is attained. The question has arisen, Why may not the same be true of Hdriocrinus? By following out this idea, these seemingly distinct forms were reduced to two whose difference is simply i in the development of the basals, which in one group are inconspicuous and in the other are enlarged to form the prominent ring or band men- tioned above. The explanation of these varying specimens is as followin No. 2 and No. 5, instead of being monocylie, are dicyclic, the ‘infrabasals, which are the largest, being fused with the basals. No. 3 shows infrabasals and basals, the latter being very promi- nently developed. No.6 has-infrabasals and fractured radials, but no brachials. This conclusion has been reached by com- paring opposite sides of the same specimen. Though on one side there seems to be a short radial followed by two short brachials in each ray, the other side shows no such division ; and it is evident that the apparent brachials are due to the transverse breaking of the radials. This view is supported by the fact that the anal plate is as high as the radials and the apparent brachials combined. No. 4 shows all the plates of the calyx and furnishes’ the clue to the others. The promi- nence of the basalsis hardly a specific characteristic and these specimens are all left in the original species, 2. pocilliformis. In the Yale collection, there is one example of /. sacculus which gives faint indications of the presence of infrabasals, though none of the specimens show any thickening of the basal ring. In regard to classification, these forms certainly cannot belong with the genus Agassizocri inus in the family Astylo- oo er inidee, where Ldriocrinus was placed provisionally by Wachs- 22 Talbot—New York Helderbergian Crinoids. muth and Springer,* because there are no supplementary anal plates in the calyx, as is the case in Agassizocrinus. Bather lists the genus provisionally under the order /7exibilia,+ an order with no anal plate in the cup; but, as Hdriocrinus has such a plate, the genus cannot be so referred. The calyx structure is that of the Cyathocrinide but there are differences that prevent the reference of Hdrvocrinus to this family. The absence of a column is one of these differences and the manner in which the rays divide is another. In Cyathocrinus, which is the most representative genus of the family, the arms in branching spread out irregularly, and the joints are generally - higher than wide; while in Zdriocrinus the joints are very short, and the arms branch as do those of Ichthyocrinus, the divisions remaining in contact and curling inward. The arms, however, do not form a part of the calyx as in the last named | enus. Family description.—Calyx elongate. Base dicyclic, prob-. ably five fused plates in each order. JRadials with facets for the insertion of the brachials extending across the whole width. Arms incurved, seemingly without pinnules, divisions remain- ing in contact ; joints much wider than long. Column wanting, the attachment being by the infrabasals in the young stages ; mature forms unattached. Genus, Hdriocrinus Hall. Edriocrinus Hall. Edriocrinus Hall, Nat. Hist. N. Y., Pal., vol. iii, 1859, p. 119; 15th Rept. N. Y. St. Cab. Nat. Hist., 1862, p. 115.—Meek and Worthen, Geol. Rept. I1., vol. iii, 1868, p. 119.—Wachsmuth and Springer, Rev. Paleoer., Pt. I, 1879, p. 21, Pt. III, 1885, p. 10, and 1886, pp. 192, 265, 286; Proc. Phila. Acad. Nat. Sci., vol. xxxi, 1880, p. 244, vol. xxxvii, 1886, p. 232, and vol. xxxviii, 1887, pp. 116, 189, 210; N. Am. Cri. Cam., vol. i, 1897, pp. 59 and 145.— Zittel, Handb. d. Paleontol., I Band, 1880, p. 350.—P. H. Carpenter, Ann. Mag. Nat. Hist., May, 1883, p. 333.—Bather, Rept. Brit. Assoc. Adv. Sci. for 1898, p. 928; A Treatise on Zoology, 1900, Pt. III. The Echinoderma, p. 191. Amended generic description.—Caylx directly cemented, either throughout life or only in the young stages, the attach- ment being by the large infrabasals. The cicatrix very large in some specimens and in others obliterated, by the accumula- tion of calcareous matter on the outer surface of the calyx plates. Infrabasals large, their height being from one-half to two-thirds that of the cup as ordinarily found, completely fused so as to destroy suture lines and to make the number of plates uncertain. Basals five, height varying in proportion to that of. * Rey. Palzocr., Pt. III, p. 192, 1885, or Proc. Phila. Acad. Nat. Sci., vol. xxxviii, p. 116. +Rept. Brit. Assoc. Adv. Sci. for 1898, p. 923 ; also The Echinoderma, p. 191, 1900. Talbot—New York Helderbergian Crinoids. 28 the infrabasals, generally so fused as to show no suture lines on the outer surface, although they are often seen on the inner side. Upper margin scalloped for the attachment of the radials and the anal plate. MRadials five, large, rectangular, the upper margin excavated slightly for the attachment of the brachials and the lower curved to fit into the concave upper margin of the basals. An anal plate half as wide as the radials and a small plate above it furnish all that is known of the anal area. Ventral surface unknown. Arms known in only one species, Ey. sacculus, where they consist of very short transverse plates and bifureate several times, but show no trace of pinnules. Genotype, £. pocilliformis Hall. Edriocrinus pocilliformis Hall. Plate IV, figures 1-6. Edriocrinus pocilliformis Hall, Nat. Hist. N. Y., Pal., vol. iii, 1859, p. 121, pl. v, figs. 8-12.—Meek and Worthen, Geol. Rept. Ill., vol. iii, 1868, p. 370, pl. 7, figs. 5a and 5b.—Wachsmuth and Springer, Rev. Paleeocr., Pt. IIT, 1886, p. 266; Proc. Phila. Acad. Nat. Sci., vol. xxxviii, 1887, p. 190.—Keyes, Geol. Surv. Mo., vol. iv, 1894, p. 221, pl. xxx, fig. 7. Amended specific description.—Infrabasals present but so fused that their number is uncertain. Height from one-half to two-thirds that of the cup as ordinarily found. Basals five, completely fused with each other and with the infrabasals or distinguished from the latter as a narrow protruding band. Suture lines sometimes apparent on the interior. Upper mar-: gin scalloped for the attachment of the radials and the anal plate. Height about half that of the nfrabasals. adials five often as high as the infrabasals and basals combined, and, like them, fused to form apart of the cup. In most instances, how- ever, the suture lines between the radials are plainly discernible. As arule, the union between the radials and basals is not so strong as that of basals with infrabasals; and the cup is gener- ally broken off at the top of the basals. Since in no specimens are brachials preserved, the union of brachials with radials must have been still weaker. Anal plate as high as the radials, but only half as wide. Radials and anal gently convex, sloping i in all directions from the center of the plate. Arms and ventral disk unknown. The attachment scar is visible on a number of specimens, and in some is a short distance up on the side of the cup, rather than on the bottom. Horizon and locality.—Throughout the New Scotland lime- stone in Helderberg Mountains. Cotypes in the American Museum of Natural History. Order, Camerata Wachsmuth and Springer. Family, Thysanocrinide Wachsmuth and Springer. Genus, Thysanocrinus Hall. Thysanocrinus arborescens n. sp. Plate I, figure 2 ; text-figure 1. Although, in America, no members of this Late have been reported above the Niagara, a number of crinoids that must 24 Talbot—New York Helderbergian Crinoids. be referred to this genus is found in one of the beds of the Coeymans limestone at North Litchfield. The generic features, as given by Wachsmuth and Springer,* are well marked—the subelobose calyx, urn or bell-shaped; infrabasals five, small, barely protruding beyond the column; basals five, the posterior one truncated by a large anal plate; radials five, considerably larger than the costals; costals two ; arms ten or twenty, rather strong and biserial; pinnules long; first interbrachial large, followed by smaller ones; anal side wider, first anal plate fol- lowed by three in the next row. The specimens under examination lack the ridges which are so conspicuous in marking the rays in most of the species of Thysanocrinus ; their plates are smooth, instead of being sculptured as is generally the case in this genus, and the column is pentangular, while in most of the species it is round. The specimens resemble 7. dzlzzformis more closely than any other species, but differ from it in the pentangular column and the absence of the ridges on the radial series of plates. Not enough is known about the bifurcation of the arms in 7”. lilaformis to make comparison. Specific description.—Calyx subglobose. Surface of plates smooth. Infrabasals five, small, projecting slightly beyond the col- umn. Basals five, large, hexagonal, the es posterior one heptagonal and _ truncated Q above to receive the anal plate. NRadials ia g five, somewhat larger than the basals, pen- 6) tagonal. Oostals two, half as large as the cy Se radials, hexagonal, the second smaller than the first and supporting on its slopmg upper OLIO margins the two rows of distichals, the | lower three of which are larger than the ee succeeding ones and are embodied in the epeased = Dia. calyx. Interbrachials, two ranges of large gram of Thysanocrinus Plates followed by smaller ones. Anal arborescens showing plate large, followed by three much smaller position of the anal ones in the next row. Arms biserial. Each plates and first bifur- ; : cation of the aums . L2y. Difureates’ on the second’ costaly ama again on the fourteenth distichal. A third bifurcation occurs, seemingly only on the inner branches and at different intervals in the different arms, varying from the four- teenth to the twenty-third palmar. Pinnules found on the fifth distichal and continuing to the tips of the arms. Column pentagonal. Near the calyx, the joints alternate in size; but farther down the stem every fourth joint is larger. Ina speci- men in which the crown is 29™™ in length, the ‘column, though incomplete, is 40™ long. *N, Am. Ori. Cam., vol. i, p. 190, 1897. Talbot—New York Helderbergian Crinoids. 25 This species is associated with Jelocrinus nobilissimus, M. pachydactylus, Mariacrinus beecheri, and Cordylocrinus plu- MOSUS. Horizon and locality.— Upper third of the Coeymans lime- stone at North Litchfield. Holotype in the Yale University Museum. Family, Melocrinide Roemer. Subfamily, MJelocrinine. Genus, Mariacrinus Hall. In re-diagnosing the genera Mariacrinus and Melocrinus, Wachsmuth and Springer recognized.the fact that the arms of the former remain apart and do not form the tubular append- age which is so conspicuous in Melocrinus. The only species in the Yale collection that shows this characteristic of darza- crinus is a new species, JZ. beecheri, in which the proximal end of the ray forms a tube while the distal end is divided, the arms diverging conspicuously. The species is thus seen to hold a position intermediate between J/arzacrinus and MMelo- crinus. As the features of the former are more strongly developed, this species is referred to that genus. Genotype, JZ. plumosus Hall. Mariacrinus beecheri n. sp. Plate I, figure 3; text-figure 2. This species bears a resemblance to Melocrinus nobilissemus but differs from it in features other than the division of the rays. The auxiliary arm, instead of being comparatively incon- spicuous, as in Meloerinus, is strong and prominent and hes alongside the tube. The joints of the rays are longer than those of WT. nobilissimus, so that, although the arms Hee. are given off more ‘fr equently than in the last Ay named species, they seem to take origin at greater intervals. Asin J. nobdlissimus, the stem joints alternate in size, but they are so very thin in all parts of the stem, and especially so near the crown, that there is no difficulty in determining this form by the column alone. The column is also much larger in proportion to the size of the calyx. Specific description.—Calyx small, elongate, | Text-figure 2.— once and a half as long as wide, the increase in Tea fogs 2 oe width being very gr feat Basals wider than with a, 6 and c as long, pentagonal, not forming a projecting the last of the anal cup, but continuing the width of the column. ee De a Radials five, four heptagonal and one hexag- Hi onal. Costals two, the first hexagonal, more ‘than half as lar ge as the radials, and the second smaller, pentagonal, and support- 26 Talbot—New York Helderbergian Crinoids. ing two rows of distichals, three ineach row. The last distichal supports two rows of palmars, whose first two plates are con- nected. Above this point, the palmars separate, those on the outside of the ray forming an auxiliary arm which lies alongside the ray but is not connected withit. The inner row of palmars joins corresponding plates from the other row of distichals to form a tubular appendage which extends for a short distance only, when the divisions separate.and remain apart to the end of the ray. On the outer side ot the ray, arms arise from every fourth or fifth joint; but, on account of the length of the joints, the arms are quite far apart. The arms are biserial to the end. The first interbrachial is large, hexagonal, followed by a double row of alternating hexagonal plates. Anal inter- radius wider and ending in a short thick tube or sac, composed of numerous plates which seem to have been hexagonal orig- inally. This sac is seen in but one specimen, where the plates are very poorly preserved (text-fig. 2). Column cireular, with diameter large in proportion to the size of the calyx. Distally the jomts alternate in size, but near the calyx they are very thin and of uniform thickness. Horizon and locality.—U pper third of the Coeymans lime- stone at North Litchfield. Cotypes in the Yale University Museum. Genus, Melocrinus Goldfuss. Genotype, Mariacrinus nobilissimus Halil. Melocrinus nobilissimus (Hall). Plate IT. Mariacrinus nobilissimus Hall, Nat. Hist. N. Y., Pal., vol. iii, 1859, p. 105, pl. 2, figs. 1-5; pl. 2A, fig. 1. Melocrinus nobilissimus Wachsmuth and Springer, Rev. Palaeocr., Pt. II, 1881, p. 122; Proc. Phila. Acad. Nat. Sci., vol. xxxiii, 1882, p. 296; N. Am. Cri. Cam., vol. i, 1897, p. 295; Atlas, pl. xxiii, figs. la, 2 and 3.—Bather, A Treatise on Zoology, 1900, Pt. III. The Echinoderma, p. 161, text-fig. Ixy! : Sixteen individuals of this species have been added recently to the Yale collections ; yet, since the type specimen is so nearly perfect, very little additional knowledge has been gained from this new material. Attention, however, may be called to a few points. One specimen shows a row of three or four small plates between the auxiliary arm and the tubular appendage. These plates appear in the figures given by Wachsmuth and Springer, but no mention is made of them in the deseriptions. They seem to be interpalmars, though it is possible that they belong to the ventral disk. The domelike extension of the anal series of plates, which is also figured by Hall, is seen indis- tinctly in one specimen. One crown has a column attached, over 21° in Jength; while another column on the same slab, and to all appearances of the same species, is over 69™ long and gives no indication of proximity to either calyx or distal end. Talbot—New York Helderbergian Crinoids. 27 At North Litchfield, this species was found associated with Mariacrinus beecheri, Melocrinus pachydactylus, Cordylo- crinus plumosus, Thysanocrinus arborescens, Homocrinus scoparius, Lepocrinites gebhardi, and Dalmanites sp. The crowns are not numerous, but judging from the associated frag- ments of stems this spot must have been very favorable to the growth of Jelocrinus nobilissimus. On one slab about four- teen inches long (pl. 11), four crowns were found with columns belonging to forty-six more. The only other fossils on this slab are one Conularia and two Bryozoan fragments. Horizon and locality.—Coeymans limestone at Litchfield and North Litchfield. Cotypes in the American Museum of Natural History. \ Melocrinus pachydactylus (Conrad). Plate I, figure 1. Astrocrinites pachydactylus Conrad, Ann. Rept. Pal. N. Y., 1841, p. 34.— Mather, Geol. Rept. N. Y., 1843, p. 347; text-fig. 6 on p. 345. Mariacrinus pachydactylus Hall, Nat. Hist. N. Y., Pal., vol. iii, 1859, p. 107, pl. 3, figs. 1-4. Mariacrinus paucidactylus Hall, ibid., p. 109, pl. 3, fig. 5. Melocrinus pachydactylus Wachsmuth and Springer, Rey. Paleocr., Pt. IT, 1881, p. 122; Proc. Phila. Acad. Nat. Sci., vol. xxxiii, 1882, p. 296; N. A. Cri. Cam., vol. i, 1897, p. 296, pl. xxiii, figs. 4 and 5; pl. xxiv, figs. 4a and. 4b. Melocrinus paucidactylus Wachsmuth and Springer, Rev. Palzocr., Pt. II, 1881, p. 122; Proc. Phila. Acad. Nat. Sci., vol. xxxiii, 1882, p. 296; N. A. Cri. Cam., vol. i, 1897, p. 296. Actinocrinus polydactylus Bonny, Schenectady Reflector, 1835. Although this species heretofore has been considered a rare fossil, it is now represented in the Yale University Museum by thirteen specimens. Little additional knowledge of the calyx, however, has been gained. In all cases where the distichals can be distingushed from the other plates, their number is two, instead of three. The former number agrees with all previous figures; yet, in their description, Wachsmuth and Springer make the distichals three in number.* One of the rays, though incomplete, shows nineteen arms, which are plainly seen to be uniserial, not biserial as previously described and figured.t The actinal sideof the rays and arms shows the ambulacral groove. As to the number of brachials in the successive orders of the plates of the rays, careful exam- ination of the specimens at Yale yields results different from those reached by Wachsmuth and Springer.t Brachials of the fourth, fifth and sixth orders have seven plates, and the subsequent orders seem to alternate with six and seven to the *N. Am. Cri. Cam., vol. i, p. 296, 1897. + Nat. Hist. N. Y., Pal., vol. iii, 1859, p. 108, pl. 3, figs. 1-3 and 4a; N. Am. Cri. Cam., Atlas, pl. xxiii, figs. 4 and 5; pl. xxiv, figs. 4a and 4b, 1897. ¢ Ibid., vol. i, p. 297. 28 Talbot—New York Helderbergian Crinoids. end of the ray. In one specimen, small dome-like interpal- mars show between the auxiliary arm and the tubular append- age, occupying the same position as in JZ. nobilissimus, but differing in form. Stem joints alternate in size near the calyx, but farther down the column every fourth one is larger. One individual has a stem 19°" long, which makes a loop at the distal end about 2°5°" in diameter. Another loop not more than 1°" in diameter has two complete whorls. M. pachydactylus is found at Jerusalem Hill with Lepocri- nites gebhardi and many specimens of [omocrinus scoparius; at North Litchfield with Mariacrinus beecheri, Melocrinus nobilissimus, Thysanocrinus arborescens, and Cordylocrinus - plumosus. Wachsmuth and Springer regard J. paucidactylus and JZ. pachydactylus as synonyms, ‘but give no reasons therefor. Hall’s distinctions are the narrower calyx and the fewer and more distant arms of the former. The specimen figured on pl. I, fig. 1, is very narrow, proving the width of the calyx to be variable. The greater distance between the branches of the arms cannot, in itself, be considered a specific difference ; and there seems to be no reason for referring these narrow specimens to another species. Horizon and locality.—Near the base of the Coeymans limestone at Schoharie ;* in the upper third of the same lime- stone at Jerusalem Hill and North Litchfield. Family, Platycrinide. Genus, Cordylocrinus Angelin. Cordylocrinus plumosus (Hall). Plate II, figures 2 and 4; text- figure 3. Piatycrinus plumosus Hall, Nat. Hist. N. Y., Pal., vol. iii, 1859, PP. 113 and 148, pl. 4, figs. 1-5. Platycrinus parvus Hall, Nat. Hist. N. Y., Pal., vol. iii, 1859, p. 114, sale 4, figs. 6-9. “Cordylocrin us plumosus Wachsmuth and Springer, Rev. Palaeocr., Pt, I, 1881, p. 61; Proc. Phila. Acad. Nat. Sci., vol. xxxiii, 1882, p. 230; N. Am. Cri. Cam., vol. ii, 1897, p. 737; Atlas, pl. lxxv, fig. 20. : Cordylocrinus parvus Wachsmuth and Springer, Rev. Palaeoer., Pt. II, 1881, p. 60; Proc. Phila. Acad. Nat. Sci., vol. xxxiii, 1882, p. 234; N. Am. Cri; Cam,, vol, 1; 1897, -p. (ar Clematocrinus plumosus Jaekel, Zeit. d. deutsch. Geol. Gesell., Band xlix, 1897, Verhandl., p. 47. Clematocrinus parvus Jaekel, Zeit. d. deutsch. Geol. Gesell., Band xlix, 1897, Verhandl., p. 737. In the Yale Museum, there are many hundreds of specimens of this species; and at first glance it seemed that substantial additions could be made to the descriptions already given. Closer examination, however, revealed the fact that in only a * Nat. Hist, N: ¥., Pal, vol aii, p..109; 1859) Talbot—New York Llelderbergian Crinoids. 29 few specimens could the plates be distinguished. It also seemed that there were two species, the fossils differing so much in size, gibbosity and general appearance; but further study failed to reveal any real differences. Some of the forms have a hemispherical calyx, and arms only three or four times as long as the cup, while others have a flat cup and arms five or six times as long; and yet the plates of the calyx, the joints of the arms,.the pinnules and the cirri seem to be the same in the two varieties. In the material from North Litchfield, the lower bed has much the larger forms, all of which are compressed. The upper bed has an abundance of the smaller ones, a few of which have the ealyx gibbous, not flattened. The speci- Q mens from Jerusalem Hill are uncom- pressed and small. Wachsmuth and Springer consider C. parvus the young of C. plumosus; and it may be that it was these small, uncompressed specimens from the upper crinoid bed that Hall had under observation when he described the former species. If this assumption can be proved, it may be well toregard ea uae C. parvus as a variety of C. plumosus, of Cordylocrinus plumo- as these small forms occur ata slightly sus. a, right postero-lat- higher geological horizon. eae ee Be left postero- From a study of the specimens in the 3.01 series of oe oe Yale University Museum, the following new data may be given: In no ease does the length of the column exceed once and a quarter that of the crown, which varies from 5™™ to 32™. DyoLe Ve Pirsson: Introduction. Definition of the province. Consanguinity shown by minerals. Augite. Biotite. Hornblende. Feldspars. Absence of minerals. Consanguinity shown in textural habit. Chemical evidence of consanguinity. General law of the province. Application to the region. Geographical arrangement of magmas. Bearing on differentiation. Regional progression of types, Introduction. Tue fact that mm certain areas of the world’s surface the igneous rocks have common characteristics, which serve to ally them together and to define them from the rocks of other areas, is now well recognized by petrographers. These com- mon features are sometimes expressed in the minerals, some- times in the chemical composition of the magmas and some- times in peculiarities of texture, but usually in a union of these qualities. In some cases these features are clearly marked, in others they are but slightly developed; neverthe- less, like those indescribable characters which define a man as belonging to one nation rather than to another, they are easily recognized by the experienced eye. The formulation of this principle, that the rocks of a given region may be thus genetically related, we owe to Judd,* and it has since been elaborated and applied with fruitful results to yarious regions by Iddings,+ who developed it under the expression ‘consanguinity of .gneous rocks.” Since then the idea has been applied to various regions by other petrograph- ers ; so, forexample, Lacroix in a recent very interesting memoir on the alkalic rocks of northwest Madagascar, calls attention to the great belt of types rich in soda that stretches along the eastern coast of Africa.t Of all the various areas, however, where the consanguinity of igneous rocks has been studied and these relationships pointed out, there is probably none better known or more thoroughly investigated than that of South * Quar. Jour. Geol. Soc., 1886, vol. xlii. p. 54. fs eee of Igneous Rocks. Bull. Phil. Soc. Washington, xii, p. 128, + Roches alcaline de Prov. Petrograph. d’Ampasindava Nouv. Arch. d. Muséum, 4™¢ Ser., vols. i et v, 1902, 1903. 36 Pirsson—Petrographic Province of Central Montana. Norway, and our knowledge of this region we owe for the greatest part to the keen perception and untiring labors of Brogeer, who has given the results of his work in that fine series of monographs which have become classics in the litera- ture of petrography. The fact that the outlying mountain groups east of the main chain of the northern Rocky Mountains are composed of rocks of a special character rich in alkalies, was pointed out by Iddings* in the work already referred to, although at that time little was known about them. Since then investigations and studies in the field and in the laboratory by a number of workers have thrown a flood of light upon this region. In the Black Hills of North Dakota the work of Caswell,+ Jaggar,t Irving§ and the writer| has shown a prevalence of types: rich in alkalies with soda dominating the potash. In Montana, the most southern of the eastern outlying groups fronting the great plains, is the Crazy Mountains, some of whose interesting rocks of alkalic types are known through the researches of Wolff.4 North of this come the various groups studied by Mr. Weed and the writer; the Castle Moun- tains ;** the Little Belt Mountains ;{+ the Judith Mountains ;¢¢ the Highwood Mountains ;8§ the Bearpaw Mountains;]|| the Little Rocky Mountains 4 and lastly, on the border line between Canada and the United States, the Sweet Grass Hills,*** the last of the outliers. While some of these have been rather thor- oughly investigated, there yet remains much to be done. The few types that have been described from the Crazy Mountains by Wolff, and its mapttt showing the vast complexity of the Z Opacity spol + Microscopic Petrography of the Black Hills, 1876. U.S. Geog. and Geol. Surv., Rocky Mts. region, J. W. Powell in charge. Rep. on the Black Hills of Dakota, pp. 469-527, Washington, 1880. { Laccoliths of the Black Hills, 21st Ann. Rep. U. S. Geol. Surv., Pt. iii, pp. 163-290, 1901. § Geology of the northern Black Hills. Ann. N. Y. Acad. Sci., vol. xii, No. 9, pp. 187-840, 1899. | Phonolite Rocks from the Black Hills. This Journal, 3d Ser., vol xlvii, pp. 841-346, 1894. “| Bull. Geol. Soc. Amer., vol. iii, pp. 445-452, 1892. Bull. Harv. Mus. Comp. Zool., vol. xvi, pp. 227-238, 1893. ** Bull. No. 139, U. S. Geol. Survey, 1896. t+ 20th Ann. Rept. U. S. Geol. Surv., 1900, Pt. iii. p. 062. This Journal, od Ser., vol. 1, pp. 467-479, 1895. ae 18th sey Rept. U. S. Geol. Surv., 1898, Pt. iii, p. 437-616. § Bull. 237 U. 8. Geol. Surv., 1905. Bull. Geol. Soc. Amer., vol. vi, pp. 389 "422, 1895. This Journal, vol. ii, pp. 315-5238, 1896. || This Journal, 4th Ser., vol. i, pp. 283-801, 351-362, and vol. ii, pp. 136-148, 188-189, 1896. “|"| Jour. of Geol., vol. iv, pp. 339-428, 1896. *** This Journal, 3d Ser., vol. 1, pp. 809-3138, 1895. +++ Little Belt Mountains Folio, Montana. U.S. Geol Surv., Geol. Atlas of U. S., No. 56, 1899. Pirsson—Petrographic Province of Central Montana. 37 dikes and sheets surrounding the main stocks of granular rocks, only serve to awaken general interest as to the character and relations of these rock masses, and it is to be greatly hoped that Professor Wolff will be able to continue his studies upon this interesting material and publish his results for the benefit of petrographers and for the understanding of the revion. In the Bearpaw Mountains the researches of the writer upon the material collected during a hurried trip through them by Mr. Weed, which brought out such a variety of novel types of alkalic rocks, can only serve to demonstrate that this relatively large area must afford a fruitful field of study in the future; one whose complete investigation will add much to our knowl- edge of theoretic petrology and yield many interesting rock types. The same must in large measure be true of the Sweet Grass Hills. The material studied by the writer gave types much like those of the Judith Mountains with hints of alkalic ones accompanying them, and the appearance of some specimens forwarded to Mr. Weed would seem to indicate that rocks of tinguoid habit occur there. Adding these facts to Dr. Daw- son’s* descriptions, it would seem as if they might consist largely of laccoliths probably with accompanying sheets and dikes similar in character and in rocks to those of the Judith and Little Rocky Mountains and the Black Hills. Definition of the province. That part of this great region which has been studied by the writer, and with which he is therefore most familiar, lies in the center of Montana and embraces as its foci of igneous activity the Castle, Little Belt, Judith, Highwood, Bearpaw and Lit- tle Rocky Mountains. Since the general reader cannot be expected to be familiar with the geography of this region and the disposition of these groups, their arrangement with respect to one another and to the main chains of the Rocky Moun- tains is shown on the accompanying sketch map. It w vill there be seen that they le in a roughly oval area stretching from the northeast towards the southwest, about 150 miles long by about 100 broad, in the middle of iNaearie and shown on the map by the dotted line. It is this region which it is here pro- posed to define as the petrographic province of central Mon- tana; the consanguinity and general family relations of whose rocks it is intended to describe. This paper then may be considered as a general summation along the line just mentioned of the work of the writer on these different mountain groups, presenting the broad petro- logic features they possess In common. For the separate * Rep. Canadian Geol. Surv., 1882-4, Pt. C, pp. 16 and 45. 38 Pirsson—Petrographic Province of Central Montana. details the reader is referred to the series of memoirs upon them whose list is given upon a foregoing page. The evidences of consanguinity are to be seen in two ways, in certain mineral peculiarities and in the chemical composition of the magmas, the first being dependent upon the second in conjunction with the physical conditions attendant upon erystallization. iy | S27Z LUT AY NUS LW Mt AN wt! » HM 1772 > 4 ' OF eo ae OTT \ ANID \ at \ White Si phur rr 7 8 @ Ro SW ree AL Bros CASTLE ~ < (ut \ ; wy INS NUL NN mat | Ry iN UTS Kain TH AR TY WN &, ERS Cea MWe Wp, 1, A RANG, RUZ RA . K 771 MAW es Uy Ci NR, (UP My Wee Alte Hi aT Tt “/\ Wy YM ali “UMN yyy Wy So MADISO ‘K 7 pein KANG wyeil bi HMA Mee Map of Central Montana showing arrangement of mountain groups in petrographic province. Consanguinity shown by minerals. Augite.—One of the most marked features in regard to the mineral composition of this composite geographical rock fam- ily is to be seen in the augite. This has been already pointed out by Iddings,* but its application to this province is worthy of special mention. The augite is of a distinct green color, very rarely pleo- chroic. It varies from very pale to a deep green. brown or purplish augites are rare. They do occur in some of the lamprophyric dikes and flows but are exceptional, so that in a great preponderance of the rocks the green augite distinetly rules. Moreover this applies through the whole series from * Op. cit., p. 131. Pirsson—Petrographic Province of Central Montana. 39 the most salic to the most femic types, the depth of color usually increasing somewhat towards the ferromagnesian pole. It is commonly supposed that the purplish color of augite is due to the titanic oxide it-contains; and while this perhaps is true, it should nevertheless be pointed out that one of these green pyroxenes from the shonkinite of Square Butte, ana- lyzed by the writer, contained over a half per cent of titanic oxide. It is also to be noted that titanic oxide occurs in all of these rocks, gradually increasing with the iron towards the ferromagnesian pole, yet the rocks towards this end still have the strong green color in the pyroxene. This is especially noticeable in the shonkinites of Yogo peak in the Little Belts, in the various occurrences in the Highwoods and in the Beaver stock and elsewhere in the Bearpaws, the TiO, ranging from 0-75 to 1°50 per cent, the silica falling as low as 46 per cent in the latter case. The occurrence of this green augite through the whole series is more strongly marked in the Highwoods than elsewhere and this local peculiarity did not escape Lind- gren’s notice and he makes especial mention of it,* not only for the Highwoods but for the other groups of the region with which he was acquainted. ‘There is no notable exception to this rule in any of the Highwood rocks numbering several hundred occurrences studied by the writer, no matter how salic or temic the types may be. This green augite is a marked feature then of this petro- graphic province, and in this respect it appears to differ from many other well-marked provinces of alkalic rocks. In the exceptional cases mentioned above, the augite is pale brown, strong purplish colors not having been noted, so far as the writer can recall, in the whole province. In the salic rocks rich in alkalies, aegirite-augite appears : this is a marked feature of those of tinguoid habit; aegirite itself is rare. This seems to be due to the dominance of potash over soda, as will be shown in the discussion of the chemical peculiarities of the province. It is possible that the characters of the pyroxene, including its green color and non- pleochroism, are also due to this general chemical character of the magmas. Liotite.—Throughout the province the biotites are the brown, strongly pleochroic variety—ordinary biotite. The red- brown biotites of the theralite rocks found in the Crazy Moun- tains to the southward do not occur, nor the pale phlogopites of the rocks rich in potash of the Leucite Hills in Wyoming as described by Zirkelt and Cross.t In some exceptional * 10th Census United States, vo]. xv, p. 726, 1886. + Micro. Petrog. 40th Parallel Surv., vol. vi, p. 261. ¢ Igneous Rocks of the Leucite Hills and Pilot Butte, Wyoming; this Journal, vol. iv, 1897, p. 120. 40 Pirsson—Petrographie Province of Central Montana. cases of the lamprophyric dike rocks the biotites have darker borders, otherwise they are very uniform in all classes alike. Hornblende.—This mineral is, comparatively speaking, of limited occurrence. It is found in an aikalic type in Square Butte syenite (pulaskose), and in the trachyandesite (adamel- lose) flow on North Willow creek in the Highwoods, and in some of the porphyries composing the laccoliths in the various mountain groups and in vogesite dikes in the Castle and Little Belt Mountains; but, with these exceptions, when it occurs it is clearly uralitic after augite. In this province augite rules in the vast majority of cases and even in the quartzose rocks (quardofelic types) it appears rather than hornblende. Leldspars.—It cannot be said that there is any specially marked evidence of consanguinity to be seen in these miner- als so far as the author is able to detect. They do not present, for instance, any such remarkable features as those seen in the feldspars of the alkalic rocks of south Norway, shown in their greatest development in the phenocrysts of the rhombic porphyries. It is to be noted, however, that on account of the tendency for potash to dominate soda in the magmas, that orthoclase or soda-orthoclase is commonly the chief feldspar. Albite is of rare occurrence, even in the strongly alkali types free from plagioclase, the one instance which is an exception to this—the porphyry of Lookout Butte* in the Little Rockies —being a notable exception. On the other hand, it is an inter- esting fact that in spite of the strong pr edominance of potash feldspar in so many occurrences of all kinds, microcline may be said to be absolutely wanting in the province. It is prob- ably due to the comparatively recent and hypabyssal character of these rocks and the fact that they have not been subjected to dynamic pressures. Absence of minerals.—The characters of a petrographic province are shown as much by the absence of some minerals as by the presence of others. In this one it is shown by the rarity or absence of minerals caused by the groups of rare earths—as they have, somewhat infelicitously, been called,— that is minerals marked by the presence of zirconia, thoria, cerium, lanthanum, didymium, columbic oxide, ete., ete. Even titanite is a rather rare mineral and zircon uncommon. Expe- rience would seem to show that it is chiefly magmas rich in soda which these oxides accompany and that the potassic domi- nance in the magmas of the central Montana province tends to exclude them and to produce rocks lacking in the interest- ing minerals they give rise to. * Jour. of Geol., vol. iv, p. 422, 1896. Pirsson—Petrographic Province of Central Montana. 41 Consanguinity shown in textural habit. In some cases the consanguinity of the rock family is shown in the repetition of certain textural habits. Thus the pseudo- leucite basalts of both the Highwoods and the Bearpaws closely resemble each other and ‘both of them differ in habit from the leucitic rocks of other regions, from those of Italy for example. A most marked instance is also seen in the minettes of Highwood type (phyro-biotitic-cascadose). These occur not only in the Highwoods but thirty miles to the north- east on the Missouri River Mr. Weed collected similar rocks and they occur doubtless in the Bearpaw Mountains.* One of these from the Missouri River so exactly resembles the occurrence on Williams Creek on the south slope of the Highwoods and described in the memoir on the Highwood rocks} that hand specimens of the two cannot be distinguished from one another. So too, while each occurrence of shonkinite in the region, in the Little Belts, the Highwoods and the Bearpaw has its own peculiarities, yet taken together they form in sum total a well marked family group. In the salic, feldspathic types, on account of their simpler composition these evidences of family relationship are less distinctly marked, and yet in the porphyries composing the laccoliths in all the groups of the province, there appears to be a tendency towards the repetition of a type with a certain tex- tural habit difficult to describe but easily recognizable. It appears to be largely conditioned by a certain abundanee, size and disposition of phenocrysts. There are many wide excep- tions and variations of this, nevertheless the rule holds. Chemical evidences of consanguinity. The strongest evidences which show that the rocks of these various groups belong to a common family are to be found in comparing their chemical compositions. For this purpose a sufficient number of analyses are available for the Castle, Little Belt, and Highwood Mountains. For the Bearpaws there are enough to show the general character of the magmas, though more would be desirable. For the Judith and Little Rocky Mountains there is only one for each, but the general similarity of the types, shown by their petrographic study, is sufficient to indicate that they must agree in essential chemical characters, and as the rocks are of very similar nature and of simple types the two analyses must supplement each other fairly well. The Moccasin Mountains, which are two compound laccoliths, are practically a part of the Judith Mountains, and their rocks, * Judging from Dawson’s description (op. cit., p. 46) it seems probable that the same type also occurs in the Sweet Grass Hills. + Igneous Rocks of the Highwood Mts., Bull. No. 237, U. 8. Geol. Surv., 1904, p. 142. 42 Pirsson—Petrographic Province of Central Montana. as shown by Lindgren’s* brief description and by specimens collected by Mr. Weed, are mainly composed of feldspar por- phyries similar to those of the Judith Mountains. In all 58 analyses have been made and published of rocks of this province, 15 by H. N. Stokes, 12 by W. F. Hillebrand, and one by W. H. Melville in the laboratory of the United States Geological Survey; 16 by the author and 14 under his direction by Messrs. H. W. Foote and E. B. Hurlburt in the laboratories of the Sheffield Scientific School in New Haven. It is not Castle Mountains Magmas. 1 2 3 + 4) 6 di pop O Fvnaihiabs Sea nba ORES 74:9 65°9 61:9 56'8 46°5 45-1 42°53 Als Ogg a erie sino 13°6 16°8 17-3 18°3 10°5 15°4 12°0 Beg @ sete ce nail 1-6 2°3 1°6 4-4 2°8 3:2 HeO Se ee eer oy th 9) 1:2 2°4 5°6 1Gs 5°6 5°83’ Meu ees hee tr 1°5 16 3°6 10°6 6°5 12°4 Ca Oe series 6 2°6 3°2 5°38 9°5 8°8 12:1 ING One ce Se eee ests 4°2 A] o°2 4°3 yall 2°8 1:2 UGG iaie A ER 4:6 31 3°8 3°3 1°5 2°8 27 Little Belt Magmas. 8 9 10 iv 12 13 14 15 16 are 18 SiO.-_-. 73:1 69:7 686 64:9 61:6 62°2 544 523 48:3 484 49-0 Al,O3 2: 14°38. 15:0 16-1 10°4* ocd) dd°8.914:3 14-0. ASS Genes Fe.O3 _- 9) "A 2:25, 270 2:0. 18 88 2°84 A sie HeO Res: 3 3 A EG R28 Ba a 4k 32S On aaoges Mo Onn "2 nid oe 236 BT BO) OL 852) \ Oia aaa CaO we. DLS 21 4 8 26 eA i le OO ie Na,O_2... 34 3:4 4:4 4:2 4:3. 3:9 134 2:8 Sore cme K,0O 2... 4:9 4:4 4:9 2 3°9° 4°0) SOC) 4:2 30. nO eon meee Highwood Magmas. Ny) 20 21 22 23 24 25 26 27 28 29 30 a1 SiO. --_ 65°5 59°2 58:0 57:2 56:4 55:2 51°9 51°7 49°6 478 48:0 460 461 Al,Os --. 17°38. 18°38 17°22. 18°5 201 18:3 15°38 14:0, 14:0" AB GE tocol Fe.Os _- 7% 65 2:5 38'S 18 2 AQ A OL Bo eerie Al ee a FeO_-.-) Li 14 1:2. 11 °4:4 2:17 3:2" 03°60) od), 2a ea MeO ie HOS 4S a le8 oi 6 1:8) BO 46 62 TO OG Alen CaO__-- .-1°9 56) .8°5) 2°3) | 21 86 6:10 70) 950 2 SiR Oram at Na.O.-- 9°95 31: 84. 45:°56 4:0. 345 2:9 3°> 2 4 as Om reels K,0.-.. 936 42 101 86. V1 C4077 “7:6 obs 3:2 onl omen Little Rocky and — Bearpaw Magmas. Judith Magmas 32 By) 34 30 36 Bye 38 39 40 SiOg>=.2 2 66°2 68°3 575 52°8 51°9 50°0 46°5 68°7 57°6 Al,O3 --- 16°2 15°3 15°4 15°7 20°3 $3) itilete) 18°3 17°5 Fe.Osz -- - 2°0 iy) 49 31 3°6 3°O 76 6 3°59 HeOlee se "2 ‘8 2) 4°8 Lie 5°0 4°4 all 1°2 MoQi ee ‘8 5 1:4 5:0 "2 11°9 47 el 2 CaQ 2 322 1°3 y) 2°6 76 16 (ay 74 10 1:4 Na,0_- 6°5 55 5'd 3°6 85 24 24 4°9 5°8 1765, jaa et 0°8 56 9°4 4°8 98 5:0 87 A°7 9.2 *This Journal, 3d Ser., vol. xlv, pp. 286-297, 1893. Pirsson—Petrographic Province of Central Montana. 43 worth while, however, to give all these analyses,* for some of them from a given area, for our purposes, would be merely repetitions of one another, and in this case only one is selected to represent this variety of magma. Only the essential ele- ments are given and consequently the summations are omitted. Inexamining these tables of analyses the first thing that is evident is that in general, high silica, alumina and alkalies go together’ and are opposed to lime, iron and magnesia. ‘This is of course merely a general truth applicable to all igneous rocks and not a special character of the province. The special and most obvious feature which distinguishes this district is in the rela- tion of alkalies to one another and to silica. The potash domi- nates over the soda. General law of the province.—Definitely stated it is this; The petrographic province of central Montana is char acterized by the fact that im the most siliceous magmas the percentages of potash and soda are about equal; with decreasing silica and increasing lime, tron and magnesia, the potash relatively mmereases over the soda, wntil in the least siliceous magmas it strongly dominates. An inspection of the tables will show that there are but a few partial exceptions to this law in the 40 analyses given, and since all which are exceptions are given, the 18 omitted analyses would merely add the weight of additional figures to the truth of the law. Application to the region.—lt will be of interest now to examine this more in detail with respect to the various moun- tain groups. The Castle Mountains lie on the extreme southern border of the province; their next neighbor to the south is the Crazy Mountains group, and an examination of analyses from: that districtt shows that in the magmas soda strongly dominates the potash throughout the series. The writer has already shownt that the general Castle magma was one of a very salic charac- ter, in fact that of a granite, and that femic rocks play but a small role. Thus in the siliceous types we see the influence of the nearby Crazy Mountains’ magma; the soda here slightly dominates the potash; we are on the edge of the province and the rocks are transitional. The relation to it is seen however in the most femic type, since here potash dominates the soda. As we go northward from here into the province the Little Belt rocks came next and its characters become more evident. In percentages potash begins to rule even in the siliceous types and in the extreme femic types this peculiarity is strongly marked. Only two exceptions are noted, both of which are given and both of which are narrow dikes. Their exceptional * They may be found in the works previously cited. + Bull. U. S. Geol. Surv. 168, pp. 120-124, 1900. { Geology of the Castle Mountain Mining District, Bull. U. 8. Geol. Surv. 139, p. 138, 44. Pirsson—Petrographie Province of Central Montana. position in the Little Belt series has been already disecussed.* In the siliceous rocks the dominance of potash as one goes toward the center of the province sometimes expresses itself more strongly ; thus in the granite porphyry of Wolf Butte on the extreme northern edge of the Little Belts the relations are K,O: Na,O = 4:4:3°3, as shown in an analysis not quoted above. Next north beyond the Little Belts and near the center of the province are the Highwoods, and here its characters reach their highest development. The most siliceous type has equal percentages of soda and potash; there is only one occurrence, the syenite of Highwood peak ; in the others the potash strongly dominates, increasing towards the femic end until in missourite itis4to1. There is only one exception in this group which has been found, No. 28 in the table. It is the analcite basalt of Highwood Gap, occurring in narrow dikes. The general line through the province to the northward now bends to the east toward the Bearpaws. Between the Bearpaws and the Highwoods there is an occurrence of igneous rocks in small stocks and dikes on the Missouri river. They have not yet been throughly investigated, but the study of sections made from specimens collected by Mr. Weed shows them of femic types like those occurring in the Highwoods, They are in fact largely minettes of Highwood type, as previously mentioned, and all their characters show that they closely conform to the general law and that potash rules. Next beyond these come the Bearpaws, in which the general law of the province strictly holds. It is to be noted that towards the southern side the rocks are femic, and as we pass through the group we find on the northeast, on the edge of the province, salic. (quardofelic) types occurring in laccoliths again as in Eagle Butte, the analysis of whose rock is shown in No. 32. Again here as on the southern edge the soda rises until it slightly exceeds potash. Southeast of here, defining the edge of the province in that direction, are the Little Rockies, another laccolithie intrusion of salic ty pes. There is some variation and the rocks pass into a tinguaitic phase. Exactly of the same character are the Judith Mountains; also a boundary group on the edge of the province, of salic types running into tinguaites. There is as yet only one analysis from each of these groups, one of a granite por- phyry from the Little Rockies No. 39 and one of a tinguaite (judithose) from the Judith Mountains. Thus they supple- ment each other and the general law holds true, the potash increasing as the silica falls. It is also to be noted that in neither of these boundary groups do any femic types occur. * Petrography of the Little Belt Mts., 20th Ann. Rep. U.S. Geol. Surv., Pt. IL, a0: Pirsson—Petrographic Province of Central Montana. 45 So far as is known, the Sweet Grass Hills off to the northwest agree with the last two groups, but the absolute confirmation of this must await future exploration and study. Moreover they are somewhat outside of the general area under discussion. In this connection however there might be mentioned Big Snowy Mountain, south of the Judith Mountains. — This is evi- dently a large laccolithic uplift. and from the Judith Mountains the heavy white Carboniferous limestones dipping away from it are clearly seen. Intelligent mining prospectors, who have searched the mountain for ore deposits, have assured me that it is all limestone on top and that no porphyry is exposed. The lacco- lithic roof has evidently not yet been eroded away, but consid- ering all the facts of structure and occurrence in this province, there can scarcely be reasonable doubt that a concealed body of feldspathic por rphyry lies underneath the limestone dome. Geographical Arrangement of Magmas. From what has been already said, it is now evident that there is a rather orderly arrangement of magmas in the province. Around the outer edge they tend to be strongly siliceous, low in lime, iron and magnesia, and with the percentages of soda about equal to those of potash, and these magmas have usually marked their upward movement and intrusion by the formation of laccoliths. One can say in truth that the boundary of the province on the south, southeast, east, north and if it be extended to include the Sweet Grass Hills, on the northwest, is defined by laccoliths or groups of laccoliths of a rather definite type of magma. On the west the boundary is not yet well known and is perhaps not so clearly defined. At all events, in this direc- tion it is eventually cut off by the main ranges of the Rocky Mountains, whose magmas, so far as we know them, are of a quite different char acter, belonging in fact to the gr anite-diorite series whose surface equivalents are rhyolites, andesites and basalts. On the border line of the province thus defined femic types are rare or wanting. When they appear, however, they tend to assume the regional character and potash rises. As we approach the center of the province they become more numerous and ot larger, volume and as silica sinks the potash rises. This is shown by the occurrences of monzonite and shonkinite in the Little Belts and on its northern edge. Finally, in the central portion of the province in the Highwoods, on the Missouri river and in part of the Bearpaws the magmas are distinctly femic and rocks rich in ferro-magnesium components form the largest masses, are most numerous in occurrence and distinctly rule. There is still a recurrence of salic types, but they are of small volume and of diminished importance. 46 Pirsson—Petrographic Province of Central Montana. This arrangement of magmas over the region is of course not so well displayed as if there had been outbreaks of lava and igneous intrusions in every ten square miles of it and all con- forming to the rule, but it is believed by the writer that one who reads car efully the facts previously stated and studies the map must be struck with the disposition of the magmas about a common center as shown in the mountain groups. It does not appear as if this could be mere chance; on the contrary, it certainly seems to point to an orderly arra angement of things according to some definite cause, whether we are able to discover the latter or not. Bearing on Differentiation.—lIt will be noticed that this arrangement is contrary to what obtains in most cases of local differentiation such as those of Yogo and Bearpaw peaks and Square butte in this province, in which the border zones are femic with a concentration of the salic components towards the center. Washington* has shown that at Magnet Cove the contrary is the case, the outer zone being more salic and the inner part of strongly femic types. An instance of this also occurs in this province in the “diorite” intrusion of Castle Mountain near Blackhawk,+ which at the center is a monzonoid rock with 56 per cent of silicaand grows steadily more siliceous towards the periphery until it becomes a quartzose porphyritic type. Other examples are described by bréggert in N orway and Ramsay and Hackmann§ in Lapland. Washington in discussing these cases| is inclined to view them as results of processes of solution and crystallization in which the magma, composed of silica, alumina and allcalies, is the solvent, the others being the solutes, and the solvent being in excess tends to crystallize first at the outer margins. ‘This might explain such cases of Jocal differentiation as are seen in laccoliths like Square Butte, but it is clear that it could not be applied to whole regions. For granting for the moment that a parent body of homogeneous magma can form diverse smaller bodies by some process, it could not do so over wide areas by one of solidification ; the facts demand that the cleaved products should remain liquid though these secondary bodies, after intru- sion into place, might yield diverse products by crystallization, The writer is not inclined to believe, on the other hand, that pure molecular flow, which Becker has shown must take place with great and increasing slowness, can be a sufiicient hiGa ee Complex of Magnet Cove, Bull. Geol. Soc. Amer., vol. xi, p. 407, + Bull. 189, U. S. Geol. Surv., pp. 184 and 140, 1896. { Brogger : Zeit. f. Kryst., vol. xvi, 1890, p. 45. S Ramsay and Hackmann: Fennia, vol. xi, No. 2, 1894. | Loc. cit., p. 408. *; Chis Journal, vol. iii. p. 21, 1897. Pirsson—Petrographic Province of Central Montana. 47 modus operandi on a vast scale. But within limits and with sufficient time it may be a factor of importance, and com- bined with convection currents and forced movements on a small scale due to the passage of heated gases, and on a large one to dynamic movements of the crust, a variety of agencies may be brought into play which may be sufficient. to render a body of homogeneous magma, even if of considerable size, quite diverse in its parts. And it is to be clearly noted that this is quite independent of the question as to whether the magma shell of the molten earth (supposing there ever was such a thing) was ever homogeneous or not. ‘This is purely an academic question and may always, as at present, remain a matter of mere speculation. ‘Whether it was or not, the magmas under- lying the crust are different now in differ ent regions, and this is the basic fact with which the petrographer has to concern himself. On the other hand, it is the writer’s belief that it would be unreasonable to throw away such indications as those afforded above, that the distribution and occurrence of igneous rocks are not due to mere chance ; to deny they are governed like other things in nature by definite laws and processes ; to affirm that they are caused by mere haphazard heterogeneity of the under- lying magma, and to thus dispose of the subject by relegating it to chaos. In regard to local as distinguished from regzonal differen- tiation, we know something of the conditions and occurrences most favorable for its operation of the magmas in which it is most likely to occur and to predict some of the probable results of its operation. But in regard to the latter it does not seem to the writer quite reasonable to assume that agencies and pro- cesses that would be operative on a small scale would be nec- essarily applicable to vast bodies of magma underlying great regions. The writer has already discussed this phase of the subject in another place and it is not necessary to repeat it here.* But in the writings and speculations of many petrog- raphers a good deal of confusion on this subject appears and the differentiation of huge “magma basins”’ presumably cover- ing hundreds and even thousands of square miles, is discussed in the same terms and with appeal to the same supposed agencies as has produced visible results in a single dike, lacco- lith or other relatively small rock body. In the writer’s opinion this is wrong and will only tend to throw discredit upon what has so far been produced that is of real value. It must be confessed that at the present time so little is known and so much remains to be discovered that any attempt, from the *Tgneous Rocks of the Highwood Mountains, Bull. 237 U. S. Geol. Surv., p. 183. 48 Pirsson—Petrographic Province of Central Montana. datain hand, to solve the problem of general differentiation over wide regions must be a mere speculation. While the physical chemist therefore is attacking the problem on one side, it remains for the petrographer, on “the other, to gather data regarding the occurrences of igneous rocks and their interrela- tionships and characters over definite areas, and the present article can be considered as a contribution towards this end. The Regional Progression of Types. It is desired to call attention here to a phase of the occurrence of rock types in the province in the hope that petrographers may observe if it is of general application in different petro- graphic provinces. In lieu of a better name it may be called the regronal pro- gression of types, and the idea involved in this term is as follows : while in a given province there are certain family characters serving to bind the various rock types into one clan, yet from place to place within its limits the magmas may vary greatly from each other, and there may be, as in the Montana prov- ince, a number of centers with complexes of their own. It is so to speak that the clan is made up of a number of families each of which consists of individuals. In traversing the area from one family to another, the observer will note that certain types which are rare or sporadic in the first will become numerous or even dominating ones in the second. Long before the second family is reached its types begin to appear, and as the areais approached they are likely to become more numerous, then attain their greatest frequency and die away beyond. Thus there is an overlapping of types and the rare one of a given cen- ter of igneous rocks becomes the common one of a neighboring center. It of course depends upon the gradual change in the character of the magmas. Some instances of this which have been observed in this prov- ince are as follows. In the Castle Mountains a single sporadic case of a monchiquoid rock was observed.* Going northward into the Little Belts they begin to be more common, and the author has described them under the name of “analcite basalts ae while still farther to the northward in the Highwoods these are exceedingly common rocks. They occur for the greater part in dikes but in the Highwoods in flows also. In the midst of the Castle rocks this type appeared out of place when consid- ered only by itself, but if we consider it not as a member of the Castle family, but of the central Montana clan, its occur- rence talls into order. *Bull. 159 U. S. Geol. Surv., pp. 68 and 114, 1896. + Petrog. Little Belt Mts., 20th Ann. Rep. U.S. Geol. Surv., Pt. iii, p. 543, 1900. ~z Pirsson—Petrographic Province of Central Montana. 49 Again, in the Castle district the stock at Blackhawk described under the name of “ diorite” has its central portion developed as a monzonoid facies, as may readily be seen by reference to its description and analysis.* In the Little Belts to the north monzonite occurs in a considerable mass at Yogo Peak,t and in the Highwoodst and Bearpaws§ it is a prominent type. Or again, shonkinitic facies of rock masses are found at Yogo Peak in the Little Belts and in the stock at the head of Beaver Creek in the Bearpaws, while in the Highwoods this rock is the common type and found in numerous masses. So also rocks of tinguoid habit do not occur at all in Castle Mountain or in the Little Belts im the southern part of the province. They first begin to appear in the Highwoods in the middle part; here they are rare, only a few occurrences being noted, but in the Judith,| Little Rocky4] and Bearpaw** Moun- tains which define the northern part of the province they are very common rocks. - Other instances might be cited but these are sufficient to indicate the idea involved. It is in intrusive rocks of various kinds of occurrence, and perhaps more noticeably in dikes, that this progression of types is seen. The extrusive rocks do not oceur so generally in this province that it may be observed among then. It would be a matter of interest to know if this progression of types is a peculiarity confined to this province and occasioned by the local distribution of magmas or whether it is of more general application. The writer has observed indications of it in other places, as, for instance, in central New Hampshire, where at Red Hill and Mount Belknap centers of alkalic mag- mas occur. ‘These are indicated at long distances by sporadic dikes of bostonoid and camptonoid habits, the latter becoming very numerous at the actual centers. but outside of the cen- tral Montana area the writer has not that timate acquaintance with other broad petrographic provinces which is necessary to be able to apply this idea to them, and it must be left to others. It would seem as if south Norway and central Italy and per- haps the Bohemian Mittelgebirge, which is being so ably inves- tigated by Hibsch, might afford good examples. Sheffield Scientific School of Yale University, New Haven, Conn., December, 1904. * Bull. 139, p. 89. + Petrog. Little Belt Mts. p. 475. ¢ Bull. 237, p. 76. § This Journal, vol. i, p. 355, 1896. | 18th Ann. Rep. U. S. Geol. Surv. Pt. iii, p. 566, 1898. *| Journal Geol., vol. iv. p. 419, 1896. ** This Journal, vol. ii, p. 189, 1896. Am. Jour. Sci1.—FourtH Series, VoL. XX, No. 115.—Juxy, 1905. oa 50 T. Holm—Croomia pauciflora. Arr. VI.—Croomia pauciflora Torr. An anatomical study ; by Tuxo. Horm. (With one figure in the text drawn by the author.) Many years ago Croomia was considered a member of the Berberrvdacee and an ally of Lerberis, Caulophyllum, Diphyl- leia, Jeffersonia and Podophylium,* with the admission, how- ever, that the examination of a single seed did not disclose whether the plant was dicotyledonous or not, and that the affinity in either case remained obscure. Several years later the mistake was corrected by Gray himself and the genus referred to the Roxburghiacee.t Besides the American species there is another one in Japan: C. Japonica Mig., but these two are the only ones, known so far, of this singular genus. The monotypic Stechoneuron and the small genus Stemona Lour. (Roxburghia Banks) are with Croomza the only representatives known of. the order. MHabitually these genera are quite distinct, Stemona being a tall climber, the others low herbs; the floral structures have been carefully described by Gray, Bentham and Hooker, and Engler and Prantl. A more detailed account of the morphology of the flower as well as the anatomy of the vegetative organs of some species of Stemona has been given by Mr. Lachner-Sandoval.t In regard to Croomzia Gray poimted out some few peculiarities in the stem-structure, sufficient to prove that its systematic position would have to be sought among the Monocotyledones, but otherwise the genus has not been studied from this particu- lar point of view. Having received some fresh material from Alabama, we have examined the internal structure of the vegetative organs of Croomia, and the following notes may be considered as supplemental to those of Mr. Lachner-Sandoval for illustrating the comparative anatomy of this peculiar little order. A few remarks upon the rhizome may, also, be inserted at this place. As stated above, Croomia paucifiora is a low herb with a few green leaves and two- or three-flowered inflorescences near the apex of the single stem. ‘The rhizome represents a sym- podium; it is slender, horizontally creeping with stretched internodes and scale-like, mnembranaceous leaves. The termi- nal bud produces the aerial, flower-bearing stem surrounded at the base by three scale-like leaves, while a bud from the axil of the lowermost of these pushes out ito a horizontal, sub- terranean branch, which continues the direction and growth of * Gray, Asa: Genera flore Americe bor.-orient. ill. Vol. i, 1848, p. 90. +Same: On the genus Croomia, and its place in the natural system. (Mem. Amer. Acad. Sce., Ser. 2, vol. vi, 1859, p. 453, plate 31.) t Beitrag zur Kenntniss der Gattung Roxburghia. (Botan. Centralb., vol. 1, 1892, p. 69. T. Holm—Croomia paucifiora. 51 the mother-rhizome. Dormant buds may be observed in the axils of some of the leaves of the horizontal portion of the rhizome. The roots are white, somewhat fleshy and sparingly ramified; they develop mostly below the nodes or, sometimes, a little above these. In the Japanese species C. Japonica Mig. the habit of the plant is the same, but the flowers are single in the axils of the leaves, and the rhizome has no stretched internodes. The internal structure of the vegetative organs of our species of Croomia shows several points of interest, when compared with the allied orders, and we will begin with the roots. The roots. The secondary roots are storage-roots with no contractile power ; they are nearly glabrous and the thin-walled epidermis persists covering directly the cortical parenchyma, no hypo- derm being developed. The cortex consists of about fifteen layers of thin-walled, starch-bearing cells, constituting a rather compact tissue. There is an endodermis of exceedingly small cells with thin walls and the Casparyan spots barely visible ; it surrounds a continuous similarly thin-walled pericambium. Inside this are seven narrow rays of hadrome with one to two protohadrome vessels and about twenty, much wider, around a narrow central group of moderately thickened conjunctive tissue. The thin, lateral roots show a like structure, but the cortex contains no starch and the endodermis is large-celled with the spots plainly visible; these roots were diarchic and the position of the protohadrome vessels was like that in the mother-root, inside the pericambium. The rhizome. When we examine one of the stretched, horizontal inter- nodes we notice a smooth cuticle and an epidermis with the outer cell-walls moderately thickened. The cortex consists of about twenty-five layers, of which the peripheral two or three are collenchymatic, the others thin-walled and starch-bearing. No endodermis was to be observed, and the mestome-bundles possess only a weak support of stereome, which does not form a continuous ring around these. It is merely represented by a few, one to two, layers on the inner face of the mestome- bundles or on the sides of these; on the leptome-side this tissue is, also, poorly developed, sometimes entirely wanting. The structure of the mestome-bundles is very irregular. They are apparently arranged in one or two circles, but several of these having fused together so as to form several groups of leptome and hadrome in immediate connection with each other, their number or real position could not be ascertained. As will be seen later, the mestome-bundles of the stem above 52 T. Holm—Croomia paucifiora. eround are leptocentric, and this structure is, also, to some extent to be observed in the rhizome, but much less regularly ; the following variations were noticed. A few were collateral with the leptome and hadrome radially opposite each other and supported by stereome on both faces, the outer and the inner. Or the leptome was found to be surrounded by the vessels on the sides, and these bordering again on other groups of leptome with or without some support of stereome; in others the lep- tome constituted but one group with some vessels on the sides, while inwards it was separated from the pith by layers of stereome. Near the periphery of the cortex were observed two isolated, collateral and, in transverse section, orbicular mestome-bundles. The central portion of the rhizome is occupied by a thin- walled, starch-bearing, solid pith. A much more regular structure exists in the short, vertical internode below the uppermost of the three scale-like leaves which surround the base of the flowering stem. In this inter- node the fourteen mestome-bundles are located in an almost regular circle ; nine of these are leptocentric and much larger than the remaining five, which are collateral and orbicular in transverse section. In regard to the disposition of these two forms of mestome-bundles, there is usually a small one between each two of the larger ones. They all are surrounded by stereome and separated from each other. The stem. A like structure was observed in the iong internode of the stem above ground. This stem-portion is cylindric and solid ; the cuticle is wrinkled and covers a small-celled epidermis, which is moderately thickened, perfectly glabrous and almost destitute of chlorophyll. The cortex consists of about four- teen strata of which the peripheral three or six are collenchy- matic, and the innermost layer did not show the characteristie structure of an endodermis. A circle of thirteen mestome- bundles is imbedded in the cortex; each of these are com- pletely surrounded by two to three layers of moderately thick- walled stereome, which enters into the leptome as a separate group in the larger bundles or merely as a bridge im the smaller ones (fig. 1). ‘he mestome-pundles are all more or less oval in transverse section and contain a very large group of small-celled leptome, completely surrounded by a ring of scalariform and spiral vessels. A pith of large, thin-walled cells without starch occupies the center of the stem. By continuing our investigation to the structure of the axis of the inflorescence, we notice here the same arrangement of the tissues, but the structure is somewhat weaker. There is T. Holm—Croomia paucifiora. 53 only one peripheral layer of collenchyma and the stereome is reduced to a few isolated cells on the leptome-side instead of forming a closed ring around the bundle. The mestome- bundles themselves occur in a smaller number, from eight to nine, and are strictly collateral ; they border on a narrow cen- = rons ee: G RS =6Y, See, Sirs sine & 2 si o ) 7an§e seemless 7 > CRYSTALLIZED STANNITE A small consignment direct from the silver mines in Bolivia affords new and unique examples of the following. They have been identified by high mineralogical authority. Stannite, Splendent crystals, grouped in cavities of the massive mineral. The crystals are most perfectly developed and very rich in planes. The species has been known to science for over a century, but until recently only in the massive form, thus leaving its crystallographic form in question. Hence the present find is peculiarly welcome. ; Andorite. A silver and lead sulph-antimonide. Massive and in small but characteristic crystals of highly modified orthorhombic habit. Some- times associated with the crystallized Stannite. : RARE MINERALS Specimens are supplied by us to students and chemists for purposes of comparison and investigation ; in commercial quantities for industrial uses. SYSTEMATIC COLLECTIONS OF TYPICAL SPECITIENS Se In sets of twenty-five up to fifteen hundred specimens. Prices $5.00 upwards per set, the average price for students’ specimens being about twenty cents. We have supplied the leading institutions for thirty years, having lately completed a single order for over 60,000 specimens. Our material is the accepted standard both as to correct labeling and high quality. Free Collection Catalog, containing lists and illustrations of General Mineral Collections, Series of Ores for Prospectors, Sets of Crystals, Series illustrating Hardness and other Physical Characters of Minerals, with Price List of Laboratory Material and Individual Specimens. FOOTE MINERAL Cee Established 1876, by Dr. A. E. Foote. W. M. FOOTE, Manager. DEALERS IN MINERAL SPECIMENS AND COMMERCIAL RARE MINERALS, 1317 Arch Street, Philadelphia. mee te = a Pak! tS > fae “ i. ‘S “ee Se a Ee me eee eae ‘ THE AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES. ] Arr. X.— On the Mechanical Equivalent of the Heat Vaporization of Water; by R. H. Hoven. Tue object of this investigation is the development of a method for the determination of the mechanical equivalent of the heat of vaporization of water directly in ergs: i. e., of a method not involving the use of the calorie. This equivalent, which will be designated in what follows by L, is usually expressed in terms of the calorie varying from 536 C. to 540 C., depending on the particular calorie taken as the unit and the particular method pursued. Regnault’s is the only classic determination. He defined the calorie as the amount of heat to raise a kilogram of water from 0° to 1°, and worked out the following formula : L=606°5 —0°695¢—0:00002¢? — 0:0000003¢3 At standard pressure this gives the value 536°5, which is generally used by physicists, notably by Joly in the reduction of his determinations with the steam calorimeter. In close agreement with this value is the 536-2 of Berthelot, whose method was much less complicated. The empirical formula of Griffith :* L=596'73—0°601¢ gives the value 536°63 in terms of the calorie from 14°°5 to 15°°5 centigrade. This agreement is only apparent, and a more just value of L is obtained by following Callendar,t who estimated L from the observations of Joly and Barnes. Jolyt * Griffith, Phil. Trans. A., 1895, p. 261. + Callendar, Proc. Roy. Soc., lxvii, 1900. t Joly, Proc. Roy. Soc., xlvii, 1889. Am. Jour. Scl—FourtH Series, VoL. XX, No. 116.—Aveust, 1905. 6 82 Hough— Mechanical Equivalent of the determined the mean specific heat of water from 12° to 100° in terms of the calorie at 20° using the following relation : wL= Ws(t, —t,) observing w, W, ¢,, and ¢,, and taking Regnault’s value of L, 5365. Callendar substitutes in this relation Joly’s observa- tions of w, W,Z,, and ¢,, and Barnes’* determination of the mean specific heat of water from 12° to 100° in terms of the calorie at 20° and solves for L. This gives the value of 540-2 in terms of the calorie at 20°. Callendar prefers this value to that of Regnault and uses it in his work on the properties of steam. It is probable that even this value is low, since Barnes’t+ values for the specific heats of water from 40° to 100° are almost parallel to but much lower than those of Regnault. There is much uncertainty as to the value of L in calories. There is as yet no absolute determination of Lin ergs. It may be expressed in ergs, however, as the product of L in calories into the mechanical equivalent of heat. Using 540-2 as the most probable value of L in calories at 20° and 4:18410' as the most probable value of the mechanical equivalent at 20°, this being an average of the values due to Barnes, Rowland, Griftith, Schuster, and Moorby, gives 2°26 10" against 2°24 10" from Regnault’s value. The sources of error for this value are many. It can not be stated that the mechanical equivalent of the heat of vaporiza- tion of water is known with certainty to one per cent. In any method of calorimetry involving the use of the calorie, no greater degree of accuracy can be attained than that of the calorie itself, But the determination of the value of © involves the use of the thermometer and all the errors incident to the measurement of temperature. That these are greater and more varied than is commonly supposed, and can only be corrected for by the exercise of the greatest care and skill, is definitely shown by Rowland{ in his work on thermometry. He sees visions of careful, painstaking observers conscientiously reading with telescope and micrometer eye-piece to the thou- sandth part of a degree, unconscious of the fact that variations due to internal and external pressure, apparent friction and previous history, to say nothing of those due to the sectional calibration and the fundamental points, are many times as great as the usual errors of parallax and estimation. Griffith,s who is not so caustic though quite as vigorous, eee “ The difficulties with regard to the measurement of temperature are not peculiar to the electrical method of inves- * Brit. Assoc. Rep., 1889. + Phil. Trans. A., 1902, vol. exevii. ¢ Rowland, Proc. Am. Acad. of Arts and Sciences, vol. xv.- § Griffith, Phil. Trans., 1898. Heat Vuporization of Water. 83 tigation, and therefore I need not dwell upon them. I would, however, venture to add my expressions of astonishment to those of Rowland, that so many enquirers attach so little im- portance to this point: many investigators, whose methods have otherwise been of a high order of accuracy, having satis- fied themselves with the mercurial thermometer as a standard.” Rowland* rejects, as having no weight, previous determinations in which the thermometer readings were not reduced to the air scale. As to the dificulty of this reduction, and. to the general uncertainty of the apparent readings of the ordinary thermometer, a very instructive object lesson is to be found in an article by Cole and Durgan,t entitled “An Example in Thermometry.” It is the record of a systematic calibration of a Gerhardt thermometer, made in a concise and thorough manner. The mere statement of the corrections made, the record of the observations, and the results of the calculations, stated as briefly as consistency with clearness would permit, occupy twenty pages. In his determination of the mechanical equiva- lent of heat, Rowland made the most involved and elaborate corrections on his thermometer readings, and only brought his results and those of Joule into agreement by making the same kind of corrections for the latter’s thermometers. Without _ raising the question of the soundness of such corrections, it is evident that a method for the determination of any fundamen- tal heat constant independent of them is desirable if only to serve as a check: for the only way to minimize their effect is to extend the range of temperature, which is sure to increase the errors due to radiation, conduction and the calorimeter con- stant. The error due to the latter constant need not be large, provided only that the water equivalent of the calorimeter be small compared with that of the substance under observation, and this can usually be accomplished without much difficulty. The error due to the water vapor in the steam is only present in methods of steam calorimetry and is almost entirely eliminated by the differential method. The errors due to radiation, convection and conduction are more serious. Reynoldst remarks on Joule’s determination “that notwithstanding the greater facilities enjoyed by subse- quent observers owing to the progress of physical appliances, the inherent difficulties remained: the losses from conduction and radiation could only be minimized by restricting the range 5 of temperature and this ensured thermometric difficulties, par- * Rowland, Proc. Am. Acad., vol. xv. + Cole and Durgan, Phys. Review, vol; iv, 1896. ¢ Reynolds, Phil. Mag., 1897. 84 Hough—Mechanical Equivalent of the ticularly with the air thermometer which does not admit of very close reading.” In fact, so uncertain are the corrections for radiation and conduction, that Griffith* asserts as “the general principle on which he proposed to work, that of elimi- nating the effects of radiation, conduction, ete:, rather than that of ascertaining the actual loss or gain due to such causes.” He eliminates these effects by maintaining the walls of the chamber enclosing his calorimeter at a constant temperature and gradually raising the temperature of the calorimeter from some point below to some point above that of the jacket, such that the gain and loss by the calorimeter are equal. This he calls the null point and determines it experimentally. The correction for convection by this method is doubtful. Row- land, who also bunches the losses due to radiation, convection and conduction, estimates the loss by convection to be more than 75 per cent of the total losses from these causes. He likewise corrects empirically. Obviously a better plan would be to eliminate not only the effects but the cause of these errors by maintaining the calorimeter, the jacket and the inter- vening medium at the same constant temperature, if a method admitting of such a process is possible. In fact, the principle of elimination of source of error is fundamental to all physical measurements since minimization and correction formule can never be more than a series of successive approximations. The grounds then for a new method of determining L are: (a) the absence of any authoritative determination ; (6) the absence of any absolute method; (c) the inherent sources of error in the present indirect methods. These are suffi- cient but there are weightier considerations: i. e. the advan- tages resulting from the use of L as the primary heat unit. Much can be said in favor of L instead of C as the primary heat unit, especially since the development of steam calori- metry by Bunsen and Joly. The substance under calorimetric observation may be in a thermo-dynamic or in a thermo-static condition. The tempera- ture may be changing, or it may be constant. In the first case the thermometer should be accurate, delicate and sensitive. That is to say, that not only should all corrections to reduce its readings to the air scale be definitely known, but it should respond to small variations of temperature in a readable degree and respond quickly. In this case the readings must be taken rapidly and are necessarily limited in number. In the second case the thermometer should be accurate and delicate but not necessarily sensitive to a higher degree. The readings may be taken more leisurely, with greater precision, and are only * Griffith, Phil. Trans., 1883. Heat Vaporization of Water. 85 limited in number by expediency. It is apparent that methods necessitating observations of the first class are, other things being equal, inferior to those involving readings of the second class only. The determinations of Regnault, Joule, Rowland, Moorby, Griffith and Barnes all involve observations of the first class. In Joly’s method of steam calorimetry, however, the temperature readings are made while the substance is in a state of thermal equilibrium which may be maintained almost indefinitely.* In this respect his method is unsurpassed. An absolute determination of L substituted for the Regnault value used by Joly would enhance the value of his work many fold. His differential method is unquestionably the best general method of calorimetry yet devised, the use of an uncertain constant being, as Joly himself pointed out, its weakest point. Barnes’ curve for the heat capacity of water from 0° to 100° will never be changed much except, perhaps, by shifting the origin along the axis of specific heats. Rowland determined only a small portion of this curve, which from 10° to 20° is practically parallel to Barnes’ but lower in value. Regnault determined the portion of the curve between 40° and 100°. It is also practically parallel to Barnes’ but much _ higher. This indicates the presence of constant errors—but where ? ‘ In this particular work the men are to be given almost equal weight. A constant error in Rowland’s work, whose results agree among themselves most perfectly but for which he only claims an accuracy of two parts in a thousand, is hard to locate. It is possibly due to the sensitiveness of his thermometers not being great enough for observations on a substance in a ther- modynamie condition. Regnault’s constant error is likely due to several causes, including radiation, while Barnes’ is possibly due to the position of his thermometers, as this is a source of error common to all continuous methods and very hard to elim- inate or to correct for. In some preliminary work on the ratio of L to C, an attempt was made to develop a continuous method of steam calorimetry. It was abandoned for a time at least because the results, while agreeing very well among them- selves, were fouud to be a function of the position of the ther- mometers placed in the ingoing and outcoming water. It would not be safe then to decide which of these curves, agree- - ing so well in all but their positions, is nearest to the true one. The substitution of the true value of the heat of vaporization of water in Joly’s determination of the mean specitic heat of water from 12° to 100° in terms of the calorie at 20°, would give a value by which Barnes’ curve could be shifted. In this way much of the work of previous investigators in calorimetry * Joly, Proc. Roy. Soc., vol. xlvii, 1889.. 86 Hough— Mechanical Equivalent of the would be enhanced in value. Hence an absolute determina- tion of the mechanical equivalent of the heat of vaporization of water is a thing to be desired in itself. The present method aims at the elimination of the errors due to thermometry, the calorimeter constant, the water vapor in the steam, radiation and convection, and a rigorous correc- dh tion for conduction. The devices used to attain these ends will be described in detail followed by a discussion of the prin- ciples involved. The general plan of the machine and the relation of its parts is best shown by the photograph (fig. 1) and the conventional diagrams (figs. 2 and 3). It consists essentially of (a) a vertical feat Vaporization of Water. 87 shaft to which power is supplied: (6) a friction brake of pecu- liar design to convert the mechanical energy into heat: (¢) a controlling device to maintain a convenient constant load: (d) 2 B Vp, pros LLL Ye “WY PB WIN Wi Via SO ASS Ys § SS YZ SSS SSS SAN W SY ESS SS S RS SSS SS wi IN" /; SS OES q O00 e[ SSAA NW ASSASS RS N = AGES = Gr: LEON ieee Ee = : po) 88 Hough—Mechanical Equivalent of the a cup suspended from the arm of a balance to hold the water to be evaporated: (e) two bent levers to balance the friction against gravity: (7) a recording device to plot the variation of the mechanical force with the number of revolutions: (g) a counter to register the number of revolutions: (A) a clutch for throwing the recorder and the counter in or out of gear at will: (z) a double-walled jacket: (7) a shield between the cup and the jacket to prevent radiation: (£) a steam supply to furnish the steam bath. The vertical shaft consists of a hollow steel tube turned and fitted to accurately bored brass boxings. Power is communi- cated to this through the bevel gearing at the top from the horizontal shaft which is driven by a motor. To this horizon- tal shaft is geared, through the clutch by which they are operated, the recorder and the counter. The counter is a Veeder and gives excellent service. The recorder consists simply of a horizontal drum whose angular velocity isa linear function of that of the brake. This carries the paper verti- cally under the marker. The marker is moved horizontally by means of the famil- iar device for parallel motion imvented by Watt. The parts of this registering apparatus are very light and accurately centered on hardened steel cone bearings. They communicate directly with one of the bent levers so that the position of the marker at any instant is a linear function of the mechanical force. The marker is RSS SASSAS SL SASS SSNS SS SS 1s) KSASAASSSSSSSSSS SSS ¥ g Z Z Z Z Z Z g Z Y Z Z y Z g Z Z Z Z g Z g g y g Y Z Z 4 Wy Tea c notin continuous contact with the paper, Ate YA but only for an instant at regular inter- iH j tt vals when struck by a bar which is actu- il Y i Db ated by a cam geared directly to the AL j Ht —e drum. The bent levers consist of accu- | LY Y it ed, Dee pee with aes Ay | i réie>s—-ei4__r dened steel knife edges bearing on har- eZ. Y eh g dened steel surfaces. Attichedie these WV Y} 4 pulleys are pendulum bars and _ bobs. Hf Y Y} D> i Flexible steel tapes with swivel joints HY j q Y; i transmit the moment from the disk of i j 5 Yj Ht H the friction brake to the pulleys of the We ] a HV bent levers. The friction brake consists HH j = ] i of a bobbin threaded to the vertical shaft 05: K and rotating with it. Two rubbers, Saw by EE) quadrant sections of a turned steel tube, N N N N N N Heat Vaporization of Water. 89 are hinged to the bobbin by means of a double joint which permits radial motion. A toggle joint, operated by a rod inside the vertical shaft, connects the two rubbers diametrically through an opening in the bobbin. This rod is foreed upwards by means of a strong spiral spring in the bottom of the bobbin, and draws the rubbers in toward the center at the same time. When pressed down by the controlling device at the top, it forces the rubbers out radially. Surrounding the rubbers and accurately turned to fit them, is a cylinder supporting a torsion disk at the top. As the shaft rotates the rubbers move with it, and on account of the friction drag the cylinder and the tor- sion disk at the top with them. This motion is communicated by the tapes to the pulleys of the bent levers and the pendu- lums are displaced until their moment is equal to that of the friction. The double hinged joints are the important feature of this device. They permit the rubbers to seat themselves perfectly in the cylinder and the resulting friction is very uni- form. In fact, the small periods of its variations are so short compared to that of the long bent levers that they are com- pletely integrated by these levers, the record being almost a straight line. The controlling device consists of a hand screw to force down the rod operating the toggle joint. This pres- sure is transmitted through the ball-bearing since the rod is rotating with the shaft. The manipulation of the machine is quite simple. A steam bath is allowed to flow through the chamber from the boilers throughout the experiment, maintaining all parts inside the shield at the temperature of the bath. The motor is started and the load is gradually increased by the control to the desired constant. When the water is evaporating freely and the ther- mal conditions have been maintained constant for some time, the weights are adjusted a little ight and, as the water in the cup evaporates and the pointer comes to the zero, the clutch is operated throwing the counter and the recorder in gear, the weights and the counter having been observed and recorded. After any convenient period the weights are again adjusted a little ight and the clutch again operated just as the pointer comes to zero, the counter and the weights being observed and recorded. The calculation of the mass of the water evaporated is made in the usual way, but that of the mechanical energy may need a word of explanation. Since the ordinates and the abscissas of this curve are linear functions of the friction and the num- ber of the revolutions of the rubbers, the following relations hold: W =F ds “0 90 Hough— Mechanical Equivalent of the =Of y.de 0 ce OH but Nt IX ay = O,8y where ¥ = average ordinate S = 2.27r where n = no. of revolutions 7 = radius of disk W, Y W, where w, = total weight of paper w, = weight of A ¢ =width of paper Cag te where g = acc. due to gravity G = mass in grams to displace marker 1°™, We SG. Daa Ww, The constant G is determined by empirical calibration, for which four steps are necessary: the calibration of the seale of one of the pendulums in grams per centimeter by suspending weights in a pan from the pulley: the adjustment of the mass of the other pendulum bob to the same grams per centimeter : the adjustment of either tape until any deflection of the disk gives the same displacement on both scales: the calibration of the marker in terms of these scales. This determines G. A second set of observations is taken using a second cylinder of different conductivity capacity from that of the first and L is determined from the following relation : W=u,+4,+R where wv, = heat in ergs to the rubbers wu, = heat in ergs to water R = heat in ergs radiated if the temperature of the shield approaches that of the cylinder Rh. 0 and W=u,4+4, but u,=(m+m')L where m = apparent loss of water in the cup m'= water deposited on cup due to water vapor in steam and if m =O u,=mL WwW = u,+mL. Heat Vaporization of Water. 91 Let W =u, +u, for the first cylinder W'=w',+w’, for the second cylinder and t= W wim Vs u, W—-mL Pe Nl: But —du, = ts a (T, —T, )dt Di a (T, — T, )d Ui _ AA i/ du’, 5 ee ee i is du = a A pe — T’)dt where & = specific conductivity A = cross section of con- ductor £ =length of conductor T = temperature } == time integrating t fe ak (r= ja : 0 t yes ae (T, —T, at ; th ee as (T’, — T’,)d¢ me = f (T!, —T ae and ae Jat oe (TT de he, Ay uy (T, = T, at HA, Se. _— 1" )dt 92 Hough—fleat Vaporization of Water, ete. where R= ratio of the conductivity capacities of the two cylinders. aes e b ae hee NR pe Nh See m W—mL mR a Wa : L— W i 1 i a ) m ley Th 70 m’ In the second term of the right hand member the two fac- tors always have opposite signs. The correction is therefore a negative quantity. By reducing the conductivity capacity of the rubbers and increasing that of the cylinder, this correc- tion is reduced. to a minimum. Only the ratio of the conduc- tivity capacities is demanded by this formula, not the specific conductivities. This ratio is determined by the method of cooling. : The advantages of this method are: the measurement of all quantities involved to a high degree of accuracy, depending only on the skill of the mechanician: the elimination of all errors due to thermometry, the calorimeter constant, the water vapor in the steam, radiation and convection: the minimiza- tion and rigorous correction for conduction. Preliminary tests of the most rigorous type show that all the factors that enter into this result are entirely within control. A long series of observations are to be made during the coming year, from which it is confidently expected that a value accurate to at least one part in a thousand will be obtained. University of Pennsylvania. Baskerville and Lockhart—Zine Sulphide. 93 Art. XIl.—The Phosphorescence of Zine Sulphide through the Influence of Condensed Gases obtuined by Heating Rare-Harth Minerals ; by CHaRLes Baskervit1e and L. B. LockHART. Hetium has been shown to be a product of the disintegra- tion of radium emanations; it is also obtained from minerals which contain thorium and uranium. It has been shown by Afanassiew, Mme. Curie, Crookes, Strutt, Hoffman, Basker- ville, and Boltwood that minerals containing these elements are radio-active. It seemed to be of interest to ignite these minerals and con- dense the gases given off and note their effect upon phospho- rescent zine sulphide. The method of procedure was essen- tially that described in the preceding paper, except that the pulverized mineral was placed in the closed tube of hard glass instead of aradium preparation. Screens of Sidot’s. blende were prepared in strips for the purpose. The glowing of the screen was assumed to indicate the condensation of the emana- tion. - No final conclusion could be drawn from the experiments, which were distinctly qualitative. It appeared, however, that those minerals which offer the richest sources of helium gave the greatest amount of emanation. Most of the minerals were obtained by purchase, but we are indebted to Dr. Geo. F. Kunz for some of them, to Dr. H. 8. Miner, of the Welsbach Lighting Co., for others, and to the Nernst Lamp Co., Pitts- burg, Pa., for still others. In addition to the minerals we made some experiments with uranium compounds*, commercial thorium oxide, and the frac- tions of that element obtained in our laboratory. The list of minerals, and observations follow: Mineral, Locality. Result. Aeschynite Hitteré, Norway Fair glow Allanite (orthite) | Amherst Co., Virginia No glow Allanite Amherst Co., Virginia No glow Annerédite Norway No glow Auerlite Henderson, N. C. Fair glow Bastnisite Manitou Springs, Col. No glow Brookite Arkansas ~ No glow Carnotite La Salle Creek, Mont. Co., Colorado Fine glow Carnotite Utah No glow Catapleiite Brevig, Norway No glow Cerite Bastnis, Sweden Fair glow Cleveite Moss, Norway Fine glow Columbite Amelia Co., Virginia No glow Crytolite Bluffton, Texas Faint glow with Fergusonite (Llano) * For which we are indebted to Dr. S. A. Tucker, Columbia University. 94 Baskerville and Lockhart—Zine Sulphide. Mineral. Crytolite Euxenite Euxenite Fergusonite Gadolinite Gummite Hjelmite Monazite sand Monazite Monazite sand Mixite Orangite Norway Orthite Arendal, Norway Pitchblende (Uraninite) 7 Pechurane Bohemia Samarskite Mitchell Co., N. C. Steenstrupine Urals Thorite Langesund, Norway Thorite (Orangite) Brevig, Norway Thorogummite Bluffton, Llano Co., Tex. Tritomite Brevig, Norway ‘Tyrite Tromsé, Norway Uraninite North Carolina Uraninite Joachimsthal, Bohemia Uranite (rare) Uranophane Spruce Pine, Mitchell Co., N. C. Xenotime Hitteré, Norway Yttro-tantalite Ytterby, Sweden Zeunerite S. B., Germany Substance. Prepared by Result. Uranium Tucker No glow carbide Uranium oxide Tucker No glow Uranium nitrate Purchased No glow Thorium—-X Miner. From 100 gals. wash-water Thorium oxide Same as for - Fair glow : Welsbach burners Berzelium* Irwin. Monazite No glow oxide sand Thorium* Davis. Monazite Fair glow oxide sand ! Carolinium* Skinner. Monazite No glow oxide sand Locality. S. Co., Texas Spangercid, Norway Arendal, Norway Ytterby, Sweden Fahlun, Sweden Mitchell Co., N. C. Karapfvet Brazil, 8. A. Norway Mitchell Co., N. C. Joachimsthal, Bohemia Fair glow Result. No glow Fair glow © Good glow Fine glow No glow No glow No glow Fair glow Medium Very faint glow No glow Medium No glow Medium glow (below fair) Strong glow Fine glow No glow Fair glow Fair glow No glow No glow No glow Fine glow > Very faint glow No glow Fair glow Faint No glow No glow Remarks. Not expected from our knowl- edge of the activ- ity of uranium. Slight glow with tiffanyites. No glow with solid willemite. No glow with tiffan yites. Slight glow with tiffanyites. *These preparations were made in our laboratory, University of North Carolina. the electrical and photographic methods. All showed some but not the same radio-activity when tested by » Baskerville and Lockhart—Radium Emanations. 95 Arr. XII.— The Action of Radium Emanations on Min- erals and Gems ;* by Cartes BaskervitLe and L. B. LockH ART. Kunz and Baskervillet have made some interesting observa- tions concerning the action on minerals and gems of radium compounds of the highest activity enclosed within glass, as well as of mixtures of weaker preparations with a limited num- ber of minerals, especially diamonds, willemite and kunzite.t{ Rutherford used willemite most satisfactorily$ for demonstra- ting to a large audience the condensation of the emanations by means of liquid air. It was thought advisable to subject other minerals, found by Kunz and Baskerville to be fluorescent or phosphorescent, or both fluorescent and phosphorescent under the influence of ultra-violet light, to similar treatment. We wish to express our obligations to Dr. Geo. F. Kunz, who gen- erously provided us with most of the minerals, all of which were authenticated. The method of testing was as follows: About 0°25 gram of radium chloride, 7000 ‘uranies| strong, was placed in a hard glass tube 2" in diameter, sealed at one'end. This was bound to a glass tube, provided with a stop-cock, which was bent so as to reach through one of the two holes in a rubber stopper to the bottom of a test-tube 2™ wide and 15™ deep. Through the other hole was passed a bent tube so that it just projected below the rubber. This tube was provided with a glass stop- cock and connected with an ordinary vacuum pump. The material upon which the action of the emanation was to be determined was placed in the wide test-tube. The tube was then dipped into liquid air contained in a suitable unsilvered Dewar bulb. On opening the cock next to the pump while it was in operation a good vacuum was produced in the container tube. When this cock was closed, the radium chloride was heated to low redness. The cock between this and the test-tube was opened. ‘The emanations were swept in and condensed. In every case the tube and contents were allowed to remain in the liquid air until they were assumed to have obtained an uniform temperature. All experiments were carried out in the dark and observations were made only after the eves had become accustomed to the conditions. * Read before the Washington Section of the American Chemical Society, April 6th, 1904. + Science. t Patent applied for. § Address before the American Association for the Advancement of Sci- ence, St. Louis, Mo., Meeting, Dec., 1903. | By an ‘‘uranie” is meant the radio-activity of metallic uranium, which is taken as the standard. 96 Baskerville and Lockhart—Radium Emanations. It was learned that tiffanyite diamonds are quite as sensitive to the action of the emanations as the phosphorescent zine sulphide. We did not have enough diamonds to change for each experiment, so in each trial a strip of Sidot’s blende screen was inserted. This served to show that the emanations had been condensed. We were much surprised to learn on frequent repetition of the experiment that kunzite, which is so responsive to radium, neither fluoresced nor phosphoresced when the emanations were condensed thereon. fore, responsive to the beta- and gamma-rays only. Before giving the results of the observations, which follow in tabulated form, it may be well to relate the results obtained in several experiments, the bearing of which upon the question In mind is apparent. It is, there- The cooling of zine sulphide to the temperature of liquid air does not cause it to glow, with or without vacuum. A good vacuum and a sudden releasing of the same does not cause zine sulphide to glow. but warming it to ordinary temperature after removal from liquid air does cause it to glow brilliantly. Chlorophane and kunzite cooled in liquid air show no phos- phorescence. : Action of emanations from radium chloride (7000 activity) on : Mineral. Locality. W ollastonite Harrisville, Lewis Corn eN yea White wollaston-|Morelos, Mexico ite(withidocrase and pink garnet) Wollastonite Franklin Tunnel, Pectolite Havers, N. Y. Pectolite Paterson, N. J. Pectolite Guttenburg, N. J. Spodumene Paris, Me. Spodumene (hid- denite) Spodumene Spodumene (kunz- ite) Willemite Greenockite Hyalite Colemanite Chlorophane Tiffanyite Alexander Co., N. C U.G., Brazil, S. A. Pala, Cal. Franklin, N. J. Yellowstone Park Mono Lake Amherst, Va. 5 Dutch diamonds, QV k. With Ultra-Violet Result. Remarks. Tent Slow to phosphor-/Tribo-luminescent|Phosphorescent esce, faint Glows brilliantly |Loses glow in less|Phosphorescent than five min- utes Glows brilliantly |Loses glow quick-|Phosphorescent ly Nothing Phosphorescent Nothing Phosphorescent Nothing Phosphorescent Nothing Nothing Nothing Nothing Nothing Nothing Very slight re- Phosphorescent sponse Glows well Not so sensitive as|Fluorescent and Glows strongly Nothing Nothing Nothing Glows very easily and. brilliantly | zine sulphide, or tiffanyite ; glows with emanations from commercial thorium oxide Goes away quickly Lastsseveral hours phosphorescent. Fluorescent Fluorescent _|Phosphorescent Phosphorescent Phosphorescence prolonged J. L. Kreider— Behavior of Typical Hydrous Bromides. 97 Art. XIII.—Zhe Behavior of Typical Hydrous Bromides when Heated in an Atmosphere of Hydrogen Bromide ; by J. Lenn Kreiper. {Contributions from the Kent Chemical Laboratory of Yale Univ.—cxxxvii. | In former papers from this laboratory* the results obtaimed in the dehydration of certain hydrous chlorides in air and in an atmosphere oi hydrogen chloride have been studied and compared. In the present paper the effects of treating typical hydrous bromides in air and in an atmosphere of hydrogen bromide are described. | Hydrous barium bromide has been taken as a type of hydrous salts which when heated in air lose their water with- out much further decomposition; hydrous magnesium bromide as typical of salts which lose part of their water without much further decomposition and the remainder with simultaneous evolution of hydrogen bromide; and hydrous aluminum bro- - mide as typical of salts which lose their water ony with simul- taneous loss of hydrogen bromide. The method of experimentation was very sinter to that followed by Gooch and McClenahan+ in their experiments with hydrous chlorides. For these experiments two combustion tubes 30™ in length and 2°" in diameter, set horizontally side by side in a tubulated parafiine bath, served as heating chambers. Each tube was fitted with a thermometer. Portions of the hydrous bromides to be treated were weighed into porcelain boats. One of these boats was inserted in each tube about midway in the bath (heated to a regulated temperature) and below the bulb of the thermometer, so that the temperature to which the material in the boat was submitted might be indicated by the thermometer as accurately as possible. Through one tube was drawn slowly a current of air purified by sulphuric acid, and through the other was sent a slow current of purified hydrogen bromide, generated in a flask by the action of bro- mine on a heated solution of naphthalene and kerosene, and passed through a purifying apparatus consisting of a tower containing successive layers of red phosphorus and glass wool and a wash bottle charged with a saturated solution of hydro- bromic acid. At the end of a definite period, the boat was withdrawn, placed in a desiccator for a suitable interval to cool, and weighed. The residue in the boat was dissolved in water, and the bromine was precipitated by silver nitrate, the silver bromide being weighed on asbestos. In this way it was *Gooch and McClenahan, this Journal [4], xvii, 365. McClenahan, this Journal [4], xviii, 104. t Loe. cit. Am. Jour. Scl.—FouRTH SERIES, VOL. XX, No. 116.—AtGUST, 1905. 7 98 J. L. Kreider—Behavior of Typical Hydrous Bromides. possible to determine the loss of water and hydrogen bromide from separate portions of the salt under examination, during definite intervals and at fixed temperatures, both in an atmos- phere of hydrogen bromide and in air, and to find for each portion under examination what proportion of the total loss was hydrogen bromide and what was water. The tabular statements and the diagrains show the course of decomposition of the various salts for the temperatures indicated. Hydrous Barium Bromide. For the experiments with hydrous barium bromide, a well crystallized specimen was prepared by taking commercially pure barium carbonate, dissolving it in hydrochloric aeid, precipitating by ammonium carbonate, washing the precipitate, dissolving in hydrobromie acid,-and crystallizing and drying the crystals by pressing between filter papers. The analysis of different portions of this salt showed a definite composi- tion, corresponding very closely to theory. Found. Theory. Pais ea ee 41°69% 41°60% 10S eat ch Ana MS ioe SS 48°05 47°95 DEO) yi coe ate aoe 10°26 10°45 100:00% 100°00% The progress of the decomposition of this salt in air and in hydrogen bromide when submitted for a half hour to the tem- peratures indicated is shown in the accompanying table and diagram. | Here may be noted a gradual loss of water from 70° C. to 160° C., at which point the water is entirely expelled, without an appreciable loss of hydrogen bromide, either in air or hydrogen bromide, and that hydrogen bromide influences the process of dehydration in no marked way. ‘There is nothing to show that any part of the water sustains a peculiar relation to the salt. Hydrous Magnesium Bronide. Similar experiments were performed with hydrous magne- sium bromide, prepared by dissolving magnesium ribbon in hydrobromic acid and crystallizing the salt over sulphuric acid. The analysis of the salt gave a definite constitution correspond- ing fairly to theory. Found. Theory. Nonny bye a SNe 08°68 08.27 ES eee as eae Mea Re 54°61 54°69 GE Oe Ce mags 36°71 37°04 100°00 100°00 J.L. Kreider—Behavior of Typical Hydrous Bromides. 99 Dehydration of Hydrous Barium Bromide. BaBr.* 2H.0. eee na Bromine in residue. ae Time. oo ees fea ere iP 10n er . BE oa Las orm. sas from eae hrs ture. theory. : HBr} -2377 |-0165|06-94| -1130/47°56|—0°39 | 06°55 : 0° Air | °2364 |:0107/04°52| -1141/48°24/+0°29 | 04°81 2 : 2} HBr} -2309 |:0147|06°36| -1105/47°82|—0°13 | 06°23 y 80°C Air | °2247 |-0125/05°56| °1082/48°16| +0-21 | 05°77 2 : ( HBr| -2299 |-0121/05-26] -1115|48-49/+0°54| 05°80 é 90°C ] Air | -2311 |-0127|05-49] -1115|48-26/+0°31| 05°80 | 2 4} HBr}| -2413 |-0121/05-01| -1173/48°58|+0°63 | 05°64 4e aa Air | -2416 |:0134/05°54) -1167|48°33|+0°38 | 05°92 2 : 2 HBr} -2399 |-0118/04°91| -1157/48°26) +0°31 | 05°22 iS lineG Air | °2342 |-0128)05-50; °1162/48°09| +0°14) 05°64 2 : : HBr| :2296 |:0115/05-:00) -1111/48°41/+0°46 | 05°46 1 |190°C Air | °2287 |:0127/05°50) -1095/47°91|—0°04 | 05°46 2 5 ; 7 HBr| -2501 |-0157|06-27| -1208|/48°32| + 0°37 | 06°64 Asses Air | °2472 |-0135/05°46] -1195|48°37|+0°42 | 05°88 2 ; #2 8 | HBr} -2389 |-0198|06°58/ -1147/48°03/ +0-08 | 06°66 th eners Air | -2304 |-0148/06-42/ -1113/48°31/+0°36 | 06°78 2 P : HBr} -2491 |-0258/10°36| °1190/47°78) +0°17 | 10°19 tee Air | :2438 |-0251/10°29| -1167/47°86|—0:09 | 10°20 DA le ag 2 i HBr 2460 |-0269|10-93| -1190/48°37| +0°42 | 11°35 y heote ) Air | -2416 |-0242/10-01| -1166/48-26] +0°31 | 10°32 2 : Ba Bra. 2hr 9. ° © e gio £0 90 yoo WO sao 190 140 /50 160° 100 J. L. Kreider Dehydration of Hydrous Magnesium Bromide. MgBr,° 6H.O. Behavior of Typical Hydrous Bromides. From these results Weight Loss on Bromine in | HBr |Water ee Atmos- | taken. | heating. residue. lost. | lost. "| Temper- ae grm. | grm. cea a peat bent ae ES — , | HBr) -2389 | 0000; 00-00) °1310) 54-81) 00-12) 00°12), ie | Air | +2400 | 0000} 00-00] 1319) 54:98] 00-29] 00-29; 2 ; § HBr| +1370 | °0000! 00°00) -0751| 54°85) 00:16] 00-16), 80° “(| Air | °13880 | 0012} 00°86] 0752) 54:55] 00:14| 00-72) 2 3 ) HBr) 1259 | 0019) 01°50) 0693) 54-96] 00°27) 01-77), 90° | Air | *1482 | 0023) 01:95] 0813) 54°87) 00°18} 02-13) 2 4) HBr) 1220 | 0053) 04°34) -0664| 54:46) 00°28) 04-11; 4 1002 | Air | +1448 | -0048) 03-31] 0797] 55-07] 00-38) 03-69} 2 5 ) HBr} 1345 | 0070) 05-20/ 0736) 54-76] 00°07) 05°27, Tae | Air | *1384 | 0103] 07:44) -0752| 54-33] 00-36] 07-80) 2 6} HBr} -1374 | -0115] 08°36] -0751| 54°69) 00-00] 08°36) 120° Air | *1853 | 0120) 08°86] -0737| 54°50) 00-19] 08:67} 2 , § HBr) -1331 | 0114) 08°56) -0724/ 54-41/ 00-28) 08°28), |, 400 | Air | °1817 | -0128] 09-71] -0714| 54-28) 00-41) 09-30) 2 s | HBr) -1345 | 0140) 10°40) -0784) 54-58) 00°11) 10-29) , | 1,50 Air | -1369 1/0183) 13:37) -0733| 53269) 01-11) 10-36) 2 g § HBr} 1375 | 0216) 15°71) 0748) 54°47) 00°22/ 15°49) [1.00 | Air | 11358 | 0251) 18°48) -0727| 53°58] 01°12) 17°36) 2 10| HBr] +1313 | 0140) 10°66) 0714) 54-46/ 00-23) 10:46, 4 |, goo Air | °1311 | 0824] 24-71] 0698) 53°26] 01:44) 23-27, 2 11 § HBr) +1366 | 0170) 12°44) -0744/ 54-39) 00-31) 12°13) 4 [400 | Air | °1379 | -0159| 11°53| 0744) 54-00] 00-69] 10-84) 2 yg) HBr} +1399 | -0220) 15°72) 0760) 54:35) 00°34) 15°38) | ape (Aur | 21858 )/50232)17 08/0723) 58.27) Oul-43 a5 -Giaimeee 19 ) HBr) 1285 | -0217) 16°89) 0696 54-75) 00-06) 16:95), | 990 | Air | *1324 | :0275| 20°77) 0702) 53-06] 01°65] 19°12)? 14) HBr} -1345 | 0284) 21-11) 0731) 54:42) 00°27) 20°84, |oop0 | Aim} <1382 | -0312) 22-57))-0730) 52287| 0184) 20-7) gee 15 ) HBr] 1349 | 0282) 20:90/ 0731) 54:19/ 00°50) 20°40, J5 4 46 | Air | 13850 | -0331) 24°51) 0704) 52°20] 02°51) 22-00) 2 1g) UBr) °1837 | 0297 22-21) 0722) 54-01) 00°68 21°53) 1 |oag0 | Air | °1820 | -0379) 28°71] -0686) 52-03| 02°69] 26-02) 2 17) HBr] -1354 | -0340) 25°11) 0731) 54-02| 00-67) 24-47, |oane “) Air | -1373 | -0606| 44-13] -0649| 47-43] 07-35] 36°78}. 2 |" 1g ) HBr] 1376 | -0401) 29°14) 0740) 53°69) 01-01) 28°13] 4 [a4 50 | Air | °1360 | :0555| 40°80} -0665| 48°93] 05°80| 35°00) 7? |~ it appears that approximately a third of the water may be removed from the hydrous magnesium bro- mide, submitted at once to the temperatures indicated, either in air or in an atmosphere of hydrogen bromide, without consid- erable simultaneous loss of hydrogen bromide from the salt, the trifling loss being somewhat less in the atmosphere of J. L. Kreider—Behavior of Typical Hydrous Bromides. 101 hydrogen bromide than in air, Thereafter the loss of hydro- gen bromide when the salt is heated in air increases generally with the temperature and is inhibited, as is the loss of water, 9 ~ eA i bose fave, : é Jers a Prornide Ong Pan. L442? a7 J Sosa pov ay - in Sens it I. 4 eS rs 7 0° Zo" 90° wo? 0° s20° 130° 140° 75:0" 160°17 08° 90° 90" gab 2 10H 02 Zb2KO” CL by the atmosphere of hydrogen bromide. It appears that about a third of the water of magnesium bromide bears a relation to the salt different from that of the remainder. 102 J. L. Kreider— Behavior of Typical Hydrous Bromides. When submitted at once, without preliminary heating, to a temperature of 170° in air and 160° in hydrogen bromide, the hydrous salt melts and in the melted condition loses water less rapidly than the solid salt at a somewhat lower temperature. This is what makes the break in the curves which indicate the losses of water and hydrogen bromide. When the salt was heated successively, for intervals of a half hour, at tempera- tures varying by ten degrees, the progress of dehydration was more uniform, as is shown in the accompanying diagram, all the water being lost at 160° in air and 220° in hydrogen bro- mide, the inhibiting action of hydrogen bromide upon the dehydration being more marked as the temperature rises from the point at which the first third is lost. | 3 3¢- i HBT aoe. - . — : . 7A 60 FO s0O NO 129° 440140 4F0 160 120° 180 °/90" 209° 240°220 6 Hydrous Aluminum Bromide. The hydrous aluminum bromide used was prepared by dis- solving pure aluminum chloride in water, precipitated alumi- num hydroxide by ammonium hydroxide, filtering off the aluminum hydroxide, and washing until free from impurities. This precipitate was then dissolved in hydrobromic acid, and J. L. Kreider— Behavior of Typical Hydrous Bromides. 103 the solution thus formed allowed to crystallize by evaporation in yacuum over sulphuric acid: the crystals thus formed were of nearly normal constitution. Found. Theory. (SAORI Ota ta 07:254% 07:20% Wipe PRA ss. ee 63.90 63°95 Ge eek «hen 28°85 28°85 100-00 100-00 The coarse of dehydration of hydrous aluminum bromide in air and in an atmosphere of hydrogen bromide is shown in the accompanying table and diagram. Dehydration of Aluminum Bromide. AI1Br;° 6H20. Weight Loss on Bromine in | HBr | Water Time Atmos- | taken. heating. residue. lost. | lost. Temper- SES ee ee Melee 1306 | 00000000) 6-2 eS. Peli Se 2S ‘ 0 1) Air eee ooo ooo | et 10" Peete 1394.) 0000) 00/00) 9) 2: We. pecs) eal), 30° mere ots ST) 0000) 00-00) 22.2) | lek |e pect) 1344-0000) 00°00) - es) 37-2) 2 =2-y 90° { Air | °13860° | -0000/ 00°00) ----| Be les he Wee Sono 7 7 HBr} +1317 | 0008) 00°60) 0831 63°13) 00°83| 00:23), | jo 90 Air | 1316 | -0014| 01:06! 0842) 64:05! 00°10} 01-16! ? 5 ) HBr) +1364 | -0008| 00-58) 0861) 63°60) 00-35) 00°23) 4 | 1490 | Air | -1378 | °0024| 01-74] -0865| 62°83| 01:13] 00°61; ? g ) HBr) 1381 | -0008| 00°57) 0878 63°62) 00-33| 00°24) 4 | Ja90 | Air | +1367 | -0054) 03°95) 0834 61:07] 02°91) 01:04) * | ,) HBr) 1297 | -0006/ 00°46) 0825 63°64) 00°31/ 00°15) 4 | 1599 | Air | 1318 | :0106| 08-04| -0767| 58°24] 05°78] 02°26| .? g ) HBr +1379 | -0011) 00°79) 0831 60°82| 03°67 02°89) | 4440 | Air | -1357 | -0127| 09°35) 0771) 56-85| 07-18] 02-17; 2 g ) HBr, +1366 | 0113) 08-27) -0730 53°46) 10°62) 02°35) 150° | Air | 1355 | :0549) 40°51) °0485| 35°84| 28-46| 12°05)? 10 | HBr| -1308 | -0121) 09:25] 0740) 56°57| 07°43) 01°82 4 | 160° Air | °1348 | :0668| 49°55| 0413) 30°66] 33:71| 15°84| * 11) HBr -1580 -0270| 19°56| -0703) 50-94| 13°17| 06°39], 10° | Air | -1390 | 0919) 66-11] -0281) 20°72] 43°77| 22°34| ? 19) HBr| -1378 | -0561| 40°71] °0477| 34°65| 29°67| 11°04) | 180° ( Air | -1346 | -0949/ 70°50} 0232) 17°28] 47-26] 23°24] ? 13 ) HBr! 1331 | 0850) 63-86) 0280) 21-09) 43-40/ 20°46 | J 990 | Air | -1307 | 0976! 74°67| -°0193| 14°80] 49°72| 24°95, * 14) HBr} -1362 | 0938) 68-87| .0232) 17-08| 47-46 21:41 + | 200° ( Air | 13877 | °1090| 79°15} -0197| 14:33] 50°24| 28°91 15 § HBr) -1355 | °0958) 70°70; 0246 18°04) 46-49 24:21, 4 | 41 G0 | Air | -1345 | -1040| 77°32| -0158! 11-76 Q4-A7\ 52°85 104 J. L. Hreider—Behavior of Typical Hydrous Bromides. 4 52 BEF},3. 6420 Ja ar, v ba WA aoe 2 ete 4) A. e o 100° 770° yao’ 739° Jyo° 180° 160° 470° 480° 90° 200° 9 jo ~HWEG CN PO From these results it appears that, at 100° and higher tem- peratures, hydrous aluminum bromide loses water and hydro- gen bromide simultaneously, both in air and in an atmosphere of hydrogen bromide; but that the loss of water, as well as of hydrogen bromide, from the salt is retarded by the atmos- phere of hydrogen bromide: At the highest temperature recorded, 210° C. the salt still retained bromine. ‘There is nothing to indicate that any part of the water possesses a dif- ferent relation to the salt from that possessed by any other — part of the water. Discussion of Results. In correlating the phenomena noted, Cushman’s hypothesis of inner and outer linkages of water relative to the molecular complex, upon the assumption of quadrivalent oxygen, seems applicable. =e J. L. Kreider— Behavior of Typical Hydrous Bromides. 105 Thus the symbol r= 0G Ba if Fe NBr=0C 93 for hydrous barium bromide, showing two molecules of water externally attached, suggests the observed easy removal of all water without simultaneous loss of hydrogen bromide, and indicates, as was observed, that concentration of hydrogen bromide in the system is not likely to affect the course of dehydration. The symbol Ho i Ho 0 Be oct ee eee HL ii \b = 6 Br =0¢4 Heer for hydrous magnesium bromide, in which two molecules of water are externally attached and four internally, shows why one-third of the water may be removed at a moderate temper- ature, without much loss of hydrogen bromide; why the remaining two-thirds of the water require a higher tempera- ture for their removal with simultaneous evolution of hydro- gen bromide; and why increase in the concentration of hydrogen br omide in thé system retards both the loss of water and hydrogen bromide, after the first third of the water has been expelled. The symbol | ee as | Sa Boe eae ae ca | : erp lee elles eae nestor ha | eg Sk = a bors. rg we Se ee 106 J. L. Krevder— Behavior of Typical Hydrous Bromides. for hydrous aluminum bromide suggests the observed impossi- bility of evolving water without simultaneous loss of hydro- gen bromide, the salt tending on continued heating to go over to the oxide. With asalt showing this constitution the natural effect of the concentration of hydrogen bromide in the system would be to retard the dehydration of the salt, as was observed. So it appears that the phenomena of dehydration of the hydrous bromides under discussion admit of explanation upon Cushman’s hypothesis of the molecular attachment of water within and without the complex. The author is greatly indebted to Prof. F. A. Gooch for advice and assistance throughout this work. E.. T. Mellor— Glacial Conglomerate of South Africa. 107 Arr. XIV.—The Glacial (Dwyka) Conglomerate of South Africa; by Epwarp T. Mettor. [Communicated by permission of the Director of the Geological Survey of _the Transvaal. | Introductory.—Few rocks have aroused so widespread and so sustained an interest as the Glacial Conglomerate occurring at the base of the Karroo System of South Africa, generally known as the Dwyka Conglomerate. From the time when attention was first directed to it by Bain in 1856, down to the present, the Dwyka Conglomerate has continued to be a source of almost continual discussion. In the first instance, this interest was in a great measure due to the very different views held by various geologists as to the nature of the conglomerate, and especially to the opposition offered by many to the theory of its glacial origin—a question which one may venture to regard as finally settled by the accumulation of evidence in recent years. This establishment of the glacial character of the deposits included under the term Dwyka Conglomerate, which occur over thousands of square miles in South Africa, and which correspond closely with similar formations of corresponding age in India, Aus- tralia, and South America, lends a newer and perhaps more widely spread interest to the study of this series, and of the conditions under which it was formed. To the South African geologist the rock derives additional interest from the fact that it affords the only geological horizon common to the various colonies yet established with any degree of certainty. Nomenclature.—The term “ Glacial Conglomerate ”’ was used by E. J. Dunn on his map published in 1873* for the northern outcrops of the conglomerate, while he still retained for the more southerly occurrences an old name, “ Trap Conglomerate,” used by Wyley. In the second edition of his map,+ two years later, while retaining the term Glacial Conglomerate for the northern outcrops, Dunn applied the term “ Dwyka Conglom- erate” to those of the southern parts of Cape Colony and Natal. The term Dwyka is derived from a river of that name in Cape Colony in the neighborhood of which the conglom- erate is typically developed. The name is now frequently applied to the glacial conglomerate at the base of the Karroo System generally throughout South Africa. It might perhaps be more appropriately restricted to the southern type, which, as will be pointed out, differs in some important respects from the more northerly occurrences, especially as the intermediate *E. J. Dunn, Geological Sketch Map of Cape Colony, London, 1873. + E. J. Dunn, Geological Sketch Map of South Africa, London, 1875. 108 £. 7. Mellor—Glacial Conglomerate of South Africa. phases have not yet been fully worked out. For the northern conglomerates the original term ‘‘ Glacial Conglomerate ” is as appropriate as ever, and I have preferred it in various descrip- tions of these conglomerates in the Transvaal. Lnistribution of the Conglomerates.—The main area occupied by the Karroo System covers a large part of South Africa, including the major portions of Cape Colony and Natal, nearly the whole of the Orange River Colony and most of the south- eastern Transvaal. This area would be included roughly between lines drawn from a point on the south-east coast of Cape Colony, near to the mouth of the Gualana River, W. to near the head of the Doorn River beyond Matjesfontein, NNE. to the Lange Berg on the southern border of Namaqualand, NW. by Prieska and Kimberley to Middelburg and Belfast in the Transvaal, SSE. by Amsterdam to Vryheid, and SSW. to the coast of Cape Colony at the mouth of St. Johns River. This area includes most of the higher portions of South Africa, and almost the whole of it lies above a level of 3000 feet. In the Drakensberg the uppermost portions of the Karroo System attain an elevation of over 8000 feet. The series of glacial deposits at the base of the system crop out almost continuously around the margin of the vast area occupied by it, following approximately the lines given above. Along their southern margin the Karroo rocks, particularly the Dwyka Conglomerate, have been affected by the intense folding characteristic of the southern portions of Cape Colony. There the lowest Karroo Beds are frequently highly inclined, and their outcrop is correspondingly reduced in width, but over the whole of the remainder of the area occupied by them the Karroo rocks are practically horizontal, and the outcrops of the various divisions occupy broad tracts of country. This is especially the case with the Glacial Conglomerate and lower portions of the system, which in many places form extensive outliers around the margin of the main area as above defined. Lelationships and Age.—In its southern portions in Cape Colony the Dwyka Conglomerate grades downwards into a series of greenish shales (Lower Dwyka Shales) some 700 feet in thickness, which in turn lie comformably upon the quartzites of the Witteberg Series. These, together with the Bokkeveld Beds. and the Table Mountain Series, constitute the ‘*‘ Cape System” of Cape Colony. Passing northwards, the Dwyka Series overlaps the lower divisions of the Cape System, which thin out in that direction, and comes to lie unconformably upon various much older systems of rocks. In all the more northerly localities where the conglomerate has been studied, it lies uncomformably on the older South African rocks, the surfaces of which are fre- E, T. Mellor—Glacial Conglomerate of South Africa. 109 quently glaciated. In Cape Colony and Natal the Dwyka Conglomerate passes upwards into the Ecca Shales, a series of shales and mudstones identical in character with the shales occurring with the Dwyka Conglomerate, and in composition corresponding with the finer portions of the matrix of that rock. The Ecca shales are succeeded by a very extensive series of sandstones and shales, attaining a maximum thickness of some thousands of feet, and including on at least two dif- ferent horizons seams of coal. These, together with the Ecca Shales and Dwyka Conglomerate, constitute the Karroo System of South Africa. Intrusive sheets of diabase occur throughout the Karroo rocks, and the uppermost portion of the system consists, for the most part, of a succession of lava-flows usually amy odaloidal and of basaltic composition interbedded with sandstones containing much fragmental material of volcanic origin. The Bokkeveld Beds of Cape Colony have yielded a numer- ous assemblage of fossils related to the Devonian fauna of Europe; the Witteberg beds which succeed them and underlie the Dwyka Series have so far afforded only a few imperfect ~ specimens showing general Carboniferous affinities. With the Keca Shales in Cape Colony and with the beds associated with the Coal Seams of the Transvaal, which sometimes, as at Vereeniging, lie immediately above the Glacial Conglomerate, a fossil flora is associated of Permo-Carboniferous age* having a number of genera in common with the lower part of the Indian Gondwana System, and the Coal Measures of New South Wales. Compared with the southern and eastern margins of the Karroo area, the northern outcrops of the Glacial Conglomerate and associated beds show a considerable diminution in thick- ness, a feature shown also by the other divisions of the Karroo System. In the southern outcrops in Cape Colony the Dwyka Conglomerate has a thickness of about 1000 feet; on the north of the Colony, in the neighbourhood of Prieska, it is stated not to exceed 500 feet.t In Natal and the eastern Transvaal the thickness of the conglomerate is about 300 feet,t while on the northern border of the formation, in the central portions of the Transvaal, it rarely reaches 100 feet, and may be locally absent altogether. As will be seen from the descr iptions given below, this difference in thickness corresponds with differences in com- position, and in general characters dependent upon variations in the original conditions of deposition in the different localities. * A.C. BO Notes on the Plant Remains from Vereeniging, Q. J. G.S., vol. liv, pp. 92-93. London, 1898. tA. W. Rogers, The Geology of Cape Colony, London, 1905. ¢G. A. F. Molengraaff, Geology of the Transvaal, Edinburgh, 1904, p. 73. 110 #. 7. Mellor— Glacial Conglomerate of South Africa. Description of the Dwyka Uonglomerate in the Southern Outcrops.—The earlier. studies and descriptions of the Dwyka conglomerate were confined to its occurrence in the southern portions of Cape Colony and in Natal. In the southern examples especially the conglomerate has certain characteristics which led to much controversy as to its origin. Its appear- ance in the Dwyka locality was thus described by Mr. E. J. Dunn:* “The conglomerate consists of a bluish grey base so fine that its constituents are not resolvable, except under high magnifying power, and then no crystals are disclosed; it appears to be a very fine indurated mud; in this base are enclosed bowlders, pebbles, angular fragments, and grains of a great variety of rocks, such as granite, granulite, gneiss, mica, and other schists, quartz rock, hard sandstone, jasper, hornfels, quartz, small pieces of felspar, ete.” The included fragments, which range in size from mere grains to bowlders several feet in diameter, are distributed in the matrix without definite arrangement. The rock as a whole is very hard and fractures pass indifferently through matrix and bowlders alike. By weathering it frequently produces a yellowish clay, through which the hard rock fragments and bowlders of the original conglomerate are scattered. Besides the conglomerate beds, other shaly beds occur devoid of included fragmeits. Individual beds persist over long dis- tances, maintaining at the same time their distinctive litho- logical characters. The conglomerate beds vary from a few inches to hundreds of feet in thickness. In the southern parts of Cape Colony the conglomerate often shows a_ schistose structure resulting from the earth movements which have affected that area—to the effects of which is probably also due in part the extreme hardness of the southern rock as compared with its northern representative. Various theories concerning the Origin of the Dwyka Con- glomerate.—The dark green color of the conglomerate, its rich- ness in minerals not usually abundant in rocks of sedimentary origin, including much chloritic material, its extreme hardness, its crystalline appearance and the frequent absence of bedding through great thicknesses of rock, disposed almost every observer, including many geologists of wide experience, to attribute to the conglomerate an igneous origin. Expressive of these views are the following names applied to the rock at various times by different workers: ‘‘ Claystone-porphyry,” “ Trap-conglomerate,” ‘‘ Melaphyre-breccia,” ‘‘ Volcanic-brec- cia,” “ Trap-breccia.” Many and various were the theories advanced at different times and by different observers to *E. J. Dunn, Report on the Camdeboo and Nieuwveldt Coal, p. 7, Cape Town, 1879. EL. T. Mellor—Glacial Conglomerate of South Africa. 111 account for the peculiar characters of the conglomerate. A. G. Bain, “The Father of South African Geology,’ who first described the Dwyka Conglomerate in 1856, suggested that it represented a flow from an immense volcano. Prof. A. H. Green thought it to be a “coarse shingle formed along a reced- ing coast-line,’ while from Green’s specimens Sir A. Geikie and Dr. F. H. Hatch considered it had the aspect of a volcanic breccia. The majority of South African geologists favored the igneous theory, accounting for its peculiar characters and occasional stratification by referring its origin to submarine volcanoes. A glacial origin was first attributed to the conglomerate in 1868 in a paper on the Geology of Natal by Dr. P. C. Suther- land,* who had previously regarded the rock as a lava-flow. Sutherland, who was familiar with the conglomerate in Natal, where the rock has more the features of a terrestrial glacial deposit, and rests in places upon striated rock surfaces, clearly stated the reai character of the rock. The glacial view received early support from Stow,t who, however, referred the glaciation to a much later period, and subsequently from Schenck.t Dunn, who did so much to work out the main fea- tures of the distribution of the Glacial Conglomerate as shown in the various editions of his “Geological Sketch Map of South Africa,” regarded the rock as largely due to the action of floating ice, an agent which no doubt had much to do with the southern deposits. It is only quite recently, however, that owing to the accumu- lation of evidence§ from various localities in South Africa the glacial origin of the Dwyka Conglomerate has received anything approaching general acceptance. Lecent Studies of the Glacial Conglomerate——In 1898 Dr. Molengraaff| published a description of the Dwyka Conglom- erate, and overlying Ecca beds, as developed in the Vryheid district of the Transvaal, to the north of the Natal border (now included in the latter colony). In the Vryheid: district the Dwyka Conglomerate averages about 300 feet in thickness, and lies unconformably upon an old land surface composed mainly of the hard quartzites and shales of the Barberton formation— the surfaces of which are frequently polished and striated. Both the conglomerate and succeeding Ecca Shales offer good * P. C. Sutherland, On the Geology of Natal, Pietermaritzburg, 1868. +G. W. Stow, On some Points in S. A. Geology, Q. J. G. S., vol. xxvii, pp. 497-548. London, 1871. tA. Schenck, Die Geologische Entwickelung Siidafrikas, Pet. Mitt., Band xxxiv. Gotha, 1888. $See recent reports of the Geological Surveys of Cape Colony, Natal and the Transvaal. |G. A. F, Molengraaff, The Glacial Origin of the Dwyka Conglomerate, Trans. Geol. Soc. S. A., vol. iv, 1898. 112 #. 7. Mellor—Glacial Conglomerate of South Africa. opportunities for study in the many sections exposed in the deeply cut valleys of the eastern rivers: In this district the Dwyka Conglomerate includes both unstratified and stratified - portions, in each of which facetted and striated bowlders are abundant, together with many angular and sub-angular rock fragments. The stratified beds are sometimes almost devoid of bowlders and pebbles, and include mudstone and shales, the latter indistinguishable from the overlying Ecca Shales into which the Dwyka Conglomerate gradually passes. In 1899 Messrs. Rogers and Schwartz* studied the Glacial Conglomerate in the Prieska district in the north of Cape Col- ony. They found the Conglomerate here to present all the features of a true ground moraine, with abundance of facetted and striated bowlders; and lying unconformably upon all the older rocks of the district, fragments of which occur in the conglomerate and which afford fine examples of ‘‘roches mon- tonnées” and striated surfaces: The direction of the striz and distribution of the bowlders point to a movement from the north southwards. In his report for the same year Dr. Corstorphiney summed up the results obtained in the north and south of Cape Colony and elsewhere, and compared the features of the northern and southern deposits, contrasting the northern Glacial Conglom- erates, possessing the characters of a ground moraine, with the southern Dwyka, which is to be looked upon as “a sediment formed under a probably inland water, into which there floated the icebergs calved from the front of the glacier or glaciers on the northern shore.” The identity in character of the Glacial Conglomerate with a true ground moraine, seen in the northern parts of Cape Col- ony, comes out with even greater clearness along the northern edge of the main area occupied by the Karroo System in the Transvaal. : The Glacial Conglomerate in the Transvaal._—In the cen- tral portions of the Transvaal, and particularly in a district lying along the eastern railway line from Pretoria to Middel- burg, I have recently mapped many outliers of Karroo rocks isolated by the progress of denudation from the main body, which covers extensive areas to the south and south-east. These outliers sometimes include portions of the sandstones, grits, and shales associated with coal-seams which form the upper portion of the Karroo System as developed in this part * Rogers and Schwartz, Ann. Rep. of the Geol. Commission, 1899. Cape Town, 1900. Onthe Orange River Ground Moraine. Trans. Phil. Soc. S. A., vol, xi, part 2, 1900. G. S. Corstorphine, Ann. Rep. of the Geol. Commission, 1899. Cape Town, 1900. (Full references to the previous literature will be found in this paper.) E. T. Mellor—Glacial Conglomerate of South Africa. 118 of the Transvaal. They are, however, frequently reduced to patches consisting almost entirely of the Glacial Conglomerate and associated beds. The copious sandy drift shed by these outliers frequently renders their examination dificult, but in some cases they offer more than usually good opportunities for the study of the Glacial Conglomerate and its relationships to the underlying rocks. In the district here more especially referred to, the glacial deposits consist for the most part of a conglomer ate showing all the characters to be expected in one formed beneath an extensive ice-sheet. This conglomerate is i! very irregular in distribution, and varies greatly in thickness within short distances, partly in consequence of its original deposition on an irregular land surface, and partly as a result of subsequent denudation. Its average thickness is about fifty feet. In depth the rock is sometimes greenish in color, but at the surface it is usually light yellow, and crops out in char- acteristic humpy masses (see fig. 1). The matrix is a sandy- looking material consisting of sharply angular fragments of quartz “and of various rocks—quartzites, hard shales, felsites, granophyres —common in the district. These angular frag- ments vary in size from the smallest particles to pieces several inches in diameter. Irregularly distributed through the matrix, and with a conspicuous absence of any sort of arrange- ment as to size or orieritation, occur abundant pebbles and Am. Jour. Scl.—¥YouRTH SERIES, VoL. XX, No. 116.—AvgGust, 1905. 8 1l4 #. 2. Mellor—Glacial Co Nepean of South Africa. bowlders of very Secelane ome composition, and ranging in size up to a diameter of eight or ten feet. These pebbles and bowlders are frequently facetted, and those of very hard mate- rials are always lighly polished, while bowlders of somewhat softer nature, especially if fine in grain, such as hard shales and weathered felsitic rocks, frequently show striations. A network of cracks in some cases divides the pebbles into a num- ber of fragments which have been again cemented into a whole. In any particular locality there is always a prepon- derance of bowlders derived from rocks which locally underlie the Glacial Conglomerate, associated with others easily recog- nizable as derived from more distant sources, which are always to the north of the present position of the bowlders. Thus along the eastern railway line, to the south of an area mainly occupied by the Waterberg Formation and the Red Granite, the Glacial Conglomerate contains an abundance of bowlders derived from these rocks, South of the outcrop of the hard white Magaliesberg quartzites, fragments of the white quartz ites are very abundant. Those lying nearest to the ridge from which they were derived are angular and frequently of huge dimensions, so that when weathered out and lying on the sur- face they are conspicuous objects at a distance of two or three miles. On the eastern Witwatersrand the conglomerate con- tains many bowlders derived from the Rand Series together with others formed of the hard cherts of the Dolomite to the north. Except quite locally, the lower and more massive por- tions of the conglomerate rarely show any traces of bedding, but are occasionally traversed by irregular partings dividing the rock into rude sheets with undulating billowy surfaces. Towards the upper portions of the conglomerate, lenticular beds of fine-grained massive sandstone frequently occur, together with patches of white and cream-colored shales and mudstones. The shales appear to have been formed in local pockets below the ice. They consist of the finest glacial mud. The examination of a district of some hundreds of square miles in extent leads to the conclusion that at the termination of the period during which glacial conditions obtained, the country was left covered with an almost complete mantle of glacial deposits, quite similar in character and distribution to those remaining in other parts of the world from extensive glaciation of more recent date. After the cessation of glacial conditions the conglomerates and associated deposits appear to have suffered a certain amount of sub-aérial erosion and denu- dation, during which materials derived from the glacial deposits underwent re-arrangement and re-deposition, giving rise in some cases to beds of conglomerate very similar in com- position and general appearance to those of glacial origin, with E. T. Mellor— Glacial Conglomerate of South Africa. 115 which they are lable to be confused, but differing in the more orderly arrangement of their materials, including a definite orientation of the pebbles and bowlders. These secondary conglomerates occasionally occur at the base of the purely sedimentary series which succeed the true glacial deposits, and by which as a result of a period of long continued subsidence the latter were ultimately entirely covered. This sedimentary series included the succession of beds constituting in the Trans- vaal area the upper portion of the Karroo System. Later formations were also possibly represented but of these no ves- 2 bw = coll tige has hitherto been discovered in the Transvaal. Raised subsequently to an average elevation of 5000 feet above the sea, the Karroo System has been again subjected to denuding forces and the removal of the overlying sandstones, shales,‘and grits of the Coal Measures has laid bare extensive areas of the underly: ing Glacial Conglomerate. Although modified by the double process of denudation‘to which it has been subjected, it still presents in its distribution a striking similarity to that of more recently formed glacial deposits. Following the contours of the land surfaces upon which it was originally laid down, it ranges within distances of a few miles through variations in elevation of three to five hundred feet. It is frequently well developed on one slope of a hill and entirely absent from the other. When protected 116 £. 7. Mellor—Glacial Conglomerate of South Africa. from erosion it fills preéxisting valleys, and is usually espe- cially abundant below ancient escarpments of the older rocks, and in such places bowlders often of very large size, attaining in some cases eight or ten feet in diameter, are exceptionally numerous, After the complete weathering away of the matrix the bowlders remain abundantly scattered over areas previ- ; ously oecupied by the conglomerate. (See fig. 2.) Glaciated Surfaces below the Conglomerate Direction of Ice-Movement. denudation of the Glacial Conglomerate around the margins of the areas now occupied by the Karroo System and its outliers continually lays bare fresh portions of the underlying old land surface. Where these include outcrops of hard and moderately fine-grained rocks, the latter frequently present excellent examples of elacially striated surfaces,* some of which are represented in the photo- graphs reproduced in figures 3, 4, 5. Striated surfaces of this kind were long ago described by Sutherland in Natal, by Griesbach in the same colony, by Dunn and Schenck in the neighbourhood of the Vaal River, and more recently by Molen- graaff in the South-Eastern Transvaal, and by Rogers and Schwartz in the Prieska district in the north of Cape Colony. While working on an area lying about 25 miles east of Pretoria in 1903, I found the surface shown in figure 3, and later those in figures 4 and 5. These latter occur on the edge of an out- her of Karroo rocks some 25 miles further east, which includes the coal seam worked at the Douglas colliery. I have since met with many similar surfaces distributed over an area of some 300 square miles. The striation in most cases is exceed- ingly clear, and the direction of ice-movement easily deter- mined and remarka ably consistent. In all the examples found it only varies within a few degrees from magnetic north and south, the direction of movement being ina southerly direction, which is also true in general for the other districts in South Africa where striated surfaces have been found. This con- sistency of direction over so considerable an area and in the case of surfaces lying 25 miles apart, points to the existence of an ice-sheet of considerable magnitude, rather than to that of a number of more or less isolated glaciers, a conclusion which is supported by the nature of the land surface laid bare by the disappearance of the Karroo deposits. Where the Waterberg Sandstone Formation, which occupies much of the district here referred to, has been ‘long exposed to ordinary denudation, the rivers cut deep valleys and gorges in the sandstone, giving rise to very varied and occasionally rug- *E. T. Mellor, On Some Glaciated Land Surfaces occurring in the Dis- trict between Pretoria and Balmoral. Trans. Geol. Soc. S. A., vol. vii, part 1, 1904. E. T. Mellor— Glacial Conglomerate of South Africa. 11% 3 | | 118 £. 7. Mellor—Glacial Conglomerate of South Africa. ged scenery. Where, however, the overlying Glacial Conglom- erate is only now in process of ‘removal, the country retains the rounded outlines characteristic of a olaciated landscape. Northern extension of the Glacial Conglomerate. I have recently met with good examples of the Glacial Conglomerate much further to the north than any hitherto described.* (Figures 1 and 2.) These are situated near the j june- tion of the Elands and Olifants Rivers, about 90 miles north of the latitude of Johannesburg, and are interesting for the addi- tional light they throw upon the northward extent of the coun- try subjected to glacial action in early Karroo times. EXPLANATION OF FIGURES. Figure 1.—Glacial conglomerate near the junction of the Elands and Olifants Rivers, Transvaal (75 miles NE. of Pretoria). FicurE 2.—Weathered-out Glacial conglomerate, same locality. The figure stands upon grits of the upper Karroo formation. FicuRE 3.—Glaciated surface. Elands River Vailey (25 miles E. of Pre- toria). Ficures 4 and 5.—Glaciated surfaces north of Balmoral (50 miles E. of Pretoria). The striated rocks are red quartzitic sandstones of the Waterberg Series. Geological Survey, Pretoria, Transvaal. *K. T. Mellor, Outliers of the Karroo System near the Junction of the Elands and Olifants Rivers in the Transvaal, Trans. Geol. Soc. S. A., vol. vii, part 3, 1904. H. F. Cleland—Formation of Natural Bridges. 119 Arr. XV.—The Formation of Natural Bridges; by HerpmMan F. CLEenanp. Untri recently the text-books of Geology and Physical Geography have given the idea, whether intentionally or not, that natural br idges are univer sally formed by the partial caving in of a long cavern, the bridge being that portion of the roof strong enough to span the cavity.* The belief seems to be prev alent that these cavities ae for long distances, a condition comparable to that which would exist if the greater part of the roof of Mammoth Cave should fall in, leaving a small portion as a bridge. This theory is simple and logical ‘and is one which immediately appeals to the reader, but, as will be seen from the examples cited in this paper, not only i is it not of universal application but it must be exceptional rather than otherwise. The writer was led to this study by an examination of the natural bridge near North Adams, Mass. which has long been considered to be a typical example and proof of the formation from caverns. The North Adams Natural Bridge spans Hudson Brook and has been an object of more than local interest for many years both because of its natural beauty and because of the rarity of these objects. Hudson Brook is a small stream emptying into Beaver Creek, a tributary of the Hoosick River. From the dam (shown in the sketch tig. 1) to the pre-glacial valley the brook flows through a gorge 30 to 60 feet deep and from’ 5 to 40 feet wide, the average width above the bridge being from 1 to 10 feet and below from 10 to 30 feet. This gorge is cut in a coarsely crystalline marble which, because of its color and texture, presents a striking appearance. The rock is Cambro- Silurian and belongs to the Stockbridge formation. The top of the natural bridge is 44 feet above the water of the stream and the bridge itself is about 8 feet thick. The span of the bridge is less than 10 feet long and the width at present 25 feet, but at one time it probably extended a short distance farther south where it is now fallen in. It is extremely dificult to take a good photograph of the bridge because, as will be seen from the sketch, the stream turns sharply ‘both above and below. Because of this condition it was found necessary to make a drawing, in order to give a cor- rect idea of its appearance. Prot. E. Hitchcock described the North Adams Natural Bridge and published a rough drawing of it in 1841.+ Con- cerning this drawing he says, “I thought it better that a sketch * Chamberlain and Salisbury, Geol., vol. i, pp. 145-147. + Geology of Mass., by Edward Hitchcock, vol. i, 1841, pp. 287-288. 120) H. £. Cleland—Formation of Natural Bridges. should be taken by one not at all accustomed to drawing, than that no memento be left of this interesting place,” (there was danger at that time that the bridge might “be destroyed by the quarry-men.) Hovey* in his “ Celebrated American Caverns” describes this bridge but gives the locality as Adams, Mass. The explanation of the for- mation of the North Adams Natural Bridge, as given by Hitcheock and accepted by Hovey, is that it is the section of the roof of a cavern, the ends of which have fallenin. In illustration of this point, Hovey states that, “the com- bination of cave, chasm and natural bridge, on Hudson Brook, Mass. is even a better example (than that of the Natural Bridge in Virginia) of the same (thinesaeeemes “that what are now. open canons were once caves, the arch being merely a remnant of an ancient cave roof.” On examining the course of the stream and the rock in the vicinity of the North Adams Natural Bridge one is struck with the width of the joints, O___5o'_joo'_/s0" and the fact that the stream Fie. 1.—Sketch map of Hudson has, for a portion of its course, Brook, Mass., showing the position of followed the joint planes. In the natural bridge, the joint planes the upper part of the accom- A-A, and the pre-glacial valley. panying aleetrelh (fig. 1) Alte relation of the stream to the Joint planes is indicated by the dotted lines A-~A. The channel through which the stream flowed previous to the formation of the bridge is also well marked a few feet to the west at b. . eeeee 95 XIII.—Behavior of Typical Hydrous Bromides when Heated in an Atmosphere of Hydrogen Bromide; by J. L. KRBIDER (ooo i ee ee ee ee XTV.—Glacial (Dwyka) Conglomerate of South Africa ; by BR. -T, Menor 2s eer 2 ee ae XV.—Formation of Natural Bridges ; by H. F. Ciuzanp.. 119 XVIL—Quartz from San Diego County, California; by G. A. WARING 230 Oo OMe tere cee eo 125 XV II.—Radio-active Properties of the Waters of the Springs on the Hot Springs Reservation, Hot Springs, Ark.; by ; B. By BOLEWOOD |. ee 2 oe Ee eee 128 We SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Studies with the Liquid Hydrogen and Air Calori- meters. I. Specific Heats, J. Dewar, 152.—Thermo-electric Junction as a Means of determining the Lowest Temperatures, J. Dewar, 153. Geology—Geology of the Vicinity of Little Falls, Herkimer County, H, P. CusHiInG, 156.—Geology of the Watkins and Elmira Quadrangles, accom- panied by a geologic map, J. M. Cuarxe and D. D. Luter, 157.—Geologic - map of the Tully Quadrangle, J. M. CLarke and D. D. Luraer, 158,.— Contribution to the Paleontology of the Martinez Group, C. E. WEAVER; ' Faune cambrienne du Haut-Alemtejo (Portugal), J. F. N. DELGabo, 159. —Paraphorhynchus, a new genus of Kinderhook Brachiopoda, S. WELLER : Sympterura Minveri, n. g. et sp.; a Devonian Ophiurid from Cornwall, F. A. BaTHER: Ancestral origin of the North American Unionide, or fresh- water Mussels, C. A. Wurter, 160.—Thalattosauria, a group of marine rep- tiles from the Triassic of California, J. OC. Mereram: Geology of Littleton, New Hampshire, C. H. Hircucock: Vorschule der Geologie, J. WALTHER, 161.—Die Moore der Schweiz mit Beriicksichtigung der gesammten Moor- frage, J. Frvw and C. Scurormr, 162.—Study of Recent EKarthquakes, C. Davison: Introduction to the Geology of Cape Colony, A. W. RoGzErRs, 163.—Ice Erosion Theory, a Fallacy, H. L. Farrcuiip, 164.—Hanging Val- leys, I. C. RussELu, 160. —Glaciation of the Green Mountains, C. H. Hircx- cock: Ice or Water, H. H. Howortn, 166. Miscellaneous Scientific Intelligence—United States National Museum, R. RatHBun: Forestry ; Tenth Annual Report of the Chief Fire Warden of Minnesota, C. C. ANDREWS: Les Prix Nobel en 1902: Negritos of Zam- bales, W. A. Reep: Magnetic Survey of Japan reduced to the Epoch 1895-0and the Sea- level, A. TANAKADATE, 167.—Beitrige zur chemischen Physiologie, F. HorMeisTER: Du Laboratoire & V Usine, L. HOULLEVIGUE : Traité Complet de la Fabrication des Bieres, G. Morgav and L, Livy, 168. SEPTEMBER, 1905. Established by BENJAMIN SILLIMAN in 1818. THE AMERICAN JOURNAL OF SCIENCE, | | Epitor: EDWARD S. DANA. | | | . . . . ASSOCIATE EDITORS Proressors GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or Camsrwvce, Proressors ADDISON E. VERRILL, HORACE L. WELLS, L. V. PIRSSON anp H. E. GREGORY, or New Haven, Proressor GEORGE F. BARKER, or PHILADELPHIA, Proressor HENRY S. WILLIAMS, or Iruaoa, Proressor JOSEPH S. AMES, or Baitrworse, Mr. J. S. DILLER, or Wasuineron. FOURTH SERIES VOL. XX—[WHOLE NUMBER, CLXX.] No. 117.—SEPTEMBER, 1995———_ WITH PLATES V-—VIII. . 5 i} \ “4 + “ \ 4 ad — NEW HAVEN, CONNECTICUL ~ of, Se. Sa THE TUITLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. ane a ee Published monthly. Six dollars per year, in advance. $6.40 to countries in the _ Postal Union. Remittances should be made either by money orders, registered _ letters, or bank checks (preferably on New York banks). CRYSTALLIZED STANNITE A small consignment direct from the silver mines in Bolivia affords new and unique examples of the following. They have been identified by high mineralogical authority. Stannite, Splendent crystals, grouped in cavities of the massive mineral. The crystals are most perfectly developed and very rich in planes. The species has been known to science for over a century, but until recently only in the massive form, thus leaving its crystallographic form in question. Hence the present find is peculiarly welcome. Andorite. A silver and lead sulph-antimonide. Massive and in small but characteristic crystals of highly modified orthorhombic habit. Some- times associated with the crystallized Stannite. RARE MINERALS Specimens are supplied by us to students and chemists for purposes of comparison and investigation ; in commercial quantities for industrial uses. SYSTEMATIC COLLECTIONS OF TYPICAL SPECIMENS In sets of twenty-five up to fifteen hundred specimens. Prices $5.00 upwards per set, the average price for students’ specimens being about twenty cents. We have supplied the leading institutions for thirty years, having lately completed a single order for over 60,000 specimens. Our material is the accepted standard both as to correct labeling and high quality. Free Collection Catalog, containing lists and illustrations of General Mineral Collections, Series of Ores for Prospectors, Sets of Crystals, Series illustrating Hardness and other Physical Characters cf Minerals, with Price List of Laboratory Material and Individual Specimens. : FOOTH MINERAL OG. Established 1876, by Dr. A. E. Foote. W. M. FOOTE, Manager. DEALERS IN MINERAL SPECIMENS AND COMMERCIAL RARE MINERALS, 1317 Arch Street, Philadelphia. i thd MORN Nt pes SEE Ss Mths te SS a ee od PAL 5 32 jek re aes Te si a Page 2 yee; ee THE AMERICAN JOURNAL OF SCIENCE FEOUR DH SE RES. | Art. XX.— Development of Fenestella; by Epear Roscor Cumines, Ph.D. (With Plates V, VI, and VIL.) Introduction. Durtine the past two years, the writer’s studies of the devel- opment of Paleozoic Bryozoa* have brought out some very interesting points bearing upon the earliest stages of Lenestella. The present paper deals with the development (astogeny) and morphology of Henestel/a, and is based entirely upon calcified material from the Hamilton formation of Thedford, Ontario.+ This material consists of numerous bases of /enestella colonies. In these, the minutest details of internal structure are pre- served with remarkable fidelity. The method of study has been the preparation of both thin and serial sections. The latter were obtained by slowly grinding down the bases and accurately drawing each stage as seen by reflected or in some eases by transmitted light. ‘The specimens studied are in vari- ous stages of growth. Some represent the bases of adult colonies from which the adult (ephebastic) portion has been lost; others are minute bases, which in their growth never - proceeded farther than the nepiastic stage. In these nepiastic * In a former paper, a classification of the growth stages of the bryozoan - colony was given, together with a general classification of the growth stages - of any colony belonging to any group of organisms. The terms applicable to the growth stages of any colony are: Nepiastic, neanastic, ephebastic, and gerontastic, corresponding to the well-known terms nepionic, neanic, ephebic, and gerontic, applicable to the growth stages of the individual. Dr. Ruede- mann has recently proposed the term astogenetic with reference to the colony, as the term parallel with ontogenetic with reference to the individual. The astogenetic stage of a colony, therefore, corresponds with the ontogenetic stage of an individual. + This Fenestella is probably the form listed by Grabau as Semicoscinium labiatum. Am. Jour. Sci.—FourtH Series, Vout. XX, No. 117.—SEPTEMBER, 1905. 12 170 E.R. Cumings —Development of Fenestella. colonies, the zocecia emerge upon the surface; but in the older ones, the apertures of the zocecia in the basal portion are sub- merged in a copious deposit of punctate sclerenchyma. In all cases, however, there has been no resorption of the earlier zocecia, So that sections of the bases of ephebastic or gerontas- tic zoaria reveal the morphology of the earliest stages as faithfully as sections of a nepiastic colony. As an aid to the elucidation of the astogeny of /enestella, the writer studied the astogeny of Lvetepora phenicea, a recent bryozoan morpho- logically very similar to the ancient Fenestellas and Poly- oras. : In the writer’s former paper on the development of Paleo- zoic Bryozoa, the term protecium was introduced as designating the primary individual of the colony. In this sense, it would have the same signification as the term ancestrula of Jullien or primary cell of Hincks. In the Cyclostomata, as is well known, the first zocecium surmounts a hemispherical base (basal disc), which serves as the point of attachment of the young colony to the substratum. This basal disc has been shown to be the calcified wall of the metamorphosed and histolyzed embryo (Barrois and others). It is believed by the present writer that the persistence of this structure (kathem- bryonic stage) in the ancient order of Cyclostomata is not without significance, especially in view of the fact, to be shown presently, that it is a conspicuous feature in the development of the ancient Cryptostomata and possibly of the Trepostomata (Phylloporina corticosa). The basal dise is probably the ¢rwe first zocecium. In the present paper, there- fore, the term protwciwm is restricted to the basal dise or its equivalent, and the superjacent portion of the primary cell is designated the ancestrula. In many recent Chilostomata, there seems to be no distinction of protcecium and ancestrula. This may mean that the extreme acceleration of these modern types has practically eliminated the protcecium from the on- togeny. In the ancient Cryptostomata, on the other hand, the proteecium greatly predominates over the ancestrula, which is ~ often little more than an exaggerated aperture to the former. In any case, the ontogenetic stage of which the protcecium is the index is always present throughout the Ectoprocta, for by a degenerative metamorphosis they all give rise to a hemi- spherical kathembrvo, from which the adult polypide arises by a sort of budding process. Furthermore, this kathembryo becomes invested with a calcareous or chitinous ectocyst, which is the first skeletal structure of the developing individual. The proteecium is therefore very closely. analogous to the protegulum of brachiopods, the protoconch of cephalopods, etc. E.R. Cumings—Development of Fenestella. 171 DEVELOPMENT OF EF ENESTELLA. The Proteecium. Many well-preserved Fenestella bases show a minute circular pit on their basal surface. This can be seen only in colonies that were attached to a substratum which disappeared in the process of fossilization, leaving the basal surface of the colony free from all extraneous matter. Where the colony is still attached to the substratum, frequently the frond of another bryozoan, the circular pit can always be demonstrated by means of thin sections. This pit is the protceecium. As will be seen from the longitudinal sections (figs. 20, 36, 387, 59), the protcecium is separated from the substratum by a thin basal membrane. In such sections, this pit appears as a semi- circular object in the proximal portion of the colony. In trans- verse sections, it appears as a dark ring surrounded by concen- tric zones of punctate secondary sclerenchyma. That the protcecium has its own proper wall, similar to that of ordinary zocecia, is shown by numerous sections (figs. 86-38, and 59). The diameter of the protcecium is from 0-4—0°6™, or about three or four times that of the ordinary zoccia. In form and position it corresponds precisely to the basal disc of Cyclosto- mata, and there can be little doubt that it has the same morphological and developmental significance. The Ancestrula. The protcecium is surmounted by a tubular structure arising from the center of its distal surface. This is the ancestrula. In some of the earlier sections prepared by the writer, one of the primary buds was mistaken for the ancestrula, and its size and shape were therefore thought to be different from what was shown in later sections. It is considerably smaller than the primary buds, being both shorter and of less diameter. It seems altogether likely that the primary polypide never per- manently ascended into the ancestrula as in the Cyclostomata. On the other hand, the ancestrula of Fenestella is far from being the homologue of the vestibule of ephebastic zocecia. It is not built up of secondary deposits, but is composed of the same thin non-punctate substance as the proper wall of the proteecium and other zoccia. The homology of the ancestrula of Fenestella is with the tubular primary zowcium of the Cyclostomata. Figures 59 and 60 indicate the shape and appearance of the ancestrula_ as seen in the majority of prop- erly orientated longitudinal sections,* and figures 10-13, 24, 43, and 54 in transverse sections. * The zoecium marked J, in figures 19 and 20, was at first thought to be the ancestrula, since it communicates freely with the protecium. A careful study of the appearances possible in a series of longitudinal sections with 172 LE. £. Cumings—Development of Fenestella. The Primary Buds. Two lateral primary buds arise from the primary zocecium (figs. 8-7, 21-23, 40-43). There is still some question as to whether these buds arise from the protcecium or from the ancestrula. The sections figured reveal all that can be ex- pected. The question becomes one of interpretation and of analogy with recent Bryozoa. The proximal ends of the pri- mary buds are in contact with the protcecium and are separated from its cavity by a very thin ecaleareous wall, which is fre- quently broken away (figs. 19 and 20). The appearance of this wall is well shown in figure 36. Figures 3-7 and 38-40, 42 show the intimate relation of the primary buds to the pro- toeclum. From the analogy of recent Bryozoa, on the other hand, these buds might be expected to originate from the ancestrula. A median primary bud is not indicated by any of the sections. If it existed, it certainly arose from the ances- trula. The size, shape, and position of the primary buds is beauti- fully shown in figures: 38 and 39, and in the transverse sections. These buds are long and tubular, and diverge but slightly from the axis of the zoarium. There is no long vestibule as in ephebastic zocecia, but the whole aspect of the buds is that of a simple tubular zocecium, quite similar to that of the Cyclos- tomata. There is also no indication of hemisepta or any other structures within the zocecium. Secondary Buds. All buds of the second generation from the protcecium are designated secondary buds. The series of sections (figs. 1-16) seems to indicate that each of the primary buds produces a lateral and a median bud. The lateral buds are very clearly shown in such a position that they could have originated from no other source than from the primary buds (see especially figs. 5,41, and 42). The median buds belong to the second tier of zocecia. They are designated //,, and //,, in figure 13. The shape of the secondary buds is quite similar to that of the primary ones (figs. 37,45, 59, and 60). Figure 50 is a drawing different assumed orientation has convinced the writer that the zocecium in question is a primary bud. To test this, four different bases in which the protceecium and primary buds could be seen on the basal surface (in some cases only after slight etching) were sectioned in the direction j — 7, figure 48, which had been determined by previous inspection of the basal surface, and marked by carefully drawing a fine line through the center of the pro- toecium and as nearly as possible between the primary buds. Every one of these sections has the appearance shown in figures 59, 60, and 45. It is therefore unlikely that figures 19 and 20 (which were orientated at random) represent the ancestrula. It is needless to state that only a very small proportion of the many sections prepared in this study are figured. EE. R. Cumings—Development of Fenestella. 173 of a secondary bud, and may be compared with figure 53, which is a drawing of two zocecia of Protocrisina (after Ulrich), a eyclostomatous bryozoan from the Trenton. The resemblance * is too striking to need further emphasis. No internal zocecial structures have been observed in the secondary buds. Tertiary and Later Buds. One bud of the third generation from the ancestrula occu- pies a position in the first tier of zocecia, diametrically opposite the ancestrula (/Z/, figs. 6-18, 24, 26, 48, 54-58). The shape of this bud is well shown in figures 37, 45, 59, and 60. There | is no means of telling from which of the two secondary buds this tertiary one is derived. It may have originated now from one, now from the other. In figure 43, it is rather more inti- mately associated with 32, which was in turn derived from the right lateral primary bud. Figure 13 indicates that each of the secondary buds gives rise to a median bud lying in the second tier of zocecia. : Ascending the axis of the zoarium (figs. 17-20, 36-39), there is exhibited a series of zowcia very symmetrically arranged about the axis. In transverse sections, above the level of y, figure 17, these present a peculiar star-shaped appearance seen in 1 fioures 15, 16, and 58, as well as in figure 61 of the writer’s former paper. The order of budding of these later zocecia cannot be determined, although the writer has devoted a large amount of time and study to this point. It is probable that the order of budding 1 in these later generations is without sig- nificance. An important point shown by the sections, how- ever, is the shape and size of these zocecia. ‘This is best seeti in fiour es 17 and 38. The zoecia are tubular, but somewhat less elongate than the earlier ones. It is not ‘until the zoarium begins to expand into its characteristic infundibular form that the zocecia assume the shape normal to Fenestella. Figure 51 shows a row of zocecia from the neanastic region (base of the cone) of the specimen represented in figure 38. For com pari- son with this is inserted figure 52, showing a specimen otf Fenestella acmea from the Waldron shale of Tarr Hole, Indiana. The resemblance is striking. The adult zoccia of the Thed- ford Fenestella are shown in figure 49. Discussion and Conclusions. The morphological element of the bryozoan colony which corresponds to the primitive integument of Mollusca, Brachio- poda, ete. (that is, to the protoconch, protegulum, ete. ), is the protecvum, or basal disc, of the primary individual of the colony. The protcecium is the calcareous or chitinous wall of 174 EE. R. Cumings— Development of Fenestelta. the kathembryo. In /enestella it is very large and in every way similar to the protcecium (basal disc) of the Cyclostomata. The ancestrula is the tubular superstructure of the primary individual. It is a simple, undifferentiated, tubular zocecium. The earlier formed zocecia (nepiastic zocecia) of the Fenestella colony differ markedly in shape and size from later formed | (neanastic and ephebastic) zocecia. In every feature in which they depart from the ephebastie zocecia of Fenestella they approach the ephebastic zocecia of the Cyclostomata. From these observations, it may be reasonably concluded that Fenestella as well as the entire order of Cryptostomata is derived from the Cyclostomata. Certain other general conclu- sions, more or less speculative, are suggested by a consideration of the probable significance of the protcecium and ancestrula. The meaning of the degenerative metamorphosis of Bryozoa has always been a puzzle to students of this class. The striking analogy of this metamorphosis to the degeneration of an ordi- nary polypide and production of a brown body, together with the nearly identicai life history of the regenerating polypide or of ordinary buds and the primitive polypide issuing from the kathembryo, have more than once led to the suggestion that the primitive polypide is in the trne sense a bud. The writer is inclined to hold this view. Assuming, therefore, that the primitive polypide is a bud, the following suggestions may be made in regard to the significance of the metamorphosis and of the resulting protcecium: 1. In the primitive bryozoan, there was no histolysis of the larval organs. The development was direct and resulted in a primitive zocecium and polypide. 2. This primitive zocecium was hemispherical in shape and possessed a simple aperture in the center of its upper surface. Some ancient types of Cyclostomata retain nearly such a form of zocecium (Stomatopora of the Trenton, especially S. twrgida). 3. This primitive zocecium might now give rise to a linear adnate series of zocecia, as 11 Stomatopora, or to a series of superposed zocecia, as in the Trepostomata. By variations of zoarial habit based upon one or the other of these fundamental plans of budding all existing types of Bryozoa could have been produced. 4. In accordance with the law of tachygenesis, later in the history of the bryozoan group a tendency toward concen- tration of the early stages in development would arise. In any colony the tendency to degenerate may be supposed to have applied to the primitive polypide as well as to later ones, and finally to have become an invariable part of its life history. By the continued operation of the law of tachygenesis, the life history of the first polypide- became so abbreviated as to be E.R. Cumings— Development of Fenestella. 175 represented only by its degenerative stage, that is, by its latest growth stage, all the earlier growth stages having been crowded out or back into the larval stage. In accordance with this interpretation of bryozoan develop- ment, the large size of the protcecium in ancient types is ex- plicable and is thought to be due to a less degree of acceleration, the calcification of the zocecial wall of the primitive individual being allowed to proceed nearly to completion before the second zoceclum was superposed upon it. The probability that the first polypide remains in the protceecium in /enestella, instead of ascending into the ancestrula as in-modern Cyclostomata, may indicate a still more primitive condition. The relations of the protcecium and ancestrula in the Cyclostomata and in Fenestella suggest the normal relation of superposition of the zocecia in the Trepostomata. «It is not without interest to find evidence, in the development of Paleozoic Bryozoa, of the fundamental relation- ship of these great groups. Ulrich (Geol. Surv. Illinois, vol. vill) has already suggested such a relationship on the ground of the resemblances of such types as the early Fenestellas, Phyllo- porma and Protocrisina. The evidence presented by these adult types is greatly strengthened by the striking parallelism of the nepiastic stages of Henestella with the series of adult types named above. Paleontological Laboratory, Indiana University, June, 1905. EXPLANATION OF PLATES. Description of Figures.* Letters having the same meaning for all the figures :— a, 6, c, d, e, primary carine (except figs. 17, 24, 47, and 48). J, fenestrule. Kk, carina. 0, protoecium. s, substratum of bryozoan colony. z, z', etc., zocecia of generations later than the primary zoccia. A, ancestrula. I, primary bud. II, bud of second generation, that is, derived from a primary bud. III, bud of third generation. 2, left lateral bud. - 3, right lateral bud. 23, right lateral bud of the second generation, derived from a left lateral primary bud. 32, left lateral bud of the second generation, derived from a right lateral primary bud. * All drawings except figures 1-16 were made with the camera lucida. Figures 30-32 are after Barrois, and figure 53 is after Ulrich. All the speci- mens of Fenestella are from Thedford, Ontario. 176 E.R. Cumings—Development of Kenestella. PLATE V. Figures 1-16.—Transverse serial sections of a Fenestella base. These six- teen sections represent 1™™ thickness of rock. Fiaures 1, 2.—Proteecium (cf. figs. 40, 41, 31-35). FIGURE 3.—Section in the plane of a-a, figure 47, cutting the proximal ends of the primary buds and the buds of the second generation (secondary buds) (cf. fig. 42). FIGURES 4, 5. —Successively higher sections. Ficure 6.—Section in plane of a/-a', figure 47, cutting the proximal end of the tertiary bud (cf. fig. 45). FIGURES 7-12. —Successively higher sections between the planes of a/—a’ and c-c, figure 47, showing the development of the initial buds. Figures 10-12 cut the aperture of the ancestrula (cf. fig. 24, with fig. 12). Fiaure 13.—Section cutting the proximal ends of buds of the second tier (IIo1, I13,, IIIb, and JTIc) (cf. fig. 26). FicuRE 14.—Section just cutting the distal end of the aperture of the ancestrula. Figures 15, 16.—Assumption of the star-shaped arrangement of zocecia, characteristic of the paranepiastic stage of Fenestella. PrarnoVi: Figure 17.—Longitudinal section of a Fenestella base. cutting in the plane of e-e, figure 47, and: a-a, figure 48. This section passes through the edge of the protcecium and misses the ancestrula entirely. 6, 6’, buds of the second tier. At z and z' the zoccia are vertically above each other; at 2’, z'’ they alternate, and at the top of the figure they lie side by side. x 17. ; Ficure 18. —Longitudinal section cutting still more excentrically than that shown in figure 17, probably in the plane of b-0, figure 48. This misses the proteecium and ancestrula entirely, but their relative. position is shown atoand A. The vertical alignment of zocecia is shown at 2-2’ and the ordinary arrangement, on either side of the carina, at z". The bifurcation of a primary branch is shown at g-h (between z” and g, h). In each new branch, the zocecia first alternate and later lie side by side. Normal arrangement shown atk’ k". x17. Fiaure 19.—Longitudinal section cutting in the plane of c-c, figure 48. The section cuts a row of zoccia (z'—z") nearly iongitudinally. x17. Figure 20.—Section in nearly the same plane as in figure 19 (d-d. figure 48). This section was orientated by polishing and etching the basal surface of the colony and marking the position of the protcecium and primary buds. The section was then ground as nearly as possible in the marked direction. A primary bud is very clearly shown (J), x17. FicuRE 21.—Transverse section in the plane of a-a, figure 47. The primary buds are very distinct. x17. FIGURE 22.—Similar section of another specimen, cutting the proximal end of the ancestrula. x17. FIGuRE 23.—Transverse section of avery slender base. Section in about the same plane as 22. x17. FicurE 24.—Section in the plane of 6-0, figure 47. Ancestrula very dis- tinct, x17. Figure 25.—Longitudinal section of a base from which the substratum was absent. x17. Ficure 26.—Transverse section in the plane of d-d, rae 47, showing the proximal ends of two buds of the second tier (z, z’) (ef. fig. 13) a tcslie Figure 27.—Protecium and ancestrula of Retepora phoenicea from St. Vincent’s Gulf, Australia. x 27. Figure 28.—Ancestrula and three primary buds (7, 2, 3) of Retepora phe- WIGS KX 29’ Figure 29.—Profile view of protecium and ancestrula of another specimen of Retepora phonicea. x27. Am. Jour. Sci., Vo!. XX, 1905. Plate V. Plate VI. Am. Jour. Sci., Vol. XX, 1905. Plate VII. Am. Jour. Sci., Vol. XX, 1905. E. R. Cumings—Development of Fenestella. aN _ Figure 30.—Protecium, ancestrula, and primary bud of Tubulipora. After Barrois. x27. FIGURE 31.—Same:; seen from the under surface. x 27. FIGURE 32.—Ancestrula and primary bud of Schizoporella. x 38. Figure 33.—Protecium and primary zowcia of Phylloporina corticosa from Cannon Falls, Minnesota. x17. FiGuRE 34.—Protecium of Polypora from the Lower Helderberg of Indian Ladder, New York. x28. Figure 35.—Protecium of Thamniscus from the Upper Coal Measures of Kansas. x17. Puate VII. FiGcuRE 36.—Longitudinal section of Fenestella, in the plane of f-f, figure pear Lt: Figure 37.—Longitudinal section in the plane of g-g, figure 48. x17. Figure 38.—Longitudinal section in the plane of h-h, figure 48. This sec- tion shows remarkably well the shape of the nepiastic zocecia. x 17. FiIcuRE 59.—Longitudinal section in the plane of 7-7, figure 48. x17. Figure 40.—Transverse section in the plane of a-a, figure 47. Shows the primary buds (2, 3). x17. Ficure 41.—Transverse section of another specimen in which the primary and secondary buds have a rather unusual arrangement. x17. FicuRE 42.—Transverse section in the plane of a'-a’, figure 47 (cf. fig. 5). Same specimen as figures 54-58. x17. Ficure 45.-—-Transverse section in the plane of 6-0, figure 47. Shows the proximal end of the tertiary bud (ef. fig. 10). x17. Figure 44.—Probable interpretation of figure 39. Section in the plane of ii, figure 48. é Figure 45.—Semidiagrammatic drawing of a longitudinal section (in the plane of j-j, fig. 48) of a specimen showing the ancestrula and two zocecia, probably one of the secondary buds and a tertiary bud (UJ, lJ). x17. FiGuRE 46.—Semidiagrammatic drawing of a longitudinal section (in the plane of ‘-k, fig. 48) of the ancestrula and protcecium of another speci- men. 6. X17. Figure 47.—Semidiagrammatic drawing from figure 37, to show the position of transverse sections. Figure 48.—Semidiagrammatic drawing from figure 43, to show the position of longitudinal sections. Fieurs 49.—Ephebastic zocecia of Fenestella. Specimen from Thedford, Ontario. x17. FicuRE 00.—Nepiastic zocecium of Fenestella. Specimen from Thedford, Ontario. x17. FIGURE 01.—Neanastic zoecia of Fenestella. From the proximal portion of the cone of the same specimen as that shown in figure 88. x17. FIGURE 02.—Ephebastic zocecia of Fenestella acmea from the Waldron shale of Tarr Hole, Indiana (ef. fig. 51). x17. FIGURE 03.—Ephebastic zocecia of Protocrisina exigua Ulr. from the Tren- ton limestone of Montreal, Canada (cf. fig. 50). After Ulrich. x18. Ficures 54-%8.—Serial sections of a Fenestella base. Same specimen as that shown in figure 42, figure 54 being the next section above. Fig- ures 99-08 are successively higher sections. x17. Figure 59.—Longitudinal section (in the plane of j-j, fig. 48) of a Fenestella base, showing the shape of the ancestrula most often seen, and three nepiastic zocecia (JI, ITI, z"). x17. Figure 60.—Semidiagrammatic drawing from figure 59. {| 178 Darton—Age of the Monument Creek Formation. Art. XXI.—Age of the Monument Creek Formation ;* by N. H. Darton. Tus contribution is an account of additional evidence as to the Oligocene age of the Monument Creek formation, or at least of its upper member, afforded by the discovery of Titano- therium and other fossil bones at several localities. On the high divide between the Platte and Arkansas drainage _ basins, at the foot of the Rocky Mountains, there is an extensive | deposit of sands, gravel and clay to which F. V. Hayden gave the name of Monument Creek group.t This observer recog- nized the fact that the group overlies the Laramie formation unconformably, but apparently he included im its lower portion more or less of the beds later separated, as the Arapahoe and Denver formations in the Denver region, The opinion was held that it was of early Tertiary age, but no precise correla- tion was suggested. In 18738, Prof. E. D. Cope examined a portion of the deposit and found a few bones in regard to which he made the following statement :t “The age of the Monument Creek formation in relation to the other Tertiaries not having been definitely determined, I sought for vertebrate fossils. “The most characteristic one which I procured was the hind leg and foot of an Avrtiodactyle of the Oreodon type, which indicated conclusively that the formation is newer than the Eocene. From the same neigh- borhood and stratum, as I have every reason for believing, the fragment of the Megaceratops coloradoensis was obtained. This fossil is equally conclusive against the Pliocene age of the formation, so that it may be referred to the Miocene until further discoveries enable us to be more exact.” Doubtless Professor Cope regarded the fauna as belonging in the White River group, which is now generally considered to be Oligocene. He added nothing regarding the precise locality, or stratigraphic position of the fossils. So far as I ean find, no further paleontological evidence has since been offered, regarding the age of the formation. A brief account of the Monument Creek formation was given by G. H. Eldridge, in the “ Geology of the Denver Basin.”§ The true strati- graphic limits of the formation in relation to the underlying 5 * Published by permission of the Director of the United States Geological vey. Wprolieianey field report of U. S. Geological Survey of Colorado 4nd New Mexico, 1869, p. 40. t [7] eer ‘Report of the United States Geological and Coon Survey of the Territories, embracing Colorado, Report for 1873, by F. V Hayden, p. 480. § United States Geological Survey, Monographs, vol. xxvii, pp. 202-204. Darion—Age of the Monument Creek Formation. 179 Laramie, Arapahoe and Denver formations were recognized, and it was shown that the formation consists of two distinct members separated by a well-defined break in deposition. The lower member lies on an uneven floor of Denver formation at the north and Laramie to the southeast. It displays “marked regularity in the succession of its beds, excepting at the base, where, owing to the uneven floor, the material varies from conglomerate through sandstone to arenaceous shale. A short distance above the base are two broad bands of green shale separated by one of pink and capped by a fine grit, or sandstone, which is soft and friable and easily disintegrates.” The thickness is estimated to be about 900 feet. The sand- stones and grits of the lower member are mostly of granite debris. The upper member consists of sandstones and shales, with numerous beds of conglomerate, and between the two there are local deposits of rhyolitic tuff, in places 40 feet thick, which are quarried extensively for building stone near Castle Rock. In the lower part of the upper member many frag- ments of this rhyolitic tuff occur, a feature which is notably displayed in the breccia and conglomerate capping the butte known as Castle Rock. The thickness of the upper member is estimated by Eldridge at about 400 feet. In portions of the area, I have observed that in the lower member there are extensive deposits of massive clay, very similar in appearance and properties to the fullers earth which is characteristic of the Chadron formation, or Titanotherium beds, of the White River group in the Big Bad Lands of South Dakota and elsewhere. In the general résumé of the geology in the Monograph on the Denver Basin,* Mr. Emmons suggests that the vertebrate remains of Miocene age probably were from the lower mem- ber of the formation and that the upper member might be correlated with the Pliocene. This suggestion was based on the fact that the uppermost Tertiary deposits in the eastern portion of Colorado are of Pliocene age, and in the region north of the Platte River they lie unconformably on White River beds. Mr. Emmons recognized the fact that these beds differ somewhat from Monument Creek beds in character, yet this could be explained by the proximity of the Monument Creek formation to shore lines along the mountain front. _Two years ago, while examining the southern portion of the Monument Creek area, I obtained from the conglomerate four miles northwest of Calhan, the distal end of a large humerus which Dr. F. A. Lucas has identified as Titanotherium. This conglomerate is the upper member of the formation and caps along line of buttes and extensive plateaus. A number of | * Loc- cit., p. 39. 180 Darton—Age of the Monument Creck Formation. bones have also been collected for me along the valley of Cherry Creek, half way between Castle Rock and Elizabeth, consisting mainly of bones of titanotherium. They were obtained at many localities and all from the sandstones of the upper member of the formation. A fragment of a lower jaw of titanotherium was the most distinctive fossil obtained. It was found in the upper beds, at Kaumpfer’s ranch, 7 miles southwest of Elizabeth. In Wild Cat Canyon, 6 miles west- by-south of Elizabeth, were found fragments of a jaw and the distal ends of a titanotherium tibia and humerus. Portions of a lower jaw of hyracodon, apparently nebrasenszs, were found in a well at Anderson’s place 6 miles south-southwest of Eliza- beth, together with various turtle bones. All of this material appears to have been obtained from the upper beds and it cor- relates these beds with the Chadron formation of the White River group, or Oligocene. No evidence was obtained as to the age of the lower member, but the fullers earth, as before mentioned, is similar to that which is so characteristic in other areas. The presence of the unconformity between the upper and lower members suggests that the latter may be of Wasatch or Bridger age. The nearest locality to the Monument Creek area, at which Oligocene deposits occur in eastern Colorado, is im the vicinity of Akron and Fremont’s Butte, where titano- therium remains occur in abundance. Farther north, in the region about Pawnee Buttes, there are well-known localities of the titanotherium and overlying beds. In the low inter- vening area, east and southeast of Denver, Oligocene deposits: are absent, but it is probable that originally they extended con- tinously from the vicinity of Akron to the foot of the Rocky Mountains in the Monument Oreek area. There is much evi- dence throughout the Great Plains region that the Ohgocene deposits were originally of wide extent, for outliers occur along the mountain slopes and in many widely separated areas. ‘They have been. subjected to extensive degradation in Miocene, Pliocene and later times and probably removed from large districts, especially in the wider valleys. In my recent report on the Great Plains,* there is given a map showing their pres- ent distribution and probably former great extent. * United States Geological Survey, Professional Paper No. 82, pl. xliv. tue Moody—Todometric Determination of Aluminium. 181 Arr. XXI1.— The lodometrie Determination of Aluminium in Aluminium Chloride and Aluminium Sulphate; by S. E. Moopy. [Contributions from the Kent Chemical Laboratory of Yale Univ.—cxxxviii. | A process for the gravimetric determination of alumina in salts of aluminium has been described by Stock,* who bases the method upon the reaction represented by the following equation : Al,(SO,), +5KI+KIO, +3H,O = 2Al(OH),+3K,SO,+61 This equation would show that iodine is liberated when potassium iodate and potassium iodide are together added to a solution of aluminium sulphate. It was found, however, that in the action of the iodide-iodate mixture upon a solution of potassium alum, only about two-thirds of the iodine corre- sponding to the aluminium salt is accounted for ; and this sug- gests that the reaction is not completed according to the equation, and that the precipitate formed is not the simple hydroxide. Upon ignition the precipitate yields, however, the total amount of alumina present; and, since the character of the precipitate is good, the process is easily managed and gives, as Stock has said, an excellent gravimetric method for the determination of alumina. Taking aluminium chloride, AlCl,.6H,O, and proceeding in the same manner, similar results are obtained, and after dis- solving the precipitate in sulphuric acid and adding silver nitrate to the dilute solution, a decided precipitate of silver chloride is observed, which upon washing, drying and weigh- ing is found to be about one-third of the amount of that sub- stance corresponding to the original aluminium chloride. This indicates that it is an oxychloride which is formed on the addi- tion of potassium jodide and iodate ; moreover, upon removing by sodium thiosulphate the iodine first set free in the action and allowing the mixture to stand, progressive hydrolysis takes place as shown by the return of color due to iodine, and this change can be still further hastened by heating after adding an excess of sodium thiosulphate to take up the iodine as lib- erated. The attempt was made, therefore, to complete the reaction between the iodide-iodate mixture and the aluminium chloride, or alum, by heating the solution in a Voit flask through which steam or, still better, hydrogen was passed, as an aid in the transfer of the iodine liberated to a receiver * Ber. Dtsch. Chem. Ges., 1900, xxxiii, i, p. 548. 182 Moody—lLodometric Determination of Aluminium. charged with a solution of potassium iodide. The iodine col- lected was titrated with = sodium thiosulphate. Table I gives results obtained by this method. The details of the experiments in which steam was used as an agent to force the iodine over are given in section A, while those of the experiments in which hydrogen was employed are indicated in section B. TABLE I, Approx. a | ae Alumini x 2 chloride | HIOs.| KI. | Time in APPYOX. jo/calculated| Diff. solution. grm. grm. minutes. | Na.S.Os. grm. grm. em?. em?, A 25 O°3 1°0 25 25°05 0°0427 |—0°0007 25 0°3 1°0 90 25°15 0°0428 |—0-°0006 B 25 0°3 1°0 20), 25°05 0°0427 |—0-0007 25 0'3 1:0 15 25°10 0°0428 |—0°0006 25 0'3 1:0 15 25'00* | 0:0426 |—0-:0009 25 0°3 1°0 ae li) 25°00* | 0:0426 |—0:0009 In each of these experiments the iodide-iodate mixture was made by exactly neutralizing iodic acid with potassium hy- droxide, adding a minute erystal of the iodic acid, introducing the potassium iodide in solution and taking up with a drop or two of sodium thiosulphate the iodine set free. This mixture was put into the Voit flask together with the aluminium chlor- ide, and the whole was heated in the current of steam or hydrogen. | Applying the process to a solution of potassium alum the results recorded in the following table were obtained. TABLE II. n Approx. 10 noone A a: Aluminium |KIO;.| KI | Time in * 10} caleu- | Al,O; Diff. potassium | grm.|grm. | minutes. | Na2S.O3. | lated. ‘found. grm. alum. emi: erm. | grm. cm?, | 25 One ca tc Ohs 30 24°55 |0°0410,0°0414; —0°0004 25 O73) lc 0 30 24°60 |0°0411/0°0416| —0°0005 25 Oar Oe 25 24°50 |0°0409/0°0414; —0:0005 25 Oo akon 30 24°70 =|0°0413/0°0416) -—-0°0003 25 Or3 te ae@ 35 24°50 |0°0409,0°0415| —0°0006 25 O34 120 30 24°55 = |0°0410/0°0415) —0°0005 29 VN Osa ae lO 25 24°50 |0°0409'0°0415| —0°0006 * New standard. Moody—TLodometrie Determination of Aluminium. 183 In Table III are shown results of the application of the process to an ammonium alum. TaBLeE III. Approx. Ss n Ammonium | KIOs. Ge Time in APDEOS: 10 alum. grm. grm minutes. | Na.S2Oz. em?. eme: 25 0°3 1:0 20 25°20 25 0°3 1°0 15 25°17 25 0°3 1:0 20) 25°10t 25 0°3 1°0 25 25°20 95" 0°3 1:0 12 25°70+ 25s 0°3 1:0 12 24-65+ 25 0°3 1:0 20 25°20f 25 0°3 1:0 25 25°15] 25 0°3 2°0 25 25°20f 25 0°3 2°0 20 25°15} Al 2 O 3 calculated erm. 0°0429 0°0429 0°0427 0°0429 0°0421 0°0420 0°0430 0'0429 0°0430 0:0429 Diff. germ. +0:°0007 + 0°0007 +0°0005 | +0:°0007 —0°0001 —0°0002 + 0:0008 + 0°0007 + 0°0008 +0°0007 These results proved to be too high and led to the conclu- sion that the ammonium sulphate was acted upon by the iodic mixture, liberating an additional portion of iodine. ments with ammonium sulphate verified the supposition, and the process is, therefore, less accurate in the presence of ammonium salts. the solution. Experi- In fact, ammonium sulphate in the amounts taken may be completely hydrolyzed in the course of three hours, about one-half of the iodine liberated by the sulphuric acid formed in the hydrolysis being available for estimation under the conditions of the foregoing determinations. however, the distillate is collected in a solution of potassium iodide containing sufficient acid to combine with the ammonia volatilized, iodine is liberated in amount equivalent to the entire quantity of sulphates present, and may be titrated with sodium thiosulphate. The reaction between iodine and ammonia in alkaline solu- tion, and the hydrolysis of ammonium salts, are undergoing further investigation by the writer. The attempts to obtain a complete reaction by heating the mixture in a pressure bottle showed that the results of this pro- cedure are low, although but slightly deficient, and cannot be used for estimating correctly the amount of aluminium salt in When, The following method can be recommended as one giving constant results which correspond closely with the gravimetric * Liquid in Voit flask not clear. + New standard. ¢{ New standard. 184. Moody—TLodometric Determination of Aluminium. determinations and the theoretical amount of alumina in neu- tral aluminium chloride, sulphate, or alum, little time being necessary for a single determination : Measure 25°™* of the approximately 3 solution of the neu- tral aluminium salt to be analyzed into a Voit flask and to this add a mixture of 10°™* of a solution of neutral potassium iodate (30 grms,. to a liter) and 1:0 grm. potassium iodide. Pass a cur- rent of hydrogen through the liquid and heat for fifteen to twenty-five minutes, or until the solution is nearly colorless, collecting the iodine liberated in a Drexel flask, about half full of water, in which 3 grms. of potassium iodide is dissolved. Titrate with sodium thiosulphate the iodine in the Drexel flask and that which remains in the solution in the Voit flask, and calculate the amount of alumina, Al,O,, corresponding to the iodine, 61, liberated. The writer wishes. to thank Professor F. A. Gooch for friendly assistance during this investigation. R. A. Daly—Secondary Origin of Certain Granites. 185 Arr. XXIII.—The Secondary Origin of Certain Granites ; by Reeivatp A. Day, Ottawa, Canada. [Published by permission of the Chief Commissioner for Canada, Interna- tional Boundary Surveys. | CONTENTS. General thesis of the paper. A. The Sills of the British Columbia (International) Boundary. The Moyie Sill. Field Hypothesis. B. Occurrences in Minnesota. (a) Pigeon Point. (6) Governor’s Island. (c) Lake Superior islands and Logan sills. (d) Cook County, Lake County and other localities. C. The Sudbury intrusive sheet. Synthetic discussion. Magmatie assimilation. Summary, Asymmetry of the intrusive bodies. Magmatic Differentiation. General Application. General thesis of the paper.—Igneous rocks originate in magmas. ‘The discovery of the laws governing the immediate derivation of such rocks from their parent magmas is, there- fore, not the final aim of the geologist. He is logically com- pelled to refer rocks themselves to the yet more fundamental problem of the origin of igneous magmas. Whence come the raw materials of basalt, gabbro, porphyry or granite ? One of the earliest answers to this question has been grad- ually assuming a systematic statement in the form of the ‘assimilation theory”. This theory holds that some igneous rocks are derived from the compound magmas formed by the local fusion of solid rock in molten rock of a different chemical composition. The process can be imitated in the laboratory furnace, and has certainly operated on many igneous contacts in nature. Yet one of the very latest utterances of one of the world’s greatest petrologists reads thus: “The untenability of the ‘assimilation’ or fusion theory I regard as definitely proved.”* On the other hand, a no less well known authority claims assimilation on a large scale as a necessary stage in the preparation of the Christiania granite.t Brdgger and many of his followers hold that the contact phenomena of this granite show that the assimilation theory breaks down even when applied to a most favorable case. * **Die Unhaltbarkeit der ‘ Assimilations’- oder Hinschmelzungs-Theorie betrachte ich als endgiiltig bewiesen.”—J. H. L. Vogt, Die Silikatschmelz- lésungen, Part II., Christiania, 1904, p. 225. +F. Loewinson-Lessing, Comptes Rendus, 7th Session, International Geo- logical Congress, 1899, p. 369. Am. Jour. Sc1.—FourtTH Series, VoL. XX, No. 117.—SEPTEMBER, 1905. 13 186 R&R. A. Daly—Secondary Origin of Certain Granites. This divergence of view is, of course, due to the lack of definite knowledge of the vital conditions controlling the activities of such an intrusive body as the Christiania granite. The study of its accessible contacts can, of itself alone, furnish neither proof nor disproof of the doctrine of wholesale assimi- lation. Without the aid of other geological data the attempt - to solve the problem is like the attempt to produce graphically a complex curve of which but two points are known and fixed. Deep-seated assimilation about any magma chamber can only be finally discussed and evaluated if the complete form of the chamber and the complete composition of its rock-filling are at least tolerably known. The present paper furnishes a brief discussion of a number of cases where it is believed that magmatic assimilation on a com- paratively large scale has taken place. It is believed, further, that the geological conditions in these cases supply elements generally untouched in earlier discussions of the doctrine. The original magma had the composition of a gabbro intruded in the manner of sills; the invaded formations are ancient sand- stones, both normal and feldspathic, with associated argillites or schists; the invaded formation, in every case, is more acid than the gabbro; the product of assimilation is always a granite graduating into granophyre. The acid magma is believed, however, to have been derived indirectly from the compound magma of assimilation through a systematic kind of differen- tiation. The primary cause of the differentiation is referred to the perfect or nearly perfect density stratification of each magmatic chamber. The result of the investigation has been to conflrm the writer’s general theoretical conclusions on the subject of assi- milation where it was necessarily introduced among the tests of the hypothesis of magmatic stoping*. Assimilation and differentiation are not antagonistic processes ; both of them are involved in the secondary origin of some granites. A. The Sills of the British Columbia (International) Boundary. During the field season of 1904 the writer developed a geo- logical structure section along the 49th parallel of latitude between Port Hill, Idaho, and Gateway, Montana, the two points where the Kootenay River crosses the boundary line between Canada and the United States. It was found that the mountains traversed by the section are for the most part composed of two very thick siliceous sedimentary formations which, in all probability, are of pre-Cambrian age. The two are conformable. The lower formation has been called the Creston quartzite. It is a remarkably homogeneous, highly indurated light- to medium-gray sandstone, generally thick-platy in structure but *This Journal, xv, 269, 1908, and xvi, 107, 1903. R. A. Daly—Secondary Origin of Certain Granites. 187 oceasionally interrupted by thin intercalations of argillaceous material. The formation is generally composed of nearly pure quartz with a little mica, but some bands are feldspathic toa notable extent. The total thickness of the formation is at least 9900 feet in the vicinity of Port Hill; its base was not directly observed. Immediately overlying the Creston quartzite is the conform- able Kitchener quartzite, composed of about 7400 feet of a highly ferruginous indurated sandstone. This formation is, in the field, distinguished from the Creston quartzite not only by the rusty color of the outcrops but also by a relatively thinner bedding and a greater proportion of micaceous cement, once somewhat argillaceous. Individual beds of the Kitchener quartzite are charged with detrital feldspar, but the formation as a whole is essentially composed of cemented quartz grains. Dark-colored red, brown, and gray shales with thin inter- ealations of gray quartzite conformably overlie the Kitchener quartzite. The series, totalling 3200 feet in thickness, has been grouped under the name of the Moyie argillite. This formation appears but twice in the section and then only in comparatively small areas. This great group of formations, from end to end of the sec- tion, has been mountain-built. A few open folds broken by faults appear in the eastern half of the belt, but the deforma- tion has generally been due to the tilting of monoclinal blocks separated by strong normal faults and, more rarely, by thrusts. The tiltmg ranges though all angles up to verticality, but the average dip is less than forty-five degrees. In consequence of the deformation and subsequent denudation the edges of some 20,000 feet of well-bedded ancient sediment are now exposed for study. There have also come to light a number of thick sills of gabbro intruded at various horizons into the Kitchener quartzite and the upper part of the Creston quartzite. The intrusion and erystallization of the gabbro is believed to have taken place before the upturning of the sedimentaries. The faulting and tilting has repeated the outcrops of certain of the sills. One of the thickest of them has, along with the quartz- ites, been warped into one of the rare synelinal folds. The thickness of the sills varies from 100 feet to more than 2500 feet. The main mass of each sill was uniformly found to consist of a hornblende gabbro with essential green (primary) horn- blende and plagioclase (labradorite to anorthite, the latter in the cores of occasionally zoned feldspars). Accessory quartz, often in considerable amount, always accompanies the other accessories, which are titanite, titaniferous magnetite, and apatite with often a little biotite and sometimes a little orthoclase in addition. Epidote and chlorite are the principal secondary 188 LR. A. Daly—Secondary Origin of Certain Granites. minerals. The structure of the rock is typically hypidio- morphie-granular. Already in those sills that range from 400 to 500 feet in thickness, the gabbro is acidified near its upper contact. The change from the normal composition is seen in the great increase of biotite, orthoclase, microperthite and interstitial quartz. 2 Fic. 1. Map of Moyie Sill, taken from plane-table sheet of the Interna- tional Boundary Commission. 1. Moyie argillite. 2. Kitchener quartzite. 3. Hornblende gabbro sill. 4. Acidified (granite) zone of sill. 5. Creston quartzite. 6. Alluvium. Conventional sign for strike and dip. Scale: one inch = about one mile. Biotite and quartz then assume the proportions of essential minerals. The quartz is characteristically in poikilitic relation to all the other constituents except orthoclase and microper- thite, with which it is m true micrographic intergrowth. From this micropegmatite-bearing phase of the intrusive there is a gradual transition to the normal gabbro which thus composes the lower three-fourths or four-fifths of the sill. R. A. Daly—Secondary Origin of Certain Granites. 189 The Moyie Sill.—The acidification of the upper zone of the gabbro being generally in a direct ratio to the strength of the sill, the phenomenon is specially marked in the greatest of all the intrusions. On account of its importance both in size and character, this rock-body is called the “ Moyie Sill,” the name referring to its situation on the Moyie River. A map and sec- tion of this sill are given in figs. 1 and 2, which illustrate one of the fault-blocks so characteristic of this part of the Boundary belt.* The sill is rather more than 2500 feet in thickness. It follows the bedding of the Kitchener quartzite, which here dips about sixty degrees to the eastward. The intrusive mass 1s seen to be cut off at its northern end by a master-fault which has dropped the Moyie argillite down into contact with the Fie. 2. Section of Moyie Sill, along line of the International Boundary. gabbro. This faulting is believed to have occurred after the sill-intrusion. There is a complete lack of contact metamor- phism in the argillite where it adjoins the gabbro. Since the Moyie sill, throughout the six miles of linear out- crop studied, is in intrusive contact with the Kitchener quartz- ite alone, the other sedimentary formations need not here be described in detail. The Kitchener quartzite is, on the whole, a homogeneous terrane. Ona fresh fracture the rock is seen to be a fine-grained, vitreous, light to darkish gray, well-bedded but tough, metamorphic sandstone, splittmg with some readi- ness along the darker colored layers. The rusty color of the joint-surtaces and bedding planes is due to the leaching out and subsequent deposition of the iron contained in the pyrite, mag- netite, etc., disseminated through the rock. Under the microscope the rock is always seen to be essen- tially a fine-grained aggregate of interlocking quartz grains, seldom showing any direct traces of their detrital origin. The quartz mosaic is, in every thin section, shot through with abundant crystals of biotite which is often developed in pheno- eryst-like individuals occasionally as much as one centimeter in diameter. Sericitic muscovite is seldom absent as an essen- tial, and sometimes rivals the biotite in abundance. Only * All the line-drawings used in illustration of this paper have been made for the most part by the aid of a typewriter, provided with a few special keys. The machine permits of a great saving of time in the preparation of the manuscript drawings. Cf. this Journal, vol. xix, 1905, p. 227. 190 &. A. Daly— Secondary Origin of Certain Granites. rarely is feldspar essential; in one slide it seems to compose ten to fifteen per cent of the rock. So far as observed, the feldspar of the staple quartzite is orthoclase. No sodiferous mineral has been certainly determined in the rock. Epidote, zoisite, titanite, magnetite, leacoxene, pyrite and zircon, besides chlorite, secondary after biotite, are the other, always subordi- nate, constituents. In marked contrast to the normal quartzite is the rock col- lected at a point thirty feet from the upper contact of the Moyie sill. It is a very hard, vitreous, massive, light bluish gray quartzite carrying much feldspar. The whole rock seems to have been recrystallized. The granular-mosaic strue- ture has been largely replaced by poikilitic and micrographic structures. Quartz is thus either regularly intergrown with feldspar or else encloses non-oriented individuals of the same mineral. The feldspar proved to be orthoclase, albite and microperthite, named in the order of their relative abundance. Biotite and sericitic muscovite are, as usual, in considerable amount. = Oligoclase, Abb; Anis.) 222-24 Sol a) 2 one Soda-bearing ortho- } es a1 ae clase | Microperthite -2-. .:2- Je ORAS og ea Sows Quartye See oo ee 4:0 6°38. 22°8 1l-7 57-1) AGsOReasleG NGUISCOR ANG. 4 Ses oak el oe kl ele) 8B ae ace AWavilewen = ee men or Oe cnn ieee tint ac Sic 3) "2 Titanite 3.2 2. lec 4, 80 8 Ie a MBenenler =a we 2 an ee ei SN 281) Dee Chionitesa see eee wee LEO cee a eer Calette tir ce ee | Ue bt a Se eee A) eae Total is 100 in each ease. 1, Normal unacidified gabbro from sill about eleven miles east of the Moyie sill. Nos. 2 to 7 inelusive are types from the Moyie sill, specimens taken thus : | 2. Thirty feet from lower contact. Two hundred feet from lower contact. Two hundred feet from upper contact. Fifty feet from upper contact. Forty feet from upper contact. Fifteen feet from upper contact. ~T O> Or H CO Table II was constructed by the use of the Rosiwal method for the determination of the relative quantities of the different constituents. The values are only approximate, owing to the difficulties of exact measurement and identification of the min- eral grains. No account was taken of the sometimes abundant grains of epidote, occasional grains of calcite (measured in one instance), and often rather abundant scales of kaolin which occur in the slides. These minerals are products of the altera- tion of the feldspars, that alteration affording another diffi- culty in using the Rosiwal method for this suite of rocks. The proportions of the micas are probably too high on account of their not being even approximately equidimensional. Though these rocks do not lend themselves to a very satisfactory employment of the method, and though the table cannot be considered as accurate, the strong contrasts between the acid and basic phases of the sill are clearly evident. * Verh. Wien. Geol. Reichs-Anst., vol, xxxii, 1898, pp. 148 ff. R.A. Daly—Secondary Origin of Certain Granites. 198 Since the compositions of the hornblende, biotite and soda- bearing orthoclase are not known, the chemical analyses can- not be caleulated from Table II. Direct chemical analyses of types Nos. 1 and 7 in Table II have been made by Professor Dittrich, of Heidelberg, and are recorded in Table III. TaB_eE III. 1. 2. SHOR nn kee os Cee 51:92% 71-69% HO eles Pe eo wee a Bea "83 59 POROR eco ret re oe oe, 14°13 13°29 eR OMe re gue Ne. 2°97 83 eOue . eee ra ee GROD 4°23 iO putes oe ee ee hE 14 09 Me Ome ew. 1 809 1:28 CVO) eae Ver, ete a eee Nee} 1°66 INLD O) ci ae a ra 1°38 2°48 TEC Oe ila a a ee “47 DBT tee O ibelow, 1102 'C,) 7. J - 10 14 EO (above 110% ©.) 222. LO 7 BIL PAO) ie ee ene ee 04 O07 CO... 3 ae 06 13 99°78 100°16 SDs CE ae ee 3°000 2773 1. Normal unacidified gabbro from sill about eleven miles east of the Moyie sill. 2. Acid rock fifteen feet from upper contact of the Moyie sill. The rock of col. 2 belongs to the granite family. The silica is normal ¢higher in types of cols. 5 and 6, Table I), but the total of the alkalies is extraordinarily low, namely 4°85 per cent, or °76 per cent lower than the total of the potash and soda in the least alkaline among the twenty-six types of granite analyses selected for Rosenbusch’s “Elemente der Gesteins- lehre.’ The comparatively high content of lime is probably to be referred to a not unimportant mixture of lime feldspar and alkaline feldspar in isomorphous relation, as well as to a small amount of secondary epidote. Col. 1 shows the gabbro to be a normal type in some respects, but the high content of silica and relatively low content of alumina and soda are abnormal for gabbro. These features are partly due to the predominance of hornblende over feld- spar and to the presence of free quartz. It can be seen by inspection of cols. 1 and 2, Table Ul, that the gabbro at the bottom of the Moyie sill would give an analysis very close to that of col. 1, Table III, which represents a good type of the 194 R&R. A. Daly—Secondary Origin of Certain Granites. average gabbro from the many sills of the Boundary belt. A comparison of cols. 2,4 and 7, Table II, shows that col. 4 corresponds to a rock-type intermediate between the two types actually analyzed. It is planned that a rather complete set of total analyses of the various phases of the Moyie sill will be published in the final report of the Chief Commissioner for Canada on Boundary Surveys. Partially absorbed inclusions of the quartzite occur also in the upper, granitic zone of the intrusive. Next to the peculiar granite-granophyre is a hundred-foot (thick) zone of intermediate rock which, with rapid transition, 3 Fic. 3. Photograph of specimens showing contrast of color between a basic and a normal phase of the gabbro of the British Columbia sills and between both of these and two phases of the Moyie Sill granophyre-granite shown on the left. replaces the acid rock as the section is thus carried inwards through the sill. The mineralogical composition of this inter- mediate rock is shown in Tables [ and II. The structure is again hypidiomorphic-granular with con- tinual gradations into the granophyric. The grain varies from medium to rather coarse. The intermediate rock grades imperceptibly into the normal gabbro of the internal part of the great intrusive body. The variation in mineral composition among the zones of granite, intermediate rock and gabbro are shown in Table II. R. A. Daly—Secondary Origin of Certain Granites. 195 The profound macroscopic differences of aspect are imperfectly illustrated in fig. 8, which shows the variation of color-tint. The corresponding variations in the specific gravity of speci- mens taken in the cross-section of the sill is shown in the fol- lowing table: Locality of specimen. Sp. gr. 15 feet from upper eye: Bpabee Pele teps A APs 2°773 40 66 66 Rena CCE ae Meee ee PAT Caer Mat 2°784 50 ce ce 6¢ ce emer Ae we ea lg OO (() Average for granite zone about ---- ---- 2°790 280 feet rom) upper contact... 22. 22-+ 2.2. 3°020 Average for middle of sill about _-_----- S020 au0nteet trom lower contact... _)..-.-..-+- 2°967 30 feet “ bs “ eh Tic ud Se heh GRASS A series of determinations showed in addition that the average specific gravity of the normal gabbro in all the siJls of the Boundary belt is about 3-020. Exomorphic contact action was observed at both upper and lower contacts with the Kitchener quartzite. It has taken the form of increasing the already high induration of the sedi- ments with an accompanying special development of biotite at both upper and lower contacts. Though there is evidence of the feldspathization of the quartzite at the upper contact, none has yet been forthcoming for the lower contact, where, neverthe- less, feldspar may have been similarly introduced from the magma. Doubtless on account of the chemical nature of the invaded sediments contact metamorphism is not conspicuous in the field, nor is it easy to trace its influence. The writer’s impression is that the effects are more manifest, the action haying been more intense, at the upper contact than at the lower, but additional field study will be required to test the real truth of that impression. Apart from the development of exotic feldspar in the quartz- ite, dications of true pneumatolytic action seem to be lack- ing at both contacts. Mineral veins, including quartz veins, except occasional stringers of quartz, are conspicuously absent. field Hypothesis.—The hypothesis adopted in the field to explain these rocks and their relations involved a secondary origin for the granite-granophyre zone at the top of the sill. That zone was thereby interpreted as due to the contact-action of the gabbro intrusion on the adjacent Kitchener quartzite ; digestion and assimilation of the sediments both on the main or “molar” contacts and on the peripheries of blocks shattered off from those contacts, was credited with the formation of a uew compound magma from which the highly acid and some- what anomalous granite was derived. The tact that the acid 196 &. A. Daly—Secondary Origin of Certain Granites. rock is practically confined to the upper contact-zone was explained by the collection of the products of digestion at the upper contact by gravitative adjustment in the magma. The low density of the locally formed new magma of assimilation would tend to effect its upward diffusion and the consequent cleansing of the heavier gabbro magma from such acid material. The comparatively slight acidification at the lower contact was attributed to the solution of the quartzite in the period imme- diately preceding the final consolidation of the sill; at that time the viscosity of the magma was too great to allow of the upward diffusion. A principal test for such a hypothesis is obvious. If it be true, there should be other examples among the great basic sills cutting siliceous sediments. It has already been noted that there is actually such acidification of the other gabbro sills encountered between Port Hill and Gateway, and that m them the acidification is always most marked at the upper contact. Much more striking examples have been described with unusual thoroughness in Minnesota and Ontario. The comparison of these other cases is so important that the best established types will here be sketched and illustrated in some detail. The fur- ther discussion of the Moyie sill will be postponed to later pages, in which a synthetic treatment of all the examples will be undertaken. B. Occurrences in Minnesota. (a) The very able and specially detailed memoir of Bayley on the rocks of Pigeon Point contains, doubtless, the most elaborate argument in favor of the secondary origin of some granites. A brief summary of his facts and conclusions may well be given in the works of Bayley’s own outline forming the introduction to his paper. ‘Pigeon Point is the northeastern extremity of Minnesota. It is one of a series of parallel points extending from Minnesota and Canada eastward into Lake Superior. Its backbone is a great east and west dike-like mass of a gray, coarse-grained rock that has always been called gabbro. This consists of phenocrysts of plagioclase in a diabasic groundmass of the same mineral, olivine and diallage, and consequently, it is a diabase porphyrite. ... . “The rocks through which the gabbro cuts are evenly bedded slates and indurated sandstones of Animikie age. ‘They dip south-southeast at 15 to 20 degrees, except at a very few places near the contact with other rocks, where they are more or less COMmborteds = 9.940. “The most interesting features in the geology of the point relate to the series of rocks usually occurring between the gabbro and the clastic beds. Beginning on the gabbro side the series R.A. Daly—Secondary Origin of Certain Granites. 197 comprises in succession coarse-grained red rocks, a fine-grained red rock that is sometimes porphyritic and a well-marked belt of altered quartzites. “The fine-grained red rock has all the characteristics of an eruptive. It sends dikes into the contiguous bedded rocks, and consists essentially of a hypidiomorphic granular aggregate of plagioclase, anorthoclase and quartz. The quartz and anorthoclase often form micropegmatite, while the plagioclase is in compara- tively large grains, some of which have hardly defined idiomorphic outlines. At afew places this red rock is porphyritic, with bipyra- midal quartz crystals imbedded in a red granophyric groundmass. The rock is similar to many of the augite-syenites described by Irving as occurring in the Keweenawan series, and is in structure and composition a quartz keratophyre. “The coarse-grained rocks between the gabbro and the kerato- phyre are intermediate in character between these two. The variety nearest the gabbro differs but slightly from the basic eruptive. In addition to the gabbro components it contains a little quartz and red feldspar—constituents derived from the keratophyre. As the latter rock is approached, the augite, olivine, and plagioclase disappear, while increased quantities of quartz, red feldspar, and brown hornblende make their appearance, and the rock becomes more and more like the fine-grained red rock. Finally the hornblende disappears and the keratophyre is reached. Since the intermediate rocks occur only between the gabbro and the fine-grained red rock, and since all gradations in composition between the two end members of the series are represented, the coarse-grained red rocks are regarded as contact products formed by the intermingling of the gabbro and the keratophyre mag- mas.””* After describing the compound external zone of contact metamorphism, Bayley continues : “ From the above-mentioned facts it is concluded that the con- tact belt represents Animikie slates and quartzites that have been altered near their contact with an intrusive rock. The meta- morphism of the quartzites has resulted simply in the recrystalli- zation of the quartz and feldspar of the fragmental grains, with the addition, perhaps, of a little orthoclase. “Since, in several instances, the gabbro is in direct contact with the metamorphosed rocks, while the keratophyre is not to be found in the neighborhood, it is inferred that the former rock and not the latter was the cause of the contact action.” The significant paragraph follows : “Inclusions of fragmentals in the gabbro and the keratophyre have alike suffered the same alterations as have taken place in the various members of the contact belt, with this difference, that quartzite inclusions in the basic rock are often surrounded by a *W.S. Bayley, Bull. 109, U. S. Geol. Survey, 1893, p. 11. 198 L.A. Daly—Secondary Origin of Certain Granites. rim of red rock, identical in all its properties with the kerato- phyre. This suggests that the keratophyre itself may be of con- tact origin.” Finally : “'The conclusion reached is that, in all probability, the kera- tophyre is of contact origin—that is, it was produced by the Eig Sie ke ior Fie. 4. Diagrammatic map of part of Pigeon Point, Minnesota, showing general relations among the different rock formations; after Bayley. 1. Animikie quartzites and slates. 2. Contact zone in the Animikie sedi- mentaries. 3. Olivine gabbro. 4. Intermediate rock. 5. Soda granite and keratophyre. Conventional sign for strike and dip. Secale: nine inches = one mile. fusion of the slates and quartzites of the Animikie through the action upon them of the ‘gabbro.’ The magma thus formed then acted in all respects like an intrusive magma. It penetrated the surrounding rocks in the form of dikes, and solidified as a soda-granite under certain circumstances, and under others as a quartz-keratophyre.”* *Op; cits, p, 12. R. A. Daly—Secondary Origin of Certain Granites. 199 The diagrammatic map of fig. 4 generalizes the field rela- tions as expressed in Bayley’s maps. There have been omitted from the diagram certain complexities in the maps showing the actual geology. The essential features are thus made all the more evident; at the same time it is believed that this arbitrary treatment of the maps does not introduce error in principles. A summary of the mineralogical compositions of the gab- bro, intermediate rock, granite- -keratophyre (granophyre) ‘and invaded sediments is given in Table I. The correlative dif- ferences in chemical constitution are noted in Table LY. TaBLE IV. Selected Analyses, Bayley on Pigeon Pt. A. B. C. D. E. EB. siO, ..--- 49°88% 57:98 72°42 73°85 59°71 70°31 Men. 41-19 1-75 40) “05 tr. tr. mPOr ===. 18°55 15°58 13°04 10°91 18°32 12°81 mie 2. 2-06 Sob 68 6-98 8-11 7-26 BeQ 22 - - 8°37 8°68 2°49 89 "85 8x8) i) 09 13 “O09 Rea oe ae ies Cer == 912 2-01 66 “44 1°05 “50 i "02 O04 15 ae wets eae Lo ao o°77 2°87 58 1°52 3°O4 2°03 a 68 3°44 4°97 1°39 3°43 1°90 IAL... 2°59 356 3°44 2°28 1:93 2°19 2 1-04 2°47 al 1°88 3°24 229 2 16 29 20 oF an ae 100°12 9991 100°33. 100:19 100°18 100°20 Sp. gr... 2°923- circa 2620 notgiven, not 2970 2°740 prob. ca given, 220)": prob: ca 2°75 A. Olivine diabase ; aver. of five specimens. -p. 37 Preairermediate. rock: 2.220208 Ft. aos eee OS C. Red granite ; aver. of 7 specimens._--..--. 56 'D. Unaltered quartzites ; aver. comp. -_.- -_-- 90 ip Wnithreredgnlater 28 ea oh Se se 90 F. Approximation to aver. comp. of sediments 113 The general similarity in the character and spatial arrange- ment of the rocks at Pigeon Point and on the Moyie River is apparent. The comparison of conditions is obscure only as relates to the structural cross-section. The Moyie intrusive is unguestionably a sill. The underground relations of the Pigeon Point intrusive, for lack of decisive field evidence, have not been fixed beyond the possibility of doubt. Bayley Says : 200 LR. A. Daly—Secondary Origin of Certain Granites. “The most prominent features of these gabbro masses are those of dikes. As has already been mentioned, the larger one [the one referred to in the present paper] in many places presents perpendicular walls both to the north and to the south. It occu- pies all the highest portions of the point, and these are in a straight line. It has the appearance of an intrusive mass, and is like any one of those forming the numerous points to the north of the international boundary line. It has been regarded as a dike by both Irving and N. H. Winchell. Its contact with the sedimentary rocks is only occasionally to be seen. At several of these contacts the eruptive has the appearance of having escaped from between the dike walls and thrust itself for a short distance between the fragmental beds, or having piled itself up around the dike orifice and overlapped the intruded rocks. . . . . At only two places on the north shore do the fragmental rocks appear, and at these places they are far below where they should be were they interbedded with the gabbro, and in neither case is the con- tact like that of interbedded eruptive and sedimentary rocks.” He coneludes that : “The larger mass of the Pigeon Point gabbro is in the form of a dike, which has broken through its walls at certain places and intruded itself between the strata of the surrounding rocks.”* In accordance with his view Bayley’s cross-sections show vertical contacts among all the igneous rock members and also between sediments and eruptives. On the other hand, Professor N. H. Winchell states, in a personal letter to the writer : ‘‘ All my observations bearing on the relations of the gabbro to the Animikie on Pigeon Point lead to the conclusion that the gabbro is later than the Animikie. But the term gabbro here is made to include those coarse non-ophitic dikes that resemble gab- bro and which are also allied to diabase. There are abundant places where this rock is in the form of sills in the Animikie. The great backbone of Pigeon Point, which is the most dis- tinctly gabbroid of the intrusive rocks, is simply a large example of a sill, while, as I interpret the structure, many of the dikes cutting the Animikie are only contemporary offshoots from it.” With Winchell’s view there agrees the observation of Bayley that the feldspar phenocrysts of the porphyritic gabbro are sometimes ‘‘arranged in rude layers parallel to the dip sur- faces of the quartzite. Their longer axes are usually in the . direction of the dip of the sedimentary rocks.’+ This orienta- tion suggests flow-structure parallel to contacts. Professor Bayley has, by letter, restated to the writer his conclusion that “the gabbro was intruded as a boss or huge dike, certainly not * Op. cit., pp. 22-28. quOp lt... Parco: RL. A. Daly—Secondary Origin of Certain Granites. 201 as a sill,” but adds the remark that “ while the contacts of the quartzites with the red rock and gabbro so far as they were seen are vertical, it does not necessarily follow that they are vertical with depth.” He continues: “I have no means of knowing the date of the intrusion. With respect to the tilt- ing (of the quartzites) my guess is that the intrusion was prior to the latest tilting, but later than an earlier tilting lakeward.” The possibility that the Pigeon Point eruptive is either a true sill only locally breaking across the bedding of the sedi- ments or at any rate dips as a whole to the south-southeast- ward, is further suggested by the analogy of the many undoubted sills of gabbro cutting the southerly to southeasterly dipping Animikie of Minnesota. Some of these sills have likewise zones of soda granite lying between the gabbro and the sediments on the southerly flank of the gabbro. Thus in those cases the sediments dip under the gabbro on the one side of the eruptive body and away from the granite on the other side. Bayley’s full and trenchant argument for the contact origin of the soda granite and granophyre need not be repeated. The independent origin of the acid rock is rendered highly improbable by the occurrences of the intermediate rock lying directly between the gabbro and the sediments without the intervention of the true granite or granophyre. The efficiency of contact-shattering in aiding the digestion of the slates and quartzites is strikingly manifest in Bayley’s descriptions. *“« Very close to the red rock appears a belt in which the various rocks are in the most complicated relations imaginable. In the -eastern portion of the point this belt is well seen on the southern shore, about one-third of a mile from the end of the point. (See Pl. XVI.) Here the red rock is exposed in low cliffs, and in it are small, sharp slate and quartzite inclusions, into which the red rock penetrates in every direction. The exact line of contact between the red rock and the bedded fragmentals cannot be detected, as they appear to merge gradually into one another, the latter becoming redder and redder as they approach the former, which penetrates them in veins and dikes, and finally includes numerous pieces in such a way as to yield a good eruptive breccia.” “Some of the inclusions are very sharp and but little altered, while others are partially dissolved, and are surrounded by con- centric zones, resulting from the action of the red rock upon the material of the inclusion, and the reciprocal effect of the partially dissolved inclusion upon that portion of the red-rock magma immediately contiguous toit..... Thus it would seem to be a fact beyond controversy that the red rock is the immediate cause of the alteration noticed in the fragmental rocks and of the brec- Am. Jour. Sci1.—FourtTH SERIES, Vout. XX, No. 117.—SEPTEMBER, 1905, 14 202 &. A. Daly—sSecondary Origin of Certain Granites. cia observed along its contact with them. If, however, the con- tact beltis examined very closely, itis found that although the red rock is always accompanied by a zone of this belt, there are localities in which the latter occurs without the presence of the POTMEN ey wee The metamorphosing rock seems to be the gabbro. Just as in the case of the contacts with the red rock, the quartz- ites become mottled as they approach the eruptive, and inclu- sions of the former in the latter are so frequent that there appears to be a gradual transition between the two rocks.’* Similar shatter-breccias are described at the northern con- tact. The metamorphism of the inclusions is there the same in kind as on the southern contact but is less intense.t (>) A significant discovery was made at a mining prospect on Governor’s [sland just south of Pigeon Point. The shaft started in hardened slate at the surface, then struck red quartz- ite and finally red granite where the sinking was discontinued. In this case there is no question that the sediments overlie the granite in a relation similar to that involved in the sill theory of the Pigeon Point intrusive.t (c) Parallels to the Pigeon Point case have also been found on Spar, Jarvis and Victoria Islands.§ Lawson has deseribed other examples among the Logan sills of Lake Superior, and says that the sills are repeated by step-faults gently tilting the sills to the southeast at the maximum angle of five degrees.| (7) Grant describes the great gabbro area of Cook County, Minnesota, as a laccolith in the gently dipping Animikie quartz- ites, slates and graywackes. He maps soda granites passing into alkaline quartz porphyries on the southern flank of the gabbro or init. This latter occurrence is possibly to be related to the occasional horizontal dips of the Animikie.4] N. H. Winchell maps a broad band of the red granite to the southward of the huge gabbro mass of Lake County. He states that the southern limit of the gabbro forms the northern limit of the red granite, but that there are numerous places where these rocks are intricately interbedded and in some instances isolated areas of the red rock are surrounded by gabbro.** The official atlas of the Minnesota Geological Survey indicates still other large-scale examples of the same or similar close relations of gabbro and red granite—notably those mapped in vol. vi, plates 68, 69, 84, 85 and 87. *Op--Citz. 8p. 2o-c0- +: Op cits, op. 8: t Final Report, Geol. Surv. of Minnesota, vol. iv, 1899, p. 516, and vol. v, Te op. cit., p. 30; cf. E. D. Ingall, Ann. Rep. Geol. Surv. Canada, 1888, Pt. H, pp. 45 and 49. | Bull. 8, Geol. Surv. Minnesota, 1893, pp. 80-33-42-44, *| Final Rep. Minn. Geol. Surv., vol. iv, 1899, pp. 328 and 826. ** Tbid., pp. 296-7. R. A. Daly—Secondary Origin of Certain Granites. 208 C. The Sudbury Intrusive Sheet. A still more remarkable parallel to the conditions of the Moyie sill has been rather fully described by Barlow and Coleman, following the earlier work of Walker in their respec- tive memoirs on the geology of the Sudbury District, Ontario. In the scale of the various related phenomena, in the wonder- fully systematic arrangement of the different rock-formations, and in the occurrence of valuable ore-bodies directly and geneti- Fic. 5. Diagrammatic map of part of the Northern Nickel Range, Sud- bury District, Ontario; after Coleman. 1. Granitoid gneiss, greenstones and graywackes. 2. Norite. 935. Inter- mediate rock, transitional between norite and micropegmatite. 4. Micro- pegmatite. 5. Slates, sandstones and volcanic tuffs. (The position of the sulphide ores shown by heavy black line.) Conventional sign for strike and dip. Scale: one inch = two miles. cally associated with the intrusive, the Sudbury District exam- ple stands unique in petrographical records. The latest reports of Barlow and Coleman agree in the con- clusion that the famous nickel-bearing eruptive has the form of an enormous intrusive sill of a composition exactly analo- gous to that of the Moyie sill excepting as regards the develop- ment of the valuable sulphides. It is “a vast sheet of erup- tive rock having a basin shape; a sheet nearly 40 miles long 204. Rk. A. Daly—Secondary Origin of Certain Granites. and 17 miles wide, and probably a mile and half to two miles thick on the average, tf the (average centripetal) dip (of the sheet) is 45 degrees.”* This great sheet cuts sediments and schists referred to the Laurentian and Upper Huronian. Their general field relations are summarized in the diagrammatic map of fig. 5, drawn from a part of Coleman’s official map of the “‘ Northern Nickel Range.” Again the gabbroid rock (norite) is seen to be con- centrated on the lower contact of the sheet, the acid rock, micropegmatite or granophyre, graduating into true granite, on its upper contact, while between the two is a zone of inter- mediate rock. On the ‘Southern Nickel Range” across the spoon-shaped basin, Barlow has determined the same arrange- ment of acid, intermediate and basic zones in the sheet, which there, however, agreeably with the basin theory of structure, has a northerly dip; so that in this case, the norite occurs on the south side of the sheet, the granite-granophyre zone on its northern side. On the basin theory of the structure, the volume of the granite-granophyre in this sheet is to be meas- ured by hundreds of cubic miles. All around the basin the nickel ores form a more or less continuous zone at the lower contact of the norite. The sul- phides are also to be found in especial abundance as segrega- tions in apophysal offshoots of the norite where the basic magma penetrated fissures outside the lower contact of the sheet. Coleman states that where the band of eruptive (outcrop edge of the sheet) is narrow, there is less change in the rock in passing from the lower to the upper contact, the most basic norite as well as ore being absent for the most part. He also notes the absence of granophyre or granite in the smaller intrusions of the norite which occasionally appear outside the main basin.t Eruptive breccias due to the shattering of the invaded forma- tions by the hot magma are found at both upper and lower contacts. Coleman notes that in the northern nickel range the con- tact metamorphism is more intense next the upper acid zone than next the norite. He explains this as possibly due to the fact that the rocks at the lower contact were already well erys- tallized before the intrusion took place, while the sediments along the upper contact were then capable of notable minera- *A.P. Coleman, Rep. Bureau of Mines, Ontario, 1903, p. 277. Cf. A. E. Barlow, Ann. Rep. Geol. Surv. Canada, vol. xiv, 1904, p. 72; also stereogram accompanying Coleman’s report. + 1904 report, p. 212, and 1903 report, p. 286. tA. E. Barlow, op. cit., pp. 122, 129 and plates; A. P. Coleman, Rep. Bureau of Mines, Ontario, 1904, p. 213; T. L. Walker, Quart. Jour. Geol. Soc., vol. liii, 1897, p. 54. R. A. Daly—Secondary Origin of Certain Granites. 205 logical changes. The metamorphism on the upper contact extends outward for a distance of from 1000 to 1500 feet. | There seems to be a decided lack of pneumatolytic action (other than that due to water vapor) incident to the intrusion.* Coleman has coneluded that the intrusion of the sheets antedated the synclinal warping of the region to which the present basin shape of the sheet is attributed.t The mineralogical ‘compositions of the norite, intermediate rock and micropegmatite-granite are summarized in Table L. Their chemical compositions are entered in Table V, taken from Walker’s paper, page 56. The corresponding specific grayities also show the significant homologies existing between these rocks and those of Pigeon Point and of the Moyie sill. The value of a close study of these tables will appear in the following general comparison of the rocks and of their rela- tions to one another. TABLE V. Tt 2. 3. 4, 3. 8i0, eee re AO O0G, 51°52% 64°85% 69°27% 67°76% TiO, Lhe Saree tee 1°39 ieee °78 "46 5/0 Se a 16°32 19°77 11°44 P26 14°00 Be O- Bernat ro EGeHIRS “AT 2°94 2°89 ways re ee i 13°54 6207 6°02 4°51. 5°18 MgO 3 af Eagan ala ee 6°22 6°49. 1°60 ‘O91 oO) od oe ee 6°58 8°16 3°49 1°44 4°28 Na,O cy ED ae Ee 1°82 2°66 3192 SL? eee K,O Re Sees a 2-25: [7 OR. 3°02 ano 1°19 OE eee "76 1°68 “78 “76 lee Ouk Os a) Gia See ley “10 “24: °06 "19 99°03 99°71 98°30 99°35 100°29 = Le ol 3°026 2°832 2°788 2°724 2°709 1 to 5—‘Specimens range from north to south” across the Sudbury intrusive sheet, that is, from near lower contact (No. 1) to near upper contact (No. 5). Synthetic Discussion. Magmatic Assimilation.—The secondary origin of granite has long been maintained by N. H. Winchell, who has referred to the Pigeon Point case as, among others, demonstrating the fact.t Bayley came to the same belief for the granite and granophyre of the point, but did not extend his argument in detail to cover other occurrences among the Minnesota intru- ° sives. On the other hand, the principle has not been accepted * A. E. Barlow, op. cit., p. 129. + 1903 report, p. 277. t Final Rep. Minn. Geol. Surv., vol. v, 1900, p. 62, ete. 206 Lt. A. Daly—Secondary Origin of Certain Granites. as applying to these localities even by Van Hise, whose rare knowledge of Lake Superior geology must give his opinion exceptional weight.* Even the latest text-books of geology give most inadequate treatment of the doctrine though it refers to one of the most important problems in the whole field of geology. Doubtless the majority of petrologists are to-day unfavorable to the assimilation theory of granite and its rela- tives except as it applies to a very limited, in point of volume insignificant, modification of certain magmas at their contacts. Van Hise’s chief argument against the contact origin of the Pigeon Point granite emphasizes the fact that that rock has not the chemical composition either of the sedimentary forma- tion or (as especially shown in the surplus of alkalies and the deficiency of iron in the granophyre-granite) of a direct mix- ture of gabbro and sediments.t The much quoted argument of Brogger with reference to the Norwegian granites is based on a similar fact. Many. other writers have, on a similar ground, excluded contact assimilation as playing any consider- able part in the formation of abyssal or hypabyssal magmas. In practically every case the opponents of the assimilation theory have treated of the assimilation as essentially a static phenomenon. ach interpretation of field facts has been phrased in terms of magmatic differentiation versus magmatic assimilation as explaining the eruptive rocks actually seen on the contacts discussed. Nothing seems more probable, how- ever, than that such rocks are often to be referred to the com- pound process of assimilation accompanied and followed by magmatic differentiation. The chemical composition of an intrusive rock at a contact of magmatic assimilation is thus not simply the direct product of digestion. It is the net result of rearrangements brought about in the compound magma of assimilation. In the magma, intrusion currents, convection currents and the currents set up by the sinking or rising of xenoliths must take a part in destroying any simple relation between the chemical constitutions of the intrusive and invaded formations. Still more effective may be the laws of differen- tiation in a magma made heterogeneous by the absorption of foreign material which is itself generally heterogeneous. The formation of eutectic compounds or mixtures, the development of density stratification, and other causes for the chemical and physical resorting of materials in the new magma ought cer- tainly to be regarded as of powerful effect in the same sense. A second fundamental principle has as a rule been disre- garded in the discussions on magmatic assimilation. The form * Monograph XLVII, U.S. Geol. Surv., 1904, pp. 730-733. + Op. cit., p. 733. ¢ Die Eruptivgesteine des Kristianiagebietes, Pt. II, 1895, p. 130. R. A. Daly—Secondary Origin of Certain Granites. 207 of the intrusive body, and the relation of the accessible points of its contacts to that form as a whole, must be taken into account. If, for example, differentiation of the compound magma has taken place so as to produce within the magma chamber layers of magma of different density, the lightest at the top, the heaviest at the bottom, the actual chemical com- position of the resulting rock at any contact will depend directly on the magmatic stratum rather than on the composi- tion of the adjacent country-rocks. Thirdly, the method of intrusion is of primary significance in the discussion of assimilation in a given instance. There are strong reasons for believing that the subterranean cham- bers of stocks and batholiths have been opened largely or at least in part through magmatic “stoping,” whereby magmas have made their way upward through the invaded formations by engulfing suite after suite of blocks shattered off from those formations by the heat of the intrusives.* In such a case the destructive action at the molar contact is chiefly physical, and chemical solution is subordinate. Most of the solution takes place in the complete digestion of the sunken blocks and is therefore abyssal rather than marginal. The conditions are peculiarly favorable for the systematic differentiation of the new compound magma. The chemical composition of the intrusive at any contact will thus depend on the constitution of a (possibly well differentiated) magma containing materials won from a@// the invaded formations and not simply materials won from the immediately adjacent country-rock. Brégger’s argument derived from the low content of lime in the Chris- tiania granite cutting thick limestones (themselves overlying an enormous thickness of crystalline schists, etc.) is clearly incon- clusive until it can be shown that this and the other two factors just noted have not been at work.t+ Magmatic stoping has, in all probability, taken place to some extent in the great intrusive body at Pigeon Point. The specilic gravity of the gabbro varies from 2°923 to 2°970. Molten at 1400 degrees Cent., its specific gravity, at atmos- pheric pressure, would be not far from 2°43 to 248. The specific gravities of the intermediate rock and granite are respectively 2-740 and 2620; molten at 1400 degrees Cent., they would, at atmospheric pressure, be about 2°30 and 2°19 respectively. The specific gravity of the invaded sediment varies from 2°70 to about 2°75. Blocks of the quartzite and slate immersed in any of the molten magmas and there assum- ing the temperature of 1400 degrees Cent., would at the same *Cf. R. A. Daly, The Mechanics of Igneous ees this Journal, vol. xv, 1903, p. 269, and vol. xvi, 1903, p. 107. + Cf. Loewinson-Lessing, op. cit., p. 368. 208 L. A. Daly—Secondary Origin of Certain Granites. pressure have specific gravities varying between 2°60 and 2°65. There are good reasons for believing that plutonic pressures would not essentially affect these contrasts of density. Assum- ing a certain degree of fluidity in the magma (an assumption underlying the whole of this paper and believed to be demon- strated by such facts as the patent ease of diffusion that once reigned in each of the intrusives), it appears that blocks of the sedimentary rocks must sink in the magma, whether acid or basic.* The actual shatter-breccias described by Bayley are there- fore to be attributed to the last destructive effort of the magma, which, at that time, through cooling, had become too viscous to allow of the sinking of the xenoliths. A precisely similar argument applies to the Moyie and Sud- bury examples (see table of specific gravities and table in this Journal, vol. xv, 1908, p. 277). All these igneous bodies, though ‘not intruded by magmatic stoping, yet show that pro- cess to have assisted in the production of the granites and granophyres. Whether this process has there been more or less efficaceous than molar or marginal assimilation, perhaps cannot be determined. In all these cases the stoping that did occur must clearly have tended to destroy a simple chemical identity between igneous rock and country-rock at any given contact. Summary.—lt will be useful to review the chief field and laboratory observations so far noted as favoring the assimila- tion theory when applied to the granites and granophyres described in this paper. 1. Bayley’s elaborate argument is believed to be valid except as it fails to take differentiation into account. No fact has been noted either by the writer in connection with the Moyie sill or in the descriptions of the other examples which tends to weaken that argument. 2. Belief in the truth of his conclusion is greatly strengthened by the repeated occurrence of essentially the same phenomena in widely separated regions. a. At Pigeon Point, at Sudbury and on the Moyie River there occur intrusive bodies of gabbro passing by gradual tran- sitions (as shown by chemical, mineralogical and specific gravity determinations) into the border phases of granite and grano- phyre. Both types of rock clearly belong to the same period of intrusion. b. All three igneous bodies are of relatively great thickness, which means that, other things being equal, they possessed relatively great stores of thermal energy. *Of. R. A. Daly, op. cit., 1903, p. 277, etc. R. A. Daly—Secondary Origin of Certain Granites. 209 c. In each occurrence the gabbro contains xenoliths of the more acid sedimentary rocks. These blocks are commonly more or less digested and the product of this local solution is always closely allied to, if not equivalent to, the granophyre- granite phase. d. In each case there is correspondence though not equiva- lence between the composition of the acid border-phase and the average composition of the invaded formation. This important fact is emphasized in Tables I, II, III, IV and V, in which the silica and alkalies are either directly or inferen- tially seen to be more or less abundant in the granophyre- granite according to the relative abundance of those oxides in the respective country-rocks. e. A considerable number of other examples not as yet thoroughly studied have been noted in British Columbia and Minnesota. The conditions are throughout identical or so allied as to favor one explanation common to all the occur- rences. jy. The assimilation theory is also supported by certain other facts which have already been mentioned but merit a more detailed discussion such as is attempted in the sequel. 3. The principal objection to the doctrine of assimilation, namely, the objection that chemical analyses disprove any genetic relationship between intrusiveand invaded formation at certain accessible contacts, cannot hold, because that objection allows no place for differentiation in the magma made compound by assimilation. Asymmetry of the Intrusive Bodies——There remains for particular explanation the cardinal fact that all the intrusive bodies are asymmetric. The granophyre-granite is always con- centrated on one side of the intrusive, that is, along the upper contact or the side away from which the enclosing sediments dip. In all the localities the dips of the sedimentaries and of the intrusive sheets are believed to have been flatter at the time of the injection of the magma than those dips now are. It is, indeed, possible that, in every instance, the gabbro sheet lay practically horizontal during the period of cooling and consoli- dation. In any case the granophyre-granites appears to have always overlain their respective cabbroid associates. Three possible explanations have offered themselves for this asymmetry. (a) It is conceivable that extensive assimilation occurred only on the upper contacts; or (6) the asymmetry may be due to the density stratification of magma compounded of gabbro and digested sediments ; or (c) due to a combination of both those factors. One or more subordinate suppositions are necessary if the assimilation be credited essentially to the upper contact. On 210 R.A. Daly—Secondary Origin of Certain Granites. the one hand, the invaded formations above and below the gabbro might be lithologically so different that the one above was much more subject to contact alteration than the one below. This idea is at once declared irrelevant in the British Columbia and Minnesota cases, where there is certainly no evidence of differences of digestibility. | j On the other hand, it is possible that the original gabbro was differently constituted, and thus more energetic in assimilation, along the upper contact than elsewhere. Magmatic water or other strong solvents may thus be conceived to have early con- centrated in the upper zone of each sill. In favor of this view would seem at first sight the fact that at Pigeon Point the zone of external metamorphism is reported by Bayley to be much wider on the upper contact than on the lower. (See fig. 4.) The same seems to be true in the Sudbury case, but is explained by Coleman as noted on a previous page. The writer could find no absolutely certain evidence of such differential meta- morphism about the Moyie sill, yet considers it as probable. That the conditions for the complete assimilation of the invaded formations obtained throughout the intrusive bodies is illustrated in the unmistakable digestion of xenoliths found at all depths in the gabbro. As already pointed out, the assimi- lation here belongs to the period immediately preceding the consolidation of the gabbro. A much greater volume of sim- ilar material derived from the interaction of gabbro and sedi- mentary rock must have been formed from other blocks in the hotter, more fluid, and more energetic magma of the preceding period. There seems to be no possible doubt that most of that material has diffused upward and now forms part of the grano- phyre-granite zone. The simplest and most probable cause for that diffusion is, as suggested in the field hypothesis for the Moyie sill, the dif- ference of density between the acid magma of assimilation and the enclosing gabbro. It is quite possible that the metamorphosing effect of the new magma may have been greater than that of the original pure gabbro.. The new magma would presumably carry with it the water derived from the digested sediments which, appar- ently in every case, are notably more hydrous than the original gabbro magma. The accompanying table shows the propor- tion of water (or loss on ignition) found in the analyses of the rocks of Pigeon Point. Rock. Per cent of water. Gabbro Peers fee eines ree ee RES yeaa etn Intermediate tocki 2442 2 ee meatier san OAT Red soda eramite tae vs) os Sek ea se 1°21 Seve he ae Ie es ak th sce Rape Wilh Sr, Ben CeO) ak Quartzite (loss on ignition) ..-..-..----- 1°88 R. A. Daly—Secondary Origin of Certain Granites. 211 So far as such water determinations in the crystallized rock can be considered as indicating a true condition of the magma before solidification, the table imphes that the compound magma corresponding to the “intermediate rock” still held the extra water of assimilation up to the moment of erystalliza- tion ; and, secondly, that the well differentiated magma corre- sponding to the soda granite had lost about half of the water of assimilation before final solidification. It is, accordingly, quite possible that this extra water which the upper acid zone could not hold in permanent combination, has been responsible for the unusual amount of external metamorphism in the sedi- ments south of the Pigeon Point intrusive. Similar reasoning may apply to the Sudbury example, but the required elaborate chemical study of its more complex terranes has not yet been made. In favor of this hypothesis is the fact that, so far as known to the writer, differential contact metamorphism of the kind here in discussion has never been described in connection with a sill that does not also show evidence of strong internal assimi- lation. | Finally, there is no cause yet well determined why water or other solvents should be systematically concentrated from the original magma along the roof of an intrusive sill. Such con- centration may, indeed, be the rule, but it has apparently not been announced by any worker among the thousands of basic sills described in geological literature. The conclusion seems justified that the special intensity of the metamorphism on the upper contact of certain imtrusive bodies is probably not due to the special activity of solvents in the original magma along that contact. The explanation seems to lie partly in the different lability of the roof-rocks and floor-rocks to metamorphic change, but yet more in the metamorphic effects of water vapor set free in the digestion of the invaded hydrous sediments. This water vapor may have also assisted in the solvent work of the magma at the main upper contact, and, finally, in increasing the fluidity of that magma. Magmatic Differentiation.—The development of basic, intermediate and acid zones in each of thesills is, thus, believed to have been the result of the density stratification of the com- pound magma of assimilation. The etticiency of differential density in separating out lighter acid material from the heavier basic, has been ably discussed and affirmed by Loew- inson-Lessing.* It isunnecessary to recapitulate his argument, with which the present writer isin full accord. * Op. cit., pp. 344-854. 212 Rk. A. Daly—Secondary Origin of Certain Granites. Loewinson-Lessing points out that when large amounts of foreign rock-material is digested in a magma, there is established a special tendency toward a systematic differentiation of the mixture.* Liquation will then take place when the cooling mixture reaches a certain temperature. The same author also holds that, according to the principles of physical chemistry, a magma becomes actually more fluid as a result of digestin foreign material. Differentiation is thereby faciliated Vogt’s valuable researches tend to corroborate this view.t The granites and granophyres of the Moyie sill, of the Pigeon Point intrusive, and of the Sudbury sheet are to be regarded as not directly or merely due to the contact solution of sedi- mentary rocks and schists by gabbro; they are controlled in their final composition by a common process of differentia- tion supplementary to the gravitative effect. At Pigeon Point the acid rock, whatever its structure and grain, is a rather definite mixture of oxides. This is illustrated in the analyses of granular soda granite, the “quartz keratophyre’, and the porphyry of Little Brick Island near Pigeon Point.{ For. lack of sufficient analyses the same statement cannot be made con- cerning the Moyie sill, but within limits it applies to the huge Sudbury sheet.$ The acid zone may have won some of its soda from the original magma; the gabbro may now hold some of the pot- ash with the silica derived from the micaceous and feldspathic quartzites and other sediments. It is obvious, however, that all the details of the chemical processes engaged in this type of magmatic separation (chemical affinity in magma disturbed by gravitative diffusion currents) cannot be worked out from existing data on the magmatic behavior of silicates. The intermediate rock at all three localities may be regarded as occupying zones of incomplete differentiation. Special interest attaches to the occurrence of the nickel ores along the lower contact of the Sudbury sheet. Barlow, Cole- man, Vogt and Walker agree that these sulphides are soluble in magmas. The solubility is in inverse proportion to the acidity of those magmas.| The fact suggests that the sul- phides have been precipitated from the norite which has been acidified by assimilation. The concentration of the ore, on the lower contact is again the result of differentiation through contrasts of density, the sulphides settling to the bottom of the sheet. LLoewinson-Lessing has already suggested this gen- = Opcis,, PP.-oout, + Op. cit., Part 2. t Bull. 228, U. S. Geol. Surv., p. 89 S$ See A. P. Coleman, 1904 report, P. 218. |J. H. L. Vogt, op. cit., p. 229. R. A. Daly—Secondary Origin of Certain Granites. 218 eral hypothesis to explain the segregation of sulphide-ores, without, however, connecting the concentration with oravita- tive influence. Coleman has announced the view that the ores have thus settled to the bottom of the sill, but has not connected the action with the digestion of acid rock in the norite.* The whole array of facts connected with the Sudbury intrusive is so accordant with the double theory of assimilation and differentiation through density stratification, as to single out this particular case as perhaps, of all those noted in the present paper, the most convincing and illuminating. It is necessary that brief reference be made to an alternative view of all these related phenomena. One may conceive that the granite-granophyre, intermediate rock and gabbroid rock in each of the intrusive sheets may be explained by simple differentiation from an orzginal magma through density strati- fication but without the aid of significant assimilation of the country-rocks. Lack of space forbids that this hypothesis be here discussed at length. The writer believes that the hypoth- esis is untenable or, “at least, is much less adapted to explain- ing the facts than the hypothesis of assimilation accompanied and followed by differentiation. Most of the facts on which that belief is founded have been already implied or expressly noted. Among the significant facts are the following: 1. There is a close similarity in composition between the granite-granophyre zone and rims of manifest digestion about xenoliths now surrounded by gabbro. ‘This consanguinity is inexplicable on the theory of mere differentiation within the original magma. 2. The genetic relationship between the granite-granophyre zone and the invaded sediments is further shown by certain special features already described among the structures of the acid rock in the Moyie sill, and of the overlying, metamor- phosed quartzite. For example, the development of remark- ably poikilitic quartz im the granite-granophyre and in the recrystallized quartzite (the quartz of the latter being largely or wholly indigenous) may be mentioned. This repeated occurrence of a peculiar structure finds no simple explanation on the pure-differentiation theory. 3. In the period of high temperature preceding the viscous period when the visible xenoliths were frozen in the gabbro, thousands or millions of other xenoliths were completely or in part digested in the gabbro magma. The product of their digestion can be found, apparently, i in no other place than in the existing acid zone of each intrusive sheet. * 1903 report, p. 277. 2 fi AY Daly—Secondary Origin of Certain Granites. 4. Mere differentiation of an original magma (through density stratification) cannot readily explain the slight but certain excess of silica along the lower contact of the Moyie sill. That degree of acidification is readily understood on the assimilation theory. | 5. Along the British Columbia boundary a large number of contemporaneous gabbro sills of practically identical min- eralogical and chemical composition have been found. In most of these no true granite-granophyre zone occurs. The composition of these gabbros is essentially equivalent to that of the gabbro in the central part of the Moyie sill; yet, on the pure-differentiation theory we should expect a distinct differ- ence of composition between these other sills and the basic pole of differentiation in the Moyie sill. The assimilation- differentiation theory finds no difficulty in the essential equiva- lence of composition. 6. The assimilation-differentiation theory demands that a great absolute amount of thermal energy be credited to a sill in which secondary granite has been formed; that sill must always be thick. Other things being equal, granite formed by mere differentiation from an original magma should be found also in sills of less thickness, though here again the absolute thickness must be considerable. True granite with the rela- tions described in this paper has never been found as a contin- uous zone in any intrusive sheet 500 feet or less in thickness. On the pure-differentiation theory it is difficult to understand why differentiation should afford true granite in a sheet of the strength observed at Pigeon Point, and should not afford a true granite zone in a sheet 400 or 500 feet thick. The assimilation-differentiation theory readily interprets the fact as due to the relatively enormous amount of heat required for the generation of the granite-granophyre zone, namely, an amount of heat characteristic only of thick intrusive sheets. 7. The pure-differentiation theory has to face another diff- cult question which does not arise if the assimilation-differen- tiation theory be accepted. Why was differentiation in the original magma postponed to the moment of intrusion? This difficulty is, of course, by no means conclusive against the pure-differentiation theory, but it means one more unavoidable theoretical burden weighting the pure-differentiation theory in a way which renders, by contrast, the assimilation-differentia- tion theory one of relative simplicity and, by so much, of greater strength. General Application.—In the foregoing discussion the sec- ondary origin of some granites has been deduced from the study of intrusive sills or sheets; but it is evidently by no means necessary that the igneous rock body should have the R.A. Daly—Secondary Origin of Certain Granites. 215 sill form. The wider and more important question is immedi- ately at hand—does the assimilation-differentiation theory apply to truly abyssal contacts? Do the granites of stocks aid batholiths sometimes originate in a manner similar or analogous to that just outlined for the sills? The writer has briefly noted peneral reasons affording affir- mative answers to these questions.* Gabbro and granophyre are often characteristically associated at various localities in the British Islands as in other parts of the world.t The field relations are there not so simple as in the case of the Moyie sill, for example, but otherwise the recur- rence of many common features among all these rock-associa- tions suggests the possibility of extending the assimilation- differentiation theory to all the granophyres. Harker’s excel- lent memoir on the gabbro and granophyre of the Carrock Fell District, England, shows remarkable parallels between his “laccolite’’ rocks and those of Minnesota and Ontario.{ At Carrock Fell there is again a commonly occurring tran- sition from the granophyre to true granite, and again the gran- ophyre is a peripheral phase. Still larger bodies of gabbro, digesting acid sediments yet more energetically than in the intrusive sheets, and at still greater depth, would yield a thor- oughly oranular acid rock as the product of that absorption with the consequent differentiation. This does not imply, ot course, that all granites are of this origin, but it is quite pos- sible that most intrusive granites are either of this origin or have been more or less modified through assimilation. The difficulty of discussing these questions is largely owing to the absence of accessible lower contacts in the average granite body. All the more valuable must be the information derived from intrusive sills. The comparative rarity of such rock-relations as are described in this paper does not at all indicate the exceptional nature of the petrogenic events signal- ized in the Moyie, Pigeon Point or Sudbury intrusives. It is manifest that extensive assimilation and differentiation can only take place in sills when the sills are thick, well buried, and originally of high temperature. All these conditions apply to each case cited in the present paper. The phenomena described are relatively rare largely because thick basic sills cut- ting acid sediments are comparatively rare. | On the other hand, there are good reasons for believing that a subcrustal oabbroid magma, actually or potentially fluid, is general all around the earth; and secondly, that the overlyi ing solid rocks are, on the average, crystalline schists and sediments * This Journal, vol. xv, 1903, p. 269, vol. xvi. 1903, p. 107. +See A. Geikie, Ancient Volcanoes of Great Britain, 1897. ¢ Quart. Journal Geol. Soc., vol. 1, 1894, p. 311 and vol. li, 1895, p. 125. 216 L&. A. Daly—Secondary Origin of Certain Granites. more acid than gabbro. Through local, though widespread and profound, assimilation of those acid terranes by the gab- bro, accompanied and followed by differentiation, the batholithie oranites may in large part have been derived. * True batho- hths of gabbro are rare, perhaps because batholithic intru- sion is always dependent on assimilation. The argument necessarily extends still farther. It is not logical tu restrict the assimilation-differentiation theory to the granites. The preparation of the magmas from which syenites and diorites, for example, have crystallized, may have been similarly affected by the local assimilation of special rock- formations. The development of some of the anorthosites of the Canadian and Adirondack Archean was possibly condi- tioned on the digestion of part of the associated crystalline limestones by plutonic magma. The officers of the Minnesota Geological Survey have shown that the same magma represented in the soda granite and grano- phyre of Pigeon Point forms both dikes and amygdaloidal surface flows.+ The assimilation-differentiation theory is evi- dently as applicable to lavas as to intrusive bodies. but demonstration of the truth or error of the theory will doubt- less be found in the study of intrusive igneous bodies rather than in the study of voleanoes either ancient or modern. Finaliy, the fact of “ consanguinity” among the igneous rocks of a petrographical province may be due as much to assimilation - as to differentiation. * Cf. Ac Daly, op: eit: +N. H. Winchell, Final Rep. Minn. Geol. Surv., vol. 4, 1899, pp. 519-22. S. L. Penfield and G. S. Jamieson—Tychite. 217 Art. XXIV.—On Tychite,a New Mineral from Borax Lake, California, and on its Artificial Production and its Relations to Northupite; by 8. L. Prnrrerp and G. 8. J AMIESON. Historical.—The new mineral to be described in this paper was discovered by the merest chance in 1895, when some minerals from Borax Lake, San Bernardino County, California, were being studied by one of the present writers (Penfield). At the time mentioned, word had been received from Mr. Warren M. Foote of Philadelphia that he had some unknown minerals from the Borax Lake region, and arrangement was made for their examination in the mineralogical laboratory of the Sheffield Scientific School. One of the minerals, which proved to be a new species, consisted of octahedral crystals, averaging about 3" in diameter, and concerning it Mr. Foote wrote that it was a carbonate of magnesium and sodium con- taining chlorine. The material sent for examination consisted of a large number of the octahedral crystals, and from amongst them a small one, which was perfect in form and seemed to be in every way typical of the lot, was selected for the purpose of making a. few preliminary tests. It was brought in contact with adrop of nitric acid on a watch glass and dissolved with effervescence; the solution gave the flame test for sodium, a minute drop of it gave the reaction for magnesium with ammonia and sodium phosphate, but a test for chlorine with silver nitrate gave a negative result. Think- ing over what else might possibly be present, the idea of a sulphate suggested itself, and a test with barium chloride indicated the presence of the SO, radical. Accordingly, a letter was sent to Mr. Foote informing him that there evidently was some mistake, for the mineral he had sent proved to be a sulphate and not a chloride. This elicited an immediate reply from Mr. Foote, stating that, on the contrary, the mistake was on our part, for he had always obtained the test for chlorine and had repeated the experiment with like results; thereupon the test was repeated by us, and the presence of chlorine was found in one crystal after another. The fact, therefore, was established, that in the material sent there were two minerals erystallizing in octahedrons, one containing the sulphate radi- eal, the other chlorine, and that by chance a erystal of the rarer sulphate happened to be the one first selected for making the initial examination. A preliminary notice of the chlorine compound was published by Mr. Foote,* who named the min- eral northupite after Mr.C. H. Northup of San Jose, California, * This Journal (3), 1, p. 480, 1895. Am. Jour. Sct.—FourtH Series, Vou. XX, No. 117.—SEepTEMBER, 1905. 15 218 S. L. Penfield and G. S. Jamieson—Tychite. who first observed the new mineral and supplied the material for investigation. A complete study of the chemical composi- tion and physical properties of the new compound was subse- quently made by Pratt, who found the composition to be MgCoO,.Na,CO,. NaCl, his results being published in this Journal.* Being assured of the existence of a second, new, octahedral mineral, associated with the northupite, Mr. Foote generously responded to our request to send to New Haven his entire stock of crystals im order that a search might be made for the missing sulphate. The following simple method of testing was employed, which did not in any way injure the specimens: Some dilute nitric acid containing a little silver nitrate was prepared, and with a broom-straw a minute drop of the liquid was applied to each crystal. Thus, if chlorine was present, a little silver chloride would be formed and the drop of liquid would become milky-white. In testing several hundred erys- tals in this way, only two were found which did not give the reaction for chlorine. One of these was a small but perfect octahedron, the other a small cluster of octahedrons, of some- what inferior quality: together they weighed only about 0°10 gram. It was hoped, however, that by sacrificing the speci- mens for chemical analysis sufficient determinations could be obtained for deriving the formula; but in this we were disap- pointed, for, unfortunately, the analysis met with an accident before a single determination had been made. We were thus compelled to abandon the hope of determining the composition of the new mineral until other crystals should be found in new lots of the northupite. Recently our attention was called to the unknown sulphate by observing in the stock of Mr. Lazard Cahn of New York a supply of northupite crystals which he generously loaned to us for examination, but when tested they all proved to be the chlorine compound. Likewise Mr. Warren M. Foote of Phila- delphia has been kind enough to send us his entire stock of northupite, consisting of something over four thousand crystals, among which we had the good fortune of finding one small octahedron, weighing but 0:0109 gram. Curiously enough, this was among the last ten crystals which were tested, and was found after hope of obtaining the desired sulphate had practically been given up. Artificial production.—Beheving that the unknown sulphate would prove to be closely related to northupite, and knowing that de Schultent had succeeded in making the latter arti- ficially, it occurred to us that possibly the wished for sulphate * This Journal (4), ii, p. 188, 1896; also, iii, p. 75, 1897. + Bull. Soc. Franc. de Min., vol. xix, p. 164, 1896. S. L. Penfield and G. S. Jamieson—Tychite. 219 might also be prepared synthetically. Following in general the method of de Schulten, 8 grams of Na,CO, and 34 grams of Na,SO, were dissolved in 120° of water, and to the solution 1-4 grams ‘of MgSO, were added, which immediately produced an amorphoid precipitate, presumably of some basic magnesium earbonate. The mixture, contained in a flask, loosely stop- pered to prevent evaporation, was then heated on a steam bath. By using chlorides in the place of sulphates, as described above, de Schulten succeeded in making northupite in a crystallized condition in about seven hours; in our exper- iment, however, we waited five days, the solution being heated without interruption, before any signs of crystallization appeared. In the meantime we had tried heating a similar mixture in a sealed tube at a high temperature, without definite results, and had practically given up hope ot obtaining the desired crystals. It was almost a matter of accident, therefore, that the flask contaming the mixture was left stand- ing on the steam bath for so long a time. When the crystal- lization had once started, however, it apparently proceeded quite rapidly, and the insoluble material in the flask was almost wholly converted into octahedral crystals, very sym- metrical in development and remarkably uniform in size, about Q-15™" in diameter. Having once produced a crop of crystals, we are now able, by “seeding ” or adding some of the product already formed ‘to a new experiment, to produce crystals in fifteen hours, though it still seems to take several days to complete the reaction. When examined under the microscope, it was found that each crystal contained minute inclusions, presumably of basic magnesium carbonate, but the inclusions constituted a very small proportion of the total bulk of the material. The crystals were next suspended in acetylene tetrabromide, diluted with benzol, and it was found that they all floated when the specific gr avity was 2°594, and on diluting to 2°583 almost all of the material sank. The mean of the two values, 2°588, may therefore be taken as the specific gravity of the mineral. It was found that the lighter crystals, left float- ing on the heavy solution, were preceptibly richer in inclusions than those which sank at 2°583. The crystals are quite hard and give a gritty sensation when ground in an agate mortar. They seratch calcite and probably, like northupite, have a hardness between 3°5 and 4. The crystals are isotropic when examined in polarized light. Using two surfaces which come together at the apex of an octahedron as a prism, it was pos- sible to determine approximately the index of refraction, but the surfaces of the crystal were not good enough to make the determination accurate beyond the second place of decimals: the value found was 1°510, while mn, for northupite was 1514. 220 S. L. Penfield and G. 8. Jamieson—T. ychite. An analysis of the purest material, separated by means of the heavy solution, gives the formula 2MgCO,.2Na,CO,. Na,SO,, the results being as follows: fe ie Theory. SOW Sea eter 15°08 15°06 15°38 CONS oie eee 33°55 33°45 33°72 Mio. ues eer e ses 15°83 15°77 15°33 INaeO Sree eke 30°49 35°65 35°62 99°95 99°93 100°60 The slight discrepancies between the results of the analyses and the theory are probably to be accounted for by the pres- ence in all of the crystals of the minute inclusions mentioned on the previous page. The finely powdered salt does not dissolve to any extent in hot water, nor does it suffer decomposition. Some powder, boiled with water for a considerable time, then filtered and dried, gave the following results :—SO,, found 15-21 per cent, theory 15°33 per cent. The filtrate gave only a shght reaction of a sulphate when tested with barium chloride. Name.—We have named the new and rare sulphate tychite, from tvy7, meaning luck or chance, a name which it well deserves, when itis considered that out of fully five thousand specimens examined, the very first crystal and one of the ten last crystals tested proved to be the sulphate, and only two other specimens were found, the ones lost in an unsuccessful attempt to make an analysis. Comparison of the artificial salt with the natural mineral.— Without question, the artificial salt is identical with the mineral found at Borax Lake: they both contain the same constitu- ents. They crystallize not only in the same system, but also in octahedrons. They are isotropic, although the last crystal of tychite found showed some slight action on polarized light, _ which seemed to be confined only to the exterior portions of the crystal, for fragments from the interior were wholly iso- tropic. The specific gravity of the artificial salt is 2°588, of the erystal examined by Pratt (the analysis of which was lost) 2°456, and of the last crystal found by us 2°30. The last erys- tal, however, contains numerous inclusions, which undoubtedly account for its low specific gravity. As far as can be recol- lected, the crystal examined by Pratt was very white and pure, but not equal in transparency to the artificial crystals. Both Pratt’s determination, 2°456, and ours of the artificial salt, 2°588, are somewhat higher than the specific gravity of northupite, as might be expected from differences in composi- tion: Pratt found the specific gravity of northupite to be [ S. L. Penfield and G. 8. Jamieson—Tychite. 221 2°380, and de Schulten determined that of the artificial salt as 2377. By using two of the faces which meet at the apex of the octahedron as a prism, we have succeeded in determin- ing the index of refraction of the last crystal found. The surfaces of the octahedron were not very perfect, and had to be covered over for the most part, taking the reflections of the signal from only the tip end-of the crystal, and the refraction of light through the same. The value obtained, n,=1-508, compares favorably with that of the artifical salt, 1-510, espe- cially when it is taken into consideration that the condition did not favor exact determinations in either case. A further argu- ment for the identity of tychite and the artificial salt, if any is needed, is that at Borax Lake both tychite and northupite occur together, and were formed undoubtedly under similar condi- tions, while in the laboratory either of these closely related chemical compounds may be made by only varying the condi- tions of the experiment by using sodium sulphate for the one and sodium chloride for the other. Of the four specimens of tychite thus far found, three have been very symmetrically developed octahedrons, but small, measuring not over 3™™ in diameter, and noticeably whiter than the average of the northupites. It is the small size of the erystals which favored the discovery of the new mineral, for in the original preliminary test one of the smallest and whitest specimens was selected, both because of its evident purity, and also with the idea of not using up any more material than was necessary. Those who may happen to have northupite crystals and wish to search for specimens of the new mineral, may jook for tychite therefore among the smaller crystals. We are informed in a recent letter from Mr. Northup that the chances of finding additional crystals of tychite, or of the asso- ciated minerals, northupite and pirssonite, are too remote to be seriously considered, as the old borax works are now disman- tled. ‘Tychite, therefore, promises to be a very rare mineral, unless a new locality for it happens to be discovered. The single crystal which we recently had the good fortune to find, Mr. Foote has generously presented to the Brush Collection of the Sheffield Scientific School, and both for this gift and for the interest he has taken in assisting us in our investigation we take pleasure in expressing our most sincere thanks. Comparison of tychite and northupite.—The two minerals, found so intimately associated with one another and both erys- tallizing in octahedrons, are chemically closely related, but in order to show the relation it is necessary to double the formula of northupite, as determined by Pratt. The compositions may then be expressed as follows : bo bo bo S. L. Penfield and G. S. Janvieson— Tychite. Tychite, 2MgCO,. 2Na,CO,. Na,SO,,. Northupite, 2MgCO,.2Na,CO,.2NaCl. Other physical properties are given below: Specific gravity. Index of refraction, 71,. Teens s 2°456 natural. 1°508 natural. J , 2-588 artificial. 1510 artificial. . 2°380 natural. ° 1514 natural. Northupite, §.377 artificial. Theoretical.—There seems to be far more interest connected with the present investigation than the mere description of a new species. Although northupite is somewhat slowly soluble in cold water, and is quickly decomposed by boiling water with the separation of magnesium carbonate, tychite is almost insoluble, even when its fine powder is treated with boiling water. Unlike. most insoluble substances, however, which precipitate quickly as soon as the constituents necessary for their formation are brought together, northupite and tychite are formed slowly. In de Schulten’s experiment, northupite was obtained after seven hours heating, and in ours it took nearly as many days of continued heating to obtain crystals of tychite. It would seem as though the slowness with which these substances are formed might be taken as an indication of their having a complex molecular structure, and that the element of time is necessary for the arrangement of the atoms in a state of equilibrium. Just what the arrangement of the atoms is, we are not able to determine, but the simplest and most symmetrically developed formulas which suggest them- selves are the following: Na=O02 VO=Na Oo —- C — O ge a Northupite, Na—O—C—O—Mg—Cl Cl—Mg—O—C—O—Na ae SOr ea CSO Na—O~ ~O—Na Na —O._LO—Na —- GC — O eg a Tychite, Na—O—C— O — Mg — (SO,) — Mg—O—C—O—Na RNS a on OO ON: Oe Na —O~ ~O—Na In these formulas the four carbon atoms are united by oxygen in ring formation, which it may be assumed it takes some time to establish, but, when once established, accounts S. L. Penfield and G. S. Jamieson—Tychite. 223 for the stability of the compounds. It is possible also that the assumed symmetrical arrangement of the atoms in the molecule is the cause of the crystallization of these compounds in the isometric system, for, as a rule, salts of a highly com- plex nature crystallize in some system other than the isometric. Moreover, if the above formulas are correct, it might be expected that tychite would be more difficultly soluble in water than northupite, for the SO, radical uniting the two magnesium atoms would serve, as we might say, to protect the latter from attack, while the sodium atoms could not be taken away without disturbing the equilibrium of the molecule. Perhaps also the union of the magnesium atoms by the SO, radical in tychite is more difficult to establish than the com- bination of the two chlorine atoms with magnesium in north- upite, which may account for the greater length of time required to make the sulphate compound artificially. In these compounds, two chlorine atoms in the one and a SO, radical in the other play the same role, and are isomorphous with one another in the broader sense of the term, namely, that . different constituents may enter into similarly constituted mole- cules without changing the crystalline form. In simple chemical compounds, it is contrary to all experience that a chloride and a sulphate should have the same crystalline form, or be isomor- phous with one another. In the salts under consideration, however, it is assumed that some definite arrangement of the large number of sodium, oxygen, carbon and magnesium atoms, by virtue of mass effect,* determines the crystalline form of the compounds, and that the roles played by two chlorine atoms in the one and a SOQ, radical in the other are relatively so unimportant that either of these constituents may enter into the molecule without changing the crystalline form. Whether it is possible to obtain a single crystal containing both the two chlorine atoms and the sulphate radical replacing one another as isomorphous constituents, or to obtain a single crystal with a nucleus of one salt and an external growth, in parallel posi- tion, of the other, we are not as yet able to state, but experi- ments along these lines, to determine to what extent the prin- ciples of isomorphism may be applied to so widely different radicals as Cl, and SO, under the influence of mass effect action, will be carried on and form the subject of a later communica- tion. In one experiment, in which the attempt was being made to obtain a product containing both Cl, and the SO, radical, a small] crop of octahedral crystals was formed which reacted for neither chlorine nor sulphate. In appearance * Compare mass effect action as applied to tourmaline (Penfield and Foote, this Journal (4), vii, pp. 122-124); also to the alunite-jarosite group of min- erals (Hillebrand and Penfield, this Journal (4), xiv, pp. 216-220). 224 S. L. Penfield and G. S. Jamieson —T ychite. the crystals were in every respect like those of the artificial northupite and tychite. As seen with the microscope the crystals were full of inclusions, and, in forming, had evidently enclosed an unusually large amount of amorphous magnesium carbonate precipitate. We assumed at once, and correctly, that the compound would prove to be like northupite and tychite, except in having a OO, radical in the place of Cl, and SO,, namely, 2Mg¢CO,.2Na,CO,.Na,CO,; see page 222. The analy- ' sis, made on a small quantity of the rather impure product, gave almost the theoretical percentage of CO,, but the MgO was high and the Na,O low, which was to be expected. Attempts will be made later to produce this salt in a state of purity, when it will be described more minutely. Mineralogical Laboratory of the Sheffield Scientific School of Yaie University, New Haven, Conn., July, 1905. Harrington— Modification of Victor Meyer's Apparatus. 225 Art. XXV.—A Modification of Victor Meyer's Apparatus for the Determination of Vapor-Densities; by B. J. HARRINGTON. THE ingenious apparatus devised by the late Professor Victor Meyer for the determination of vapor-densities has been in use for many years and has proved of great value for the purpose for which it was intended. It, however, has certain imperfections, being awkward on account of its height and very liable to be broken, especially in the hands of inex- perienced workers. Two modified forms of the apparatus have been devised by the writer and have proved so useful in our own laboratories that it has been deemed worth while to publish a description of them. In both cases an attempt was made to simplify the apparatus and make it more convenient and rapid to work with. The first form tried is that shown as fig. I in the accompany- ing illustration. It will be observed that the receptacle dd 1 is horizontal instead of vertical and that the long stem of Meyer’s apparatus is bent upon itself a number of times ; the apparatus accordingly occupying but little space. Instead of the long outer tube or jacket ordinarily employed, a box made of tinned iron or copper is used. In making a actual determination the space around the glass at m and & (fig. IL) is packed with a little asbestus, and it has been found advantageous to lay a piece of asbestus ‘card on the cover of the box. The weighed material in the ordinary stoppered tube or bulb is dropped in at e (fig. fT) and as it has not far to fall there is no need of the usual cushion of asbestus or sand. As soon as one operation is completed the vapour is quickly swept out of the apparatus by connecting the tube ao (fig. I) with the vacuum-pump, the water in the box 7 (fig. If) being kept continuously boiling. In this way one operation quickly succeeds another, and it has been found that students can make two or three determinations in the time required for one with the ordinary apparatus. The second form experimented with is shown at III. In this the receptacle dd of I is placed vertically, as it was thought that the vapor would be less likely to be carried into the delivery tube than if the horizontal position were adopted. The tube ¢ is somewhat longer than in the first form (1) but the curve at the bottom checks the velocity of the little tube containing the liquid and no asbestus is required at the bottom 296 Hurrington—Modification of Victor Meyer's Apparatus. of dd. Like No. I, this form is much more easily dried out than the ordimary apparatus. The metal box for No. III. is ale sM dyes bald fa forbe Pec lenfefe lade bole Lok Ve as (Gea OU Aset ~ ae h----2 4 ee e ° ena et ees oe at not shown in the drawing, but its construction can be readily understood. With both forms of apparatus é was closed with Harrington— Modification of Victor Meyer's Apparatus. 227 an ordinary cork, a correction being made for the small quan- tity of air displaced by the cork, but of course one of the improved appliances for imtroducing the liquid could be employed. So far the apparatus has been tried only for bodies with comparatively low boiling points, but it could no doubt be adapted for use with liquids with higher boiling points. The following table gives a series of molecular weight determinations kindly made for me by Mr. Douglas McIntosh, D.Sc., of this university, with the different forms of apparatus, and vives an idea of the results which may be expected. Apparatus No. II. has, on the whole, been found to give more concordant results than No. I, but the latter is simpler and less likely to be broken than the former and in either case the fioures obtained are sufficiently accurate for the purpose. They were obtained by working very rapidly and with no special precautions, and cannot therefore be fairly compared with those given by Victor Meyer’s apparatus in the last column; for in the case of the latter Dr. McIntosh states that he took every precaution in order to ensure the most accurate results possible. — MorecuLtar Weieust DETERMINATIONS MADE BY Mr. Dougias McIntosu, D.Sc. (Air=2 x 14:44) Modified Modified Meyer’s Apparatus Apparatus Apparatus. Now: Nori Methyl 35°0 Alcohol 36°0 CH,OH 36:9 32:9 (32) 36°5 33°4 34°60 33°5 31°91 36°9 33°] 31°94 34°7 34°8 37°2 Mean 35°8 Mean 33-2 Mean 31:93 Methyl 46°0 Alcohol 43°71 44°] C,H,OH 45:0 44:7 46°70 (46) 44°5 44°3 46°10 43°6 42°3 —EEEE, eee —— Mean 44:1 Mean 44:4 Mean 46:40 228 Harrington—Modification of Victor Meyer's Apparatus. Modified Modified Meyer's Apparatus Apparatus Apparatus, No. I. No. II. Acetone 59°4 CH BY cH eo 59°2 58°9 (58) 58-0 59-0 57°90 59°O 58°'5 : 5780 59°1 Mean 58°6 Mean 58°8 ’ Mean 57°85 Ether 76°1 (OME @ 80:0 (74) 82°6 Deh 74:9 75:0 75°70 82°6 al 76°90 76°1 76°2 Mean 78:7 Mean 76°5 Mean 76°30 Benzol Wulae (C.H,) 73°3 Sige (78) 80°2 80°8 79:00 tod BPI 80°4 79°20 81°7 79°77 Mean 78:7 Mean 80°5 Mean 79:10 Chloroform 134°9 CHCl, 131°5 (119°5) 122°6 126°3 P2AS 136°8 124°9 125°5 L228 123°230 125°8 L262; 123°00 Mean 129°1 Mean 124°5 Mean 123°10 McGill University, May, 1905. St tae T. C. Brown—Frauna trom Chappaquiddick Island. 229 Art. XXVI.—A New Lower Tertiary Fauna from Chappa- quiddick Island, Marthas Vineyard ;* by Tuomas C. Brown. (With Plate VIL) CuHappaguippick Island lies at the eastern end of Martha’s Vineyard and owing to the shifting nature of the sands and varying tidal currents, it is at times connected with that island, but it is for the oreater part of the time completely separated from it. Dr. Arthur Hollick has made a very careful study of the structure of this island and collections of the molluscs and plants found fossil upon it.* The fossil plants have been very fully described by him in the Bulletin of the New York Botanical Garden, vol. 1, No. 7. The mollusc material has not been described and its horizon was provisionally set as Cre- tacic by Dr. Hollick because of the similar lithological char- acter of this material with other deposits on Martha’s Vineyard containing undoubted Cretacic fossils. A eareful study of the fossils has shown that this material is not Cretacic but Eocene in age and that it contains a new and peculiar fauna, a fauna which differs considerably from that of the Eocene deposits of the southern Atlantic slope. In describing the deposits from which these molluscan remains were obtained Dr. Hollick says: “. . . the Island may be said to be composed of reassorted drift. . . . These hills in general may be described as kame-like, both in appearance and in composition. They are rounded accumulations of sand, gravel and cobble stones, with some bowlders, and were evi- dently formed by water action. In many places the sand and oravel is cemented together by limonite, forming hard lenses and strata, and ferruginous concretions and shaly fragments are abundantly represented.” + In his geological studies of Martha’s Vineyard and surround- ing islands Professor Shaler recognizes these ferruginous con- cretions and concerning them he says: “On the Island of Chappaquiddick and in the region near Edgartown, occasional fragments of a ferruginous sandstone are found which closely resemble in their general character the materials containing the Cretaceous fossils, but as they offer no organic remains I hesitate to consider them of that age.” t+ Dr. Hollick considered these concretions as lithologically identical with those contaming Cretacic molluscs and plants and set out to make a collection of organic remains that would * The investigations on which this paper is based were carried on in the Paleontological Laboratory of Columbia University and the types of these species are in the university collection. + Bull. N. Y. Botanical Garden, vol. ii, No. 7, p. 399. tN. S. Shaler, 7th Ann. Report U. S. G.S., p. 3826: 230 T. C. Brown—Ffauna from Chappaquiddick Island. substantiate the point. “A systematic exploration of all expo- sures was therefore prosecuted ; hundreds of the concretions and shaly fragments were broken open and critically examined and the result was a collection, not only of molluses but also of plant remains, a few of which were found sufficiently well pre- served for identification.’’* Upon his return Dr. Hollick submitted the molluses to Professor K. P. Whitfield of the American Museum of Natural History for a hasty examination, and concerning them Pro- fessor Whitfield spoke thus: “ I have examined the fossils you sent the other day but I cannot satisfy myself as to their age. They consist of a JModzola, which apparently does not differ from our common AZ, plicatula, of the harbor here ; an Anomia which might pass for A. gigantarva of the lower greensand marls of New Jersey, if it were not for the Modzola; also a single imperfect internal cast of a small (young?) Pectunculus not enough of it to tell the species, and a small bivalve of which I cannot yet determine the genus. These are the onl shells I can recognize, and from their evidence I should think the rocks could hardly prove to be Cretaceous.” + These fossils were also submitted to Professor Grabau of Columbia University for examination. ‘ Mr. Grabau is of the opinion that they may represent a new fauna, of more recent age than the Cretaceous, and this is quite consistent with the conditions under which they occur, so far to the south of any recognized Cretaceous outcrop. ‘The character of the matrix also, with a single exception, is notably different from that in which undoubted Cretaceous molluscs have been found else- where, being a micaceous sandstone instead of a hardened clay or greensand.” + A careful study and detailed comparison of these fossils with descriptions, figures, and specimens of the Cretacic and Eocene species shows that these fossils represent a new fauna of Eocene age. This fauna, however, differs widely from that of the Eocene deposits of the South Atlantic coast and seems to be more closely allied in general to the Eocene of England. Some of the specimens are very well preserved, while others are only represented by external and internal molds. Many of these molds are of such a nature and so well preserved that a wax impression can easily be taken and the characters of the fossil observed and compared. The following descriptions and comparisons include the best preserved and most typical speci- mens. Some of these are not perfect enough to be described as new species, but most of them can be generically placed. * Bull. New York Botanical Garden, vol. ii, No. 7, pp. 399-400. + Bull. N. Y. Botanical Garden, vol. ii, No. 7, p. 400. t Ibid., p. 401. T. C. Brown—Fauna trom Chappaquiddick Island. 231 Modiola vineyardensis sp.n. Pl. I, fig. 1. Shell strongly ventricose, with a very prominent, almost angular umbonal ridge extending from the beak to the ventral margin. Shell distinctly concave anterior to this ridge; pos- terior to this ridge it becomes flattened toward the posterior margin ; anterior end extremely short barely extending beyond the beak ; posterior margin angulate, front margin nearly straight only a slight emargination occurring, basal end rounded ; the portion of the margin from the end of the hinge line to near the point of angulation, and from beyond the point of angulation to the ventral margin, are almost straight lines. Surface with pronounced raised radii, flattened at the top and separated by spaces equal to or slightly wider than the radii ; the radii are very fine and crowded on the anterior portion of the shell, much coarser on the median and posterior region, and distinct from the beak to the margin. They increase in size progressively from the dorsal to the ventral portion of the shell, with a corresponding increase in the width of the inter- spaces. They increase in number by intercalation as well as bifureation. Fine distinct growth lines cross and cancellate the radii. This species resembles J/. alabamensis Aldrich,* from the Eocene of Maryland, but differs from it in general outline. It has a less curved anterior border and more radu, which are very distinct from the margin to the beak. The shell is shorter antero-posteriorly, and the posterior margin is more obtusely angulate. The shell of J/. vineyardensis is also more ventri- cose and the umbonal ridge more angular and more pro- nounced. | In general outline this species approaches more nearly J/. grammatus Dall,t from the Oligocene of Florida. The sur- face ornamentation is very similar, but judging from Dall’s figure his shell is less ventricose and thé umbonal ridge less angular and less distinct. But even closer than to any of these is the resemblance of this species to I. elegans Sowerbyt from the Eocene of Eng- land as figured and described by Wood among the Eocene bivalves. In general outline and surface ornamentation the resemblance is very close. JZ. elegans is, however, slightly less angulate at the postero-dorsal margin and judging from the figures is less ventricose. Compared with the modern J. plicatula Lamarck, living along the Atlantic coast, J/. vineyardensis seems to be nar- * Bull. of Am. Palaeont., vol. i, p. 68, pl. v, fig. 18. + Trans. Wagner Free Institute of Sci., vol. iii, pt. 4, p. 794, pl. xxx, fig. 2. ¢ Paleontological Soc. Monographs, London 1861-71. Eocene Bivalves, p. 60, pl. xii, fig. 5 (c). 232 T. C. Brown—Ffauna from Chappaquiddick Island. rower toward the ventral portion of the shell, while the pos- tero-ventral margin is more nearly a straight line and the radii are more numerous and finer and proportionately more widely separated. In JL. vineyardensis and M. plicatula the mode of increase in the number of the radii by occasional intereala- tion and bifurcation is very similar. In J. plicatula the umbonal ridge is less angulate and less pronounced. The shell described is a left valve with the following meas- urements: length 32°5™™, width 14™™. Several small specimens of this same species occur in other fragments of the conere- tions, as well as imprints of these shells. These smaller speci- mens correspond exactly with the growth lines of the younger stages in the larger individuals. Modiola vineyardensis mut. inornata. This mutation is very similar to the type of the species described above, except that the radii are very faintly marked. Fine, distinct, concentric growth lines mark the surface. Dis- tinct radii can be seen on the anterior and umbonal region of the shell. These radii are flattened on top and separated by very narrow impressed lines. They fade out as they pass away from the umbonal and completely disappear on the ventral portion of the shell. This mutation is represented in the collection by a compara- tively small left valve. Modiola Hollicki sp.n. Pl. VIII, fig. 2. Shell ventricose, with a prominent umbonal ridge extending from the beak to the ventral margin; shell sloping abruptly to the anterior margin and becoming flat in the postero-dorsal part; anterior end rather short; anterior end rounded, front - margin slightly arcuate, ventral margin broadly arcuate, rounded, postero-dorsal margin obtusely angulate; cardinal line straight; surface without ornamentation except for rather faint concentric lines of growth. In general outline this species somewhat resembles J/. Mitchelli Morris,* from the Eocene of England. It has a more obtuse postero-dorsal angle and is slightly narrower, with a slightly arcuate anterior margin instead of being emarginate as In that form. Represented in the collection by two specimens, one nearly perfect left valve (fig. 2) and a valve lacking the beak and hinge area. These occur together with the Corbulas (see be- low) in a fine-grained hard ferruginous lutyte concretion quite different in character from the micaceous sandstone concretions in which all the other fossils are found. * Palaeontological Society Monographs (see above), p. 68, pl. xiii, fig. 10. T. C. Brown—Fauna from Chappaquiddick Island. 233 Corbula Whitfieldi sp.n. Pl. VIII, fig. 3. Shell large for the genus, ventricose and subtriangular ; beak high and incurved ; anterior margin sharply rounded, ventral margin broadly arcuate in the median and anterior portions but sinuously emarginate posteriorly ; the posterior end of the shell is narrow, produced and abruptly truncated. Surface marked by distinct concentric asymmetrical folds or concentric wrinkles which are broadly rounded on top, with the dorsal border slightly broader and not as abruptly sloping as the ventral border. The folds are separated by narrow channeled interspaces. These folds constitute a surface ornamentation, and not lines of growth, asis shown by the fact that they increase in number by intercalation, some folds extending from the anterior to the median portion of the shell, while others extend almost to the posterior end. The principal folds extend to the posterior end and are there sharply tlexed. These folds are well defined on the ventral half of the shell and become finer and more crowded on the umbones and almost disappear at the beaks. This species approaches very closely in general outline and surface ornamentation to C. alaeformis* Gabb, trom the Tejon formation of California, but is less than one-half as large. The concentric folds become finer and more crowded on the umbonal region in the specimen from Chappaquiddick than in that figured by Gabb. This species also somewhat resembles C. swhengonatat Dall, from the Eocene of Maryland and Virginia, but differs from that species in being narrower anteriorly and more produced posteriorly, and in the absence of a subcarinate ridge extending from the umbo to the posterior margin. The concentric folds are also more crowded and less prominent on the umbonal region. The material in hand represents a right and a left valve. These specimens occur in a very fine-grained hard ferruginous lutyte concretion, quite different in character from the material in which most of the other fossils are found. Anomia simplexiformis sp.n. PI. VIII, fig. 10, 11. Shell subovate and prolonged in the region of the beak ; left valve very globose, nearly equilateral, “somewhat irregular ; beak located in median dorsal portion of the shell, submarginal, slightly projecting and incurved; surface without plications or ornamentation, except possibly very faint indications of concentric lines of growth. * Palaeontology of California, vol. ii, p. 177, pl. xxix, fig. 638. + Md. Geol. Sur., Eocene, p. 163, pl. xxxii, figs. 1, la, 2, 2a, 2b. Am. Jour. Scit.—Fourts SERIES, Vou. XX, No. 117.—SEPTEMBER, 1905. 16 234 7. C. Brown—Fauna from Chappaquiddick Island. A portion of a right valve, probably of this species, is present in a fragment of the ferruginous coneretion (fig. 11). It is very much flattened, more or less irregular, with large byssal opening and indications of very faint concentric growth lines. This species is represented by several complete or nearly complete left valves, varying greatly in size, the length ranging from ten to thirty millimeters. It resembles very closely the modern A. semplex from the shores of Long Island. It has the same general outline and shape, and approaches that species in size and in the absence of surface plications and other orna- mentations. Anomia paucistriata sp.n. Pl. VIII, fig. 12. Shell subcircular, somewhat irregular; left valve convex, nearly equilateral; beak submarginal, dorso-medially placed and not pronounced ; surface marked by a few faint radiating striations, crossed and cancellated by very fine concentric lines of growth. This species is smaller than the preceding, averaging in length about ten to twelve millimeters. It is represented by several left valves. Right valve unknown. . Glycymeris sp. ?.. Pl. VIII, fig. 138. Represented by several internal molds not preserving char- acters sufficient for specific description. The figure shows the internal characters of the shell and is drawn from a wax im- print made from a mold. | This species is smaller and more ovate in form than G. idoneus Conrad, from the Nanjemoy and Aquia formations of Maryland. In size, form and general appearance it resembles more closely Glycymeris (Pectunculus) decussatus Sowerby, from the Eocene of England. Nucula sp. ?. Represented by a few small internal molds. In one at least the dental characters are very well preserved. In general out- line these resemble very closely VV. potomacensis Clark, from the Eocene of Maryland, but do not preserve sufiicient charac- ters for specific identification or description. Turritella sp. ?. Pl. VIII, fig. 4. Shell small, spire high, angle about twenty-five degrees each whorl marked by a distinct, well-defined anterior and less prominent posterior spiral carinate ridge, following around above and below the suture, otherwise the surface is smooth and free from ornamentation ; suture distinct; whorls closely placed and rapidly increasing in size. T. C. Brown—Fauna From Chappaquiddick Islund. 235 This species resembles very closely a Turritella not specifi- eally identified from the Eocene of Whellock, Texas, in the University collection. The Whellock specimen is larger but has the same apical angle, is free from ornamentation and has the anterior and posterior carinate ridges present but faintly marked. This description is based on a wax imprint made from a very perfectly preserved external mold in a red micaceous sandstone eoneretion. The full length of the shell is not represented, so that the number of whorls and dimensions cannot be given. There are no characters of aperture and lips apparent. Terebra sp. ?. Pl. VIII, fig. 5. Shell elongate, spire elevated, whorls closely placed, rapidly enlarging, flat on the outer surface between suture, free from ornamentation or with very faint revolving lines, aperture elongate elliptical, pointed anteriorly, rounded posteriorly ; outer lip thin and broadly arcuate, inner lip smooth without eallus or ridge. | The specimen figured occurs on the edge of a small fragment of rock. The apex is concealed in the matrix and the anterior end of the aperture is slightly injured so that it does not show the minute characters. Terebra juvenicostata sp.n. Pl. VIII, fig. 6. Shell small and slender, spire elevated ; apex pointed, acute with an apical angle starting at about thirty degrees and decreasing toward the body whor] where the sides of the spire approach to parallelism; the whorls are closely placed and flattened between the sutures. There are distinct ribbings on the earlier whorls which become less distinct along the advanc- ing spire and disappear on the body whorl. : Odostomia semicostata sp. n. Pl. VIII, fig. 7. - Shell small, consisting of six or seven volutions, spire ele- vated, apical angle thirty degrees; sutures very pronounced ; volutions flattened convex between sutures; earliest whorls marked by distinct transverse plications or ribs which become almost or quite obsolete on body whorl; outer lip distinctly denticulate within. The aperture and columella of this specimen is not fully preserved so it cannot be accurately described. Length of shell as preserved 9°5™™. Odostomia crenulata sp.n. Pl. VIII, fig. 8. Shell very small, spire high and closely coiled, apex sub- acute, whorls flattened externally, faintly crenulate along the posterior margin, suture distinct, aperture and lips unknown. 236 7. C. Brown—Launa from Chappaquiddick Island. Genus? sp.? Pl. VIII, fig. 9. Shell small, loosely coiled, apex acute, whorls five or six, well rounded, rapidly increasing in size, smooth without any ornamentation: suture quite deep and distinct, character of aperture and lips unknown. This species is very similar to a Lemnaea in shape but ean hardly be one of these as it appears to be a salt water form. Represented in the collection by a small but very perfect external mold of which a wax impression was taken. Ostrea sp.? | Several small internal molds of representatives of this genus are present among the fragments of the concretion. These are not sufficiently well outlined to be specifically determined. They seem to represent at least two or three different species and all are comparatively very small. Cardium? sp.? Several casts doubtfully referred to this genus are to be found among the fragments of concretion collected by Dr. Hollick. These fossils represent a new and distinct fauna markedly different from that of any other Eocene deposits of this country. Since this fauna does not contain a single species in common with the Eocene faunas of the Atlantic slope and gulf deposits, it cannot be accurately correlated with these beds and assigned its proper place in the geologic scale. Nevertheless from the general characteristics of the contained species and their affinities to forms from widely distant prov- inces, the horizon of these deposits can be ascertained with some approximation to the truth. Considering the marine Eocene deposits of this country as a whole, we find that they naturally fall into several provinces lithologically quite distinct, and containing faunas with very few species in common. In New Jersey there is a small and isolated area known as the Shark River beds from their out- crop along that river. According to Clark, these beds repre- sent lower Eocene and rest conformably upon the Cretacic below. By early writers they were considered a part of the Cretacic, as there was no marked line of separation between them and the underlying strata. The fossils, however, were found to be of undoubted Eocene character, and although the fauna was lacking in some of the most widely distributed Eocene species, it still contained no characteristic Cretacic forms. These Shark River deposits were thought by Harris to rep- resent a higher horizon than the Eocene deposits of Maryland T. C. Brown—Fauna from Chappaquiddick Island. 237 and Virginia, and Dall, in his correlation tables of the North American Tertiaries, has placed them in the Claibornian stage or equivalent to upper Middle Eocene. The fossils of this province differ so widely from those of the regions immedi- ately to the south that correlation is very difficult, and even now there is doubt as to the exact position of these beds. A second province of the Eocene, generally known as the Pamunkey formation from its typical development along the Pamunkey River in Virginia, begins in Delaware and extends across Maryland well into Virginia. Lithologically these deposits have more similarities to those of the provinces to the south than to the Shark River beds of New Jersey, yet they are sufficiently distinct both lithologically and in their con- tained fauna to require complete separation. According to Clark these deposits “ constitute a single geological unit.” A third province embraces the Eocene deposits of the Caro- linas and Georgia and affords a far more complete series of Eocene strata than either of the more northern areas. The lower beds consist of arenaceous and conglomeratic deposits, rather sparingly fossiliferous, probably because the material by its very nature was not adapted to permit the preservation of fossils. The middle and upper beds are well developed and represented by limestones and marls containing an extensive fauna, yet quite distinct from the surrounding provinces. A fourth, the Gulf provinee, is by far the most extensive of the Eocene areas. It extends from Florida to Texas and includes the so-called Mississippian embayment, an area extend- ing well up into the Mississippi basin. All stages of the Eocene are nore fully represented, but both lithologically and paleontologically this province is very distinct from those along the Atlantic coast. Peculiar conditions in this area resulted in the interbedding among the other deposits of many lignitic strata. A fifth marine Eocene province occurs along the Pacific coast, and outcrops along the coastal range in California, Ore- gon and Washington. These deposits are generally known as the Tejon group and were originally referred to the Cretacic. Later study has shown them to be of Eocene age, and yet their fauna differs widely from those of the Atlantic and Gulf provinces. The fauna from Chappaquiddick represents a new and dis- tinct Eocene province, differing from all the other provinces but no more widely different from these than they are from one another. Although in this fauna there are several species somewhat resembling those of the provinces to the south, on the whole it would seem to be more closely allied to the Eocene of England. The genera most abundantly represented 238 7. C. Brown—FHauna from Chappaquiddick Island. in these Chappaquiddick deposits, e. g., odiola, Glycymeris, are also among the most abundant in the English deposits. These same genera, although represented in the Atlantic and Gulf provinces, are there more sparsely distributed and occur with other more abundantly represented genera that appear to be altogether wanting in the Chappaquiddick deposits. A comparison of this Chappaquiddick fauna with other Eocene faunas indicates that it is of lower EKocene age, the species most closely resembling those found in this fauna being found in the lower beds of the Atlantic and Gulf provinces, the Tejon of California and the lower beds of England. These deposits may possibly be of the same age as the Shark River beds of New Jersey, but being deposited in a region separated from this have no forms in common with it, but such correla- tion could be only conjecture. As the cor relation of the well’ known Eocene deposits is even yet very uncertain, it is unnec- essary and impossible to place these beds any more definitely than simply to say they are lower Eocene. Columbia University, New York City. EXPLANATION OF PLATE VIII. HicuReE 1.—Modiola vineyardensis 2.222252 — 228 2 p. 282 > PIgurRE(2:——Modiola hollickt \ {evs 2 eee eee p. 202 FIGURE 3.—Oorbula Wiihacks WTA ek SE A ALS ana p. 233 Fiaure 4,—Turritella a8 elias Cees SIES Gv eeu p. 234 FIGURE 0.7 erebra Spit eos -2 ee ee p. 289 FiGurRrE 6.—PLerebra juvenicostata 2s. 2522. one eee _ p. 235 FIGURE 7.—Odostomia semicostata .___...-..-.++-=.---2--- p. 2390 HicuRre 6:—Odostomia crenulata ee os ee ee ee p- 285 RicgurE: 9:— Genus? sp. ?.2oio Ses ee ee p. 236 FraurEs 10 and 11.—Anomia simplexiformis __.----------- p. 238 Fiagure 12.—Anomia POU ELES wikia SAU AG i Oe ery eh em p. 234 HiGuRE 13:—Glycymeris Sp.-2s a5 2s 2 ee eee ae p. 234 Am. Jour. Sci., Vol. XX, 1905. Plate VIII. Fie. 12, x 2. - Boltwood—Production of Radium from Uranium. 239 Arr. XX VII.— The Production of Radium from Uranium ; by Bertram B. Bottrwoop. THE A orice that-radium is a disintegration product of uranium has been greatly strengthened through the demonstra- tion of the fact that in radio-active minerals the quantity of radium is directly proportional to the quantity of uranium present.* On the basis of the disintegration theory a propor- tionality of this sort is to be expected between the parent element and its radio-active successor. Additional data on this highly important question are how- ever desirable, and a single experiment likely to further eluci- date the problem has been independently undertaken by a number of different investigators. This experiment consists in observations conducted on a carefully purified uranium salt with a view to determining whether, with the lapse of time, measurable quantities of radium will be produced within it. If radium is a direct product of uranium through the inter- mediate stage of uranium-X and if the average life of radium is approximately 1,000 years, then it can readily be deduced that, with the delicate methods of measurement at command, the quantity of radium formed in a few hundred grams of uranium salt will be readily detectable and measureable after the lapse of a period no longer than a month. If, however, one or more transition products of a relatively slow rate of change intervene between the substance uranium-X and radium, the production of radium will be so protracted that no quantity of it sufficiently great to permit its detection will be formed within a greatly extended period. The difficulties involved in the experimental demonstration of the growth of radium do not appear to be great. Uranium forms no radio-active, gaseous disintegration product, while the radium emanation affords a most convenient means of quanti- tatively estimating any radium which may be present. A solution of a carefully purified uranium salt can therefore be prepared and can be tested at intervals for radium emanation. If radium is formed from the uranium its existence will be indicated by the presence of radium emanation in the uranium solution. Three papers in which an experiment of this character is described have been published by Mr. Soddy.t In the first * Boltwood, Phil. Mag. (6), ix, 599; Strutt, Proc. Roy. Soc. Lond., Ixxvi, 88 ; McCoy, Berichte d. deutsch. chem. Ges., xxxvii, 2641. f ‘The Life-history of Radium,” Nature, Ixx, a0; seo the Gren of Radium,” Nature, lxxi, 294; “The Production of Radium from Uranium,” Phil. Mag. (6), ix, 768. Mr. ’Whetham has also published two contributions on the same general topic (Nature, 1xx, 5; ibid., Ixxi, 319) in which he states 240 Boltwood—Production of Radium from Uranium. paper, published May 12, 1904, very scanty details of the experimental procedure are given, but a summary of the con- clusions reached at that time by the author is as follows: 1. The quantity of radium which has accumulated in one kilogram of uranium nitrate in twelve months is less than 107" gram. 2. The question so far as the production of radium from uranium is concerned is practically settled. 3. If uranium changes into radium, less than one ten-thou- sandth part of the theoretical quantity is produced during the first year’s accumulation. 4, The evidence may be taken as indicating that uranium is not the parent element of radium. The second paper, published Jan. 26, 1905, eighteen months from the commencement of the experiment, is likewise lacking in a detailed account of the experimental methods, but the author states that measurements carried out at that time with the kilogram of uranium nitrate under observation indicate that it contains 15x<10~° gram of radium, a quantity which, while of considerable relative magnitude, is only one five-hun- dredth of the amount to be expected from the disintegration theory on the assumption of a direct change. The author suggests that the greater part of the radium emanation may (under the conditions of the experiment) be retained in the uranium solution and not evolved as a gas. On the basis of tbe amount of radium assumed to be then present it is deduced that the fraction of uranium changing per year is 2107". After pointing out certain sources of error likely to have exercised a disturbing influence during the elapsed period of © observation, the author adds,—“‘if the whole series of meas- urements from the commencement are recalculated, eliminat- ing the error alluded to, they are fairly consistent with there having been a steady production of radium at this rate contin- uously from the commencement.” One of the sources of error alluded to was the introduction of very considerable quantities of radium salts into the laboratory during the period when the kilogram of uranium nitrate was under obser- vation. It is stated that the presence of this radium greatly disturbed the electroscope in which the measurements were conducted. Additional difficulty had been previously experi- enced in attempting to standardize the measuring instrument with the emanation corresponding to a known weight of pure radium salt. that he also believes that he has observed indications of the growth of radium in uranium compounds. Since Whetham’s communications contain neither any account of experimental details nor any record of quantitative measurements, it is impossible to judge as’ to the value of the data on which the author’s conclusion is based. Boltwood—Production of Radium from Uranium. 241 The third and more elaborate article by the same author appeared in the June number of the Philosophical Magazine. The data briefly given in the earlier articles are here treated at greater length and a closer insight can be gained of the experimental methods and the results on which the author’s later conclusions are based. Although it is stated in this paper that observations had been taken occasionally over a period of eighteen months and that these observations indicated a grad- wal growth of the emanating power of the uranium solution, the only definite and directly comparable numbers are restricted to a total period of about three weeks (Dec. 17, 1904 to Jan. 9, 1905) and include only four measurements conducted at the close of the period of observation. Without entering into a discussion of various minor details in Mr. Soddy’s papers, it is desired to call particular attention to the followmg important considerations in relation to the experimental data submitted : First. No conclusive evidence is brought forward to show definitely how much or how little radium was present in the uranium solution at the commencement of the experiment.* Second. It appears extremely possible that the increase in the content of radium which Mr. Soddy believes he has observed in his uranium solution may in fact have been due to the acci- dental and unconscious introduction of radium salts during the tests conducted at the end of the twelve months period. According to his own statements these tests were carried out in a laboratory notably contaminated with various radio-active products, and the accidental introduction of the sub-micro- scopic quantity of material (1-6 x10~° gram.) which was after- wards detected would account for the later positive results. The liability of contamination from an extraneous source is strongly suggested by the behavior of Mr. Soddy’s electro- scope, In which the normal air leak has risen from 0:048 divi- sion per minute to 1°56 division per minute, an increase of over thirty times, during the period covered by his experi- ments. The conditions essential to the elucidation of the question of the actual production of radium in uranium compounds would seem to be: * The writer of the present paper convinced himself at the beginning of his own experiments that the method of procedure followed by Mr. Soddy in testing his solations for radium emanation is entirely unsuited for the determination in question. A concentrated solution of incompletely puri- fied uranium nitrate containing traces of radium gave up only a fraction of the total radium emanation generated within it when the solution was allowed to stand for days in contact with a small air space and air was bubbled through it. It was speedily found that only by boiling the solution vigorously for about fifteen minutes could the total emanation present be positively separated. 242 Boltwood—Production of Radium from Uranium. (a) The employment of a method for the determination of radium which gives positive and quantitative results. The method must be suitable for the determination of very small quantities of radium and must be capable of indicating the maximum quantity present at all times. (6) The preparation of a pure compound of uranium and the demonstration that-the compound is initially free from radium. (c) Proper conditions for testing and preserving the uranium salt in order to preclude the introduction of radium or radium emanation from external sources, so that if the presence of radium is noted it can be assumed with certainty that the radium found has actually been formed in the solution. It would appear that none of these essential conditions has been fulfilled in the experiment described by Mr. Soddy. The writer of the present paper has been conducting an experiment on the growth of radium in a uranium solution for the past thirteen months. The conditions of the experiment were the following: In May, 1904, a kilogram of “ purest uranium nitrate”? was purchased from Eimer & Amend of New York City. This material was tested qualitatively for radium (through the emanation) and readily detectable quanti- ties of this element were found to be present. The salt was dissolved in distilled water and the solution was filtered. The compound was then recrystallized five separate times, the con- ditions being so chosen that the separate crystals of each of the different crops were not over two millimeters in cross-section. The mother lquors were each time removed from the crystals on a suction filter, and the crystals were washed with a small quantity of ice-cold water. The final yield of purified material was a little in excess of 200 grams. Of this 100 grams were taken and dissolved in pure, distilled water. This solution was introduced into a glass bulb with a capacity of approximately 400°, diluted to about 250°, and the neck of the bulb was drawn out into a short capillary and sealed in the flame of the blowpipe. The solution was sealed up on July 8, 1904. Thirty days later the bulb was opened under conditions which precluded the escape of any portion of the contained gases and the entire gaseous contents were removed and transferred to an electroscope. In order to completely displace the dissolved gases and any radium emanation which might have been present the solution was boiled vigorously for about fifteen minutes.* *The removal and collection ‘of the gaseous contents of the bulb was accomplished by the use of the apparatus which has been described in a previous paper (this Journal, xviii, 379). The neck of the bulb containing the uranium solution having been first notched with a file, it was inserted in the rubber tube D, the point was broken off within the tube, and the gases displaced from the bulb on heating were collected in the burette D, which was filled at the start with boiling water. Bolitwood— Production of Radium from Uranium. 248 The type of electroscope used in this investigation has already been described (this Journal, xviii, 97). The ema- nation from the radium associated with 0°1 gram of uranium in a radio-active mineral caused a leak of approximately 21 divisions per minute. Assuming that the 100 grams of uranium nitrate contained 48 grams of uranium, the leak correspond- ing to the quantity of radium in radio-active equilibrium with 48 grams of uranium would be approximately 10,000 divisions per minute. The normal air leak of the instrument was 0-012 division per minute, and an increase of 0°005 division per minute could have been detected with certainty. The electro- scope was therefore capable of indicating the presence of a quantity of radium equal to 5x10~ of the equilibrinm quan- tity. The actual quantity of radium equivalent to a leak of 0-005 division per minute was 1°7<10~" gram.* On introducing the gases present} in the uranium solution into the electroscope no wncrease in the leak of the mstrument could be detected although the observations were continued over a period of eight hours. The quantity of radium present at the start was therefore less than 1-710" gram. The uranium solution in the bulb was allowed to cool, and the neck of the bulb was again sealed. At the end of six months trom the start, in January, 1905, the uranium solution was again tested under conditions identical with those under which the first test was carried out. Entirely negative results were obtained and the quantity of radium then in the solution was still less than 1-7x10~-" gram. A similar test was con- ducted on August 2, 1905, 390 days from the commencement, and no evidence of the presence of radium emanation was even then obtained. It can therefore be positively stated on the basis of sound experimental data that in 390 days the quantity of radium formed from 48 grams of uranium in a uranium nitrate solution is iess than 1-°7X10~" gram. The quantity of radium which can have been produced in the given time is therefore less than one two-millionth of the equilibrium quantity and less than one sixteen-hundredth of the quantity which would be expected from the disintegration theory if the value of X for radium is taken as 8°8x107* (year)”.t The quantity is furthermore only about one-tenth of the quantity assumed by Mr. Soddy to have been formed from an equal quantity of uranium in his solution during an interval of eighteen months. It is important to add that the whole series of measurements has been conducted in a laboratory which has been carefully * Rutherford and Boltwood, this Journal, xx, 55. + At the end of the 30-day period. ¢ Rutherford, Trans. Roy. Soc. London, (A) eciv, 215. 244 Boltwood—Production of Radium from Uranium. protected from contamination by the salts of radium or other radio-active substances, and that the electroscope used has been reserved for this particular research, its original normal air- leak having remained unaltered throughout the entire period. It has therefore been unnecessary to introduce any corrections or to make any allowances for possible errors due to known causes of any description. The experiments described in this paper are considered to indicate that the results obtained by Mr. Soddy are without significance and that one or more products of a slow rate of change intervene between uranium and radium. It is claimed, moreover, that the conclusions in Mr. Soddy’s first paper, so far as they relate to the dvrect transformation of uranium into radium, are more truly in accord with the actual facts than are those contained in his later publications. 139 Orange street, New Haven, Conn. August, 1905. Geology. 245 SCIENTIFIC INTELLIGENCE. ; Il. GroLoey. 1. Heplorations in Turkestan with an account of The Basin of Eastern Persia and Sistan. Hapedition of 1903, under the direetion of RAPHAEL PUMPELLY. 4to, 324 pp., 6 pls., 174 figs in text. Washington, D.C. (Published by the Carnegie Institu- tion of Washington. Publication No. 26. April 1905.)—This publication contains the following five papers: Archeological and Physico-Geographical Reconnaissance in Turkestan by Raphael Pumpelly ; A Journey across Turkestan by William M. Davis; Physiographic Observations between the Syr Darya and Lake Kara Kul, on the Pamir, 1902, by Raphael W. Pumpelly; A Geologic and Physiographic Reconnaissance in Central Turkes- tan, by Ellsworth Huntington; The Basin of Eastern Persian and Sistan, by Ellsworth Huntington. Professor Pumpelly states in the introduction that “At the end of 1902 the Carnegie Institution voted a grant to me ‘for the purpose of making, during the year 1903, prelimi- nary examination of the Trans-Caspian region, and of collecting and arranging all available existing information necessary in organizing the further investigation of the past and present physico-geographical conditions and archeological remains of the region.’ “The investigation was proposed because (1) there is a school that still holds the belief that central Asia is the region in which the great civilizations of the far East and of the West had their origins ; and (2) because of the supposed occurrence in that region, in prehistoric times, of great changes in climate, result- ing in the formation and recession of an extensive Asian Medi- terranean, of which the Aral, Caspian, and Black seas are the principal remnants. “it had long seemed to me that a study of Central Asian arche- ology would probably yield important evidence in the genealogy of the great civilizations and of several, at least, of the dominant races, and that a parallel study of the traces of physical changes during Quaternary time might show some coincidence between the phases of social evolution and the changes in environment ; further, that it might be possible to correlate the physical and human records and thus furnish a contribution to the time scale of recent geology. “ At my request Professor William M. Davis assumed charge of the physico-geographical part of the preliminary reconnais- sance.” In concluding he remarks that “We have shown that the recent physical history of the region is legibly recorded in glacial sculpture and moraines, in orogenic movements, in valley cutting and terracing, in lake expansions, and in the building up of the 246 Scientific Intelligence. plains, and we have made some progress in correlating these events. “We have also found full confirmation of the statements as to a progressive desiccation of the region of long standing, which has from a remote period continually converted cultivable lands | into deserts and buried cities in sands. ‘We have found, widely distributed, great and small aban- doned sites of human occupation, with evidences of great antiquity. “We have reason to think that a correlation of these physical! and human events may be obtained through a continuance of the investigation, and that archeological excavations will throw light on the origin of Western and EKastern civilization.” In the second article Professor Davis describes his observations upon the Caspian region with its abandoned shore lines up to 600 feet above the present water-level, and the traces of the Pliocene sea whose deposits, as the Russian geologists have shown, under- lie the plains of southern Turkestan. He says of the Piedmont plains that: “Since the withdrawal of the Pliocene sea, the eastern and southern borders of the plains of southern Turkestan appear to have been aggraded by the rivers that flow out upon them from the mountains. That a certain measure of such construc- - tive action has taken place is announced by the Russian geolo- gists, but itis not apparent that the full measure of river action has been recognized. Some of the strata of the plains are said to be not fluviatile but lacustrine, because they are. of fine texture and uniform structure, without the variable layers of gravel that are by implication supposed to be always indicative of river work; but this seems to be a simpler solution than the problem deserves. ‘There are many rivers that do not carry gravel, and there are many river plains whose smooth surface must receive very even and uniform deposits of flood-laid silts over large areas. Records of boring are quoted by Walther which show river muds on sand and loess to a depth of nearly 50 meters beneath the bed of the Amu River at Charjui, where the great railroad bridge was built. The record of a well boring at Askhabad, quoted by the same author, shows variable pied- mont deposits over 2,000 feet deep. It seems, indeed, as if we had in the plains of Turkestan and the Great Plains of our West one of the most striking of the many physiographic resem- blances between Eurasia and North America; and that there as well as here an increasing share may be given to the action of agerading rivers in forming the plains, as observations are extended. It is well known that the tide of geological opinion in this country has in recent years turned more and more toward a fluviatile origin for the strata of the Great Plains that slope eastward from the Rocky Mountains, and the traditional lacus- trine origin of the plains strata has been repeatedly questioned ; sO we may expect, as closer attention is given to the details of river-laid formations, that a larger and larger share of the fresh- Geology. 24:7 water strata that slope westward from the mountains of Central Asia may be interpreted as fluviatile rather than as lacustrine.” “The irregular structure of the piedmont slope, as exposed in cuts along the railroad line, is well described by Walther. There is a frequent and irregular alteration of stratified or massive loess-like clay, finely stratified sands, and coarse gravel, with many local unconformities; all this being the result of the varia- ble action of floods that sweep suddenly, unguided by channels, down the piedmont slope; now eroding, now depositing; here sweeping along coarse blocks, there depositing fine silts. Ten miles south of Askhabad, where the railroad station is 819 feet altitude, we saw, when returning by the Meshed road from an excursion in the Kopet Dagh, more abundant piedmont deposits of mountain-waste dissected to depths of several hundred feet. A great thickness of these deposits has been penetrated by the artesian boring in the suburbs of Askhabad, already mentioned, 2000 feet deep, and therefore with more than half its depth below sea level, but without securing a water supply. The whole depth, as shown in the record quoted by Walther, is in variable layers of clay, sand, and gravel, similar to the deposits seen in the barrow-pits near the railroad embankments, or in the natural sections; and all of this heavy deposit is therefore best explained by conditions and processes like those of to-day during persistent depression of the surface. The failure to secure a water supply from this deep well is in itself very suggestive of the irregular underground structures and of their torrential origin.” An excursion into the Kopet Dagh and the mountains of Persia revealed abundant evidence of sub-recent terracing in the valleys of a character to suggest a relative uplift of the heart of the chain. The desert plains from Askhabad to Samarkand are characterized by aggrading rivers. “The most notable feature of this district was the absence of valleys. The rivers have channels in which their waters are usually restrained, but there were no valleys in which the river floods were limited. The plains were open to overflow as far as flood supply held out. We were told, however, that some distance upstream (to the south) the Murg-ab has a flood-plain slightly depressed beneath the plain. This we interpreted as meaning that the river had there changed its habit from aggrading to degrading. On crossing the Amu at Charjui we saw a low bluff on the north or right of its course, although on the south the plain is not significantly above the river. “The general absence of valleys is a natural, indeed an essen tial, feature of a fluviatile plain in process of aggradation by flood deposits. It is peculiarly appropriate to rivers like the Tejen and Murg-ab, which dwindle away and end on the plain, so that every grain of sand and every particle of silt must be laid down as the water volume lessens and disappears. The absence of valleys would, on the other hand, be surprising in a 248 Scientific Intelligence. lacustrine or a marine plain, for the reason that coincidence could hardly be expected between the slope that might be given to such a plain when it is laid bare and the slope that is satisfactory to the graded rivers that run across it. It is not, however, as has already been pointed out, always the case that fluviatile plains have no valleys eroded beneath their general level. The river- made plains of northern India are now commonly somewhat trenched by their rivers. Our Great Plains, piedmont to the Rocky Mountains, are likewise in process of dissection by their rivers. The plains of Turkestan are therefore somewhat excep- tional in this respect. As a result we had unfortunately no opportunity of seeing sections of the plains in which the struc- ture of the deposits could be examined. A well on the Ozar’s estate at Bairam Ali, a modern village near Old Merv, where we were most agreeably entertained by the superintendent. Mr. Dubassof, was said to have shown nothing but ‘sand and loess.’ The desert and river deposits found by borings beneath the Amu River beds at Charjui have already been noted. The inspection of these vast plains of silt was very suggestive in connection with the problematic origin of the fresh-water Tertiary forma- tions of the western United States. Certainly no one who sees the river-made area of the plains of Turkestan can doubt the capacity of rivers to lay down extensive fine-textured deposits.” In regard to the Tian Shan mountains Professor Davis states that “ A number of the mountain ranges that we saw were of vigorous form, with sharp peaks and deep-carved valleys, in which it was impossible to recognize any trace of the original unsculptured mass ; but certain observations made in the central and northern ranges, near Lakes Son Kul and Issik Kul, and on the steppes that border the mountains on the north, led to the belief that the region had been very generally worn down to moderate or small relief since the time of greater deformation, which probably occurred in the Mesozoic age; that large areas of subdued or extinguished mountain structures are still to be seen in the low ranges and in the steppes north of the Ili River; and that the present relief of many of the higher Tian Shan ranges is the result of a somewhat disorderly uplift and of a more or less complete dissection of dislocated parts of the worn-down region. Mr. Huntington’s report shows the application of these conclusions to a large part of the central and southern Tian Shan.” The space devoted to a notice of so wide ranging a report forbids further detailed mention of the numerous observa- tions of the author upon river and glacial phenomena of the valleys of the Tian Shan. In the article by Mr. Pumpelly, an account is given of the Kara Kul, a lake of bitter salt water, and its desert shores, and also a good description of the moraines in the mountains. Indications of two long-separated ice advances were noted and signs of a feeble third. Variations of lake level and ice advance are attributed to climatic control. Hviden-e is discussed Geology. 249 to support the supposition that in early Pleistocene time the Alai mountains wasted down until a detritus-covered piedmont plain formed on the north of the range, whereupon a dislocation seems to have occurred nearly parallel to the range and north of it with sinking of the plains still farther north or with uplift of the range. The relations of the river work to this change of altitude are briefly explained. In his article on central Turkestan Mr. Huntington gives a summary of the geology and topographic development. Of the Paleozoic series he states: “In Central Turkestan a single suc- cession of strata is repeated again and again, with only slight local modifications. The oldest observed formation is an ancient white marble, shot through and through with intrusions of gran- ite. It was noticed only in the Alai Mountains in the neighbor- hood of Kok Su and Karategin. Its junction with the overlying formation was not seen, but the contact presumably shows an unconformity, as a conglomerate near the base of the covering strata contains pebbles of the marble. The granite which is intruded into the marble is of much later date, for it occurs abundantly in the Paleozoic series in the ridges of the Tian Shan plateau and along the north side of the Alai range. The main body of the Paleozoic series is a great thickness of limestones, many of them slaty, which are stated by Tchernachef to be of ‘Devonian and Carboniferous age. They are greatly folded and have been penetrated not only by granite intrusions, but also by some basaltic lavas, as may be seen, for instance, in the Sugun Valley west of Shor Kul. The folding of the Paleozoic strata is of the sort which is associated with mountain building, hence at the end of the Paleozoic era or in the early part of the Meso- zoic this part of Central Asia must have been highly mountain- ous. In evidence of this it may be pointed out that the succeed- ing unconformable conglomerates are so coarse that they could only have been formed subaerially in a region of considerable relief, and yet at the time of their deposition the old folds of limestone and slate had already suffered great denudation. ThO. + ZrO, ? By dividing Products of the Radio-active Elements. 261 Neglecting for the present the results under rx and xx, which are types of secondary uraninites, it will noticed in an examination of the numbers given in Table I that— 1. In specimens from the same general locality, viz.: from Connecticut, from Norway and from North Carolina, a rough - proportionality is shown between the content of uranium and the content of lead, rare earths, helium (nitrogen) and water. A still more striking relation appears to exist between the pro- portion of uranium in the form of the lower oxide, UO,, and the amount of helium (nitrogen). This was remarked by Hille- brand, who makes the following statement* in connection with the results obtained from the analysis of the first eighteen samples: “Throughout the whole lst of analyses in which nitrogen has been estimated the most striking features is the apparent relation between it and the UO,. This is especially marked in the table of Norwegian uraninites recalculated+, from which the rule might almost be formulated that, given either nitrogen or UO, the other can be found by simple caleulation. The same ratio is not found in the Connecticut varieties, but if the TABLE 1 : Norway. Texas. | S. Carolina. | Canada. Saxony. Xa DIA GE XIV xeVves XVI XGVALE XVIII MIX | XX XOX, XGXGITE es ense 22-04 32:00 35°54 42°71 26°81) 44-17 |. 2:2 | 41-06 59°30 46-13 50°74 43-03 43°88 48°38 24:18 4418 | 2089 _... | 8467 | 22°33 66 66 57 65 68 56 61 Bees ee 65 68 9-04 10°06 858 9:46 9:44 10°54 10°95 | 10:08 8°58 | 11°27 6:39 GUiereas 2. *- 8:98-6:68 2 45 6°39 1:65 6°41 0:0 7-62 9:03 8-43 10-48 8:09 13:42 13°87 | 19:19 10°25 10°49 0-0 meee OR 1-08) 12038 C108) a 124 O54 piso 0°86 0:02 O74 0-75 0°74 0:77. - 0°79 4-23 2. TEASE oe ks PA rnoe Sg S690 9414> 8°32 - 8-96 8:98. 7-50: ... 8:29 Raa ee Oco Of the samples from Norway XII was from Anneréd, XIII and XIV from Elvestad, XV from Skaartorp, XVI from Huggendskilen, and XVII and XVIII from Arendal. Sample XIX was from Llano Co., Texas, XX from Marietta, South Carolina, XXI from Villeneuve, Canada, and XXII from Johanngeorgenstadt, Saxony. determination of nitrogen in the Branchville mineral is to be depended on, the rule still holds that the higher the UO, the higher likewise is the nitrogen. The Colorado and North Carolina minerals are exceptions, but it should be borne in mind that the former is amorphous like the Bohemian and possesses the further similarity of containing no thoria, although zirconia may take its place, and the North Carolina material is * This Journal, x1, 391 (1890). + Excluding the insoluble matter. 262 B. B. Boltwood— Ultimate Disintegration so much altered that its original condition is unknown.” This generalization can apparently be extended to include lead also. 2. When the analyses of samples from the same actual locality are compared it will be evident that, in general, a) The content of rare earths increases with the amount of lead present. This is most strikingly shown in the groups I-V, VI-vill, xim—xtv and xvi-xvir. The simultaneous variation of thorium is also indicated somewhat imperfectly in those instances where this constituent was separately deter- mined. b) That in those specimens having the highest specifie gray- ity (v and vu) the proportion of helium compared with the lead present is greatest. It is in general to be expected that the denser and therefore less porous material would retain a greater proportion of the helium formed within it. The low proportion of gas compared with lead in x and xix might well be due to the high emanating power of the former* and the greater porosity of the latter indicated by its low density. It is moreover interesting to note that those specimens (x, XIX, XXI) containing disproportionately large amounts of water contain a relatively low amount of helium compared with the lead present. It is possible that these minerals were sufficiently porous to permit the entrance of water from with- out while at the same time a part of the helium formed has escaped from within them. It is evident that, in Table I, a lack of agreement exists between the proportion of lead and rare earths and the pro- portion of helium in the Connecticut material and the propor- tions of the corresponding constituents in the Norwegian sam- ples. In the latter the amounts of lead and rare earths as compared with the gas present are much greater than in the former. This can be explained by assuming that the Nor- wegian minerals are considerably older than the American varie- ties, and that the Norwegian specimens examined by Hillebrand have in some manner lost a large part of their helium. The geological data available on the relative ages of the American and N orwegian occurrences, while not entirely in accord with the assumption of such a oreat difference in age, would not appear to be sufficiently definite to preclude such a possibility. In considering the bearing of the results of the analyses of the two secondary uraninites, rx and xxi, on the general theories proposed in this paper, it is evident that the presence * Phil. Mag. (6), ix, 609. Products of the Radio-active Hlements. 263 of the low proportion of lead and helium, and the practical absence of thorium in 1x, is quite in accord with the geological indications that this material is of an age ereatly inferior to that of the primary uraninites. In xxir the very notable amount of lead shown by the analysis would seem to offer no serious obstacle to the theory, since this material occurs inti- mately associated with the sulphide of lead and other similar minerals, and the massive and amorphous form in which it is found would indicate that the conditions under which it was originally deposited were not favorable to the separation of a pure uranium compound. The statement of Hillebrand* that nitrogen (helium) and the rare earths were practically absent in specimens of secondary uraninite from Piibram, Joachims- thal and Johanngeorgenstadt, which he examined, is also of interest in this connection. The experience of Debiernet in separating actinium from a secondary uraninite of this charac- ter, is, however, indicative of the existence of smal! amounts of thorium in these minerals. Other Radio-active Minerals. In the table which follows (Table IL) will be ne some data compiled from various sources on the composition of a number of primary and secondary radio-active minerals. As bearing on the topic under discussion it is interesting to note the following _— i> Phe sr eatest proportion of helium with respect to the uranium and lead present has been observed in those primary minerals which have the lowest emanating power and the highest specific gravity, i. e., in the most compact and least porous minerals. Examples are furnished by thorianite, fergu- sonite, samarskite and monazite. (Of the varieties of thorite, much greater proportions of helium have been observed in the var iety known as orangite, which has also the greatest density.) 2. Greater proportions of lead and helium with respect to uranium are found in those primary minerals which occur in the oldest geological formations. This point is well illustrated by thorianite, which is found in Ceylon in a geological forma- tion which is probably of the Archean period. 3. The primary minerals containing the greatest proportion of thorium are in general the most hydrated. In considering the secondary radio-active minerals certain probable conditions must be recognized. Where these minerals are formed by the alteration of primary minerals in place, namely, where the primary mineral is acted on by underground * Bulletin U. S. Survey, No. 78, p. 72. t Compt. rend., cxxx, 906 (1900). PbO 264 B. B. Boltwood— Ultimate Disintegration TABLE II. PRIMARY MINERALS. Species. Loeality. ThO2 UO, Thorite, Hitterd, Norway ---- 48°66 900’ Mackintoshite, Llano Co., Tex. 45°30 22°40 Yttrialite, Llano Co., Tex. .__- 10-85)" wlsos Phorianite, Ceylone = ee UOISG ne eat Samarskiteen 2° Ge eee .---. 10-18% ms (2) Colorado: 22% 3°64 4:02 Anneroédite, Anneréd, Nor._.. 2°37 16°28 Hniwenite tad oo ei eee Lams EY OG Hielmite, Falun, Sweden ..-._ ? 234" Polycrase, Slattakra, Nor. ---. 3°51 18°45 Fergusonite, Llano Co., Tex... 0°83 7:05° 66 Bini ? Biofeh S Xenotime, Narest6, Sweden... 2°43 3°48? Monazite, North Carolina _.... 5°00 0°40° 126 3°74 0°80 2°59 0°72 2°40 0°92 0°21 0°92 1°43 0°16 0°08 tr. SECONDARY MINERALS. UO; Gummite: North Carolia <2" 2:25" 75-20 Thorogummite, Llano Co., Tex. 41°44 22-43 Carnotite, Colorado, 2325274 0205 2 52325 Uranophane, North Carolina.. ..-.° 66°67 1 Wis Os. 2 UO, 6°03 -+ UO; 9:07. 5°57 2°16 0°25 0°60 H.O He 10°88 X® Ae Bi Oke OFS? Nees X* 0°39% 8-1% X° [258 ieee R19 ae Ail ? 223 oe AT lee Ne 20 Oar ee OOS) Li 7 fea 0-20) Xe 10°54 ? 7°88 2 3°06) 2 Oe 1202) a Reference. Dana, p. 488 sae va vee Dana, p. 740 Dana, p. 741 Dana, p. 744 Dana, p. 742 Dana, p. 745 Dana, p. 730 Dana, p. 749 Dana, p. 892 A 5 BN. Dana, p. 699 > Hofmann and Strauss (Berichte, xxxili, 3126) state that they found both thorium and lead in samarskite and in euxenite. 4 UOs. 5 UO; and Wi@r 6 The composition of monazite given above is derived from experiments of the writer. 7 Specimens of gummite from North Carolina analyzed by the writer have been found to contain from 2 to 8 per cent. of thoria. 8 It is stated by Adams (this Journal, xix, 321 (1905) that helium is absent from this mineral, which is to be expected since it is highly porous and of recent formation. ’ Samples of this material have been examined by the writer in which no thorium could be detected. X* Helium has been found in the variety of thorite known as orangite. X> Hillebrand’s experiments suggest the presence of helium in this mineral. X¢ Including this species among the primary minerals is possibly open to objection. from 1°¢ to 2° of helium per gram. Hillebrand’s experiments would seem to indicate that it contains X4 The analyses of Dunstan ‘and Blake (see Ref.) do not indicate the presence of water, but several tests made by the writer, on samples kindly supplied by Mr. Geo. F. Kunz, suggest the presence of water in quite notable quantities. X° The occurrence of helium in samarskite, hielmite, polycrase, xenotime, monazite, orangite, and other radio-active minerals is described in papers by Ramsay, Collie and Travers (Jour. Chem. Soc. Lond., Ixvii, 684) and Ramsay and Travers (Proc. Roy. Soc. Lond., 1x, 442). A, W. F. Hillebrand, this Journal, xlvi, 101 (1893). — A» Hillebrand, this Journal, xiii, 195 (1902). As Dunstan and Blake, Proc. Roy. Soc. Lond. (A), Ixxvi, 253 (1905). A, Ramsay and Travers, Proc. Roy. Soc. Lond., lii, 316 (1898). A; Hillebrand and Mackintosh, this Journal, xxxviii, 480 (1889). A, Hillebrand and Ransome, this Journal, x, 120 (1900). Products of the Radio-active Klements. 265 waters, etc., with the removal of certain constituents and the substitution of others originally dissolved in the waters, the resulting hydrated residue will in some cases consist of a mix- ture of several different chemical compounds and its general composition will not correspond to any definite formula, but will depend on chance and the accidental local conditions. An excellent example of a secondary product of this character is afforded by the mineral known as gummite, which occurs as an alteration product of the North Carolina uraninites. Sam- ples of this mineral from the Flat Rock mine have been exam- ined by the writer, in which great variations in the proportions of lead, thorium and uranium present were observed in samples removed from different parts of the same comparatively small specimen. The mineral known as uranophane from the same locality shows corresponding variations in composition. Both these substances are amorphous in structure but very frequently occur with a crystalline form as pseudomorphs after the origi- nal uraninite. It is obvious that these facts must be considered in attempting to arrive at any conclusions from a chemical examination of these materials. In other cases the percolating waters undoubtedly dissolve the more readily soluble components of the primary minerals and deposit them again as definite, crystalline compounds of a relatively high degree of purity. Examples of this sort are afforded by such minerals as torbernite [Col UO) EO, 3H,01, autunite [Ca(UO,),P,O,°8H,O,], uranocircite [Ba(UO,),P,0, 8H,O], zeunerite [Cu(UO,), As,O,-8H,O], uranosphaerite [(BiO),U,O,°3H,O], and a considerable number of others. The examination of minerals of this character will probably afford data of considerable value on the nature of the ultimate disintegration products of uranium. The mineral mackintoshite is quite possibly of secondary origin, but owing to some doubt in the matter it has been placed among the primary minerals. It may represent a variety of thorite, containing originally a considerable propor- tion of uranium, ‘which has under gone alteration owing to the radio-active processes which have taken place within it. The evidence is strongly in favor of the view that the thorogum- mite has been formed from the alteration of the mackintoshite through external causes. Any definite conclusions at present as to the formation of carnotite are quite impossible. Its composition and occurrence are both so unique that little or no analogy with other known uranium compounds can be detected. It seems highly probable, however, that its age is not relatively very great and its general composition, ¢. g. the low amount of lead present and the practical absence of thorium and helium, is quite in accord with such a conclusion. 266 B. B. Boltwood— Ultimate Disintegration An interesting radio-active mineral has been described by Danne.* This substance is stated to be a phosphate of lead, or pyromorphite, containing quantities of radium equivalent to about 6 per cent of uranium. It is asserted, however, that no uranium is present in the mineral, although considerable deposits of uranium minerals are known to exist at no very great distance in the same region where it occurs. According to Danne, the pyromorphite is found in fissures through which underground waters containing radium salts are constantly per- colating, and he suggests that the radium contained in the min- eral is derived from the water. It might also be conjectured that the lead of the mineral has resulted from the disintegra- tion of radium, the radium itself having been formed from the disintegration of uranium in the neighboring deposits. Occurrence of Minerals. It would seem possible that some general data on the disin- tegration products of radio-active substances might be derived from the study of the conditions under which the radio-active minerals occur in nature. The following suggestions may perhaps be of interest im this connection. The primary min- erals found in the pegmatitic dikes include uraninite, thorite, fergusonite, aeschenite, euxenite, columbite and monazite, all of which, with the exception of columbite,t probably contain thorium in greater or smaller proportions. The theory gener- ally accepted by geologists is that the pegmatites were formed under conditions of so-called hydro-igneous fusion, involving high temperatures and the presence of considerable water vapor which was prevented from escaping by the high pressure due to incumbent masses of rock of great thickness. Assum- ing the prior existence of considerable deposits of uranium compounds at great depths, it would appear probable that in an upheaval of deep-lying material, with the intrusion of the plastic magma into the upper layers from below, the conditions would be favorable to the separation of the various constituents of the already partially disintegrated uranium with the pro- duction of new minerals representing new combinations of the various elements present. Thus some of the uranium might separate out as the oxide (uraninite), either quite free from other elements or with admixtures of other isomorphous oxides (thorium oxides and other rare earth oxides), while the thorium might be greatly concentrated in the form of such minerals as thorite and thorianite, containing mixtures of variable propor- * Compt. rend., cxl, 241 (1905). +The very common association of radio-active elements with niobium, tantalum, etc., in minerals is possibly significant of some ultimate relation between them. Products of the Radio-active Hlements. 267 tions of uranium and the rare earths. Others of the rare earths present might be themselves concentrated to form such minerals as allanite and gadolinite, compounds containing but relatively small proportions of the radio-elements. When uraninite is found in metalliferous veins the general indications point to its transportation hither from greater depths by thermal waters and its deposition at a temperature considerably lower than that existing in the plastic pegmatite. The association of the secondary uraninites with the sulphides of iron, copper, lead, bismuth and other metals is indicative of conditions of deposit unfavorable to the simultaneous produc- tion of rare earth minerals, which have never been observed to occur under similar conditions in any locality. The mode of occurrence of radio-active minerals would therefore appear to offer certain valuable data on the processes taking place in the radio-elements and the products formed by their disintegration. | Origin of Elements. If it can be ultimately demonstrated that lead, bismuth, barium, hydrogen and argon, or any one of them, actually result from the disintegration of uranium, an interesting ques- tion which naturally arises will be: Have the quantities of these chemical elements already existing been produced wholly in the same manner? Any discussion of this problem at the present time would certainly be premature, but the time may not be very far remote when this question will deserve serious consideration. Summary. Various data have been presented which are interpreted as indicating that the ultimate disintegration products of the radio-elements may include lead, bismuth, barium, the rare earths, hydrogen and argon. | The writer is fully conscious of the meagerness of the data upon which the hypothesis of the production of these substances is founded, but the suggestions are made in the hope that the attention of other investigators may be directed to the possi- bilities offered by a careful study of the composition and occurrence of the radio-active minerals, and that their interest may be sufficiently awakened to induce them to independently undertake the experimental investigation of the theories which have been suggested. 139 Orange St., New Haven, Conn. August 16, 1905. 268 Llora—Lstimation of Cadmium taken as the Sulphate. Art. XXIX.—The Use of the Rotating Cathode for the Esti- mation of Cadmium taken as the Sulphate; by Cuarizs P. Fora. [Contributions from the Kent Chemical Laboratory of Yale Univ.—cxxxix. | In a recent paper from this laboratory* has been described the application of the rotating cathode to the rapid estimation of copper, silver and nickel; and, in a later paper,t its fitness for the estimation of cadmium as well as several other metals has been shown. The object of the present investigation has been to more thoroughly study the conditions under which cadmium may be estimated by this means. The apparatus used was that described in the previous papers. Since it had already been shownt{ that cadmium taken in the form of the sulphate can be estimated by deposition of a solution shghtly acidulated with sulphuric acid, this formed the natural point of departure. I. In Solutions containing Sulphuric Acid. A solution of cadmium sulphate was prepared, containing approximately 16°6 grams of the salt to the liter of water. Portions of this solution were carefully measured from a burette, diluted to the desired volume, a few drops of dilute sulphuric acid (1:4) added, the proper connections made and the electrolysis conducted as previously described. The fol- lowing were the results obtained upon two different solutions : SOLUTION A. No. Sol. H.SOQu.. Cur’t E.M.F. Cd. of taken. (1:4) Time. read. = N.D.i00 approx. found Bxp..* 4em?-> drops. , min. amp amp. volts, erm. Ike 15 a) 18 0°4—1:0 1°2—3°0 8 O- aut 2. 15 9) 10 0°4—0°5 dl ed 8 0°1090 3. 15 9) 16 0°4—0°9 ]°2-2°7 8 O°1115 4, 15 7 35 0°5-1°0 PSB) 8 071117 5. 15 12 2D Oleg) 3°0—4°5 § O-1115 6. A 10 35 1°0-1°5 3°0-4°5 8 01120 le 15 18 30 1°5—2°0 4°5—6'0 8 Ould ig 8. 15 15 25 1°5—2°0 4°5—6°0 8 Ov LEG 9. 30 ~=Indef. 15 3°O0—4°0 9°0-—12°0 8 0°2235 10, 20 12 35 2 0-3°0 6:0—-9:°0 8 0°1491 EL. 15 io Andel e260 2°0 6°0 8 Ol E20) In experiments numbered 1 to 4, the liquid at the end of the period indicated showed traces of cadmium remaining, but in the seven experiments following these the cadmium was all deposited upon the cathode in a satisfactory condition. These * Gooch and Medway, this Journal [4], 320 (1908). + Medway, - is Journal [4], xviii, 56 (1904). 1 Toc. eit: Flora—Estimation of Cadmium taken as the Sulphate. 269 results were therefore taken as indicating the standard of the solution used, the mean of the series showing the presence of 0-007454 germ. of cadmium in each cubic centimeter of the solution. In a second solution which it became necessary to standard- ize the following results were obtained : SoxLvuTiIon B. No. Sol. Curt Cd. of taken. Time. read. = N.Dioo E.M.F. found. Exp. cm. min. amp. amp. vts. erm. qT: 20 27 1:0-—1°5 2 O0-- 4° 7:9 00816 2. 25 30 2°0—3°0 6°0— 9:0 7:9 0°1018 3. 25 dd 2°5—4°0 7°5—-12°0 7°6 01019 4, 30 25 2°0-2°5 6°0- 775 12°0 0°1224 dD. 30 20 1:0 3°0 Lo 0°1228 6. 30 10 1°5—2°5 A°5— 7°5 7°8 0°1226 The mean of these six experiments gives a value of 0°10194 grm. of cadmium for every 25°™* of the solution, or 0:0040776 grm. for each cubic centimeter. This value was taken as the standard whenever this solution was used. One point which was not mentioned in the former paper* on the estimation of cadmium by this method, but which is of much importance, is that of dilution. The earlier experiments in this work were performed at a dilution of from 65°™ to 75°™, Much trouble was experienced, however, at this dilution; for the last traces of the metal were driven from the solution only with extreme difficulty and with much loss of time, as may be noted by comparing the time interval of most of the experi- ments with the shorter interval of the last two experiments of the second series, where the dilution was 45 to 50™. More- over it was found advisable, in order to avoid mechanical loss, to deposit not more than 0:2 grm. to 0°25 grm. of the metal upon the cathode, while even smaller quantities are to be preferred. The current density must also be kept within the limits indicated; for otherwise a spongy deposit may result. Cadmium seems to be especially liable to the formation of these spongy, unweighable deposits, and the greatest difficulties experienced in this investigation have come from this behavior of the metal. The best condition, therefore, may be briefly summarized as follows: Cadmium sulphate, equivalent to not more than 0-2 to 0°25 grm. of the metal, is dissolved in 45°™* to 50°™ of water ; ten to fifteen drops of dilute sulphuric acid are added ; and the proper connections made and the solution subjected to electrolysis as described, fifteen minutes being sufficient time tor the complete deposition of the metal upon the cathode. It * Loe. cit. Am. Jour. Sci.—FourtH Series, Vou. XX, No. 118.—OcToBER, 1905. 19 270 =Flora—Lkstimation of Cadmium taken as the Sulphate. is not necessary to heat the liquid, as the passage of such large currents soon heats it sufficiently. When electrolysis is com- plete, the excess of sulphuric acid may be destroyed with a shght excess of ammonia water, the current broken, and the cathode removed, thoroughly rinsed with water and alcohol, and dried by waving over a free flame. If the deposit is not spongy the drying is a matter of only a few nioments, and there is no danger of oxidizing the metallic deposit. If it is preferred, the current may be reduced by interposed resistance, the rotation stopped, and the liquid readily siphoned without danger of injuring the metallic coating. Il. Ln Solutions containing Acetates. The next method to be studied in its application to the rotat- ing cathode was the use of solutions containing acetates, as recommended by Edgar F. Smith. Originally, Smith used a solution obtained by dissolving cadmium oxide in acetic acid* but later found that the electrolysis proceeded equally well in solutions containing the nitrate, chloride or sulphate of cad- mium with an excess of sodium acetate.+ In the study of the application of this method to the estimation of cadmium, taken as the sulphate, upon the rotating cathode, two methods of proceeding were followed, both of “which had been previously used by Exner} in his work upon the rotating anode. In series A, of the experiments following, measured amounts of cadmium sulphate solution were run off from a burette, the indicated amount of sodium acetate was added in solution, a small amount of potassium sulphate was added to increase the conductivity of the solution, the whole diluted to the desired volume and electrolysis conducted as with the solution contain- ing sulphuric acid. In series B, the cadmium in the measured solution was precipitated as the hydroxide with a sodium hydrate solution, the precipitate dissolved in a very slight excess of acetic acid, potassium sulphate added as before, and the solution subjected to electrolysis. SERIES A. Cd. Cur’t taken. NaOC.H;0. K.SO,.. read. = N.Dioo. E.M.F. Time. (Cd. fd. Error. No. grm. grm. grm. amp. amp vts. min. grm. grm. i 0°1864 1g) 0°5 2°0 6°0 8°0 : ($) aes 2. 0°1491 2°0 1:0 1°5 4°5 8:0 15 (||) eee 3. Of TNS 0°5 ue) 1:0 3°0 8°0 20 Osi + 0°0005 + 0°1491 1°5 0°5 0°9 Dei 8°0 15 0°1494 + 0°0003 5. 0°1491 1°5 0°5 0-9 742) 8°0 15 0°1496 + 0:°0005 6. 0°12238 1°5 0°5 0°75 22 US 20 0°1237 +0°0014 * Ber., xi, 2048 (1878). + Am. Ch. J., ii, 41 (1880). + J. Am. Ch. Soe., xxv, 896 (1903). & Did not weigh, as precipitate was non-adherent. cadenium present. | Donen spongy and blistered. Too much electrolyte present. Current too high for quantity of DOT eR Oo pO Flora— Estimation of Cadmium taken as the Sulphate. 271 SERIES B. Cd. Cd. ‘taken. NaOH. K.SO,. Cur’t= N.Djoo E.M.F. Time. found. Error. erm. grm. crm. amp, ~ amp. vis. min. grm. erm. 0°1491 excess 0°5 12554 29) 830 10s yO 4 96 + 0°0005 61491 0°2 OS real Vets) 2°4 80 15 Ost 491 + 0°0000 071491 O'2 O5d>- 9-058 2°4 8°0 15 =—0°1493 + 0°0002 0°12258 0°5 O22 ese Oe nOme eb 22Ots Oe O41 225 + 0°0000 0°1223 0°5 Uae tl 0) SneOr erm Os Ze 021225 + 0°0000 0°1223 0-2 0°5 2 eo a 75 OO ale 2 +.0°0004 In both series the volume of the solution was about 60° to 65°. The sixth experiment in each series will indicate the result when the greater concentration of 45°" to 50°™* was tried. In these cases the precipitate showed a tendency to sponginess, which was more noticeable in series A. At the greater dilu- tion, the deposition of the cadmium proceeds rapidly and satis- factorily ; ; the deposit is rather crystalline, fairly compact, and easily washed, so that the method forms one of the very best where the cadmium is taken in the form of the sulphate: the chloride and nitrate behave differently and will be treated later. The second modification seemed to give deposits more satisfactory than the first. Certain cautions, however, are to be observed. Not more than 0°1500 grm. may safely be esti- mated; the normal current density should not exceed 3-0 amperes if a spongy deposit is to be avoided; and, for the same reason, a large excess of electrolytes is to be avoided. II. Ln Solutions containing Cyanides. The deposition of cadmium from a solution of the double cyanide has always been very satisfactory, and the results with the rotating cathode were in complete accordance with previous work on this method. The range of conditions of current and quantity of electrolyte is broad, the deposit is a beautiful silvery plate, so compact as to be rubbed off only with dift- culty, which dries very quickly; and although the complete deposition of the metal is not so rapid as it is from solutions containing sulphates or acetates, it is sufficiently rapid. Care should be taken to avoid foaming of the solution, as this retards somewhat the deposition of the final traces of cadmium. Generally, a volume of 65° to 70° was found most satisfac- tory. The solution was run off into a beaker of convenient _ size, the cadmium precipitated with sodium hydroxide, and the precipitate redissolved in potassium cyanide. The following results were obtained : 272 =Flora—Lstimation of Cadmium taken as the Sulphate. i Cur’t Cd. taken. “NaOH... KCN: read) NaDion. E.M.F. Time. found. Error. No. grm. erm. grm. ~ amp. amp. vts. min. grm. erm, PONT AOL 1B Oe 5 8 35 0'1498 +0-0007 2. O-1491 1°0 05 25-45 7:5-13°5 8 30 0°1490 —0-0001 3. 0°1223 1°5 1:0 2°5 bus EO 8 35 0°1225 +0°0002 IV. In Solutions containing Pyrophosphates. Brand* has recommended the use of a solution containing sodium pyrophosphate for the electrolytic estimation of metals, among others, cadmium: and the fitness of this solution for use with the rotating cathode was now studied. In each ease the cadmium was precipitated with the indicated amount of sodium pyrophosphate, the precipitate dissolved in an excess of ammonium hydroxide (series A), phosphoric acid of 1:7 specific gravity (series B), sulphuric acid (series C), or hydro- ehloric acid (series D), and subjected to the action of the eurrent. The volume of the solution was 60°’. While fairly accurate results may be obtained, the method is neither so accurate as those previously described, nor are the conditions so flexible. Particular care must be used to avoid too large a current, as a spongy deposit may result. The following were the results obtained : SERIES A. Cd. Cd. taken. NaiP20O;. Cart = 2 N. Dios: )) Ee Mow Times one Error. No. )grm.. grm. NH,O8. amp. amp. vis. mins. ono germ. 1, 0°1491 05 15°7°(1:4) 1°0-1°5 3°0—-4°5 8 15 0°1498 +0-0007 2 OAD Oxo e€XCess. 0°4 1:2 8 15 01489 —0:0002 8. 0°1864 9 0:5), 15°" " (cone) 4.9,0°7 Za 8 15 0'1869 +0°00005 SERIES B. H;PO, (1°7 sp. gr.) 4, 0°1491 1:0 oe 1:0 30 8 30 0'1496 +0 0005 5. 0°1864 1°0 IOs 1'0-1'°d = 8" 0—-4°5 8 30 O°1857 —0:0006 6, -OMAOK 410 AO 0) 3°0 8 30 01493 +0°0002 SERIES C. H.SQx. TOON SaeO Qemr(d: A) © 2:0-2°5 2-0-7") 8 30: O'1501. + 0;0010 8. 0°1864 0°5 excess. 1:0—2:0. 3:0-6°0 8 30 ©0°1862 —0°0002 SERIES D HCl. 9. 0°1491 0°5 sit. excess. 1°0 3°0 8 37 O°'1499 +0°0008 10, O:L491G. 0-55,-25* Fs: 1:0 3°0 8 36 0°1486 —0°0005 *Z. anal. Ch. xxviii, 581 (1889). No. a PSD MTS oR PP Flora—Estimation of Cadmium taken as the Sulphate. 278 In Nos. 1 and 3 a small amount of dilute sulphuric acid was added to increase the conductivity of the solution, but the time was not reduced thereby, while the resulting deposit was slightly spongy. In No. 3 the cadmium was not quite all precipitated. In No. 7 the precipitate was spongy. V. In Solutions containing Phosphates. The use of a solution contaming the orthophosphates dis- solved in phosphoric acid has_been recommended by Smith*, and this solution was next tried. The following results will show the scope of the modifications tried: Cd. H.PO, Total Od: meen PNAS OO, (1:7). vol. Cuart.=- N.Dioo.3-E.M.F. Time. -fd. Error. grm. erm. ems 726M. >" amp. amp. Vis. Min: > erm. grm. 0°1491 0°5 1°0 (ose OS) 2) 30-136 8 20 071477 —0:0014 0°1491 0°5 5-02 foe -2°0=2°5 -6°0=- 725 8 23 0°1496 +0°0005 0°1491 0°5 a:08 756 -2:0-1:5 > 6:0= 45 8 30 071508 +0:°0017 0°1491 0°5 3°0 75 2°5 4D 12 25 0O°1502 +0°0011 . 071491 0°5 4:0 75 2°5 iD 8 25 01485 —0:0006 071491 9°5 2°0 75 2°5 75 12 0d =0°1501 +0°0010 0°1864 0°25 2°0 75 2°0-3°0 6:0-— 9:0 12 40 0°1861 —0-:0003 0°1491 0°3 1°5 Ths 3°5 10°5 2 30 6071502 +0°0011 0°1019 0°25 2°0 75 2°5-3°0 7°5—-9:0 TS. 2305 0-024: + 0:0008 071019 0°25 5°0 75 3°0-3°5 9°0-10°5 78 30 0°1027 +0°0008 0°1225 0'2 5°0 75 3°0-3°5 9°0-10°5 78 40 071221 —0°0002 No. 1.—Not all out. No. 2.—Slight yellow color with H,S. No. 8.—Shght yeliow with H,S. No. 4.—Spongy. No. 5.— Not all out. No. 6. —Spongy. No. 10.—Spongy. No. 11.— Slight test with H,S. From this series of experiments it may be seen that the method may be made to give fair results if the following con- ditions are closely adhered to: for a total volume of 75°™*, the cadmium is precipitated with 0-25 erm. of hydrogen disodic phosphate, 5™* of phosphoric acid (sp. gr. = 1:7) added, and the solution electrolyzed with a current of about 8 volts poten- tial. If the normal current density does not exceed 9 amperes, the deposit will be fair, and complete in about 30 minutes. VI. In Solutions containing Oxalates. Much work was expended upon the oxalate method, but in spite of this, a satisfactory deposit could not be obtained. When ammonium oxalate was present, even in small amounts, the deposit was very spongy: while the use of sodium oxalate alone, when carried down even to the smallest excess possible *Am. Ch. J., xii, 329 (1890). 274 Llora—kstimation of Cadmium taken as the Sulphate. to give a soluble double oxalate, The dissolving of the pr ecipitated oxalates in various reagents furnished no solution to the problem. The results in the following table will show the scope of the work done: gave results much too high. | Cd. Am. Pot. nittiel coy Ae 1a eOAOhae No.| tkn. oxalate. oxalate. Solvent mes nee ie a8 fd. Errem | grm. erm. erm. P P. cope = = grm. seer 1. 01491) excess none none 3-0 9:0 |12 | 2 | 0-1558 | +0-0067 2. 0:1491 | slt. excess none ie 25 73d |12 | 80 | 0°1506 | +0-°0015 3. | 01491 2° 0°5 z 2-0 -2°5 | 6-0 — 7:5) 8 | 25 | 01531 | +0-0040 4. |0-1491| none 0:5 his aps 255 75 |8 | 30 | 0-1476| —0-0015 5. | 0°1491 « 8-0 none 2-0 6-0 6-2} 20 | 0-1507 | +0°0016 6. | 0°1118 4:0 none “ 1°5 4:5 6-1} 20 | 0°1272 | +0-0154 aol | § NH,OHs ) ux ; . Lee none 5:0 \ fewom., | 20 -8°5 | 6-0 -10°5| 6-8] 18 Q Q 8. 01118 “ | 5:0 | none 1:5 4:5 6-0) 15 | 0:1129 | +0-0011 9. 6°1019 “ | 4:() «6 2:0 -3:0 6-0 — 9-0 8 | 20 | 0-0995 | —0-0024 10, 01019 «“ | 6-6 “ 0-5 -1°5 | 1-5 — 4:5) 4-6] 35 | 0-0982 | —0-0037 11. | 0:1019 « 5-0 « 1-0 3-0 5°5| 55 | 01033 | +0-0014 12. | 01019 «i 7-0 ve 0-5 15 0=6«| 4 «| «58 | 01080 | +0-0011 13. | 0-123 2-0 8-0 z 0-1 -0°15 0°3 -0:45| 4 | 60 | 0-1286 | +0-0013 14. | 0:1223 2-0 8:0 0:02-0:10 -06- 0°34 | 76 | 0-1229 | +0-0006 Keses KOH H.SO, dil.| § ae acid i : ‘ ; 15. 0°1228 user. tie aoe 3-0 90 | 8 | 40 | 0-0876| —0-0347 ; aya |b Olek oxalic acid 16. | 071228 |) 0-95 orm SEs yee Sas Bier are <3 17. |o1019| 2 Ye ee Bea io 3-6 | 8 | 35 | 0-1088| +0-0014 ; | § oxalic acid : G 18. | 0°1019 is a hos ei 3 9 8 | 60 | 0:1018 | —0-0001 Of these, numbers 1, 2, 8, 6, 8, 9, 14, 15, 17, and 18 gave very spongy precipitates, while No. 7 was so spongy that it could not be satisfactorily dried, and so was not weighed. In experiments numbered 4, 9, 10, 14, 15, and 18 the “cadmium was not all precipitated in the time allowed. In No. 10, also the precipitate was non-adherent. In No. 16, the oxalate was precipitated and was not broken up by the current. VIL. Balachowsky* obtained good results by the electrolysis of solutions containing, in addition to cadmium salts, urea and various aldehydes. These solutions were found to offer no difficulties with the rotating cathode, when cadmium sulphate is taken, as may be seen from the following results. deposits ‘were gray, compact, and quickly dr ied. The solution was diluted to about 60° or 70, and the best current poten- tial was found to be that given by six storage cells connected in serles—approximately 11°8 volts. * Compt. rend., cxxxi, 385 (1900). In Solutions containing Urea, ete. Tes Flora—Estimation of Cadmium taken as the Sulphate. 275 SeRtes A.—Urea, 3 grms. Cd. Cd. taken. Cur’t = NDipo.* Lime. found. Error. No. grm. amp. amp. min. grm. grm. f= “01019, ©0°25-0°5 (0°75 =1°5 35 01018. —0-0001 2. 071223 0-2 0°6 395 0°1223 +0°0000 3. 0°1223 0°25—-0°5 0°75-1°5 30 0°1230 +0°0007 Series B.—Formalin, 2°™°. fe 0-bO19). 0:1. 1-0: +03) 3-0 30 071018 —0:0001 Pee Ost 223. - 025-10" 2-0°6 = 30 30 0:1224 +0°0001 ee 501993"... 02 —1:0.- 0°6.=8°0 30 0:1225 +0:0002 Series C.—Acetaldehyde, 2°™°. ee Os10 19%) Ost —0'8 0:3 9-4 "85 071022 +0-0003 ee 21293. O21 0°83 2-4 30 0°1228 +0:0005 Beet 29S) 0° 078 =. 0°38. 284 30 0°1222 —0-0001 Since the conductivity of the solutions containing urea and the aldehydes is comparatively low, the effect of adding elec- trolytes was tried. The rate of deposition was very much increased, but the precipitated metal showed such tendency toward sponginess that this procedure is not to be highly recommended. ‘The following were the tests tried: * SeRreR A.—Urea, 3 grms.; Time, 20 min.; EMF., 78 volts ; Current read, 0°5 amperes; N.Djoo, 1°5 amperes. Cd. taken. Cd. found. Error. No. grm. Electrolyte. grm. erm. Notes. 04019 -K'SO0...0°5 orm. -/0°103 1 * + 070022. “spongy. 2. 0°1019 same 0'1031 + 0°0022 af { H,SO, (1:4) i ; 3. O°1019 ys drps. 0°1023 +40°0004 good ppt. 4. 0°1019 same as 3 0:1027 +0°0008 — slt. spgy 5. 01019 eo) 0°1027 + 0°0008 “ 1 8 drps. Series B.—-Formaldehyde (formalin), 2°5°™; Time, 20 min.; E.M.F. 7:9 volts. Current started at 0°5 ampere and rose to 1:0 ampere at the end of the process (N. Dio. = 1°0-3'0 amperes.) In each case the precipitate was good. Ten drops of dilute sulphuric acid were added to increase the conductivity of the solution. Cd. taken. Cd. found. Error. No. erm. grm. erm. i 0°1019 0°1020 +0-0001 2 0°1019 0°1019 + 0:0000 VU. Ln Solutions containing Formates. The use of the solutions containing potassium formate and a slight excess of formic acid has been recommended,* but I * Warwick, Z. anorg. Ch., i, 285 (1892); Avery and Dales, J. Am. Ch. Soc., xix, 380 (1897). ‘A o (OTH TRON 276 Flora— Estimation of Cadmium taken as the Sulphate. was unable to adapt this method to the rotating cathode. When potassium formate was present in even the smallest amounts the precipitate was spongy, and non-adherent. From solutions containing formic acid alone, however, the metal is deposited in a satisfactory form, but only after long passage of the current. The following results will show the limit of applicability of the process, experiments numbered 8 and 9 seeming to represent the most desirable conditions: In the experiments numbered 5, 6 and 7 the cadmium was not all precipitated in the time indicated, as was shown by testing the solution with hydrogen sulphide. IX. in Solutions containing Tartrates. Solutions containing ammonium tartrate were also tried, but failed to give satisfactory deposits, the deposit in each case being spongy. If the solution contain only tartaric acid, how- ever, in place of its salts, fairly satisfactory results may be obtained, as shown by the following table: Ca Tartaric Cd. KCHO, Cd. tkn. sat. sol. Curt — N. Ditton OHM BP Pimes gee ede Error. erm. em®. HOCHO. amp. amp. vis. min. grm. grm. 071019 2 ae 1:0 —2°0 3-6 8 17 | Not weighed ; Ort D2 3-20 <9 hp O°4 122 8 .. | “ppt. blistered 071223 075 a 0°4 1°2 8 ae r and dropped O12 3e 2 O25 sh 0O°4 1:2 8 ay off. 0-223 oe 15 dps. 0°25-0°8. 0°75-2°4 2 25 0°1228 +0:°0005 Ome ae RS orto 0°25-0°8 0°75—2°4 8 25: 0°1212 —0-0010 0°1223 was ie eas 075 -1°5) «1°95 -—4%5) «612-16 35 ©0°1202 —0°0021 0°1019 ae fee! 32 2075 1-0 less 330 12 60 0O°1022 +0°00038 0°1223 pe: Heeest 0-4 -—1°0 1°2 —3:°0 12 55 0°1218 —0°0005 tkn. acid. Cur’t = N.Diono. © E.M.F. Time. (Cd. fd. Error. Tests with hydrogen sulphide showed that the cadmium was No. grm. erm. amp. amp. vts. min. germ. grm. 1, 0°1223 3 0°5-1°0 = 1°5—3:0 8 20 071212. —0o-0011 2. 071223 2 0°5 1°5 8 30 0°1216 —0-°0007 3. 0°1223 2 0°5 1°5 8 50 0°1215 —9:0008 4, 01019 3 1°5 Rae ER Seal es) 0°1022 +0°00038 not all precipitated in the tests numbered 1, 2 and 3, which ~ were performed at a dilution of 70°™*; experiment 4 was per- formed at a dilution of 50°. It will be noted that, as in the sulphate process, the last traces of cadmium are thrown out of the higher state of dilution only with extreme difficulty. A. J. Moses—Crystailization of Luzonite. 277 Arr. XXX.—The Crystallization of Luzonite ; and other | Crystallographic Studies ; by Atrrep J. Moses. 1. The Crystallization of Luzonite. Tue reddish bronze, fine-grained variety of Cu,AsS, which is found in the copper veins of Mancayan, Luzon Island in the Phillipines, has been generally accepted as dimorphous with enargite, but the minute erystals, “tiny individuals of unrecognizable form,”* observed in the cavities growing from the granular mass have not been measured but rather referred to as ‘indistinct, uneven, striated crystals not rhombic but monoclinic or even Peclince! et Recently Mr. Maurice Goodman, senior field assistant in the Bureau of Mines, Manila, collected a number of |uzonite speci- mens showing these crystals in cavities, from which I selected and measured the crystals here described. Crystals No. 1 and No. 2.—A mass of typical luzonite, free from all visible columnar blackish enargite, showed a number of cavities the walls of which were crystallized ; that is, little detached fragments of the walls under the microscope were seen to be facetted by minute crystals which projected very slightly and the faces of which could be traced down until they merged in the bronze-colored mass. They were not implanted on or enclosed in the mass, but distinctly suggested that the mass on solidifying for med little facets such as form on the cooling of a fused mass of pyromorphite. Itis curious and probably of genetic significance, that the terminal planes of these crystals are decidedly lighter in color and of less brilliant luster than the side planes, the latter suggesting the dark gray otf enargite or stibnite and the former a reddish steel-gray not very different from the tint of the massive luzonite. In more than one instance in which a fracture extended across a crystal into the massive material it was impossible to see any difference in the color or character of the surfaces. Two little crystals were mounted for measurement. No. 1, shown in fig. 1, was only 4 to 4™™ in any direction, but was attached to a fragment of the mass from which it had devel- oped. Signals were obtained in the two-circle goniometer from seven se but were a little blurred. Crystal No. 2, shown in fig. 2, was the largest crystal I observed as a cay ity wall facet, and its terminal face was approximately a rhomb of 14x3"". It also yielded signals from seven faces and a series of sionals from a curved triangular surface. In both crystals the terminal faces were reddish steel-gray and the vertical faces dark gray. Taking the terminal faces * Weisbach, Tscher. Min. Mitth., 1874, 257. + Frenzel, ibid., 1877, 303. 278 A. J. Moses Crystallization of Luzonite. as ¢ = 001, the consideration of the angles in the vertical zone suggested an orientation for comparison with the common orms of enargite as follows: forms of gite as foll 1 2 Enargite Crystal No. 1. Crystal No. 2. Form. Face. Signal. Measured 4d. Face. Signal. Measured @. C001 Al. Sain Bis ae 1 Fair ease (=O Ionlole ale” Bl" 2 Double 49° 02’ 6 eC AOS BG) 5 Fair 48° 44! (2 elinreds e438: 130: é Wes on Jo NW20) eves ee ‘Spa 3 Faint Bl” AL == 1SX0) 2 Fair OO Dy! u Eee Pee 4 Double 19> 17" E tA Sa b=010 3 ho Om tiou e ee is aS C= 100 ae See or i Double 89° 46’ hol x Merits Gat curved Series 90° 0’ The comparison of the averaged angles is: Enargite @. Crystal No. 1 ¢. Crystal No. 2 9. Wm==48° 59! 47” 48° 48’ 48° 56’ N= 29 rod aS), NS. 31° 44 1=20° 58’ 38” DO? DO! eee Et) On 5 eee a=90° bynes 89° 46’ That is, ad? the angles are those of the common forms of enargite within the limits of accuracy that the measurement of minute crystals with rather dull ¢ faces and somewhat striated vertical faces would permit. Crystals No. 3 and No. 4.—The relatively simple erystals from the cavity walls connect directly with the two other more highly modified crystals here described. Upon another specimen and so in contact with the massive luzonite as to be, In my opinion, developed from it, were a number of little, bright, highly modified crystals which like Nos. 1 and 2 are much lighter colored on the terminal faces than on the side or prism faces. Crystal No. 3 was the best of these found, and as shown in fig. 3 it proved to include all the forms of Nos. 1 and 2 as well as those of the later described crystal 4. Its size was 4X$x1™ in the directions @, 6, ¢ respectively. A. J. Moses— Crystallization of Luzonite. 279 From still another specimen of massive luzonite, but resting upon it rather than growing from it, was a little group of black lustrous crystals, the best of which, crystal No. 4, shown in fig. 4, measured 3; X$X2™™ in the directions a, b, c, which, while differing from all described enargite crystals in the pres- ence of a pyramid, P= 223, as its most prominent terminal form, connects directly with crystal No. 3 by the fact that this pyramid and all the other forms of the crystal are prominent on crystal 3. Both crystals were measured in the two-circle goniometer. Crystal No. 3 yielded good to fine signals from twenty-one faces and poorer ones from four others, while crystal No. 4 yielded good signals from twelve faces and poorer from two others. The average results tabulate as follows: Numb Measured angles. Computed enargite angles. Number Form. Crystal.| faces. @ eka @ | p ¢ (001) 3 1 ae 0° Bee aoe e 4 1 acre oe eee | Baa b (010) 3 is Oe 89° 492" 0° 90° a (100) 3 2 90° 044! 90° 90° 90° « Ae 2 Approx. 90-90 *| ee penne 7m ©) 4 AGO rk 90r a. ABS 59°47" | 90° Bees oc 4 4 Sea ace | aoe h (120) 3 4 99° 564’ U0 1.29754 13" 90° + 1 30° 03’ SO as eee pees ei (or ele Z (130) 3 1 MeO ele 907. 20; 58: 38> 5 gue s (011) 3 bree eee SOC 0. aie BO. SOleMG. O51 3 os 0° bn6 18. 0° iG BAO! PATOL): | -:3 Do dy SOM gy wes | 43° 39) 902 [243° B45 6" . 4 2 Bi eA ASAD! si iid otha P (223) 3 4 49° 12’ | 39°54 | 48° 59' 47" | 40° 387 16" 4 4 PS oe es. Ochi een sh Oe * Poor signals. 280 A. J. Moses—Crystallization of Luzonite. The Calculated Angles of Hnargite—The axial elements calculated by Dauber* in 1854 are @:b:¢= 0°8711 31: 0°8248 based upon angles of mm = 82° 7’ and cs = 89° 81’. In 1895 Fletchert calculated new elements a:b: ¢=0°8694 : 1: 0°8308 based upon angles mm = 82° 03’ and ck = 48° 42’ This value of mm is the average of so many measurements that it cannot well be questioned and it is not far from the angles here obtained since the mean of twenty ¢ angles of 110 and 223 is 48° 574’ and Fletcher’s mm = 82° 04’ yields @ of 110 = 48° 592". . Fletcher’s value for ¢, however, considers only the faces k= 101 and is the mean of some fourteen values of ck. The new pyramid, P = 223, is represented on crystals Nos. 3 and 4 by eight good faces and the readings especially in crystal 4 are close. The angles ¢ and p of 223 in crystal 4 vield a:b: ¢= 8698: 1:°8241, essentially those of Fletcher in the case of @ but not so near in the ease of ©. . I have therefore used in my calculation an intermediate value for ¢ of -8274, which is also an approximate mean be- tween the ¢ values of Fletcher and Dauber. In conclusion, these results show that the crystals which form at the solidification of luzonite and those which form possibly later on luzonite have the angles of enargite. In other words, “luzonite ”’ is not an independent species but merely a variety of enargite. 7 I base this claim principally on the angles here recorded for the small and relatively simple crystals Nos. 1 and 2, which are types of the cavity-wall crystals so connected with the massive material that it is impossible to doubt that they are the results of its solidification. Crystal No. 3 I believe to have formed in the same manner but under more favorable conditions, while crystal No. 4 is evidently secondary. The new form P = 223, prominent in both, connects them however. The observed color difference on the terminal faces and ver- tical faces of the cavity-wall crystals, and crystal No. 8, prob- ably has genetic significance. The recorded analysis by. Winkler is of practically pure material, which makes inadmis- sible a theory of crystallographic regularity in elimination of impurities. The comparative dullness of the basal plane in Nos. 1 and 2 might suggest a light effect explaining the color, * Pogg, Ann., lxxiii, 383, 1854. + Mineralogical Magazine, xi, 73, 1895. aang A. J. Moses—Crystallization of Luzonite. 281 but in erystal No. 3 ¢ is bright and the color is still reddish steelgray. Tarnish does not seem to explain it, as enargite usually tarnishes a blue-black, and finally the possible deposi- tion of a thin layer of dark-colored enargite observed on pyrite associated with Morococho enargite seems not to explain, since the cleavage on No. 3 is also of the dark gray color. 2. Crystallized Wolframite from Boulder Co., Col. Mr. Morris K. Jones, of Boulder, Colorado, sent me a sack of tungsten ore from different lodes in the property of the Great Western Exploration and Reduction Co., situated about twelve miles west of the city of Boulder. The mineral, which varies in the percentage of manganese in the different lodes, occurs in most of the specimens as the cementing material of a brecciated rock composed chiefly of fine-grained quartz and partially decomposed feldspar. The spaces between the angular rock fragments are filled with the 5 erystalline black ore, the crystals often crossing the crevices. Occasionally the ore thickens to a considerable mass. On breaking the specimens numerous black brilliant little crystals were found, rarely exceeding 1$ to 2™™ in their longest dimension. So far as observed none of the crystals is doubly terminated in the direction of the 6 axis, but all have grown out in that direction from the mass. This and the frequent existence at the visible end of a rectangular face or cleavage b= 010 suggests at first examination a simple com- bination of the three pinacoids. ‘The actual form, however, is that shown in figure 5. Two crystals, each ending in a 6 cleavage and essentially alike in habit, were measured. Crystal No. 1 was 4 x 4 x 3" in the directions a, b, c, and crystal No. 2 a trifle larger. The forms identified by the measurements were: Prismatic zone—l = 210; m=110; 6=010; all yield- ing good signals from bright decided faces of both crystals. In addition a signal was obtained from both erystals which closely corresponded to d = 310. It was, however, evidently a second element in the striations upon the faces / = 210. 282 A. J. Moses-—Crystailization of Luzonite. The largest face in each crystal was undoubtedly ¢= 001, but these faces were so striated’ that the series of images gave values for p each side of the correct position through four or five degrees, and probably involved various indeterminable domes h Olandh 01. The remaining forms determined were ¢ = 102 well devel- oped; A = 112 minute but bright, and a new form, p = 214 occurring as a narrow truncation. The comparison between the measured and computed coordinate angles for 2, m, ¢t, A and p is: Measured. Calculated. Face. i) p iy p U 67° 362" 90° 67° 344 90° Mm SOS Ney 90° 50° 274! 90° t 89° 524’ 28° 103' 90° UiS har A lle Be, 34° 104! a0 7 bs 347 20% p 68° 45’ 302 17, 68° 6! 3 OreiGen Upon a few of the specimens there were small yellow sphalerite crystals and small crystals of seheelite not suitable for measurement. 3. New Faces on Sylvanite Crystal from Cripple Creek, Col Some three or.four years ago Mr. F. C. Hamilton purchased some telluride specimens from a dealer at Cripple Creek, Col., and presented them to Columbia University. Among these was a mass of 3% 0z. in weight which consisted almost entirely of large erystals and crystal bunches some of them 20 & 5™ in length and breadth. Nearly every one of these was partly coated with a thin layer of chalcedony, but many brilliant faces and cleavages were visible. There were a few smaller crystals upon the mass which were nearly free from chalcedony; one of these was so symmetrical that it was measured under the impression that it was ortho- rhombic and possibly a highly modified krennerite. The angles, however, quickly proved its identity with sylvanite. The dimensions of the erystal were approximately 1x 1 2™™ in the directions @, b,c. For better adjustment the crystal mas mounted in the two-circle goniometer with the large = (010) face parallel to the vertical circle, and centered by ie face and the faces of the zone [100 001]. The results were then transformed. Twenty-six forms were identified, 5 which twenty have been previously described by Dr. Chas. Palache* on erystals from Cripple Creek; two others, J/=101 and p=112, are recorded forms not previously noticed on the er ystals from this locality * This Journal, x, 419, 1900. A. J. Moses—Crystallization of Luzonite. 283 and four are new domes H=102, 7=103, /=203 and L=208. The following angles give the proofs for these previously un- recorded and new forms: Measured p. Calculated p. SA) ETH ON |e SES Nee tel rea eater 34° 354 Byers ase ed GO aye a eee ahs ee PO TQ! 19° 124! SOUT ea eee en ease 24° 508’ 94° 55’ eel enter. Ae ea yee, oe 13° 164 138° 41’ rane ene, eet tee OAS NS! WO Ts For the pyramid p=112 Measured angles-.------ o—3 il 18! p= 33) wade Calculated angles...---- Goo 0s Sh N= OA oOo s : The oceurring forms and their relative development may be judged by the following tabulation. The forms in the first column are in most cases composed of fine relatively large faces; the largest, however, being the three pinacoids and the three domes 102, 101, 303. All of these domes are new to the locality. Faces yielding Faces yielding Faces yielding Type. fine signals. good signals. faint signals. Pinacoids__ 001, 010, 110 ea cei ieee See 110. 210 310 bee (3) See cere O11 ce 1 ee ei 102, 101 203 | 103 Ld 2a eg Meee LOR203 aes ae OO aa ee 121, 321 ET 112, 123 TA) 3258 O24 Oh 2a era Rts ie 23 521 4. Hematite Purting from Franklin Furnace, N. J. A mass of ore from Franklin Furnace, N. J., weighing about two pounds, consisted principally of hematite with a very marked rhombohedral] parting. With the hematite was calcite also showing a parting (parallel to 0112) and enclosed within the calcite was a broken crystal about one inch in diam- eter which consisted of a well-defined crust of hematite with the parting, the red streak, the very feeble manganese and very weak magnetism; and a core of franklinite with different luster, no parting, brown streak, decided manganese reaction and decided magnetism. For record the nearly cubical parting was measured. The signals are not bright and there is a little calcite between the parting surfaces. “Two angles of a fragment yielded respect- ively 94° 35’, 93° 52’ or an average of 94° 13’. The unit rhombohedron angle of hematite is 94° 0’. 284 A. J. Moses—Crystallization of Luzonite. Mr. John Crawford, Jr., made triplicate analyses for me of selected material for total iron and for I'eQ, the result being: Fe per cent. » FeO per cent. 67°15 Shy) 67°07 1°66 67°22 1°54 67°15 average. 1-72 average. Dedueting the 1°34 Fe equivalent to 1°72 FeO leaves 65°31 Fe present as Fe,O, or 94:00 per cent. The total analysis becomes: Tnsoluble se oe ee CaO calculated to CaCO, ------ 2°85 B60 Ve os ea HeQ ex SiG ei 8 ed a ee ei 100°07 Or recalculating the Fe,O, and FeO to 100 per cent. Ke OX se ie eee 98°20 per cent. HeO ss 3 he eee ean as 1°80 z Columbia University, June, 1905. Wright—Optical Character of Birefracting Minerals. 285 Arr. XXXI.—The Determination of the Optical Character of Birefracting Minerals; by Frep. Eveenr Wriaut. MineErRALs are recognized in the thin section chiefly by their crystallographic properties and by the effect they have on transmitted light. The more important optic features used in their microscopic discrimination are color, pleochroism, refrac- tive index, birefringence, optical orientation, angle between the optic axes (2V),* and optical character. Of these the latter two are determined in convergent polarized light and are well adapted for general application. They furnish exclusive data as to the nature of a given raineral, and can be accomplished by ordinary petrographic microscopes. The optical character of a mineral, whether positive or nega- tive, depends by definition solely on the value of the bisector of the acute angle between the optic axes; it is, therefore, inde- pendent of the erystal system and pertains to all biretracting minerals. The usual methods available for its determination, however, apply in practice only to uniaxial minerals and to those biaxial minerals for which the angle between the optic axes in air (2E) is less than 80°; if 2K exceeds this limit, the traces of the optic axes lie outside of the microscopic field and give rise to uncertainty as to the position of the acute bisectrix, thereby seriously affecting the results. There are several methods, however, which, although not novel in prin- ciple, are scarcely recognized iD literature, and which practi- cally obviate this difficulty. They are based on phenomena observed in convergent polarized light with nicols crossed and apply equally well to uniaxial and biaxial minerals. A general consideration of microscopic mineral determina- tion shows conclusively that the optical character of minerals is one of their most useful traits for practical determination since the means employed are simple and of easy application. The following paragraphs aim to present these methods from a working standpoint, the necessary theoretical data appearing in fine print. The crystal sections of birefractng minerals, from which decisive interference figures can be obtained, are those cut exactly or nearly perpendicular to the bisectrices of the optic axes, to the optic axes, and parallel to the plane of the optic axes. These sections and the methods applicable to them can be discussed tor all birefracting substances if uniaxial minerals are treated as a limiting case of biaxial minerals. * The use of the term optic binormal in place of ‘‘ optic axes” as pro- posed by Mr. L. Fletcher in his treatise on The Optical Indicatrix may be an improvement on the original term, but since the distinction implied by the words uniaxial and biaxial is in use in all languages, is convenient and causes no confusion, it is probable that the original designation will remain. Am. Jour. Sci.—FourtH SERIES, Vou. XX, No. 118.—OcToBEr, 1905. 20 286 Wright—Optical Character of Birefracting Minerals. The figures 1—6, used to illustrate the methods, were obtained in part by graphical and in part by mathematical means based on the law of Fresnel, that the planes of polarization for rays traveling in any direction bisect the angles between the planes containing the ray and the two optic axes respectively ; in other words, the directions of extinction for any face bisect the angles between the projections of the optic axes on the face. Plates cut perpendicular to the acute bisectria. For birefracting minerals in which 2E is less than 80°, the methods ordinarily described in text-books are applicable and satisfactory. Both optic axes appear then in the field, and the optical character can be ascertained in convergent polarized hight by observing the change in position of the lemniseatie interference curves in alternate quadrants on the insertion of a quartz wedge or a plate showing the interference-color red of the first order, or a quarter-undulation mica plate. The numerical value of 2E can also be measured on the same sec- tion by the Bertrand-Mallard* method described below. For minerals whose 2E is greater than 80°, a method deseribed by Michel Lévyt for determining whether the section is perpen- dicular to the obtuse or the acute bisectrix can be used to advantage. It consists in observing the angle of revolution of the stage from the position where the black achromatic curves of the interference figure form a cross to that at which they are tangent to a given circle (usually field of the microscope). From this angle 2E can be determined, and from it in turn the true optic axial angle (2V), if the medium index of refrac- tion of the substance be known. It can be proved both mathematically and graphically that the dark achromatic hyperbolas, which form during the revolu- tion of the stage, pass through the traces of the optic axes and recede from the field along the diagonals of the principal planes of the nicols.’ Practically, the course of procedure is to find a plate cut perpendicular to the bisectrix, to record the angle of revolution of the stage from the point where the dark hyperbolas intersect to that at which they are tangent to a given circle within the field of vision. From this angle the corre- sponding axial angle in air can be obtained by using fig. la, provided the Mallard constant of the microscope has been pre- viously determined. If the medium refractive index of the mineral is also given, it is possible to convert 2E into 2V_ by means of fig. Lb. * EK. Bertrand in Mallard, Miner. physique, 11, 418. E. Mallard, Sur la mesure de l’angle des axes optiques. Bull. Soc. miner., 1882, page 77 et seq. + Michel Lévy, Minéraux des Roches, 94-95. 287 Wright— Optical Character of Birefracting Minerals. Rire: : \ \ \ \ | « - x - \ + ‘ > oh 4 \ { \ \ | | jana ae pe A UU peak) rts ithe bebe aay | ecksh be hig cede Pe wlll di i 5 SAN id 6. SASSO WE i SEDATE: INE g as 288 Wright—Optical Character of Birefracting Minerals. Bertrand-Maliard method for measuring the optic axial angle (2K) under the microscope. Mallard has shown that the distance of the trace of an optic axis from the center of the interference figure is proportional to the sine of the angle which the optic axis makes with the axis of the microscope; that, if the distance D be measured by means of a micrometer ocular, the angle E can be figured from the formula sin HK = uD Kk in which I< is the constant of the microscope to be determined once for all on a substance whose 2E is known. By drawing a circle of radius Kt in fig. la (once for all), the angle EK corre- sponding to any number of divisions of the micrometer ocular is then the angle in the figure included between its base and the radius passing through the intersection of the are K with the horizontal line at the distance D from the base line. To convert 2K into 2V use fig. la, which was derived from the formula } sin EK aia Ws nN mM nm,, being the medium refractive index of the substance. The angle 2V is then the angle on the degree circle included between the base line and the horizontal line which passes through the intersection of the radius EK and the given refractive index are. Michel Lévy method. Michel Lévy has developed a formula from which approximate values of the axial angle 2K can be cal- culated, provided the index of refraction of the objective lens in which the interference figure is observed be known. As this, however, is not generally the case, a modification of the formula by introducing Mallard’s constant in place of the refractive index is better suited to actual practice. ie, Be adoeaecthsoce meee eas a =x In fig. 2* let the plane of the paper represent the section per- pendicular to the bisector of the acute optic axial angle and the figure itself the achromatic lines observed in convergent polar- * Compare Preston, Theory of Light, 3d ed., pp. 400-401. Wright— Optical Character of Birefracting Minerals. 289 ized light ; A,A,, the projection of the optic axes, and P that of any ray in the achromatic hyperbola. Fresnel’s law states that the planes of polarization of rays traveling in any direction P are the bisectors of the angles between the planes A,P and A,P. ' For small angles of incidence, the traces of the planes of polari- zation of the rays will approximately coincide with the bisectors of the angle A,PA,. Since P is a point of the achromatic curve, the bisector of the angle A,PA, must be parallel to one of the principal planes of the nicols. ‘The triangle FPA, is then isos- celes, and the triangles PFD and PDA, are similar. Therefore Depa Oe ete ; eres Ia (2) Y= Ya YT Y, wy = ay, (2) the equation of an equilateral hyperbola. In order that this curve be tangent to a circle, its tangent must be perpendicular to the radius the equation for which is Y= —— & (3) Soe d By substituting the value of ~ from (2), (83) becomes BG wy (4) which shows that the hyperbolic curves are tangent to the circles along the diagonals of the nicols. For these points (2) reads C= HY, (5) Transforming (5) to polar codrdinates, we find == SIN (6) From Mallard’s method above, it is evident that 7=K sin E and p=K sin O sin O Therefore, sin EK = — ‘/ sin 2h (7) where sine O is the constant of the circle used and to be deter- mined once for all by the Mallard method. For any given angle of revolution (¢) the corresponding E can be found by finding in fig. 1b the intersection of the horizontal line at the distance sine O from the base line with that are which corresponds to the angle @. 2E can then be reduced to 2V by fig. la, if the medium index of refraction be known. Owing to the width of the achromatic curves, the results attained by this method are only approximate but of sufficient accuracy to be useful in many instances. The angles ¢ can also be figured for sections not exactly perpendicular to the bisectrix ; they possess, however, only slight practical value. 290 Wreght—Optical Character of Birefracting Minerals. The mathematical formula above is only an approximate one, while a graphic method can be applied which is theoretically correct and by which more accurate results can be obtained. The method has been used by Michel Lévy, Viola,* von Fedorow and others in their feldspar studies and is well adapted for general use in the study of optical phenomena. The lines along which any face will extinguish can be found by passing planes through the normal to the face and the optic axes respectively, and bisecting the traces of these planes on the face. In order to do this readily, a stereographic projection of the optic axes In any desired position should first be made. By a revolution about each of two horizontal axes in the principal planes in the nicols, any face normal can be brought to coincide with the pole of the projection and the face with that of the paper. The bisectors of the angles between the straight lines drawn through the pole of the projection and the optic axes in their new positions are then the desired directions. The achro- matic black hyperbolas of the interference figure correspond to those face-normals whose extinction lines are parallel to the axes of revolution of the projection. In the projection the achro- matic lines, however, do not appear as they do when observed under the microscope, for its interference figure can be considered with shght error as an orthographic projection of the rays on a sphere, as shown by Mallard’s formula above. ‘The curves of the stereographie projection must therefore be replotted by making the polar distance sine E instead of tan 9 28 it Is in the stereographic projection. The general aspect of the curves is not changed by this transformation. The graphic method has been applied to the methods below with satisfactory results. (Figs. 4 and 6.) The interference figure from the section perpendicular to the obtuse bisectrix differs from the above only in the wider optic axial angle, which can be measured by the same methods. Plate perpendicular to an optic axis. The interference figure obtained from this plate consists ordinarily of a black achromatic bar which revolves in a direction opposite to that of the stage. In general the bar is a straight line only when it is parallel to the planes of polari- zation of the nicols ; in the intermediate positions it is more or less convex, depending on the angle between the optic axes. If 2K, however is equal to 90°, the curve is a straight line in all positions for the usual microscopic field of vision. * Michel Lévy, Sur la determination des feldspaths, 1894, pp. 15-20. C. Viola, Zeitschr. fiir Kryst., xxx, 232, xxxi, 484, xxxii, 305. E. von Fedorow, Zeitschr. fir Kryst., xxxi, 579, xxxii, 246. 291 racting Minerals. ler £ a “Ve WAM Re abteay DAY nV AD jy RASS Dal ete tay en IG eo: Wroght— Optical Character of B ent Eee ff 5 A vp nfl ALMA DearS A | 2 etromn rie mrimslailebtpiniabure Ay SI AtOE SD Hifi N > Vv or 2a ri Wee Zoe OA Sr alt Ta i J i NG Z ‘ \ Ms : oe na A Tine seTaSeSSS - ss Tere. 292. Wright—Optical Character of Birefracting Minerals. An examination. of the curves for the several optic axial angles, figs. 3 and 4, indicates clearly that the convex side of the bar in the diagonal position points toward the acute bisec- trix. The determination of the optical character is effected then most readily by means of the red of the first order plate. If the achromatic bar of the interference figure be placed in the position of fig. 8 with the convex side of the hyperbola pointing to the northeast and the arrow of the plate (direction of least ellipsoidal axis) also in the same direction, the convex side of the curve will show a blue interference color if the mineral is optically negative; the blue spot will be on the con- cave side of the curve if it is optically positive. This method can always be applied if the convexity of the curve can be discerned. In certain plagioclase feldspars, the limiting case of 2V =90° is encountered occasionally and there the bar is in fact a straight line. For most of the feldspars, however, the curvature is sufticiently marked to enable a deter- mination of their optical character. The result can be checked by extinction angles of the section and adjacent twinning lamellee after the method of Michel Lévy. The approximate formula for the achromatic curves from this plate can be derived from a discussion of fig. 5, which is based on assumptions similar to those obtaining for the curves of the eee perpendicular to the acute bisectrix. Hic. 9d. vara = ak (1) Wright— Optical Character of Birefracting Minerals. 293 the equation of an equilateral hyperbola passing through the zero coordinate point with asymptotes parallel to the X and Y axes. For the special case under consideration where x, = y,, the formula becomes igs gs (3) From (3) the curves of fig. 3 were plotted in gnomonic pro- jection. For «, =, equation 3 becomes cay the equation of a straight line passing through the zero point at an angle of 45° with the codrdinate axes. If the formula of Mallard were exact, x, could not assume a value greater than 1 (sine 90°); since it is approximately correct only for small angles, the above remark does not obtain. The gnomonic projection was, therefore used in fig. 3 instead of the orthographic. In such limiting cases “the graphic method gives more satisfac- tory results and is in general better suited to the study of optical phenomena. In fig. 4, the stereographic plat with curves for Somevaxial angeles 0°, 15°, 45°, 75°, and 90° is given. ‘Their course in the vicinity of the pole of the projection only is repre- sented since it corresponds to that portion which is seen under the microscope. Plate parallel to the plane of the optic axes.* In the uniaxial minerals this plate corresponds to any section in the prism zone. The interference figure from the section can be recognized by the fact that in the position of darkness the entire field is practically dark and that a small revolution of the stage (5°) will cause the faint hyperbola to recede entirely from the field of vision. In the diagonal position the colored interference curves have the form of hyperbolas. Since ordinary approximate methods of calculation do not apply to this section, the graphic method with the stereographic projection plat as base was adopted. The result, as depicted by the curves of fig. 6, shows that the recession of the dark achromatic lines for the optie axial angles 2V=0°, 10°, 80°, and 90° after a revolution of 1° of the stage is very marked, and that, except in the limiting case of oY = o Ome thie dark hyperbolas pass out of the field most slowly in the direction of the acute bisectrix. For 2V=90° the hyperbolas in all quad rants recede from the center with equal rapidity. In fig. 6 * Compare F. E. Wright, this Journal, xvii, 387-391. 294 Wreght— Optical Character of Birefracting Minerats. the lines between the outer and inner circles represent the actual position of the bisectrices and optic axes under the conditions stated. | Owing to the fact that for this section the angles of extine- tion are very low for all rays whose angle of incidence is small, the intensity of the rays adjacent to those of the achromatic curve is also low, since it varies with the square of the sine of Fie. 6. the angle p between the planes of polarization of the nicols and that of the section according to the formula Des sila 2 sin : (o—e) The black curves are therefore indistinct and require careful scrutiny to be observed at all. The colored hyperbolic interference curves which appear in the interference figure most sharply in the diagonal position of the section can also be used to locate the direction of the acute bisectrix. It can be proved in several different ways that the acute bisectrix is generally direction of less birefringence than the obtuse bisectrix. The birefringence of any section can be figured approximately by the formula y'—a'=(y—a) sin 6, sin 6, where y’ and a’ denote the maximum and minimum refractive indices of the given section, y and a those of the mineral, 0, and @, the angles between the normal to the section and the Wright— Optical Character of Birefracting Minerals. 295 optic axes respectively. The formula indicates clearly that, except in the limiting case of 2V=90°, the birefringence for sections in the alternate quadrants containing the acute bisec- trix is less than that for corresponding sections in the two remaining quadrants. The rule resulting from this fact is that the interference colors for points in the quadrants containing the acute bisectrix are lower than those for corresponding points in the direction of the obtuse bisectrix. After the direction of the acute bisectrix has been found by one of the above methods, its value (c or a) can be readily ascertained by ordinary methods either in parallel or conver- gent polarized light. Summary. In the practical determination of minerals under the micro- scope advantage is taken chiefly of those properties which are definite in character and which can be readily ascertained. Of these the optical character is one of the most useful since it apples to all birefracting minerals and can be determined in convergent polarized light on plates cut along one of several different directions : 1. On plates perpendicular to the acute bisectrix, by obsery- ing the direction of movement of the curves of the interfer- ence figure on the insertion of a quartz wedge, mica plate, or plate showing the interference color red of the first order. If the loci of the optic axes lie outside of the field, determine whether the plate is perpendicular to the obtuse or acute bisec- trix by measuring the optic axial angle in air by the modifica- tion of the Michel Lévy method described on page 288. The reduction of the observed optic axial angle to that in the crys- tal ean be accomplished only when the medium refractive index of the substance is known and then easily by fig. 1. 2. On a plate perpendicular to an optic axis by noting that, when the black achromatic bar lies in a position diagonal to that of the principal planes of the nicols, its convex side points toward the acute bisectrix and that on the insertion of a plate showing the interference color red of the first order, the con- vex side will be colored blue if the arrow (nv) of the inserted plate les in the plane of the optic axes and the mineral is optically negative ; if the blue spot lies on the concave side of the bar and the arrow of the plate still liés in the plane of the optic axes, the mineral is optically positive. This method is applicable whenever the curvature of the achromatic bar can be observed. ‘The section is moreover easy to find because in parallel polarized light with nicols crossed it remains nearly dark for all positions of the stage. 296 Wreght—Optical Character of Birefracting Minerals. 3. Ona plate parallel to the plane of the optic axes the direc- tion of the acute bisectrix can be located by two different methods: a. Kevolve mineral from the position of darkness through a small angle and note that the direction in which the faint dark hyperbolas recede from the field is that of the acute bisectrix. 6. In the diagonal position of the interference figure observe the interference colors of corresponding points in adjacent quadrants and note that the points i the direction of the acute bisectrix show the lower interference colors. In both cases the value of the acute bisectrix (c or a) can be determined either in convergent or parallel polarized hght by the usual methods and thus the optical character of the mineral be ascer- tained. U.S. Geological Survey, Washington. Barus— Groups of Efficient Nuclei in Dust-Free Air. 297 Art. XXXIL—On Groups of Efficient Nuclei in Dust-Free Air; by O, Barus. 1. Dustfree Air.—By this term I refer to atmospheric air ae with extreme slowness (through large wide filter of packed cotton) and thereafter left without interference for two or more hours. Such air shows a high fog limit. In the fog chamber used the coronal condensation begins at a pressure difference of about 69 = 26 em., rain-like condensation at eg — 21 cm. Tn the present experiments all tests are made at 6p = 41°5 cm., at a pressure difference therefore much above the fog limit, and probably approaching the condensing power of the appa- ratus. The number of nuclei computed from the coronas observed is an approximation merely, as the constants needed for the very large range of variation in question are not avail- able. Nev ‘ertheless, if the same 6p is used throughout, the nucleations obtained are immediately comparable. With these reservations* the number of nuclei found in the dust-free air and at the 6p in question is about 880 x 10° to 460 x 10° per em*. Itisobvious, more- over, that these nuclei %0-5p & 30 3 4G 45 are excessively small, much smaller than ions, smaller even than those | which would respond to — g9— smaller exhaustions, exceeding 6p = 26 em. In fioure Ll have. given an vexample of these eelamons. Between’ , dp = 21 and 26 (for this apparatus: condensation probably takes place largely on ions above that on the nuclei of dust-free air. The upper dotted line shows the limit of value found, the latter being variable because (as will appear more clearly below) the ionization of atmospheric air is essentially variable. Though relatively small in number, the ions from their larger size probably cap- ture much of the moisture. 2. Lifect of Radium.—Now let the fog-chamber (fig. 3) be subjected to the radiations from weak “yadium (10,000 x, 10 mg.) contained in a thin hermetically sealed aluminum tube. As the walls of the fog-chamber are °3 em. thick and 300; * The nuclei are supposed to be removed by exhaustion, faster than they can be restored, either by radiation or by the molecular mechanism. 298 Barus—Groups of Lficient Nuclet in Dust-Free Atr. the end (bottom) toward the tube nearly 1 em. thick, y-rays only will penetrate into the inside apart from the secondary radiation there produced. In figure 3 /’is the cylindrical fog- chamber, / the radium tube at an axial distance ) from the nearer end. In addition to this the radium was also tested at T (top) in the figure, where it is nearest the body of dust-free air under experiment. Within the fog-chamber the coronas are everywhere normal and of the same size, in spite of the axial length of 45 em. available. This is a singular result when contrasted with the marked positional effect observed for the case of radium placed at the different distances D outside of the chamber. The data investigated are shown in the curve (fig. 2), where the abscissas are the distances and the ordinates the number of epicrent nuclei per em’. = oom 40 6 80 It follows from the graph that as the radium is brought in an axial direction from o to the end of the fog-chamber, the number of efficient nuclei in the dust-free air contained is gradually but enormously reduced to a minimum for D = 25 em. (about), after which the number again increases to the maxi- mum at 2 =0. Curiously enough, if the radium is further approached to the body of the air by being placed at 7, the number of nuclei does not increase; in some observations it even diminishes. If the radium is enclosed in a long thick lead tube (60 em. long, walls °5 em. thick), the nucleation is but moderately reduced (see fig. 2, crosses), showing that y-rays are in question. 4. Cause of the minimum.—This is easily explained since the ions are relatively large bodies and relatively few in num- ber as compared with the nuclei of dust-free air for the same 6p. Hence the ions virtually capture the moisture more and more fully as their number, with diminishing distance D, becomes greater. At D == 25 em. probably the whole of the Barus—Groups of Lifficient Nuclei in Dust-Free Air. 299 moisture is condensed on ions, and as their number increases as D vanishes, the minimum in question results. In fact it was shown elsewhere, that below the fog-limit of air, the nucleation observed and due purely to radium at differ- ent distances D, is for example (6p = 22) ji =e 0 10 30 50 100 INES Ara eee Salt 2X0) 50 32 20 12 Ke, agreeing in character as far as may be expected with the data here in question. These data multiplied by 4,1.e., 4 < 10™, are also given in figure 2 for comparison. Hence the ions caught at 69 = 41°5 are about four times more numerous than at 6p = 22, and correspondingly smaller. They are, therefore, markedly graded, but nevertheless, as a group, throughout much smaller than the nuclei of dust-free air so long as the radiant field is appreciable. Whether the latter are agglomer- ated under the influence of radiation to make the ions (as would seem more probable), or whether the ions are made from the molecules themselves so that the ions and the nuclei of dust-free air are present together, is a question beyond the scope of the method. While the number of nuclei continually grows smaller, with diminishing J, the efficient or capturing nuclei may nevertheless increase again below a certain JD, seeing that the nuclei in dust-free air are enormously in excess, only a few of which are caught even in the absence of radium. 4. Cause of the maximum.—lt is more difficult to account for the result that the same nucleation is observed wherever the radium touches the elongated fog-chamber. In other words, radium at the end of the chamber produces at least the same nucleation as when at the top, although the distances from the center of mass of the glass are as 3 to 1. The same kind of explanation already given in $38 may possibly hold. The radium tube when placed on the top (7’in figure 3) and in contact with thinner glass, may act with sufficient imtensity to admit of the formation in turn of a group of nuclei larger than ions. This is what actually occurs in the case of X-rays. But it is more probably connected with the uniform distribution of nuclei within the chamber (§ 1) and in some way reterable to secondary radiation evoked within the chamber. Secondary radiators added on the outside are quite without effect. 5. General Conclusion.—The occurrence of a continuous succession of groups or gradations of nuclei in the curve of figure 2, each of which groups constitutes a condition of chem- ical equilibrium for the given radiating environment, is sugges- tive. In the first place, it may be recalled that the nuclei of dust-free air are an essential part of this body as much as the molecules themselves. Such nuclei if withdrawn by precipi- 300 LBarus—Groups of Lfficient Nuclec in Dust-Free Air. tation are at once restored. Again air left without interference for days shows a maximum of this nucleation for the given conditions of exhaustion when all foreign nucleation must have vanished. Indeed the molecules themselves may be treated as a continuous part of the nucleation in question, the frequency of occurrence being a maximum for the molecular dimensions. Furtbermore in the presence of radium the character of the phenomenon is the same, only the nuclei are larger. If with- ee by precipitation, they are at once restored. They are an essential part of the air in the new environment. it is natural to compare the particular nuclear status intro- duced in the latter case by a particular kind of radiation (y rays), with the former case of dust-free air in the absence of recognized radiation. In other words, quite apart from the details of the mechanism, chemical agglomeration might be considered referable to an unknown radiant field, but be other- wise essentially alike in kind to the much coarser nucleations observed in the known radiant fields of the above experiments. But the effect of radium, however distant, is always virtually an increase of the size of the air nuclei and a decrease of their number. Hence if we were to fancy that the nucleation (not the ions, of course) of non-enereized dust-free air BESO 8 to its own radiant. environment, this radiation would have to be special in kind. Returning to the case of the gamma rays, fig. 2, (or of the X-rays coming from a distance,) let me recall that the effective radiation within the fog chamber is everywhere the same and the same in all directions. Hence whether the radiation be corpuscular or (in other cases) undulatory, the interior is noth- ing less than an ideal Lesage medium; and there must there- fore be a tendency at least to agglomerate the colloidal nuclei of dust-free air into fleeting nuclei or ions, so long as the radiation lasts. When it ceases the ions are free to fall apart, so far as external influence goes, as they actually do. Further- more since the pressure so obtained would increase with the number of corpuscles per cubic em. and with the square of their velocity, it is conceivable that with increasing electrifica- tion this pressure would become strong enough to bring about permanent union of the aggregates, corresponding to the observed continuous transition of the ions into persistent nuclei, produced by the X-rays. Again a different nucleus would presumably correspond to the bombardment of the negative corpuscles as compared with the residual positive quantities. Finally, if any physical or chemical process like combustion or ignition or electric charge, or the case of phos- phorus, ete. is accompanied by intense ionization, one would for the same reason anticipate the presence of nuclei in such a field. Brown University, Providence. T. Holm—Studies in the Cyperacee. ION Arr. XX XIII.—Studies in the Cyperacee ; by Toro. Horm. XXIV. New or little known Carices from Northwest America. (With 18 figures, drawn from nature by the author.) Wirn the object of preparing a treatise of the genus Carex as represented in the northwestern part of this continent the writer has examined several very extensive collections, con- taining a vast number of specimens, among which some few have been observed as imperfectly understood or as hitherto undescribed. Inasmuch as the treatment of the genus in a subsequent paper will be from a geographical point of view, we prefer to publish the diagnoses of the new species sepa- rately with some remarks upon their affinities. These species are: Carex limnca sp. n. (figs. 1-8). Rhizome vertical with ascending shoots and light brown, fibrillose leaf-sheaths; leaves a little shorter than the culm, narrow, but flat, glaucous, scabrous along the margins; culm about 60™ in height, erect or slightly curved above, very slender, triangular, scabrous, phyllopodic ; spikes 3 to 5, but mostly 4, the terminal staminate or, sometimes, androgynous, the lateral pistillate, the uppermost contiguous, the lowest remote, sessile to shortly peduneled, erect, not very dense- flowered, cylindric, about 2°" in length, subtended by sheath- less, foliaceous bracts, the lowest one often exceeding the inflorescence ; scale of staminate spike lanceolate, light pur- plish-brown with green midvein; scale of pistillate spike oblong, obtuse, black with hyaline apex and greenish midvein, shorter than the perigynium ; perigynium stipitate, slightly spreading, narrowly elliptical; granular, plano-convex, prominently many- nerved on the outer (convex) face, three-nerved on the inner, pale green with a black, entire and very distinct beak ; stig- mata 2, style long and exserted. Oregon: Crater Lake National Park, Cathedral spring, col- lected by Mr. F. V. Coville, September, 1902 (No. 1456); Four- mile Lake, Klamath County, in meadows, and between Dia- moud and Crescent Lakes, Gascade Mountains. The graceful habit of this species reminds us more of @. rhomboidea than of C. vulgaris, but when we, nevertheless, prefer to place it nearer C. vulgaris it is on account of the structure of the perigynium, narrowly elliptical and promi- nently many-nerved. Am. Jour. Sci.—Fourts Serizs, Vou. XX, No. 118.—OcToBER, 1905. 21 302 T. Holm—Studies in the Cyperacee. Carex brachypoda sp. n. (figs. 4-6). Rhizome short with ascending shoots and persisting, dark brown leaf-sheaths; leaves shorter than the culm, relatively broad (about 5™™) and flat, deep green, scabrous along the margins and lower face, glabrous above; culm about 35™ in height, erect, stiff, triangular, scabrous, phyllopodic; spikes 3 to 4, mostly 4, the terminal staminate, the lateral pistillate, somewhat remote, sessile or the lowest one short-peduncled, erect, dense-flowered, cylindrical, from 1 to 2° in length, sub- tended by sheathless bracts with narrow blades much shorter than the inflorescence; scale of staminate spike lanceolate, red- dish brown with pale midvein; scale of pistillate spike ovate, obtuse, black with green, not excurrent midvein, a little shorter than the perigynium; perigynium minutely stipitate, erect, almost orbicular, granular and denticulate along the margins above, compressed, nerveless, pale green, the minute beak dark purple with the orifice entire, papillose; stigmata 2. Oregon: Orater Lake National Park, Cathedral spring, col- lected by Mr. F. V. Coville, September, 1902 (No. 1455). The affinity of this species 1s with C. gymnoclada, but it differs from this by the perigynium for instance, which is more roundish, denticulate and very shortly beaked. Carex pachystoma sp. n. (figs. 7-8). | Rhizome ceespitose with strong roots and persisting, reddish leaf-sheaths; leaves almost as long as the culm, quite broad and flat (0:5), glabrous, hight green; culm from 30 to 56™ in height, erect, somewhat slender, triangular, scabrous, phyllo- podic; spikes 4 to 6, the terminal and uppermost lateral stami- nate, the others pistillate, remote or the uppermost contiguous, all, especially the lower ones, slenderly peduncled, erect or spreading, dense-flowered except at the base, from 3 to 5™ in length, subtended by sheathless, leafy bracts about as long as the inflorescence or a little longer; scale of staminate spike lanceolate, obtuse, purplish brown with green midvein; scale of pistillate spike lanceolate, mucronate, deep purple with broad, green midvein, narrower, but longer than the perigy- nium; perigynium sessile, slightly spreading, elliptical, granu- lar, compressed, nerveless, green or purplish-spotted above, the beak short and thick, sparingly denticulate, the orifice very narrow, slightly emarginate on outer face; stigmata 2. Oregon: Crater Lake National Park, Anna Creek Canyon, near the falls (No. 1862) and near Odell Lake, Klamath County (No. 520), collected by Messrs. Applegate and Coville. Washington : Springy places, northern slope of Mt. Adams, and Falcon Valley, W. Klickitat County (No. 2959), by Mr. W. Suksdorf. T. Holm—sStudies in the Cyperacee. 3038 The species may be placed between C. variabilis and C. lenticularis, although it shows some approach to C. acutina, though merely in respect to its habit. We have examined a number of specimens and are unable to refer the plant to either of those mentioned above. Carex Nebraskensis Dew. Habitually and in several other respects this species seems inseparable from the dZicrorhynche, but we have placed it* as one of the most evolute types of these on account of the biden- tate beak of the perigynium. It is excellently described by. Boott,t and well marked by the strong stolons covered by brown scale-like leaves, which are never shining, by the pale, glaucous leaves and especially by the perigynium with its prominent ribs and bidentate beak. In the extensive collec- tion of Mr. Suksdorf we found several specimens, which were somewhat like this species, but a careful examination of the spikes convinced us that these could not safely be referred to the species, nor ought they to be considered as simply varieties, hence we prefer to describe them as two distinct species: C. eurycarpa and C. oxycarpa. Carex eurycarpa sp. n. (tigs. 9-10). Rhizome stoloniterous with persisting, brown leaf-sheaths and strong roots; leaves half as long as the culm, narrow (3™™), carinate, light green, scabrous along the keel and margins; culm 60™ in height, erect, slender but somewhat stiff, scabrous, triangular, phyllopodic; spikes 8 to 5, mostly 5, the terminal and, sometimes, the uppermost lateral staminate, the others purely pistillate, all remote; the pistillate short-peduncled, erect, dense-flowered except towards the base, until 5™ in length, cylindric, but relatively thin, subtended by narrow, sheathless bracts, about as long as the inflorescence ; scale of staminate spike oblong, obtuse, ight brown with pale midvein and narrow, hyaline margins; scale of pistillate spike lanceo- late, acute, blackish with pale, not excurrent, midvein, nar- rower, but about as long as the perigynium; perigynium sessile or nearly so, erect, roundish, granular, slightly plano- convex, prominently many-nerved on both faces, brownish, the beak short, emarginate ; stigmata 2. Washington: W. Klickitat County, Falcon Valley, collected by Mr. W. Suksdorf, June, 1886 (Nos. 1284 and 2962). Carex oxycarpa sp. n. (figs. 11-12). Rhizome stoloniferous with strong roots and _ persisting, brown leaf-sheaths; leaves a little shorter than the culm, nar- * The author: Greges Caricum. (This Journal, vol. xvi, p. 457, 1903.) + Ill. gen. Carex, vol. iv, p. 175 and plate 592. 304 LT’. Holm—Studies in the Cyperacec. row (4™™), carinate, light green, scabrous; culm about 75™ in height, erect, slender, but somewhat stiff, triangular, scabrous, phyllopodic; spikes 4 to 5, the terminal staminate, long- peduneled, the lateral pistillate, contiguous, seldom remote, short-peduncled, erect, dense-flowered, cylindric, from 2 to 4™ in length, subtended by sheathless, narrow, foliaceous bracts, the lowest one exceeding the inflorescence ; scale of staminate spike oblong, obtuse, light reddish-brown with pale midvein ; scale of pistillate spike lanceolate, acute, blackish with pale, not excurrent midvein, narrower, but about as long as the ‘perigynium ; perigynium sessile, broadly elliptical, granular, compressed, prominently 3-nerved, brownish, prominently denticulate along the margins from near the base to the short, emarginate beak; stigmata 2. Washington: W. Klickitat County, meadows near the Co- Iumbia, collected by Mr. W. Suksdorf, June, 1885 (No. 816). Of these C. eurycarpa is a very slender plant and much more so than any of the numerous specimens of C. Vebraskensis, which we have studied. The broad perigynium with the beak merely emarginate constitutes, also, a good distinction. In the other, C. oxycarpa, we have, also, a plant of slender habit, but the spikes are relatively heavy, and the perigynium is here merely 8-nerved and with the margins quite prominently denticulate from the base to the emarginate beak. The affinity of these two species is unquestionably with C. Nebraskensis Dew., next to which they should be placed in the system. Carex campylocarpa sp. n. (figs. 13-15). Rhizome with short stolons and purplish, persisting leaf- sheaths; leaves shorter than the culm, narrow, but flat, sca- brous along the margins and on the lower face; culm about 40™ in height, erect, stiff, triangular, scabrous, phyllopodie ; spikes 3 to 4, mostly 8, the terminal staminate, the lateral pistil- late ; remote, sessile or nearly so, erect, dense-flowered, short ‘eylindric to ovoid, from 4 to 1° in length, subtended by sheath- less, foliaceous bracts, shorter than the inflorescence; scale of staminate spike lanceolate, obtuse, purplish brown with pale midvein; scale of pistillate spike ovate, obtuse, blackish with the midvein faintly visible and the margins narrow, hyaline, much shorter than the perigynium; perigynium shortly stipi- tate, spreading, elliptical-oblong, granular and prominently denticulate along the upper margins, turgid, nerveless, pale green with purplish spots and streaks, the beak quite promi- nent, excurved, the orifice entire; stigmata 2, style not ex- serted. Oregon: Crater Lake National Park, Cathedral spring, coi- lected by Mr. F. V. Coville, September, 1902 (No. 1457). EF. Holm—Studies in the Cyperacee. 305 The systematic position of this species seems naturally to be among the MZzcrorhynche, but as a deviating type on account of the excurved beak of the perigynium, and if it were not for the distinct marginal denticulation of the perigynium and its slender shape the species would resemble C. scopulorum to some extent. A perigynium of this kind is somewhat unusual within the representatives of the grex, but is, as we remember, very characteristic of the Spzrostachye ; in these, however, the beak is generally bifid and more distinctly differentiated from the body. The species may be placed next to C. scopulorum. Carex cryptochlena sp. n. (fig. 16). Rhizome espitose with purplish, persisting leaf-sheaths ; leaves about half as long as the culm, broad (about 1™) and flat, glabrous except along the margins; culm from 70 to 90™ in height, erect and stiff, triangular, scabrous along the edges, phyllopodic; spikes from 4 to 7, the terminal and frequently the uppermost lateral staminate, the others pistillate or andro- gynous, contiguous or the lower ones remote, sessile or short- pedunceled, erect or spreading, seldom nodding, dense-flowered, subtended by sheathless, foliaceous, broad bracts of which the lower ones exceed the inflorescence; scale of: staminate spike elliptical-oblong, acute, light reddish-brown with pale midvein ; scale of pistillate spike lanceolate, sharply pointed, deep pur- plish with broad, greenish midvein, exceeding the perigynium ; perigynium almost sessile, erect, broadly elliptic to roundish, nerveless, pale green, granular, sparingly denticulate near the minute, entire beak ; stigmata 2. Alaska: Kussiloff, on sands with Elymus, collected by Dr. Walter H. Evans, July, 1898 (No. 618), and Seldovia near mouth of Cook inlet by Prof. C. V. Piper, August, 1904 (Nos. 4818 and 4819). This species is somewhat remarkable on account of its resem- blance to Carex cryptocarpa, so far as concerns the structure of the spikes, the deep-purplish, lanceolate scales and the broad pale-green perigynia. But it shows, on the other hand, a strik- ing contrast to this species, C. cryptocarpa, not only by the almost sessile and mostly erect pistillate spikes, but also by its very broad leaves, the basal and the bracts. Habitually the species does not resemble C. cryptocarpa, but, to some extent, Drejer’s C. hematolepis or certain very robust forms of C. salina ; it appears, however, to be distinct from these, and as a type intermediate between the true Saline and C. crypto- carpa Mey. Carex luzulefolia W. Boott var. strobilantha nob. (fig. 18). Taller and more robust than the typical plant; the spikes thick and very compact-flowered ; scales of staminate and pistil- 306 T. Holin—Studies in the Cyperacee. late spikes mostly mucronate; perigynium glabrous through- out, famtly nerved on the inner face, nearly sessile, roundish in outline and terminated by a very distinet, bidentate beak. California: Above Donner Pass in Placer County, in a subalpine meadow, where snow-drifts lie late, and usually near granite rocks, collected by Mr. A. A. Heller, August, 1903 (No. 7187). In the specimens of this new variety the rhizome is densely matted with ascending shoots and covered by dark, brownish fibers from the old leaf-sheaths. The leaves are very broad, but much shorter than the culms. The heavy, deep-brown spikes remind of small cones, hence the name ‘“‘ strobilantha,” and there is quite a variation in respect to their number, posi- tion and the distribution of the sexes. We noticed the follow- — ing instances in 26 specimens: 2 Se AD DENG ue 3 pistillate she in 14 qeehne i 3 5 9 66 (45 4. 66 66 oy) 66 1 66 (19 4 (54 (74 9 6¢ oy) (14 66 9 66 (55 l 66 3 (45 (44 3 (74 66 I (9 3 Re “ l androgynous “ 1 ee In some specimens the pistillate spikes were borne on very long peduncles overtopping the terminal, and several of these were observed to be more or less decompound.—The structure of the perigynium is very characteristic and differs essentially from that of the typical plant, which, as described by W. Boott,* is: “oval to lanceolate,” ‘“slenderly nerved, slightly serrate on the upper margins, longer and ‘broader than. the scale.” The accompanying figures of the perigynia show the distinction very plainly, a distinction, however, which appears to the writer as merely varietal. Brookland, D. C., May, 1905. *S. Watson: Botany of California, vol. 2, p. 250, 1880. Figure 1. Perigynium and scale of Carex liinnea. “e ce es 2. Perigynium of same, inner face. 3). Perigynium of same, outer face. 4-6. Perigynium of Carex brachypoda, outer face. 7,8. Perigynium of Carex pachystoma, outer face. . Perigynium and scale of Carex eurycarpa. 10. Perigynium of same, outer face. 11. Perigynium and scale of Carex oxycarpa. 12. Perigynium of same, outer face. 13. Pistillate spike of Carex campylocarpa. 14. Perigynium of same, side view. 15. Same, outer face. 16. Perigynium of Carex cryptochleena, outer face. 17. Perigynium of Carex luzulefolia, outer face. 18. Perigynium of C. luzulefolia var. strobilantha, inner face. All figures magnified. 308 Schneider—Overthrust Faults in Central New York. Arr. XX XIV.— Preliminary Note on some Overthrust Faults in Central New York; by Putte F. Scunerper. My attention was recently called to two unrecorded over- thrusts in the limestones of this vicinity by Mr. Charles E. Wheelock, who discovered the same, and at whose request this preliminary notice has been prepared. In company with Mr. Wheelock the writer recently visited the locality and this description is largely confirmatory of Mr. Wheelock’s observa- tions, which will be given in full in a future paper. These disturbances in the horizontally stratified Paleozoic rocks of central New York, where for so many years it was thought they could not exist and where the first announce- ments of such occurrences were received with such incredulity, are not yet sufficiently common to permit them to pass unre- corded. The faults are furthermore important because of the relation between them and the well known peridotite intrusives and the probability of the identity of the causes producing the same. Both of the faults brought to ight by Mr. Wheelock occur in some thinly bedded limestones which he correlates with the Bertie dolomite as described by Clarke in his recent report on the formations in the Tully Quadrangle,* or with the lower layers of the Waterlime of Vanuxem,t Geddes,t Schneider,§ and Luther.| The faults can be easily studied in the gorge of Butternut Creek, near Dunlop’s station, one and one-quarter miles north of Jamesville. In the east cliff, a few vards to the south of the stairs leading from Fiddler’s Green to the gorge of the creek, the thrust plane of the southernmost of the faults (Fault IV. Dunlop’s) can be easily distinguished as it extends upward from the base of the cliff through its entire height, a distance of nearly thirty feet. At this point the cliff 1s comparatively free from talus. The dip of the fault plane is 28° to the north- east, N.40° W. Thisnortherly dip of the thrust planes of both of the faults located by Mr. Wheelock is interesting inasmuch as they seem to belong to a series of faults extending in an east and west direction across the country, which hade to the south- ward.4 It is ae surprising as they occur about mid- * Bulletin 82, Y. State Museum, 1905, J. M. Clarke. + Rept. 3d Dist. N. Y. 1842. & Notes on Gail of Onondaga Ge. N. Y. 1893. || Econ. Geol. of Onon., 15th Ann. Rept. N. Y. State Geol. 1895. “| This refers to the overthrusts in the Helderberg limestone series only and not to the slips and slides which are so common in or near the gypsum beds, and which can be explained by the expansion. due to the formation of the gypsum, or to the solution of the gypsum or salt immediately underneath. Schnecder— Overthrust Faults in Central New York. 309 way between Gifford’s and Russell’s faults, the two disturb- ances showing the greatest amount of displacement and practi- cally in a straight line with them. The layers have been sharply bent along both sides of the thrust plane and secondary erystals of calcite have been formed in the numerous fractures in and between the layers, but not as abundantly as at Hast Onondaga and Marcellus. Mr. Wheelock believes the amount of displacement is about four feet, but it is impossible to deter- mine the thrust accurately because of the marked similarity of the layers of limestone. The continuation of this fault may be seen in the west wall of the gorge, where it is not as easily accessible nor as readily studied because of the accumulated material. The bending and buckling of the layers is even more pronounced here than on the east side of the stream, although the displacement was apparently less. Following the direction of the fault to the eastward, a cut on the trolley line just north of Dunlop’s station is reached, showing some disturbance and a marked anticlinal fold. The fractured and disturbed condition of the layers in the entire eut and especially at the fold, which is directly in the line of the strike of the fault, makes ‘it difficult to determine whether the faulting has reached upward to this point. The’fracturing and shattering of the layers resembles somewhat that produced in certain of the layers overlying the gypsum, and lends color to the belief that Fiddler’s Green marks the position of the gypsum deposit. A study of the gypsum ledge to the north- eastward indicates that the gypsum occurs either just above or just below the cut showing the shattered layers, while a com- parison of the altitudes of the adjoining gypsum deposits shows that it should occur at the Fiddler Green locality. Neverthe- less it has not yet been noted there. However, the gorge of the creek lies below Fiddler’s Green, hence it is hardly possible that the faults just described can occur in the Bertie limestone which is described by Clarke as overlying the gypsum. How- ever, according to Clarke’s map the Bertie occurs in the gorge of the creek at this point. Fault II1. Dunlop’s—Following the gorge to the north- ward for a hundred yards or more, the folding and buckling of the layers give evidence of another disturbance. At this point the force seems to have exerted itself mainly in the bending of the layers, and without any large amount of displacement. The thrust plane of the fault is plainly visible, dipping at an angle of 23° to the northward. The displacement is not more than two feet. Fault III occurs in the same formation as that already described, the Bertie dolomite (?) The fault cannot be seen on the west side of the gorge because of a change in the 310 Schneider—Overthrust Faults in Central New York. course of the stream here, which change in the direction is no doubt due in part at least to the existence of the fault right here. ‘ault [. Dunlop’s.—In making a cutting for Jamesville branch of the Suburban railroad about two years ago, two some- what similar faults were exposed in the caleareous layers occurring three-eighths of a mile farther north. These layers may be continuously traced to the northeastward until they are found underlying the gypsum, They undoubtedly correspond with the limestone ledge mentioned by Clarke as containing the Leperditia Scalaris Jones, which occurs in the Camillus shale near the base of the Heard gypsum quarries. Inasmueh as there is only a difference of five feet in elevation between the altitude of these layers at faults I and II and faults © III and IV with practically horizontal layers between the localities, it leaves little question but that faults II] and IV occur in this same Leperditia Scalaris limestone and not in the Bertie. Fault No. I may be seen in the first cut show- ing the limestone, which is about 150 yards south of the cross- ing of the trolley and the Jamesville and Orville turnpike. The thrust plane of the fault cuts these somewhat thinly Jami- nated layers, and dips at an angle of 35° to the south. The layers show little disturbance except at the fault line. Second- ary calcite crystals occur in the fractures of the hmestone, near the fault. | Fault LI. Dunlop s.—Oceurs in the same formation and in practically the same layers twenty yards south of fault No. I. The thrust plane dips south 32° and the layers are bent for several yards to the southward. ‘The slickensided surfaces are well shown, also a slight tendency toward slaty fracture. Cal- cite crystals are lacking. The displacement is slight, probably not more than three feet. The fault maintains its character throughout the entire height of the cut. Owing to an accumn- lation of talus and the dense vegetable growth the faults have not been located on the west side of the stream. Other evidences of slight faulting are noticeable farther north in this cut, also some shearing of the layers with the for- mation of calcite crystals. The overthrusts now known and described in central New York are— (a) Russell’s Quarry at East Onondaga, fault plane cuts the Manlius, Lower Helderberg, Oriskany, and Onondaga forma- tions. Displacement for ty-two feet. Also shown in Hibbard’s and adjoining quarries. Rocks affected for over a mile to the eastward as shown by the marked slaty cleavage in the finer grained limestones of the Corniferous. Luther, “ Econ. “Geol,or Onon.,” 15th Ann. Kept. Nea Schneider— Overthrust Faults in Central New York. 311 State Geologist, 1895. Schneider “Science Series, No. 2.,” Onon. Acad. Science publication, 1899. (db) Maylie’s Quarry at Marcellus, cuts Corniferous and Seneca layers of the Onondaga. Displacement, three feet. Shown in adjoining quarries for over one-half mile. to west- ward. Thrust plane dips 17° to N. See preceding references. (c) Gifford’s Glen, two miles west of Manlius. Cuts the Onondaga and Marcellus groups. Decidedly interesting because of the remarkable manner in which the heavy layers of Onon- daga limestone have been arched and bent. Thrust plane not visible. Luther makes the elevation of the limestone sixty feet, but says it is due to bending. (d) Fillmore’s Corners, one-half mile west of preceding. Cuts Onondaga and Marcellus groups. Displacement, fifteen feet. “Geological Fault at Jamesville,” Schneider. This Journal, vol. iii, 1897. (e) Indian Reservation Quarries. Two faults cut Onon- daga formation. Dip 23° $8. Total displacement of the sev- eral faults about six feet. Schneider, “‘Science Series No. IV,” Onon. Acad. Sei. 1905. (7) Dunlop, No. I, cuts Sealaris limestone in Camillus shale. Displacement, three feet. Dip, 35° S. (g) Dunlop, No I], Scalaris limestone. Displacement, three teen Dip, 32°. 8. (A) Dunlop, No. ILI, cuts Bertie dolomite (?) Displace- ment two.teet... Dip, 23° N. (2) Dunlop, No. IV, Bertie dolomite (?) Displacement, four jee Dip. 238° Noi. (7) Heard’s gypsum’quarry. A small overthrust in the Camil- lus shale occurs here, apparently more deeply seated than the displacements so common in the gypsum quarries due to the formation and subsequent solution of the gypsum. The writer also has MS. notes and drawings of several small _ faults occurring in the Camillus shale near the peridotite dikes which were temporarily exposed during the trenching of that region for city water. At the Solvay quarries at Split Rock in the Manlius and Onondaga formations and in some of the adjoining abandoned quarries several sharp folds and some slickensided surfaces occur which tell of further disturbances. Similar evidences occur in Madison Oo. in the vicinity of Chittenango Falls to the east of the described localities, while to the westward they may be seen in the same ledge about Auburn in Cayuga Co. Cleland* mentionsa faultin the outlet of Keuka Lake, still far- * «A Study of the Fauna of Hamilton formation of the Cayuga Lake sec- tion in central New York,” H. F. Cleland, Bulletin No. 206, U. S. Geol. Survey. 8312 Schneider—Overthrust Faults in Central New York. ther west, but no facts are given, while the folds in the higher formations are well shown in long arch at Cayuga Lake, and similar undulations in strata at Seneca. Disturbances are also noted by Lincoln in his account of the geology of Seneca Co.* Inasmuch as most of the above mentioned disturbances occur in or near the Helderberg escarpment, composed in the main of heavy limestones aggregating several hundred feet in thickness, and the persistence of the faults across central New York, it would seem that all are the result of some considerable force capable of affecting this entire region. In a general way the solution of the salt from the Salina formation which imme- diately underlies the Helderberg series has been regarded as an explanation for all the disturbances in this vicinity. Mr. Wheelock believes that the solution of all of the saline ingre- dients of the Salina rocks together with the slight dip of rocks of central New York is a sufficient explanation for the fault- ing, as any settling of the layers must shorten the length of the hypothenuse of the triangle and thus produce the force which crumpled and fractured the rocks. The fact that the softer shales sandwiched between the hmestone bands are some- ~ times bent and sheared while the harder layers are not affected, and that the larger throws all occur in the more resistant layers, he believes willfavor his explanation. This, however, would be true irrespective of the cause, provided of course that it were compression. It has also been suggested{t that expansion due to the formation of gypsum would explain the faulting. While considering the causes of the faultsit would be well to keep in mind that there is a series of widely known intrusives which parallel north and south this series of faults, and which extend across the state from Little Falls on the east to Ithaca on the west, and it is not impossible that both faults and dikes owe their origin to the same general disturbance. The considera- tion of this question, however, will be left to another paper. Syracuse, N. Y. * “© Geol. of Seneca Co.,” Rept. N. Y. State Geologist, 1894. +E. H. Kraus, verbally. FN. Guild—Petrography of the Tucson Mountains. 318 Art. XX XV.— Petrography of the Tucson Mountains, Pima Co., Arizona; by F. N. Guitp, University of Arizona: (With Plate IX.) | Tue Tucson Range of mountains is located directly west of Tucson and is about twenty miles long with an average width of about seven miles. It consists of a series of jagged peaks extend- ing nearly north and south, the higher ones of which are esti- mated to have an altitude of 4000 feet above sea level. The approach to the main line of peaks is over a series of low-lying rounded knolls devoid of all vegetation except a few cacti and other stunted growths characteristic of an arid region. Petrographically quite a variety of rocks are represented which are almost entirely eruptive. There occur, however, in places, remnants of the original quartzites and limestones through which this great mass of lava has broken. On the west side of the range and southwest of Tucson is an elevated plateau of an area of one hundred square miles or more, con- sisting entirely of these limestones and quartzites. It is quite level and the beds are exposed only along its edges and near the center where the uplifted strata form two small buttes, consisting almost entirely of crystalline limestone tilted to an angle of about forty-five degrees. It is the purpose of this paper to describe from a _ petro- graphical standpoint the eruptive rocks without discussing their geological relations. The question of names and classification is not taken up, the writer considering descriptions of more importance. With this introduction, they will be described in the order of their relative abundance. Lhyolite—The main line of jagged peaks referred to above is made up of this rock, varying in color from a dark red to nearly white. Phenocrysts are inconspicuous, not very abun- dant and rarely exceed three millimeters in diameter. They consist of quartz and less abundant orthoclase. Under the microscope, the quartz is found to occur in-rounded masses corroded by the groundmass with frequent inclusions of the latter in the form of bag-shaped inlets. Black dust-like inclu- sions and glass with gas bubbles are common. The feldspar, although much decomposed and containing opaque inclusions, still shows the characteristic cleavage of orthoclase. The groundmass in the darker varieties is too much altered to show any characteristic texture. In the southern portion of the district, however, material is sometimes met with of sufficient freshness to admit of satisfactory study, and here it is found to have a cryptocrystalline structure. Specimens of it are frequently found possessing faint flow lines, sometimes vis- 314 2. WV. Guild—Petrography of the 7 ‘weson Mountains. ible to the unaided eye, but usually requirmg a microscope to be seen. (Fig. 1, Plate IX.) Occasionally dark shredded masses occur which may have been originally mica. Al vari- eties contain irregular inclusions of varying size, sometimes two inches across, of a red jasper-like substance or of sandstone or quartzite. Frequently also rounded patches are met with which under the microscope are found to be made up of quartz and feldspar in equidimensional erystals, which may represent areas of more complete crystallization of the groundmass. Associated with these more typical rhyolites are large masses usually of a light yellow to butt color, lacking all phenocrysts. They correspond to rocks which have been variously called felsite, felsophyre, granophyre, etc. They sometimes break with conchoidal fracture, but are more often too coarse-grained to show this characteristic. Under the microscope, quartz, feldspar and sometimes shreds of mica can be seen in the coarser varieties. The finer-grained types are made up entirely of eryptocrystalline material in which none of the constituents can be determined. Lehyolitic Tuff.—-Associated with the outflow forming the main rhyolite peaks, there were probably formed masses of voleanic ash. The greater portions of this have been washed away, but occasionally where geological conditions have been favorable some of this material has become consolidated into a compact rock of hght gray color of sufficient strength to be used extensively in building. Underlying Sentinel Peak in places there are small masses of it which have been held in place by the basaltic outflow. Under the microscope it is found to be made up of fragments of quartz, feldspar, glass, ete. i 1s interesting: to note that the quartz has the same kind of inclusions as the quartz phenocrysts in the rhyolite described above. Andesites.—Several types of this rock varying greatly in appearance and texture occur. They may be grouped rather roughly as follows: 1. Light-colored andesites containing phenocrysts of mica or hornblende or both and of feldspar. 2. Dark-colored andesites of non-porphyritic texture. 38. Vitrophyric andesites. The first variety covers an area only slightly less in extent than the rhyolites and constitutes the material of the low-lying knolls previously referred to. It has usually a mottled appear- ance not unlike that of some granites. The feldspar is pure | white and the groundmass varies from white to greenish gray. The chief variation is in the black ferro-magnesian minerals, which are most often biotite, but in some localities hor nblende predominates, while in still others the black phenocrysts are quite inconspicuous. Under the microscope the feldspar is clear, F. N. Guild— Petrography of the Tucson Mountains. 315 usually striated, and as shown by the extinction angles on the twinning plane appears to be an acid plagioclase. The biotite is quite fresh and of the usual dark yellow-brown color. Horn- blende has become darkened by alteration and is often quite opaque. The groundmass is crystalline and made up mostly of feldspar with some magnetite and shreds of the dark silicates. The second yariety is found in the northern portion of the district near the edge of the mountains about twelve miles from Tucson. It varies greatly in both megascopic and microscopic structure in different parts of the same mass. Its general appearance is more like that of a diabase except in portions where phenocrysts of feldspar appear. It is very dark with a slight green tinge weathering red. Por- phyritic texture is not conspicuous and may be megascopically absent. Under the microscope, however, the rock is found to consist of crystals of plagioclase, pyroxene, and biotite in a variable groundmass. In some portions the distinction between groundmass and phenocrysts is very marked, the groundmass being typically andesitic, while in other parts there is com- paratively little difference in size between the constituents of the groundmass and the phenocrysts. The pyroxene is light yellow-green in color with high extinction angle and non-pleo- chroic and rarely occurs in crystals longer than one millimeter. The plagioclase phenocrysts are usually somewhat larger, ordinarily clear but sometimes opaque from decomposition. The biotite appears in rather small crystals compared with the other phenocrysts and is of a light yellow-brown color with darker borders. In altered specimens the dark-colored con- stituents have decomposed into yellow non-pleochroic masses. The third variety, or vitrophyric andesite, is also found in the northern portion of the district as a low rounded ridge not more than one hundred feet above the surrounding country. It is also a pyroxene mica andesite, and is distinctly porphyr- itic, the phenocrysts occupying fully one half of the entire mass of the rock. Black mica and feldspar are very conspicu- ous and occasionally orthoclase crystals eight millimeters in length showing well-formed Carlsbad twins occur. The groundmass varies from a nearly black to hight gray transpar- ent glass. Under the microscope the feldspar is found to be of plagioclase and of an unstriated variety. It frequently possesses zonal structure and is often much broken, appearing in angular fragments. The biotite is in fresh hexagonal plates and irregular shreds. Pyroxene is light green and shows high extinction angle. Magnetite is present in the usual quantities. The groundmass is isotropic and filled with what appear to be small fragments of the phenocrysts and very small crystallites. 316 EF. NV. Guidld—Petrography of the Tucson Mountains. Fine flow-lines and perlitic cracks occur in places. This variety of andesite is very common in southern Arizona and frequently possesses flow-lines of remarkable beauty. (Fig. 2.) In the upper part of the andesite the groundmass has become entirely opaque through devitrification. Basalt.—Outflows of this rock oceur at various intervals along the edges of the mountain range especially west and south of Tucson. They vary greatly in character and may be grouped into the following varieties : 1. Fine-grained olivine basalt. 2. Porphyritic basalt. a. Containing phenocrysts of feldspar and aungite, in a coarse-grained or doleritic groundmass. | 6. Containing porphyritic crystals of feldspar only in a basaltic groundmass. e. Containing feldspar, augite and olivine in an andesitic groundmass. 3. Quartz basalt. One of the most prominent of these basaltic outflows is one mile west of Tucson in the form of a symmetrical cone-shaped © mass called Sentinel Peak. Immediately northwest of this is another irregular dome-shaped mass of the same rock. It is further represented in two promontory-shaped outflows south- west of the San Xavier Mission and ten miles south of Tucson. These elevations are made up chiefly of the fine-grained type of basalt in which none of the constituents can be recognized with the naked eye. It is usually compact and free from cavities, but occasionally is. found quite cellular and even scorlaceous in structure. The cavities are sometimes rounded in outline with a diameter of one half inch or more, but are more often drawn out by movements of the mass when ina molten condition into irregular channels. These cavities are usually empty, but are sometimes filled with gypsum or ara- gonite. The predominating color is black, but deep red varieties are met with, especially in the San Xavier outflow. In places pressure and movement of the mass have developed a schistose structure, the laminations frequently being nearly vertical. This is especially noticeable on the dome-shaped mountain mentioned above near the Carnegie Desert Botanical Laboratory, and leads to the conjecture that the vent through which the basalt escaped is located under it. Microscopically the rock is made up of numerous feldspar rods crowded together and frequently arranged in flow-lines, large amounts of magnetite and rather small quantities of olivine. Glass is pr esent in greatly varying quantities. The olivine occurs mostly in rounded grains with a dark red halo FN. Guild—Petrography of the Tucson Mountains. 317 of ferritic material and occasionally the interior of the crystal has been reabsorbed leaving skeletons filled by the groundmass. (Fig. 3.) The accompanying illustrations (figs. 3, 4) will show the most important variation in this type of basalt, In fig. 4 the constituents are more porphyritically dispersed than is usual in these outflows and the groundmass contains more isotropic material, yet to the unaided eye they all appear nearly identical. Porphyritic basalt of the first type (a) is found underlying the compact basalt of Sentinel Peak at its southern extremity. That this. represents an outflow distinct from the compact variety is shown by the sharp contact between the two, where there is a layer of dark red basaltic tuff and breccia from two to six feet thick. A rock practically identical with this is found at the San Xavier Mission in a small cone-shaped eleva- tion. This variety appears to be made up of large plagioclase erystals constituting nearly one half of the mass, and frequent black lustrous crystals of augite in a groundmass varying from coarse crystalline to compact. The feldspar crystals are some- times over one half inch in length and frequently broken. Under the microscope they are found to be quite fresh, twinned plagioclase filled with dark inclusions of the groundmass. The pyroxene is light yellow with parallel cleavage cracks and high extinction angle. The groundmass, where it can be made out with the microscope, is mostly feldspar and augite with but small amounts of glass. Olivine is not at all abundant and in some slides is absent. (Fig. 5.) The second type of porphyritic basalt (6) is found as ocea- sional outflows south of Sentinel Peak, the largest yet observed bemg about seven miles from Tucson. The color of the rock is medium dark gray, and the only minerals which can be determined in it by the naked eye are feldspar and occasionally magnetite. The feldspar is rarely over one quarter inch in length and is more rod-shaped than in the foregoing type. It becomes very conspicuous only as the rock weathers. Under the microscope the feldspar is like that in the first type. The groundmass can hardly be resolved by the microscope but seems to consist mostly of feldspathic material and magnetite. The third type of porphyritic basalt (¢) occurs in a very small mass not more than one hundred feet in length at the southern base of Sentinel Peak. Portions of the mass are amygdaloidal and much decomposed. The amygdules are sometimes six inches in diameter and filled with agate, usually in concentric rings of varying translucency, or with calcite or siderite. Sometimes there is an outer shell of agate, the interior being filled with calcite. Geodes of brilliant smoky quartz have also been found. In places the rock is sufficiently Am. Jour. Sci.—FourTH SERIES, VoL. XX, No. 118.—OcToBeEr, 1905, 22 818 LN. Guild—Petrography of the Tucson Mountains. fresh for satisfactory study. To the naked eye the porphyr- itic character of the rock is not at all apparent. Under the microscope, however, it is found to be made up of distinct porphyritic crystals of abundant feldspar, considerable pyrox- ene, and much less olivine in a semicrystalline groundmass consisting of a felt of magnetite and dark matter which reacts feebly under polarized light. Quartz basalt.—This unusual type of basalt is found in the extreme southern end of the mountain range as a portion of the promontory-shaped hill two miles southwest of the San Xavier Mission. The greater portion of the outflow consists of the compact basalt already described with cellular and scoriaceous modifications. On the extreme eastern slope an abundance of the quartz-bearing variety appears. Quartz is the only mineral that can be detected with the naked eye. Aside from this porphyritice constituent the general character of the rock from both a megascopic and microscopic stand- point is the same as the compact varieties described above. The quartz occurs as rounded and semi-angular grains, rarely more than six millimeters in length. Under the microscope the quartz appears clear, much fractured and quite free from inclusions of all sorts. That the quartz is primary and not due to secondary filling of cavities is inferred from the fact that the grains each consist of but one individual as shown by the extinction. This is not the case where previously existing cavities have been filled by infiltration. Amygdaloidal fillings have been observed in this same rock and they present a struc- ture quite different from the quartz in question. Basalts con- taining quartz have been described by Diller,* Iddings} and Pirssont from various localities in the United States, and by Andreae§ and Lacroix| from other regions. By most of these writers their origm has been discussed and they have been held to be of primary origin. Some, like Lacroix, have, how- ever held them to be inclusions, quartz grains caught up from lower rocks and held in the magma. DESCRIPTION OF FIGURES, PLATE Ix. Figure 1.—Rhyolite, showing flow-lines in the groundmass, ordinary light, 18 diameters. Figure 2.—Vitrophyric andesite, near Gila Bend. Figure 3.—Basalt, Sentinel Peak, 45 diameters. Figure 4.—Basalt, near San Xavier, 45 diameters. Figure 5.—Porphyritic basalt, showing crystals of augite, 18 diameters. Figure 6.—Agate, under polarized light, showing complicated structure, found as amygdaloidal filling in porphyritic basalt (c), 45 diameters. * This Journal, vol. xxxiii, p. 45, 1887. Bull. U. S. Geol. Surv. 79, 1891. + This Journal, vol. xxxvi, p. 208, 1888. Bull. U.S. Geol. Surv. 66, 1890. t Bull. U. S. Geol. Surv. 139, p. 129, 1896. & Zeit. deut. Geol. Gesell, 1892, p. 824. | Enclaves des rockes volcaniques, Ann. Acad. Macon, vol. x, 1898, p. 17. Am. Jour. Sci., Vol. XX, 1905. Plate IX. Chemistry and Physics. 319 SCN WEE hLCeINTELELGEN CE. I. CHEMISTRY AND PHYSICcs. 1. The Gases produced by Actinium.—It is known that solu- tions of radium salts give off continuously a mixture of hydrogen and oxygen from the decomposition of water, and it has been found that this detonating gas contains a small quantity of helium which is believed to be a product of the disintegration of the radium atom. DEBIERNE has recently confirmed this behavior of radium by using nearly a tenth of a gram of Curie’s radium bromide and operating in a manner similar to that of Ramsay and Soddy. He has found, further, that solutions of actinium salts give off detonating gas containing helium, and that the amounts of these products apparently correspond to the amounts produced by a quantity of radium having the same activity. For the experiments with actinium he used the whole of his most active products, and obtained the same results with a por- tion which had been specially purified from any possible contami- nation with radium by adding to it barium chloride and removing the barium. It was found, moreover, that the barium thus re- moved did not contain an appreciable quantity of radium. It was found also that solid actinium fluoride gave off helium. Debierne states that in addition to the large quantity of emana- tion with a rapid rate of decay which is given off by solid salts of actinium, there comes from it a very small quantity of an emanation of much slower change which he has identified as identical with the radium emanation; but its quantity is too small to have produced the helium found in his experiments.— Comptes Rendus, cxli, 383. H. L. W. 2. A New Heavy Solution. — Dusorn has prepared some liquids analogous to the well-known Thoulet’s solution, one of which, at least, appears to possess decided advantages over the latter. In the place of the potassium iodide used by Thoulet, he uses sodium or lithium iodide. The alkaline iodide and mer- curic iodide are alternately added to a small quantity of water until saturation takes place, the temperature being slightly raised at the end of the operation. ‘Then the liquid is allowed to cool, and after twenty-four hours it is filtered. It was found that Thoulet’s solution prepared in this way, and filtered at 22:9°, gave a specific gravity of 3:196 and an index of refraction of 1730, while the sodium mercuric iodide solution, filtered at 24°75°, gave a density of 3°46 and an index of refraction of 1°797. The lithium solution is intermediate in its density and refraction between the two just mentioned. Analyses of the solutions showed that their compositions corresponded closely to the formulas K,HegI,, Na,HgI, and Li,Hgl,, and in each case the amount of water present was somewhat more than 10 per cent. A similar ammonium mercuric iodide solution was prepared, but this was less dense than Thoulet’s liquid. 320 Scientific Intelligence. The sodium mercuric iodide solution is of considerable interest, as it is heavier than even methylene iodide. Although water produces in it a precipitate of mercuric iodide, it dissolves with- out change in alcohol and many other organic liquids. — Comptes ~ Rendus, exli, 385. H. L, W. 3. Hydrolysis of very Concentrated Ferric Sulphate Solu- tions.—It has been observed by Recoura that a concentrated solution of ferric sulphate made by dissolving the anhydrous salt in its own weight of water is completely decomposed when it is placed in contact with acetone for several days. The products are sulphuric acid, which dissolves in the acetone, and a basic ferric sulphate which separates in the solid form. The latter is yellowish white in color, is soluble in water, and has a composition represented by the formula 6Fe (SO): Fe, O,. H,0. The same solid is formed without the use of acetone when a strong solution of ferric sulphate is placed in a well-stoppered flask and allowed to stand for a longer time. With solutions of the strength given above, the deposit begins to form after about twelve days and extends through the liquid in about a month. With stronger solutions the precipitate is formed more rapidly and abundantly, while with solutions slightly more dilute no basic salt separates. The deposit is formed most rapidly at 20°, and more slowly as the temperature is kept lower.— Comptes Rendus, exl, 1685. H. L. W. 4, Separation of Gold from the Metals of the Platinum Group.—JaNNascH and von Moyer have found that gold is precipitated quantitatively by a salt of hydrozine in any kind of solution. This reagent, however, on account of its powerful reducing action does not serve to separate gold from the metals of the platinum group, although it is thus separated satisfactorily from potassium, sodium, barium, strontium, calcium, magnesium, aluminium, chromium, ‘zine, manganese, iron, uranium, nickel, cobalt, cadmium, mercury, lead and copper. Gold is precipitated by a hydroxylamine salt in acid solution somewhat slowly, and not below a temperature of 80°. Preliminary tests indicate that hydroxylamine hydrochloride is a satisfactory reagent for the separation of gold from palladium, platinum, iridium and rho- dium, as well as from ruthenium and osmium.— Berichte, xxxvili, 2129. Hi, Le By, 5. Determination of Sugar with Fehling’s Solution.— On account of difficulties encountered in determining small quantities of sugar by Fehling’s volumetric method, LavatuE has modified this by carrying it out in the presence of an excess of caustic soda, so that the cuprous oxide produced remains in solution, and the change i in color is more readily detected. The operation is as follows: In a porcelain dish of 200° capacity are placed 5 or 10° of Fehling’s solution, 30° of sodium hydroxide solution (1: 8), and 50 or 60° of distilled water. The liquid is then heated, and when it begins to boil the solution to be tested is gradually added. The operation is finished when the last drop causes the blue color of the Fehling’s solution to disappear.— Berichte, xxxviii, 2170. H. L. W. Chemistry and Physics. 321 6. Slow Transformation Products of Radium.—An article by Prof. E. RuTHERFORD, in the September number of the Philo- sophical Magazine, closes with the following summary of the products recognized in the slow transformation of radium. “The results of the comparison of the products of radium with those contained in polonium, radio-tellurium, and radio-lead are summarized below. (Radium D=product in new radio-lead, no rays. Half transformed in 40 years. Products | Radium E = gives out B rays, separated with bis- in old | muth, and iridium. MHalf transformed in 6 Radio- 4 days. lead. | fadiaas F = product in polonium and radio-tellurium. | Gives out only a rays. Half transformed in 148 L days. The family of substances produced by hie disintegration of radium, together with the time for each to be half transformed, is shown diagrammatically in the figure. 6d. RADIUM EMAN. RADA Rav 8B RAD.C RaD-D RADE RAD.F S ey & a e e /300 Yrs. 4dys. 3mins- aimins. &8mins 40yrs. édys. (43clys. | | RADIO-LEAD RADIO -TEt_uriur,PoLonium ACTIVE DEPOSIT RAPID CHANGE ACTIVE DEPOSI/IT SLOW CHANGE It is now fully established by the researches of Boltwood, Strutt, and McCoy that the amount of radium present in radio- active minerals always bears a constant ratio to the amount of uranium. ‘The investigations of Boltwood, in particular, have shown a surprisingly good agreement between the content of radium and uranium for minerals obtained from various localities, which differ very widely in their content of uranium. This pro- portionality is a strong indication that radium is produced from uranium ; and a conclusive proof of this point of view is given by the experiments of Soddy and Whetham, who find that there is a slow growth of radium in uranium which was initially freed from radium. In addition, the actual amount of radium in radio- active minerals is of the right order of magnitude to be expected from theoretical considerations, if uranium is the parent of radium. Soddy finds that the present growth of radium from uranium is only a very small fraction of the theoretical amount. This is most simply explained by supposing that one or more products of slow period of transformation intervene between UrX and radium. The uranium-radium family and their connection with one another is summarized below. 322 Scientific Intelligence. Uranium. | V Urx. | V ? | V Radium and its family of rapidly-changing products, viz., the emanation, radium A, B, and C. | V Radium D = primary constituent in radio-lead. V Radium E. | V Radium F = active constituent in radio-tellurium and polonium. No evidence has been obtained that any further active products exist after radium EF has been transformed. If the a particle is a helium atom, remembering that five products are present in radium which emit a particles, the atomic weight of the trans- formation product of radium F should be 225—20 or 205. This is very close to the atomic weight of lead, 206°7. The view that lead is the final or end product of the transformation of radium is supported by the fact that lead is always found in the radio- active minerals in about the amount to be theoretically expected from the content of uranium, when the quantity of helium, present in the mineral, is used to compute its probable age.* A similar suggestion has recently been advanced by Boltwood.” IJ. GEOLOGY AND MINERALOGY. 1. Indiana, Department of Geology and Natural Resources, Twenty-ninth Annual Report, W.8. Buarcuiry, State Geologist, 1904. 888 pp., 34 pl.—This Twenty-ninth Report of the State Geologist of Indiana is largely devoted to the economic interests of the state, which have shown a very large increase in recent years. Thus comparing the figures for 1895 with those for 1904, although there has been a falling off in natural gas, the amount of coal produced has been more than doubled and that of petro- leum increased nearly three times, while the value of the output of building stone and of clay products has also doubled. 'Twenty- five years since the resources of the state were almost exclusively agricultural, while in 1904 the total value of the mineral resources amounted to not less than forty million dollars. The present volume discusses very fully the clays and clay industry of the * A full discussion of this question was given by the writer in the Silliman Lectures, Yale University, March, 1905. Geology and Mineralogy. 323 state, in which direction the state has been found to be very rich, the shales, particularly those of the Coal Measures, which were not many years since supposed to be valueless, now being turned on an extensive scale into pipes and tiles, bricks of various kinds and other products. An account is also given of the petroleum industry, and the volume closes with an illustrated chapter upon the insect galls of Indiana by Melville T. Cook. 2. Geological Survey of Louisiana, G. D. Harris, Geologist- in-charge.—It is announced that hereafter the biennial reports of the State Survey of Louisiana will be brought out first as Bulletins and subsequently will be bound up in part as regular volumes. Of the Report of 1905, three Bulletins have already appeared: No. 1—The Underground Waters of Louisiana; No. 2—Maegnetic Survey of Louisiana; and No. 3—Tide Gauge Work in Louisiana. These may now be had gratis by address- ing Dr. W. R. Dodson, Director Experiment Station, Baton Rouge, La. 3. Geological Survey of New Jersey. Annual Report of the State Geologist, Henry B. Kimmel, for the year 1904. .317 pp., 19 plates, 18 text figures. Trenton, 1905.—This report contains a popular account of fossil fishes and their place in paleontology, by Dr. C. R. Eastman, followed by a detailed account of the fossil fishes of the Triassic as found in the Newark formation. Dr. Weller contributes papers on the faunas and corresponding formations of the Cretaceous of New Jersey. Professor F. B. Peck has a chapter on the tale deposits of Phillipsburg, N. J., and Easton, Pa., while the molding sands are treated by H. B. Kiimmel and 8. H. Hamilton. Progress is noted in the survey of the pre-Cambrian rocks in coédperation with the United States Geological Survey, and further parts treat of well records, forest fires and mining. The work throughout the report is thorough and of high grade; it deals largely with subjects of practical value to the state. SE 4. Brief descriptions of some recently described Minerals.— BECKELITE is a silicate of the cerium metals and calcium, described by J. Morozewicz and named after Prof. Fr. Becke of Vienna. It is found in a rock of the eleolite-syenite type, called by the author mariupolite and forming one of the petrographic elements of the Azov granite table. It occurs in coarse grains of a light yellow color, optically isotropic, also in octahedrons and dodecahedrons resembling pyrochlore. The hardness is 5 and the specific gravity about 4115. An analyis yielded : si0, ZrO,+R,0O, Mn,O, CaO MgO K,O Na,O ign. 17-13 65°31 0:07 15-46" 20739-20778. * 0: 99==100°13 The rare elements forming the 65°31 of ZrO,+R,O, included the following: ZrO, 2°50, Ce,O, 28:10, La,O, 13°60, Di,O, 18°00, Y,O,+Er,O, 2°80, Al,O, 0°30, Fe,O, tr. The calculated formula is Ca, (Ce, La, Di), Si, O,,.— Win. petr. Mitth., xxiv, 120, 1905. Several new species are described by R. H. Solly, in a recent number of the Mineralogical Magazine (xiv, 72). They are 324 Scientific Intelligence. derived from the dolomite of the quarries in the Binnenthal, Switzerland. Hurcuinsonits, named after Dr. Arthur Hutchin- son, of the University of Cambridge, iS a species occurring in prismatic orthorhombic crystals, with numerous terminal faces. The color is gray to grayish black, and the streak vermilion. The crystals are transparent to nearly opaque. Hardness, 1°52 ; cleavage, good, parallel to the macropinacoid. In composition it is found by G. T. Prior to be a sulpharsenite of thallium, lead, silver and copper ; it contains nearly 20 per cent of the rare element thallium. SMITHITE, named after Mr. G. F. Herbert Smith of the British Museum, occurs in monoclinic crystals, resembling flattened hexag- onal prisms, with prominent bas alplane. Thelustre is adamantine, the color light red, and the streak vermilion. ‘The crystals are transparent to translucent. Hardness, 1°5-2; cleavage, parallel to the orthopinacoid, perfect. The surface of the crystals changes on exposure to ight from pure red to orange red. According to G. T. Prior, the composition is expressed by the formula AgAss.. TRECHMANNITE, after Dr. C. O. Trechmann, occurs very spar- ingly in minute rhombohedral crystals resembling the two species hutchinsonite and smithite in color, streak and hardness. The crystals showed portions of hexagonal prisms, with small pyra- midal and rhombohedral faces. Cleavage was observed perpen- dicular to the prism. The composition is as yet undetermined. Marriter, named after Dr. John Edward Marr of Cambridge, occurs in highly modified monoclinic crystals, usually doubly terminated. The color is lead- to steel-gray, the surface showing iridescent tarnish ; the luster is metallic, brilliant. The hardness is 3 and the fracture conchoidal; no cleavage was observed. Only a single specimen had been found at the time the descrip- tion was published; this showed some fifteen small crystals im- planted upon the dolomite, hence though the crystallographic data are complete the composition is yet to be determined. LENGENBACHITE, named after the Lengenbach, a tributary stream in the Binnenthal, occurs in bladed crystals often very thin and sometimes curled up like paper. They show a highly perfect cleavage and splendent luster; the crystals are appar- ently twinned and are inferred to belong to the triclinic system. The plates are flexible and somewhat malleable but not elastic. The color is steel-gray, often with iridescent tarnish, the luster metallic; the specific gravity is 5°80. In composition it is essen- tially a sulpharsenite of lead with small amounts of antimony, silver and copper, as determined by A. Hutchinson. BowManirex, named after H. L. Bowman of the University of Oxford, occurs in rhombohedral crystals with basic cleavage and having the form of six-sided plates, often grouped in rosettes ; the crystals show a pseudo-symmetry in the basal sections. The color is honey-yellow, the luster brilliant vitreous to resinous. The hardness is 4'5, the specific gravity 3:2. According to Bow- man it is essentially a phosphate of lime and alumina with small amounts of iron, water and possibly magnesia. COLLEGES NEED: The Zittel Paleontological Wall Charts and Keller Reconstructions of Ancient Life. Our Improved Mineral and Rock Trimmer. Common Minerals by the Pound. Biological Laboratory Supplies. Our Unsurpassed Special Dissections. The Shaler-Davis-Harris Series of Physiographic Models and Photographs. Our Geological Models and Relief Maps. Beecher Restorations of Trilobite and Stylonurus and Models of Typical Brachiopods, Clarke Restorations of Eurypterus and Hughmilleria. Casts of Famous Inscriptions—Rosetta Stone, Black Obelisk, Deluge Tablets, etc. = Our Genetic Collection of Rocks. Our Collection of Phenomenal Geology. Penfield Goniometers, for laboratory and field use. Our Regular College Collections of Fossils, Minerals, Rocks, and Invertebrate Animals. Send for catalogues or circulars of some of these. THEY WILL DOUBLE YOUR TEACHING EFFICIENCY. Ward’s Natural Science Establishment, 76—104 COLLEGE AVENUE, ROCHESTER, N. Y. CONTENTS. : Page Art. XXVIII.—Ultimate Disintegration Products of the Radio-active Elements ; by B. B. Botrwoop __-----_-- 253 X XIX.—Use of the Rotating Cathode for the Estimation . of Cadmium taken as the Sulphate ; by C. P. Frora -_ 268 XXX.—Crystallization of -Luzonite; and other Crystallo- eraphie Studies ;-by A.J “Moses = 3-2 = eee 217 _ XXXJI.—Determining of the Optical Character of Birefract- ing’ Minerals; by FE. Wrigur. 2-522...) 3 See XX XII.—Groups of Efficient Nuclei in Dust-Free Air; by C. BaRUs ss 220 be eo ee ee XXXII.—Studies in the Cyperacer ; by T. Honm_-_-_.-- 301 “XXXIV.—Preliminary Note on some Overthrust Faults in Central New York; by P. FP. Scunriper = 228 - 308 XXXV.—Petrography of the Tucson Mountains, Pima Co., Arizona ; by. KON. “Guinp 203 4 ei eee 313 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Gases produced by Actinium, DEBIERNE: New Heavy Solution, Dusoin, 319.—Hydrolysis of very Concentrated Ferric Sulphate Solutions, Recoura: Separation of Gold from the Metals of the Platinum Group, JANNASCH and von Moyer: Determination of Sugar with Fehling’s Solution, LavauLEe, 320.—Slow Transformation Products of Radium, E. RUTHERFORD, 321. Geology and Mineralogy—Indiana, Department of Geology and Natural Resources, Twenty-ninth Annual Report, W. S. BLATCHLEY, 322.—Geo- logical Survey of Louisiana: Geological Survey of New Jersey: Brief descriptions of some recently described Minerals, 323. TEN-VOLUME INDEX. An extra number, containing a full Index to volumes XI to XX of the Fourth Series, will be ready in December. Sent only to those who specially order it. Price one dollar. $25 aed > eet. ee as ae Se ode me NS Men WT. WYTUS ACIETr, Librarian U. S. Nat. Museum. MevOr xx 5 NOVEMBER, 1905. a Established by BENJAMIN SILLIMAN in 1818. ~ THE AMERICAN JOURNAL OF SCIENCE. Epirorn: EDWARD S. DANA. ASSOCIATE EDITORS = PROFESSORS GEORGE L. GOODALE, JOHN TROWBRIDGE, : W. G. FARLOW anp WM: M. DAVIS, or Camprince, Y Proressors ADDISON E. VERRILL, HORACE L. WELLS, L. V. PIRSSON ann H. E. GREGORY, or New Haven, Proressor GEORGE F. BARKER, or PHILADELPHIA, Proressor HENRY S. WILLIAMS, or Itnaca, Proressor JOSEPH S. AMES, or Battimore, Mr. J. S. DILLER, or Wasuincton. FOURTH SERIES VOL. XX—[| WHOLE NUMBER, CLXxX.] No. 119.—NOVEMBER, 1905. WITH PLATES X, XI. NEW HAVEN, CONNECTICUT. 1905 THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. Published monthly. Six dollars per year, in advance. $6.40 to countries in the Postal Union. Remittances should be made either by money orders, registered letters, or bank checks (preferably on New York banks). ‘ RARE MINERALS IN TON LOTS Correspondence solicited with miners or consumers of the ores of Molybdenum Tantalum Thorium Titanium Tungsten Uranium Vanadium Yttrium Zirconium _and other rare metals. Offers from producers should be accom- panied by samples. : | i. SYSTEMATIC COLLECTIONS OF TYPICAL SPECIFENS . In sets of twenty-five up to fifteen hundred specimens. Prices $5.00 upwards per set, the average price for students’ specimens being about twenty cents. We have supplied the leading institutions for thirty years, having lately completed a single order for over 60,000 specimens. Our material is the accepted standard both as to correct labeling and high quality. Free Collection Catalog, containing lists and illustrations of General Mineral Collections, Series of Ores for Prospectors, Sets of Crystals, Series illustrating Hardness and other Physical Characters of Minerals, with Price List of Laboratory Material and Individual Specimens. FOOTE MINERAL CO, Established 1876, by Dr. A. E. Foote. W. M. FOOTE, Manager. DEALERS IN MINERAL SPECIMENS AND COMMERCIAL RARE MINERALS, 1317 Arch Street, Philadelphia. ae Am. Jour. Sci., Vol. XX, 1905. Plate X. 4 Carapace of Toxochelys Bauri Wieland sp. nov. from the Niobrara Creta- ceous of Gove County, Kansas, as partly restored and mounted in the Yale Museum.— Actual length about 53°™, AV Tela AMERICAN JOURNAL OF SCIENCE (FOURTH SERT#S.] Art. XXXVI.—A New Niobrara Toxochelys;* by G. BR. Wreranp. (With Plate X.} None of the numerous marine, or semi-marine turtles from the Kansas chalk or Niobrara Cretaceous have proven. of greater interest than the forms included within the genus Toxochelys. For this wholly extinct American group unites carapacial and plastral characters of the Lytolomas of the Upper Cretaceous of New Jersey with Chelydra-like cranial features, and is well represented by a considerable number of specific forms and variations presenting fairly clear evidence that we have here to deal with a line which independently acquired the modifications of limb structure suiting at least some of its members to a marine habitat. | Moreover it is very significant that discrete epi-neural ossi- cles somewhat similar to those the writer supposed might be present in Archelon are borne serially either on the neural, or over the neural junctions in an order suggesting that they have an ancient history, possibly analogous to the ossicles of somewhat similar form so characteristic of the Crocodilidee and in part the Chelydride. These ossicles as noted further on were first observed in Toxochelys (serrifer) stenoporis by Case (2) and later more fully described and commented on by Hay (6, 7, 8). The character of the entire series is, however, now determined for the first time. The idea that such ossi- * The writer’s previous contributions, mainly on the marine turtles, are as follows :— This Journal, vol. ii, Dec., 1896, pp. 399-412, pl. VI. American Natural- ist (p. 446), 1897. This Journal, vol. v, Jan., 1898, pp. 15-20, pl. II; vol. ix, Apr., 1900, pp. 237-251, pl. II; vol. ix, June, 1900, pp. 413-424 ; vol. xiv, Aug., 1902, pp. 95-108; vol. xv, March, 1908, pp. 211-216; vol. xvii, Feb., 1904, pp. 112-132, pls. I-IX ; vol. xviii, Sept., 1904, pp. 183-196, pls. V-VIII.—(In Press,—Protostega ; Memoirs, Carnegie Museum of Pitts- burgh ; Plastron of Protostegine. ) AM. Jour. Sct.—FourtH Series, Vou. XX, No. 119.—NOVEMBER, 1905. 23 326 G. R. Wieland—On Marine Turtles. cles really represent a disappearing series of dermal elements is further strengthened by the writer’s observation that inter- polated ossicles also occur in the marginal series of occasional specimens of Lytoloma angusta, as will be further considered below. Despite the frequent occurrence of Toxochelyds in the Niobrara, until now no complete carapace has been described. It is, therefore, of timely interest that a specimen collected by Mr. Charles H. Sternberg in Gove County, Kansas, and very recently acquired by the Yale Museum, includes a carapace and plastron sufliciently complete to determine accurately all the details of shell structure and form. The original locality, according to Mr. Sternberg, is in a ravine about three miles north of Monument Rocks, and about four miles east of the western Gove County line. This fossil is numbered 2823 in the Yale Museum accession list, and on the basis of the analy- sis given below is referred to the new species Zoxochelys Bauri, in honor of that distinguished student of the Testudi- nata the late lamented Professor Georg Baur. As shown on Plate X, 7. Bauri, represents one of the most ornate of all extinct Testudinate species. The type consists in the follow- ing elements :— The nuchal and eight closely articulated neuralia with the ninth median or post-neural element bipartite, and followed by an antero-pygal and the pygal marginal (the postero-pygal being the only ‘median element absent) ; three epi-neural ossi- cles ‘respectively seated on the 8d and 4th, the 5th and 6th, and the 8th-10th members of the neural series; the 1st— 3d, and the 8th-11th right marginals; the 4th—6th, Sth and 10th left marginals; most of the pleur alia; also the right hyo- and hypoplastron nearly complete, and various fr agments of verte- bree with several centra and arches. Of the right pleuralia the first and seventh are complete, and the third, fourth and sixth only lack rib-tips, while the expanded plates of all the right pleurals but the distal portion of the fifth, are fortunately present. On the left side the pleurals are not so complete, only the proximal ends having been recovered, with the excep- tion of the third, which only lacks a middle portion of the Plates tc fioure 6. The hyo- and hypoplastron lack their interior digitations, but fortunately permit an approximate restoration from what is known of the plastron of several other species (ef. figure 7). The fragmentary or not directly determinable skeletal parts include two dorsal centra, 4° in length, and several caudal centra, with a few por tions of cervicals. With the exception of some of the middle and anterior marginals, which are curiously crushed from very different G. R. Wieland—On Marine Turtles. B27 angles, the various elements of the present in reality excep- tionally fine fossil do not appear to have been much displaced in their original chalk matrix. This had been removed, how- ever, and aside from the neurals, which remained for the greater part solidly articulated, any clues to form and organ- ization afforded by position im the matrix had been thus destroyed before the specimen reached the Yale Museum. Despite this crushing and dissociation of parts, as the result of a careful joint examination by the Museum preparateur, Mr. Gibb, and the writer, it has nevertheless proven possible for the former to make a very handsome and successful mount- ing of the fine carapace with the considerably restored plas- tron in its approximately natural position, as illustrated on Plate X, and figures 1-3, and 6, 7. In fact it is owing to the presence of the nearly complete hyo- and hypoplastron that we are enabled to determine the true width of the carapace, which is indicated in the corrected drawing (figure 1) based in part on the measurement thus obtained. The specimen itself is mounted more nearly as removed from the chalk matrix, the width being somewhat exaggerated by compression. For it was at once decided that it would be far better inv mounting the specimen to adhere nearly to the form that had resulted from crushing in the matrix, rather than to distort the junc- tions of the several elements in an effort to reach the elongate form Yoxochelys Bawri really had. The restoration is accord- ingly, although at first sight indicating a considerable length of shell, not nearly so narrow and relatively long as originally in life,—an interesting fact because this is almost the only marine form with a carapace suggestive of the great length seen in Dermochelys. Description of Parts. As the main features of the anatomy of the carapace appear in sufficient detail in the summary of characters and measure- ments given below, taken in conjunction with the accompany- ing figures and plates, we may pass on to a discussion of the special or unique features of interest, namely the nuchal, the epi-neural spines, and the pygal region. Nuchal.—The Trionychid-like fontanelles at the junction of the nuchal, first neural and pleurals (figure 1, 7}, are circu-. lar to slightly elliptical, and 1°" in diameter. Such have not been hitherto observed to occur outside the Trionychids, and with the general form of the nuchal suggest a certain connec- tion with original lines less distant from the Trionyehid stocks than are the Chelonine. Elsewhere the writer has suggested that the Nuchal and Epiplastra of Dermochelys, Protostega, and the Jurassic Thalassemyds may go to indicate a yet ~ 328 G. R. Wieland—On Marine Turtles. closer relationship to stocks ancestral to the Trionychide, and that there are many most suggestive indications that indepen- FiGurRE 1.—Carapace of Toxochelys Bauri Wieland, x14 nearly. (Drawn from type.) N, Nuchal; 2, 4, 6, 8, Neuralia; 9’, posterior segment of the 9th or post-neural element of the median series; A, Antero-pygal; P, Pos- tero-pygal; M, Marginalo-pygal; I-VIII, the Pleuralia; M0, 10th (rib- free) Marginal; f, the post-nuchal foramina; f!, f%, Ist and 8th pleuro- marginal fontanelles. The three Epi-neurals are not lettered. dent marine races of Testudinates, of which at least a half dozen may be enumerated, have been repeatedly developed ever since the Jurassic. G. R. Wieland—On Marine Turtles. 329 It is also of much interest that while in forms like Osteopy- gis a nether nuchal process is wholly absent, there is in the - present turtle a mere, although distinct beginning of such a FIGURE 2.—Plastral view of Toxochelys Bauri Wieland, x14 nearly. (Drawn from type).—en, Entoplastron ; h, Hyoplastron; hp, Hypoplastron ; x, Xiphiplastron ; f, f, f, the median and the lateral hyo-hypoplastral, and the hypo-xiphiplastral foramina; 4-7, the plastron-supporting marginalia. Other letters as in figure 1. process, and in Zoxochelys latiremis a vouch larger projection for actual cervical articulation. This process thus appears to have arisen in different groups rather than to have been com- 330 G. R. Wieland—On Marine Turtles. monly present in Cretaceous turtles, and may now be consid- ered to have been definitely traced to its origin in at least one genus. Figure 3.—Lateral view of the Carapace of Toxochelys Bauri Wie- land, x 14 nearly. Drawn from the type. s,s, s, the three Epi-neural spines supported respectively by the 3d and 4th, the 5th and 6th neurals, and the 8th neural and the post- neural elements; f, the post-nuchal foramina. : Lipi-neural Spines —The ser- ies of epi-neural spines taken in conjunction with the strongly carinate neurals, and the keeled marginals, give to the present fossil carapace a most ornate form. See figures 1 and 3. The earliest suggestion of the possible presence of epi-neural elements in the Testudinata was — made by the writer in his orig- inal description of the Fort Pierre Cretaceous turtle Avrche- lon given in this Journal for Dec., 1896. It appears on page 400 of that number as follows: “One of the chief features of the carapace is the arching into a heavy dorsal ridge, beginning just back of the region of the tirst dorsal vertebra, and from thence continuous, except in the sacral region. It marks the position of the neural spines and is very distinctly grooved from anteriorly to the region of the eighth dorsal vertebra. Im- mediately over the neural spines this groove inclines to widen and send out asteriations. In life these grooves were no doubt filled with horny material, and the animal may have borne w dorsal row of spines.” Two years later the spines of Toxochelys were first observed by Case,* and have been since more fully described and com. mented on by Hay, who would see in them the remnants of a former dermal series, probably onee common to all turtles, and going far to explain the homol- * Kansas Univ. Geol. Survey, vol. iv, p. 382 (1898). G. R. Wieland—On Marine Turtles. a3. ogy of the osteodermal mosaic of Dermochelys (6, 7). The present is, however, the first time that the entire series of ossicles and their relation to the successive neurals has been determined. As may be judged from reference to the several figures, the system of ossicles may really be a much changed and disappearing one. ‘The first neural bears a small but very distinct completely fused boss near its middle, and then forms the beginning of the dorsal carina. The third neural which is rather short, and the fourth which is abnormally long, support a large epi-neural spine. This occupies all the median poste- rior three-fourths of the length of the third and the anterior fourth of the length of the fourth neurals, and is doubtless the second member of the original epi-neural series. The second free epi-neural [or third of the original series] is the largest, and is equally borne by the fifth and sixth neurals. The third free epi-neural [or fourth of the hypothetical primitive series | rests over the ninth member of the neuralia, so as to project slightly forward onto the eighth and well backward over all the anterior half of the post-neural tenth. This latter epi- neural is the smallest of the three free epi-neurals. 4 s; Figure 4.—Toxochelys Bauri type, x'1/:. a, Vertical transverse section through the second neural showing the average elevation and outline of the median neural carina. b, Vertical transverse section through anterior end of the 6th neural, and the epi-neural spine (s) borne on this and the 5th neurals. Whether a fifth member of the epi-neural series was borne by the postero-pygal, which would afford the symmetrical posi- tion, is of course conjectural in the absence of this latter member of the median series. / S52 G. R. Wieland—On Marine Turtles. Whether or not the keels of the marginalia mark the fusion of a lateral series of elements, corresponding to the epi-neurals, is likewise only conjectural, although it appears that some light may be shed on the subject by Proganochelys. There is however some uncer tainty as to the number of marginals and true significance of the peculiar marginal fringe of spines in this singularly interesting turtle as so carefully studied by Fraas (4) from material recovered under conditions unfavor- able to the exact preservation of structural details. But it is also a most interesting and suggestive fact that small ossicles are irregularly inter polated between the lateral marginals of the Cretaceous Zytoloma, as small triangular elements about 15° on each side. Such are shown at E, E, E in the accompany- ing figure 5. As these epi-marginal ossicles are not equally pre-— sent on both the right and left marginals even in the same individual and certainly not always present in all specimens of Lytoloma angusta, they would at first sight appear to be of much less significance, taken by themselves, than are the epi- neurals of Toxochelr ys. Nevertheless it would now seem that they do represent a disappearing series that may once have invested the entire margin of the carapace. If so, they form one of the most impressive examples of the very last vestiges of a vanishing series. The truth of this hypothesis yet remains to be mainly deter- mined by fossil evidence, which we may hope ere long to dis- cover, 1f correct. At any rate it is extremely interesting and suggestive to find further traces of an additional osteodermal series in Lyotoloma, whatever may be the homology to that of Dermochelys. What the characteristic number of elements in this system as developed in pre-Cretaceous testudinates was, no one has yet attempted to suggest. Nor is it possible to reach a safe conclusion in the absence of further paleontologic evidence. It would appear however that the series was once at least as complex as is the horn-shield and the bony plate series, and that it had some form of alternate or imbricate relationship to both these latter systems. Also, if the osteodermal mosaic of Dermochelys arose from such an additional dermal series, such origin must therefore have been in part by a subdivision pro- cess such as was suggested to Baur by the ‘abnormal breaking up into smaller ossicles along the edges of the pleurals observed by him in Hretmochelys. Such a subdivision would of course follow the lines of the original system, and could thus very well produce the carapacial carina seen in Dermochelys. It should be especially noted in this connection that such an hypothesis for the more primitive origin of the osteodermal mosaic does not necessarily imply a more ancient origin for G. R. Wieland—On Marine Turties. aoe Dermochelys than for the Chelonide, and that its correctness would not necessarily leave Dermochelys the most primitive of turtles, but rather the most specialized, as hitherto held by Baur, Dollo, and the writer. As stated, only new fossil evi- dence can settle the very interesting questions that here arise. oor = FiGuRE ).—Carapace of Lytoloma angusta from the Upper Cretaceous Greensand of Barnsboro, Gloucester Co., New Jersey. E, E, E, Epi-mar- ginals respectively borne by the right 4th and 5th, 5th and 6th, and the left 4th and 5th marginalia. (Enough marginals are present in the original spe- cimen—No. 625 of the Yale Collection—to determine that no further epi- marginals accompanied these three, unless such were borne anteriorly to the 4th marginals.) Epi-marginals are not always present in L. angusta. The pygal region.—The neural series of Toxochelys Bauri, excluding of course the epi-neural ossicles, agrees with that of Hardella thurgi (1) in having ten elements, in the neural row, —in reality an interpolated element between the normal or S04 G. R. Wieland—On Marine Turtles. common eighth and ninth elements, or better a division of the ninth or post-neural region of the median series. Unlike Hardella, however, the pygal is not single, the post-neural region being divided into an antero- and postero-pygal, as in Osteopygzis, and in the Cheloninze. The existence and outlines of the postero-pygal are indicated by the conformation of the pleuralia and posterior marginals, together with the posterior suture of the antero-pygal and the anterior suture of the. pygal marginal, which are quite unlike. From these sutural borders it is also quite evident that the heavy median keel which evenly traverses all the length of the antero-pygal, finally ran out on the postero-pygal, where it no doubt ended as a distinet boss like that of the first neural, which would perforce repre- sent a fused fifth member of the median or epi-neural ossicular | system. The pygal marginal, in correspondence with the strong keels of the marginals, is ornately double-keeled. The organization of this region has not hitherto been determined in any species of Zoxochelys. Both Case (2) and Hay (6) have figured the posterior half of the carapace of 7. (serrifer) steno- porus type, but without determining the sutures, whether because not indicated or because of difficulty of interpretation not being stated by either. A distinct difference from the present specimen is, however, obviously present in the postero- pygal region. Synopsis of the Characters of Toxochelys Bauri (type). Carapace.—Elliptical to elongate in outline with large and persistent pleuro-marginal fontanelles; composed of 52 bony plates and 3 additional epi-neural spinose ossicles; numerical arrangement of parts combining the general alignment and form seen in the Chelonine Lytoloma angusta with the post- neural arrangement of the existing Hardella thurgt. Surface finely granulate to smooth, and horn-shield sulci not apparent, save for notches formed by the posterior border of the mar- ginal keels. (A distinctly leathery hide is not, however, sup- posed to be present.) Marginals, 11 pairs, rather narrow anteriorly, increasing very slowly in breadth to the 11th, which is still nearly twice as long as broad, outer borders all the way to the pygal marginal more and more sharply keeled anterior to the indistinet to absent horn-shield sulci, upper and nether surfaces of nearly equal area, supported by rib-ends only with ~ the pits of the plastral digits small to indistinct and extending from the 3d to about the middle of the 7th; rib-pits small, with the 10th marginal ribless, and the 11th supporting the 9th rib anteriorly as in Chelone and Lytoloma. Nuchal large and very broad, uniting by straight sutures with the 1st neural and 1st pleurals, between which are formed G. R. Wieland—On Marine Turtles. 300 posteriorly two small oval fenestrae as in the Trionychids ; with a minute (incipient) nether articular projection but no costiform processes. Neuralia 8 with the post-neural bipar- FiGuRE 6.—Toxochelys Bauri (type). ' A supplementary figure to Plate X, showing by stippled surfaces the parts of the original carapace actually recovered. (Lettering as in figure 1.) tite, oblong to hexagonal, prominently carinate and supporting the three large epi-neural spinose ossicles. Antero- and pos- tero-pygal nearly as in Lytoloma. Pleuralia more reduced than in either Chelone or Lytoloma. 336 G. R. Wieland—On Marine Turtles. Plastron.—Ot the same Chelydroid form seen in Osteopygis and Lytoloma. FIGURE 7.—Towxochelys Bauri (type). Restoration of the plastron. x+.— The stippled surface shows the portions of the hyoplastron (h) and the hypo- plastron (hp) actually recovered.—(The epiplastron and entoplastron is only known in T. latiremis, cf. figure 8, and the xiphiplastron in T. stenoporus.) Specific Relationships. The specific identity of the Zoxzochelys described in the foregoing pages with any of the known species of the genus cannot be affirmed, as appears from the following analysis.— Five species have been assigned to the Niobraran genus Zoxo- chelys as first established by Cope in 1873, namely: 7. latire- mis, the generic type; 7. serrifer, Cope, 1875; TZ. brachy- rhinus, Case, 1898; and 7. procax and T. stenoporus, as pro- posed in a recent revision of the genus by Hay (8). With 7. Jatiremis as close a comparison as desirable is not yet afforded, since but few of the elements of the carapace and plastron of this form are known. It appears, however, that the nuchal was of markedly different proportions from those G. R. Wieland—On Marine Turtles. BS of the present 7. Lauri, as may be noted on comparison with a nuchal figured by Case.* Nor is there specific agreement with the nuchal of the Yale specimen I referred to, 7. dateremis, when describing the accompanying flipper (10). This nuchal is here shown in figure 8 for the sake of more convenient reference. 8 FicurEeE 8.—Towxochelys latiremis, from the Niobrara.Cretaceous, Gove County, Kansas. (Yale accession list 2419.) x about 1. Nuchal with the attached first marginals of both sides and the proximal half of the right second marginal, together with the accompanying epi- plastron.—This nuchal bears far back nearly in line with the front border of the large curved posterior notches a large and prominent nether process for cervical articulation. Although true that the general form varies in turn from that just noted as figured by Case, the differences are more easily reconciled within specific limits. The simple fact is that in no previously described specimen of Zowxochelys, and in no other semi-marine, or marine member of the Chelonide, do we observe Trionychid-like foramina between the nuchal and first neural and pleurals. JI may add that from recent measurements given by Hay it appears that amongst the several Toxochelyds 7’ brachyr hinus is next related to 7. latiremis ; and there is a question in my mind if the former is a distinct species, the differences in cranial proportion from T. laturemis being so slight as to be of very doubtful signifi- cance in specimens so invariably crushed at more or less vary- ing angles as are the Niobrara fossils. With the skull fragments and crushed [9th] left marginal of 7. serrifer as recently figured for the first time by Hay (8), I am unable to identify the present handsome specimen. As the horn-shields of 7. serrzfer formed a very deep marginal notch leading into a pronounced sulcus (as indicated by Hay), there appear to be distinct differences. It is, of course, one of the difficulties of vertebrate paleontologists that species based * University Geol. Survey of Kansas, pl. lxxxii, figure 3. 338 G. R. Wieland—On Marine Turtles. on such meager skeletal parts accumulate in the course of time; but surely we are permitted little diffidence in applying the laws of priority and nomenclature now in vogue to a hand- some and reasonably complete fossil like that discussed in the present paper. Perhaps the day is not distant when fr agments will be merely noted within generic limits, and then numbered and laid aside for a certain number of years before being arbitrarily dignified as the types of new forms. Assuredly such a method would simplify the study of extinct faunee. The extreme difficulty of reaching accurate specific identifica- tions alter most painstaking comparisons and study of deserip- tions primarily based on fragmentary material, has been espe- cially brought home to the writer in his consideration of the Upper Cretaceous Turtles of New Jersey, and he has great sympathy with Professor Marsh’s oft repeated contention that ‘the types of extinct vertebrates ought to be mainly founded on fairly complete forms. With the isolated and imperfect skull of rather large and robust form named 7. procaw by Hay, as with that of 7. brachyrhinus, no comparisons are afforded by the material thus far obtained. From 7. (serrifer) stenoporus, finally, 7. Bawri differs distinctly. as shown by comparison with the posterior half of the carapace figured by Case.* From that and other specimens of 7. (serrifer) stenoporus the present fossil differs in being of a larger type with relatively heavier marginals and larger pleural plates; also in the much more pronounced sutural union of the postero- and marginalo-pygal, which 1s reduced to peg-like junction in 7. (serrifer) stenoporus. Systematic Position of the Genus Toxochelys. Because of the carapacial organization with much reduced — pleurals and marginals, as well as certain plastral characters, all suggesting primitive relationships to the Cheloniudee, it was first suggested by the writer on his discovery of the organization ‘of the front lege of Toxochelys latiremis, that the Toxochelyds do not justly constitute a separate family of tur- tles, as proposed by Cope and held by Hay, but are better classified as a sub- family of the Cheloniide, the Toxoghely- dinee. eee Hay, while accepting the principle that the limbs do furnish “a test of the correctness of this disposition of the genus,” interprets the evidence differently (7). He now reaches the conclusion that Wieland misinter preted the limbs of 7. latiuremus (10), and that these, as in the Trionyehid Amyda spinifera, were merely long fingered and webbed, * Kansas Univ. Geol. Sur., vol. iv, plate lxxxiii. G. PR. Wieland—On Marine Turtles. 339 and not markedly modified for marine life, so that Zoxochelys “ did not navigate the open seas.” In support of his contentions Dr. Hay uses a percentual method of comparison in which the humerus is conveniently and arbitrarily considered the unit in terms of which the length of the digits is expressed. This very effective means of comparison was first used by the writer in the case of forms in other ways related, and is, within limits, unquestionably useful in a diagrammatic sense. But Dr. Hay now mistakenly employs it in a far wider application than originally contem- plated, when he reaches direct conciusions as to the front limb of Zoxochelys by comparison with the Trionychid Amyda spiniferda, thus :— ARM. FINGERS. (a —— = Ee (eat —_—~— eN Humerus. Radius. Ulna. Ist. 2d. 3d. Ath. oth. Amyda 100 53 5) Il 69 90 98 116 98 Toxochelys 100 58 50 51 73 100+ 104+ TO One might as well go on to prove that the “ hawks-bili,” Eretmochelys imbricata, 1s unable to “ navigate the open seas”’. —For similarly : ARM. FINGERS. ee ——*~-—-— aN Car ——_ —*~—_—_ ==>) Humerus, kadwuse.. Ullnas ist 2d. 3d. Ath. oth. Amyda 100 58 Hil CD. OO GS ANG Oe Eretmochelys 100 53 FART AQT 89) 28) Obs | 44 Whence the following differences: ARM. FINGERS. =a SSS iS aS = \ Humerus. Radius. Ulna. SG 2d. SOL, Ail. Ohl Amyda aa sit +7 +20 +1 __ +11 +4854 Eretmochelys — -- ae oy as Pasa Lis It is clear that save for that short thumb and long fourth finger of Hretmochelys, were this an extinct form, no conelu- sive evidence of the true flipper development would be afforded by such measurements as the above when considered alone. For it is a noteworthy fact that the disparity between the thumb and fourth finger of Amyda is +47 as against +53 in Toxochelys, and +56 in Eretmochelys. Yet as a true indi- cation of unequal finger development, instead of disparity between only the short first and the long third and fourth “fingers, as in Hretmochelys, there was in Toxochelys strong dis- parity between the short first and second and the long third and fourth fingers. There was also ulnar disparity. All these fundamental numerical relations have been over- looked in Dr. Hay’s criticism. He entirely ignores, too, the 340 G. R. Wieland—On Marine Turtles. fact that as a merely web-footed turtle Zovochelys would have been very unlike Amyda. For these percentual results must always be considered in connection with the humeral changes in the direction of marine forms, which are indicated in the thalassoid humerus of Zoxochelys, as well as the enlargement of the pisiform to nearly the same size as in Hretmochelys. In short, it is evident that Dr. Hay overlooked important factors and that his views are untenable. When I originally described the flipper of Zoxochelys I was of the opinion that it represented the most primitive form yet discovered that could be called more distinctly marine than merely natatorial, long-fingered and web-footed ; and now that I have had the present opportunity to briefly reconsider the subject I may say that I believe this interpretation in accord with the facts.* Dr. Hay “readily grants that the fore limb of Toxochelys had entered on the early stages of those modifications which resulted in the production of flippers.” But as clearly enough indicated by the facts, much more modification had been under- gone, and the foot was more a swimming than a walking one. Whether the third to fifth fingers were encased in a leathery hide, or still retained some of their freedom of motion, as in distinctly webbed types, is open to some question ; but never- theless finger disparity, reduction of the 3d—5th claws, pisiform enlargement and humeral change had all been accomplished to such an advanced extent that the limb is to be regarded as a flipper, quite admirably fitting Zoxochelys latirenis to range the great inland Niobrara Sea. And even were the anatomical facts of less certain interpretation, the onws probands would rest on him who asserted the non-marine nature of those turtles which occur so widely distributed in as extensive a chalk forma- tion of indisputably marine origin as the Niobrara Cretaceous. It is very evident, therefore, that on the basis of limb organ- ization Zoawochelys is a member of the Cheloniidee, and that as proposed by the writer on the basis of the general organiza- - tion, limb structure, and relationships the genus is most con- veniently placed in the Chelonidan sub-family Toxochelydine. As a concluding word it may be added for the sake of clear- ness that no great diagnostic significance is attached to the presence of the epi-neural ossicles,—certainly not if they are to be regarded as vestiges of a disappearing system, likewise indicated in the genus Lytoloma of the Chelonine. Yale Museum, New Haven, Conn., Sept. 26, 1905. *TIn view of the great interest of the subject I will as early as convenient refigure the flipper of Toxochelys with all possible care. Dr. Hay is also of thé opinion that the great turtles of the Fort Pierre, and perforce the Nio- brara Protostega were likewise littoral and web-footed rather than marine. As will be incontestably demonstrated by the writer in a forthcoming Memoir of the Carnegie Museum of Pittsburgh, Protostega and Archelon were powerfully equipped for their marine habitat. G. R. Wieland—On Marine Turtles. 341 Measurements of Carapace and Plastron of Toxochelys Bauri. (Yale Museum accession list 2823. Elements disarticulated and more or less altered in form by crushing in matrix. Recovered portions as shown in the accompanying figure 6 by the stippled surfaces.) Length of carapace (estimated to within 1 or 2%™ _______- 530m Breadth of carapace (greatest, as measured across the ~ anterior sat OF GhetGLh Netikaky seas oN eC Bes 40+ (a) (0) Exact length Width measured on outer edge at notch of the of carapace. hornshield sulci. UN ilo See ie rae ok 12°0 ie bstamranomales fe pe 6:0 2°5 2d Eee nee Steere tet 5°0 2°5 3d Sete ran Mace apeeet abi st te | 5°9 2°8 4th x Be peri eae See eh Bee Lt 6°0 a 5th Se ene ie wee Tas 6°95 2°8 6th ee aes ey ee ks le 70 3°3 Basle ec ees obras WS ie 75 me Ste 2S pegs Pele Stoeger 8:0 4°5 9th Bice e Pa ea RGR Se 75 4°5 roca oe zat AO ers nce ees 7°0 4°5 PEG yp) ee 6°8 4°5 1 7 SR aA a ic aac pas 7°0 4°5 (The thickness and transverse sections of the marginals are approximately the same as in Lytoloma angusta. Owing to the crushing undergone by most of the marginals a closer approxi- mation cannot readily be given.) Length on Greatest Median line. —_~width. Brae lials atinsce eo aa ey 5°5 * 14:5 list: HEURAL Ss 2.12 32h0% ies repeat ae hohe 3°8 3°8 De ae hapa te NE Soe ne se Lae 4°4 3°7 SOS Preah Ni ea oe Ce EY ce 4-4 4°0 GG i, Ms ee eC 46 4:4 SU iper cdma nets = MOE Ses eet 3°4 4°5 Deepa hie oer acc Ate Oe | OR Bite Be 5°0 4°0 FE SIS NR Aa Sain ah i kta ok Rn ae a 4°0 3°8 2) 18 SL GAN SDS Sa ae eR oe Spare gn D5 8°5 Sphere hy eRe ae one eer apa Ibe. 3°d EOS pies iS Oe Oia sca i a Saar n 2°5 3°4 PRTEDEEO NYS les os ae syns 8 4°5 5°9 (Posterd-pygal)s2 eval (3°5) (4:0) Marginalo-pygal 22.2. 52: 2. -- 4°5 6°5 Am. Jour. Scl.—FourtTH SERIES, Vou. XX, No. 119.—NOVEMBER, 1905. 24 342 G. R. Wieland—On Marine Turtles. Ist epi-neural Ossiclel: mld gutsy 3°5 ver 2d Ba Seats Vaan Taam oe tae wea coe 4°5 2°0 3d ie sas Se eat Rg 3°9 1°5 (Thickness of 2d neural measured through carina, 1-4°™.) (The total height of the epi-neural ossicles is respectively, 15 21, and 15™™, the projection above the carina, 9, 12, and 9™™.) > (d) (a) Length of Length posterior (c) over sutural Median curvature. border. width. Nena ts ice ney a naa [10] cn ist pleurales 2s. eye 15°0 6°5 5:7 od a WNC eals uae Mane atin 18°5 8°4 On) 3d Se elt ceric OL AUR oily 19°5 8°4 5°0 Ath e Bs Oe SHON Yio 19°5 8°2 4:9 5th iG yt adhee ths a Samant te 19:0 Gil 4°8 Ctl NG een CN ae 17°5 6:9 AS 7th SES ANION pe agin Dane ai 14°5 5°1 3°8 8th SEN de Seeley mul etd Ad abs 2°2 3°8 (The average thickness of the pleurals is 50™™. The distance between the bases of the rib-capitulae of the 7th pleurals is 4°. The large pleuro-marginal fontanelles are approximately one- half, or more than one-half the length of the pleurals which bound them. The boa sulci, save for the notched marginals, are indistinct.) The Plastron. (Cf. figure 7.) (With added width of the inferior faces of the adjoining mar- ginals, or 2° < 2, this measurement yields as the approximate breadth of the carapace 40°.) Width (on antero-posterior line) of the marginalo- hyo-hypoplastral fontanelle 2250553 a od Length of hyo-hypoplastral suture __..-_-.---- Bel) Least width of the hyo-hypoplastral bridge .-.-. 11:0 10) EA. 12, G. R. Wieland—On Marine Turtles. 343 References. . Boulenger, G. A. : Catalogue of Chelonians, British Museum of Natural History, 1889. . Case, E. C.: Toxochelys. University Geological Survey of Kansas, Paleontology, Part IV. Topeka, 1898. . Cope, E. D.: Vertebrata of the Cretaceous Formations of the West, vol. 1, Rep. U. 8. Geological Survey of the Territories. Washington, 1875. . Fraas, E.: Proganochelys Quenstedtit Baur (Psammochelys Keuperina Qu.). Ein neuer Fund der Keuper- schildkréte aus dem Stubensandstein. Jahreshefte des Ver. fiir Naturk. in Wiirttemberg. Stuttgart, 1899. . Gray, J. E.: Supplement to the Catalogue of Shield Reptiles in the British Museum. London, 1870. . Hay, O. P.: On the skeleton of Toxochelys latiremis. Pub- lications of the Field Columbian Museum. Zool. I. Chicago, 1895. — On the group of Fossil Turtles, known as the Amphichelydia ; with Remarks on the Origin and Relationships of the Suborders, Superfamilies, and the Families of the Testudines. Bull. American Museum of Natural History, vol. xxi, Article LX. New York, June 30, 1905. . — A Revision of the species of the Family of Fossil Turtles called Toxochelyide, with Descrip- tions of two New Species of Zoxochelys and a New Species of Porthochelys. Ibid., Article X. . Wieland, G. R.: Archelon ischyros: A new gigantic cryp- todiran Testudinate from the Fort Pierre Creta- ceous of South Dakota. Ibid., vol. 11, Dec., 1896. — Notes on the Cretaceous Turtles Zoxochelys and Archelon, with a classification of the Marine Tes- tudinata. Ibid., vol. xiv, Aug., 1902. — Structure of the Upper Cretaceous Turtles of New Jersey: Adocus, Osteopygis, and Propleura Ibid., February, 1904. — Structure of the Upper Cretaceous Turtles of New Jersey: Lytoloma. Ibid., Sept., 1904. 344 Pirsson and Washington— Geology of New Hampshire. Arr. XXX VII.— Contributions to the Geolegy of New Hamp- shire. I. Geology of the Belknap Mountains; by L. V. Presson and H. 8. Wasuineron. (With Plate XI.) Introductory Note.—Our object in this paper and in one to follow it is to present the results of a study made im the field and in the laboratory of the occurrence and characters of a group of igneous rocks from a locality about which little is known. Our field work was done in two visits to the area and covers a period of between two and three weeks, during which it was traversed and roughly outlined and the highest peaks and ridges ascended. ‘This was sufficient to give a good gen- — eral idea of its geology and of the various rock types. In the lack of a suitable base map on a sufficient scale, upon which to make record, more detailed and careful work was not war- ranted and would have enabled us to add little of interest to the general results presented in this paper. The map used and upon which our results are given is taken from that accom- panying the Hitchcock Survey, referred to later, and which we have modified to some extent. The topogr aphy is more or less generalized and in places somewhat inaccurate, but it is the only one showing topography of which we have any knowl- edge and it has ser ved as the basis of several topographic maps since published for the use of tourists which we have also con- sulted. LocATION AND GEOGRAPHY. The Belknap Mountains form an elevated tract south and west of Lake Winnepesaukee in New Hampshire and lying in the townships of Gilford, Alton and Gilmanton. Although they are sometimes referr ed to as the « Belknap Range” they do not form a mountain range of the anticlinal type, being the irregular, eroded upper portion of a great intrusion of igneous rock of a generally granitic char: acter. In its greatest length, which is northwest and southeast, the mountain tract extends about eleven miles and its width at the broadest point east and west is about six miles. In shape the mass is triangular, the long side facing the west composed of the main ridge which ear- ries the highest summits, while an eastward extension produces the triangular shape. At the eastern end of the triangle there is an extension running southward. On the north and east sides the slopes descend into Lake Winnepesaukee; on the west and south into a much- lower, irregularly hilly country, The drainage on the west is carried off by the Gunstock River, which in its course of about six miles runs due north at the foot of the mountain slopes in a valley cut along the contact Pirsson and Washington— Geology of New Hampshire. 345 zone of the igneous rock mass. On the south the drainage is less clearly defined and is carried off through a series of small lakes which empty to the southward. On the other sides small brooks run into the lake. The mountains are quite gen- erally covered with trees andebrushwood on the steeper slopes; below these are generally open pasture fields, and the highest erests and summits are more or less bare rock exposures with small meadows between them. At the foot of the eastern and northern slopes, along the shore of Lake Winnepesaukee, runs the Lake Shore Railroad, a branch of the Boston and Maine Railway system, which ends at Lakeport-Laconia. These towns with Alton Bay at the south end of the lake and the vil- lage of Gilford are the most important places in the vicinity of the mountains, although the shore of the lake at their foot is thickly dotted with summer cottages and places of resort. Around them elsewhere is an open farming country and the high valley between the northern extension and the eastward one of Mount Straightback is also a cultivated area reached by a road over the mountains from Gilford to West Alton. Listoricul.—The only reference in the literature to the geology of the Belknap Mountains which we have been able to find is the short description by Hitchcock.* He states that the mountains are composed of eruptive syenite similar to that of Red Hill in Moultonborough. He describes briefly a few localities, and mentions that in places it is in contact with porphyritic gneiss and mica schist. He thinks that the syenite has come up through asynclinal fault. Near the contact with the porphyritic gneiss “it is brecciated and full of dark hornblendie spots. He alludes to a “trap” dike ten feet wide cutting the syenite in one place, and says that reddish feldspathic veins are common. ‘This is an evident reference to one of the lampro- phyric dikes and the felsitic ones. He also refers to a breccia which is found in one locality, the coarser syenites occurring as nodules in a rock resembling trap. The mass of diorite (camptonose) rock above the Gilford station on the lower. west slope of Locke’s Hill is not mentioned and was probably not seen by him. In Hawes’t report the rocks of this area are not mentioned, although he describes the syenite of Red Hill. GEOLOGY OF THE BELKNAP MOUNTAINS. The Belknap Mountaims are formed of a mass of granitic igneous rock, the result of the upthrust of a great body of molten magma into the rock masses surrounding it, the latter being broken and displaced to permit of its entry. In sequence to this major event there were later upthrusts of other maginas * Geology of New Hampshire, vol. ii, p. 607, 1877 + Lithology of New Hampshire, loc. cit., vol. iii. 346 Pirsson and Washington—Geology of New Hampshire. of different composition in small amounts which now appear as accompanying intrusive masses and dikes. Since then the superincumbent rocks have been removed by long-continued erosion, which has also bitten deeply into the igneous mass as well, but this has resisted better than those which surround it, and in consequence the igneous stock now projects as a rough mountain tract. Lastly, much material was removed at the time of the glacial invasion and the rock surfaces left scored and polished. : Lhe enclosing rocks.—Vhese are mostly gneisses and mica schists, rocks of metamorphic character. Although they do not especially concern us in this paper, a word or two may be added regarding them. On the eastern side the contact is with . a heavy solid gneiss, composed of quartz, alkali feldspars and biotite, and often carrying red garnets. In its texture it is rather irregular, not presenting that evenness of aspect fre- quently shown by gneissoid granites, and it is possible that detailed study in the future may show that it is of sedimentary origin. It has a wide extension in this general region and has been called the Winnepesaukee eneiss by the Hitchcock Survey. In Mount Major and Pine Mountain’ are two small masses of a porphyritic granite as shown on the map of the Hiteh- cock survey. - Intheir report it is spoken of as the porphyritic gneiss. It covers a large area to the north of this region, where we have seen and studied it to some extent. By ‘its general characters, contact modifications, etc., it is.clearly an igneous rock—a granite which carries large, often huge, phenocrysts of orthoclase. It occurs in other parts of New England and is a type worthy of especial study. It sometimes has a pro- nounced gneissoid structure which evidently is often a fluxion texture, at other times it is due to dynamic shearing and in some places it is quite devoid of any gneissic char acter. On the west and south the Belknap massif is in contact with micaceous gneisses, micaceous slates colored dark with organie matter and iron ore and with mica schist rocks evidently of sedimentary origin. The boundaries and names of these forma- tions are those given on the Hitchcock map. The lack of ro) printed symbols on this map connecting the legend with the outlined areas and the great similarity of colors makes it exceedingly difficult, in many cases impossible, to determine what some of these areas are meant to be, nor does the text afford much help in this direction. The formations are men- tioned in many places, but there is no definite description of them given in a systematic manner by which their characters may be recognized. From what is stated,* however, we con- clude that fhe rocks on the west belong to he Montalban series * Op. cit., vol. ii, p. 600. Pirsson and Washington— Geology of New Hampshire. 347 of Hitchcock, and they are so designated on the map. Where we have seen them they are mostly gray micaceous gneisses and mica schists. Geology of the main mass.—The greater part of the moun- tain group is made up of a coarse- orained hornblende syenite, a hornblende-grano-pulaskose in the new classification, whose characters will be given in a succeeding petr ographical paper. It is this rock which composes the mass of Mt. Gunstock, of Mt. Belknap the peak next north of it and of the northern extension’in Locke’s Hill. It occurs also in the ledges exposed on the higher part of the road from Gilford to West Alton, where it crosses over the mountain. It also forms the higher parts of Piper Mountain south of Gunstock Peak, and it runs over towards Mt. Straightback. In Piper Mountain it assumes a somewhat porplhyritic character. It is seen on the sides and crests of the main elevations in massive outcrops and exposures often several hundred feet across and is thus thoroughly laid bare. These surfaces show everywhere the planing and smoothing of glaciation. In none of them did we find the rock perfectly firm and unchanged. Everywhere its color ranges from a reddish to brownish, it tends to crumble under the hammer and in places it is loose and crumbling into coarse gravel. The chemical analysis shows however that this is not due to any chemical alteration of the constituent minerals, but to mechanical disintegration from the action of frost and weathering, which have tended to loosen the texture of the rock. Blasting would probably reveal excellent material at a few feet below the surface. We did not find any quarry open- ings in this rock-materia]l; it is m general too high above the zone of cultivation to have made such work necessary. In only one place did we find this type at the contact zone against the older rocks, on the southwest slope of Locke’s Hill in a little ravine where it is in contact with mica schist. It is here rather coarse, altered and not of typical composition. Contact facies of fine-grained granite.— With the exception just mentioned, in all localities examined by us, we have found that at the contact with the enclosing rocks, not the syenite but a fine-grained granite (grano-liparose) is present. The lower slopes where the actual contact lies are in general so covered with glacial drift and soil, often with a more or less dense growth of vegetation, that it is masked and rarely seen, but immediately above where it should be this granite appears and beyond and above it the syenite. This we found to be the case in a number of places on different sides of the mass, so for example at the west foot of Mt. Gunstock, the west and south slopes of Piper Mountain, the northeast foot of Locke’s Hill, at West Alton and on the southern prolongation of the 348 Pirsson and Washington— Geology of New Hampshire. mass northeast of Hills Pond. Hitchcock’s description also gives clear indications of the same thing in other localities not visited by us. One of the best localities for the study of the contact that was seen by us is at the foot of the west slopes running down from the north end of Piper Mountain in the pasture fields south of Morrill’s farm, where the path, by which Mt. Gunstock is generally ascended, begins. The mica schists and other rocks, which we infer make the formation shown by Hitchcock on his map as the Montalban, are full of pegmatite and fine granite stringers and dikelets and ‘appear to be enriched in feldspar. As the contact is approached they change to a fine dark-gray gneiss which is cut by fine granite dikes. Higher up appears the syenite itself. The attitude . and characters of the gneisses are such that they indicate quite clearly that they le, thinning out toward the mountain, upon a rising slope of the igneous rock below and that the contact plane is therefore here not vertical but dipping away from the mountain. The syenite from the slopes above is that of the main type but finer-grained. At the south end of Piper Moun- tain the bordering granite has a faint but distinct gneissoid appearance. It is also to be noted that these bordering masses of granite are generally filled with spots and streaks of varia- ble size of darkerinclusions, which are no doubt fragments of the country rock thoroughly altered by immersion in the magma. It appears to us that the best explanation for these fringing granite masses is to consider them a differentiated border facies, an endomorphie contact -modification of the mam type. They may not exist everywhere, but they have been so gener- ally found on different sides, as seen by ourselves and indicated by Hitchcock, that the phenomenon seems difficult of explana- tion on any other basis. It is true that we have not been able on continuous exposures to trace the gradual merging of the granite into the syenite, because this should be done on the lower slopes, and for reasons given above these do not afford proper exposures for this purpose. We cannot affirm then positively that these are not a series of later eruptions which have broken out around the border of the previously intruded syenite, but in view of their disposition such an explanation seems unnatural, and especially so since they do not exhibit cer- tain phenomena shown by an undoubted later intrusion of granitic magma on the western slope of Locke’s Hill, which will be presently described. From the facts at our command, therefore, we are inclined to think the first explanation the | more reasonable one and to regard the granite as a differen- tiated border mantle of the syenite. We also do not regard the granite border as having been produced by the melting up ee $8 Am. Jour. Sci., Vol. XX, 1905. Plate XI. Geologic and Topographic Map of the Belknap Mountains, N. H. = == TIMBER | == : —=3 GOVERNORS |s. LOCKE'S 1S. i, Le ox SL77 £ =| SLEEPER Ss. 4 \\ V alii xwa NO QS at Th Z7 LOOK, ELA 7m LE, 5775) a, — 4, way ole aretiitl Y . y Vs o, e, A % ‘ Gd 6 6 ; ; he 4 ve re 7, A “k / INO ORO is Eeare* ee 3-101 u he er As UI RE aaa ero z qo Ba POD ti Ops woe 2S fe VI a Sool anges: ay “Set )adifer vf S ‘f ‘se fy pe). ® WIA o/s WAX CER 2 Pe NTI NY age SEERA ne S So rol os Se) 0 SI «\- a ‘ NS “ xine yD Oe an : Leh. \ Age) =e aa NAAN S. Qa . Ory ay A ar A ks Fen) = uN PW ° oA Jue . SAYA ~ mo \/ CW QQ (HH Syenite. Granite. Diorite Porphyritic Winnepe- Montalban Rockingham and breccia. Gneiss. saukee Series ? mica schist. Gneiss. Cambrian ? Seale 1 inch = 2 miles. Contour interval = 100 feet. Pirsson and Washington— Geology of New Hampshire. 349 and absorption of the country rock with which the mass of syenite magma came in contact for two sutticient and convine- ing reasons. First, because as already shown, the surrounding rocks differ widely in character and in chemical composition in different places, while the granite border maintains everywhere essentially the same characters, and second, because in many places inclusions of the country ‘rock are to be seen in it which, without regard to size, preserve all the sharpness and angulari ity of their original fragmentary form, thus showing that, although they have been much metamorphosed, melting of them did not occur. On the highest peaks and ridges and in the deepest erosive cuts into the mass it has been worn away and the main type of syenite appears. Its thickness was quite variable, and in a few places it did not appear at all. The line between the syenite and granite as shown on the geological map is therefore to be taken as a generalized expression of the existence of the two types and not as a definite geological boundary line, since for reasons just given this could not be definitely determined. Gilford diorite aree.—On our first visit to the region we found quite abundantly distributed in the form of bowlders in the fields and stone walls of the fences along the higher part of the land from Gilford over to West Alton, a most peculiar dark rock composed of large dark-brown hornblendes poikiliti- eally enclosing ophitic feldspars. In field usage it is here ealled a diorite for purposes of geologic description ; petro- graphically it is, as will be shown later, a grano-hornblende- camptonose, or in the older systems an essexite. On our second visit an especial search was made to locate if possible the occurrence of this type, and it was found to constitute a considerable mass on the lower west slope of Locke’s Hill and not far from the Gilford station on the railway. Its area is small, probably not over half a mile in length north and south along the slope and less than that in breadth. On the north it rises in heavy ledges above a little spring drainage and on the west its lower slopes are covered with soil and debris, but above this it forms a rather well-defined bench on the lower mountain side and in rather prominent outcrops it is seen everywhere over the pasture fields which lie upon it. On the south it descends into a little ravine, a locality mentioned above in connection with the syenite, and is here in contact with the mica schists and gneisses. Its upper edge is in con- tact with the syenite, but the actual contact was everywhere covered so far as we could discover. We have traced it, how- ever, to within a few yards distance, and it is then observed that the rock diminishes very str ikingly in the size of its grain, especially so with regard to the large poikilitic hornblendes, ro) and for this reason and others to be mentioned later we Weliave 350 Pirsson and Washington—Geology of New Hampshire. it to be a later intrusion than the syenite and that it has broken up alongside of it. The upper contact with the syenite is, however, largely replaced by a remarkable breccia zone to be presently described. This rock varies considerably in charac- ters from place to place, as will be described in a succeeding petrographic paper. Breccia zone.—As just mentioned, the contact between the diorite and syenite above is occupied by a brecciated rock mass. The cement is a quartz-alkali feldspar rock much like the granite facies previously described; it has a sugar granular texture, and is of the character of rocks designated as aplites. In this are thickly scattered blocks of all sizes, which may attain an extreme dimension of four feet in length but which average perhaps a foot or two in diameter and descend from this size to minute fragments of a fraction of an inch. In some places they are so thickly crowded that their mass is much greater than that of the cement. In shape they are usually extremely angular and the sharpness of the angles has been perfectly preserved. In other cases they appear some- what rounded as if partially melted, and are surrounded by darker aureoles richer in ferromagnesian minerals. It is remarkable, on the other hand, how some of the smallest frag- ments retain in some cases all their distinctness of outline. There are several different types of rocks among these included fragments. One common one is a dense black basaltic-looking type too compact for the component minerals to be seen, in which lie phenocrysts of mica aud other minerals—a rock of well defined lamprophyrice character. Other fragments are of the diorite mentioned above, while others are obviously pieces of gneisses and schists. The determination of the relative age of this breccia and of the syenite and diorite is not easy. It would be simple to imagine that the latter is the older rock, that the syenite with its granite border broke up alongside of ‘it enclosing masses of femie rock. If this view is adopted, then the basaltic, lampro- phyric and granitic and felsitie dikes which cut the s syenite must of course be separate and later intrusions and there would be four periods of eruption, in two of which, those of the diorite and the lamprophyric dikes, similar magmas were pro- duced. The oldest rock, the diorite, is then also the most dif- ferentiated one, a fact contrary to general experience. Con- sidering these points, we are inclined to believe the syenite the first and oldest, to place the eruption of the diorite next, which would also explain the distinet endomorphic contact modifiea- tion it exhibits toward the syenite and make it contemporane- ous with the lamprophyrie dikes. Then came an eruption of granitic magmas, which also forms dikes in the syenite, one of _Pirsson and Washington— Geology of New Hampshire. 351 which broke up at the border of the diorite, involving masses of it in its various modifications and thus produced the breccia. If this view is adopted, there are but three periods of eruption and they follow the normal course commonly seen in such cases. Dikes.—l\t has been observed by us that wherever the main | types of ignevus rocks are exposed over considerable areas in this mountain tract they are commonly cut by dikes, and the same is true of the border zone of the enclosing schists and gneisses. Except, however, in the highest parts of the moun- tains, such exposures are not very common nor are they of great size. It seems probable, therefore, that only a very small part of the dikes actually pr esent in the region has been seen by us, the greater part being covered up by the heavy mantle of debris and glacial drift. _ «As is so often the case ein the dikes are found to be a throng of satellites attendant upon a large body of igneous rock, they may be readily referred to two strongly contrasted groups. In one of these the rocks are light colored, strongly persalic and therefore almost devoid of ferromagnesian min- erals; in the second the rocks are dark colored, salfemic, heavy and composed in very large, if not for the gr eater part, of ferromagnesian minerals. They are persalanes and salte- manes in the new classification or aplites and camptonites in the older ones. The persalane dikes are found cutting the main syenite in al] directions, of a generally pink color and varying in size from dikelets but a few inches in breadth to masses twenty feet wide. The bare exposed slopes and ledges of the upper part of Mt. Belknap were found cut by them in great abundance and it was here noticed that they often ended abruptly and appeared as if somewhat elongated roughly lenticular masses. They were often branched, were connected with others, anasta- mosed or formed reticulated systems, large and small together. Their small angular chippy jointing, light color on the weath- ered surface and flinty felsitic aspect clearly distinguished them from the massive granular rock they cut. These same characters were found repeated on the exposed surfaces of Mt. Gunstock and Mt. Piper, and in one place, about half way up Mt. Gunstock from Morris house, on the west slope above the spring, the ledges in a pasture field on an open shoulder of the mountain were found cut by a dike of this nature 15-20 feet in width and with north and south trend. It was also found that where the contact -zone was exposed at the foot of the mountain slopes, as along the west side in the localities described above, that both the igneous rock and the enclosing schists and gneisses were cut by dikes and stringers 352 Pirsson and Washington—Geology of New Hampshire. of persalic rock. While in the crest of the ridges and in the peaks these dikes vary in texture from dense felsitic to sugar eranular granites, in the contact zone we observed only the latter, and they sometimes pass into varieties with pegmatoid texture. With only a few exceptions all of these occurrences are on too small a seale to be shown on the map. The salfemane dikes were not nearly so numerous, but on account of the cen- trast made by their dark color appear more distinctly defined. They were also observed cutting the exposed rock surfaces on the tops of the mountains; there are several below the summit of Mt. Belknap on the southwest side and one six feet wide with porphyritic feldspars cuts the very highest point of the | peak with east and west trend. Three or four of about the same size were found on the top of Mt. Gunstock and they were likewise observed on the crest of Mt. Piper. The lower slopes of the mountains are probably cut by them also, but the masses of debris and vegetation which cover them hide the exposures in which they might be seen. They occur also in the surrounding rock masses in which the intrusion took place. Here again the exposures are difficult to find, but one place, Sander’s Neck, a small promontory on the shore of the lake north of the mountains, presents considerable areas of the glaciated gneisses, and these we found traversed by several intersecting dikes of these salfemic rocks. As usual they were but a few feet in width. They occur in the mica schists which are exposed at the foot of the lower west slopes of Mt. Gunstock and Mt. Piper, and from what we have observed around the similar intrusive mass of Red Hill north of the lake, it seems probable that a more detailed study of the surrounding region would show a considerable number of them extending to relatively long distances from the central parent mass. Some of those mentioned above are shown on the accompanying map. New Haven, Conn., and Locust, N. J., May, 1909. f\ P. EF. Raymond—Fauna of the Chazy Limestone. 358 Arr. XXX VIII.— The Fauna of the Chazy Limestone ;* by Percy E. Raymonp. INTRODUCTION. Ty several papers on the Chazy limestone, Brainerd and Seely have given sections showing the lithological characters and thickness of the tocks at various localities from Chazy, New York, south to Orwell, Vermont.t These authors have divided the formation into three parts, A, B, and ©, of which A is the base and © the top. These divisions are founded partly on lithologie and partly on paleontologic grounds. Only a few species of fossils, however, were listed; hence it has been the object of the present writer to ascertain which are the common species in the Chazy, and to learn their strati- eraphic and geographic distribution. For this purpose, detailed sections have been made at Crown Point, Valcour Island, and Chazy, and extensive collections have been obtained at other places in the Champlain and Ottawa valleys. The sections will be fully described in the Annals of the Carnegie Museum. In this place, however, only a synopsis of each is given. DisTRIBUTION. The Chazy formation was named by Ebenezer Emmonst from the outcrops studied by him at Chazy village, New York, this locality, therefore, becoming the typical one for the formation. In stratigraphic position, the Chazy overlies the Beekman- town. (Calciferous) and underlies the Lowville (Birdseye) member of the Mohawkian. It may be traced from Orwell, Vermont (along the Champlain Valley), to Joliette, north of Montreal, Canada. in the Ottawa Valley, it extends from Hawksbury west to Allumette Island, 80 miles northwest of Ottawa. The formation is seen again at the Mingan Islands in the St. Lawrence, where it covers a small area. In the Lake Champlain region, these strata are mostly lime- stone, and the thickness ranges from 60 feet at Orwell to 890 * Abstract of part of a thesis presented to the Faculty of the Yale Uni- versity Graduate School for the degree of Doctor of Philosophy. The detailed paper, with full discussion and illustration of species, will be pub- lished in early numbers of the Annals of the Carnegie Museum. For descrip- tion of the trilobites here mentioned, see Annals of the Carnegie Museum, vol. iii, p. 328, and this Journal, vol. xix, p. 377. Other new forms noted in the text are described at the end of the present paper. + Amer. Geol., vol. ii, p. 323, 1888; Bull. Geol. Soc. Amer., vol. ii, p. 300, 1891; Bull. Amer. Mus. Nat. Hist., vol. viii, p. 305, 1896. t ‘Geolog gy of New York, Pt. 2, Report of the Second District, 1842, p. 107. 354 P. EB. Raymond—Ffauna of the Chazy Limestone. soymutgy “YUE 9-€ $U07]99S uaINjeq p20 GL - yu aes [297)22f SNOLLOSZS GHL JO SNOILV14U SHL MOHS OL WVuDVIC T1IMYO ANI0d NMOYD anv1S! 9 5 UNODTIWA a AZWH9 § X193S ANY QYaNivua YSWIAV GNVISI YNODTWA G F P. EF. Raymond—Ffauna of the Chazy Limestone. 355 feet at Valcour Island. Further north the thickness is not definitely known. In the Ottawa Valley, the formation is usually from 100 to 200 feet thick and is about half limestone and half sandstone, the former u-ually overlying the latter. At the Mingan Islands, the thickness was estimated by Sir William Logan at about 300 feet, and the strata include both shales and limestone. Lake CHAMPLAIN REGION. As the typical Chazy is exposed in the Lake Champlain region, that area will be first taken up. In general, the Chazy rocks are seen as a narrow belt running almost north and south from Orwell, Vermont, to Joliette, Canada. The area is seldom more than 10 miles wide, and is not a continuous exposure, but occurs in small patches, in most cases evidently fault blocks, and the strata are usually inclined at a consider- able angle. The principal outcrops are along the west side of Lake Champlain and on the islands in the northern part of the lake. South of Willsboro Point, there are scattered patches on both sides of the lake nearly to Fort Ticonderoga. Faunal Divisions. In the Lake Champlain region, three major faunal divisions of the Chazy may be distinguished. Within these, there are again various zones which are more or less local in geographi- cal extent. Division 1. The Hebertella exfoliata Division.—The strata of this basal division are chiefly light-colored, impure, rather coarse-grained limestones and frequently have shaly partings. The thickness varies from nothing at the south end of Lake Champlain to 300 feet on Valcour Island, 8365+ at Chazy, and 225 feet on Isle La Motte. The characteristic fossils are: Hebertella exfoliata sp. nov., Orthis acutiplicata sp. nov., Strophomena prisca sp. nov., Scenella pretensa sp. nov., S. montrealensis, Paleacmea irregu- laris sp. nov., Leaphistoma immatura, and Scalites angulatus. Other species occurring abundantly in this zone are: Llastov- docrinus carchariedens, Bolboporites americanus, Zygospira acutirostris, Laphistoma stamineum, Lophospira subabbre- viata, Bucania sulcatina, and Pseudospherexochus chazyensis. Those which occur only rarely in this division, but which thus far have not been found in higher divisions, are: Lingula belli, Cyrtodonta solitaria sp. nov., Cyclonema ? normaliana sp. nov., Hunema leptonotum sp. nov., and Heliomera sol. Of the 141 species in the Chazy whose range is known, 64 make their first appearance in this horizon and 23 are found in 356 P. Eb. Raymond—Fauna of the Chazy Limestone. all three divisions. This member is further marked by the appearance of the earliest of American Bryozoa, and these, unlike most Ordovician species, range throughout the entire formation above the sandstone. Division 1 is characterized by the predominance of individ- uals and species of Brachiopoda. Fourteen of the 25 species of this group occurring in the Chazy of the Champlain Valley are found in this lowest member, while only 2 of the 16 pelec- ypods are represented. Exactly half the species of trilobites are also found here, but specimens are not common. Gastro- pods are more numerous, as half the species are represented and individuals of some forms are abundant. They do not occur in the lower strata, but are confined almost entirely to the upper part. : There are three zones in this division which are worthy of MONS == : Zone 1,, or the Orthis acutiplicata zone, is near the base of the division and is found at Valcour Island and Isle La Motte. ~ The characteristic fossils are: Orthis acutiplicata, Rafines- — guina incrassata, Lsotelus harrisi, and Thaleops ovata, all long rangers except the first. Zone 1,. The Scalites angulatus zone. The faunule of this zone is found at Plattsburg and Chazy. It is located near the middle of Division 1. The characteristic fossils are : Scalites angulatus, Raphistoma vmmaturum, fe. stamineum, Bucania sulcatina, Camarella longirostris, [llenus globosus, and Thaleops ovata. Only the first two are restricted to this horizon. Zone 1,, the Lophospira subabbreviata zone, has been found only at Chazy, but is very strongly marked. It occurs about 75 feet below the top of Division 1. The characteristic fossils are: Lophospira subabbreviata and Laphistoma staminewm, both of which are very abundant. Of less importance are the rare Schizambon? duplicemuriatus, Heliomera sol, and Clionychia marginalis sp. nov. Division 2. The Macturites magna Diwision.—The strata of this middle division are usually heavy bedded, dark blue and grey, fairly pure limestones, with an occasional layer of grey sparkling dolomite or of light coarse-grained limestone. The layers near the middle usually weather into nodular masses, and the fossils are frequently poorly preserved and difficult to extract. The thickness varies from 200 feet at Chazy to 400 © at Valeour Island, and decreases toward the south. The char- acteristic fossils are: Maclurites magna, Lafinesquina cham- plainensis, Plesiomys platys, P. strophomenoides sp. nov., Stre- phochetus, Lospongia varians, ELotomaria obsoletum sp. nov., Eccyliopterus fredericus, Bathyurellus minor, Glaphurus primus, and Leperditia limatula sp. nov. P. EF. Raymond—Fauna of the Chazy Limestone. 357 > Thus far, the following fossils have been found only in this division, and most of them in but one locality : Camarotaechia pristina sp. nov., Ctenodonta dubiaformis sp. nov., Clidoph- orus obscurus sp. nov., Cyrtodonta expansa sp. nov., Lndo- desma unduletum sp. nov., Scenelia robusta sp. nov... Raphis- toma undulatum sp. nov., Helicotoma vagrans sp. nov., Bucania bidorsata?, Trochonema dispar sp. nov., Subulites prolongata sp. nov., Holopea scrutator sp. nov., Hoharpes ottawdensis, Asaphus marginals, fsotelus angusticauda, He elus ? bearsi, Llienus punctatus, Cybele valcourensis, Cera TUS pompilius, CU. hudsont, and Pseudospherexochus ae mus. This middle division is marked by an abundance of pelec- ypods, gastropods, and trilobites, and in this respect is sharply contrasted with the lower division. Of the 16 pelecypods, 13 are represented here. Of 35 trilobites, 27 are present. Species of Stromatocerium and Strephochetus are common in these rocks, but are also abundant in the lower zone of the next division. | Zone 2,. The MMalocystites murchisont zone. Thus far, only one subfaunule has been detected in Division 2, and that is at the very base. It is best developed at Valcour, but occurs alsoon Valeour Island. The zone is characterized by the great abundance of cystid fragments. The characteristic fossils are : Glaphurus primus, Eohar pes antiquatus, Lonchodomas halla, Cybele valcourensis, Malocystites murchisoni, M. emmonsi, Glyptocystites forbest, Palwocystites tenuiradiatus, Raphis- toma stamineum, Maclurites magna, Plesiomys strophome- noides, and Camarella varians. Division 3. The Camarotuchia plena Division.—The strata of this division are rather thin bedded, light grey, coarse- grained limestones, abounding in fossils. Near the base there are always buft-colored, pure, fine-grained dolomites and heavy bedded, coarse-grained blue limestones. The only fossil which is found throughout this division is Camarotechia plena. Other characteristic fossils are: Camarotechia major sp. nov., Orthis ignicula sp. nov., Modiolopsis fubaformis sp. nov., and Glaphurus pustulatus. There is here a decided falling off in the number of gastro- pods and pelecypods, only 6 of the former and 5 of the latter being represented. There are about as many trilobites (16) in this division as in Division 1, and 8 of these are found in all three sections. The number of species of brachiopods is about the same in each of the three divisions, but they dominate the fauna in the first and third. In the former, one of the Pro- tremata (Hebertella) is most abundant, while i in the third divi- sion one of the Telotremata ( Camarotichia) predominates. Am. Jour. Sct.—FourtH SERies, VOL XX No. 119—NOVEMBER, 1905. 25 358 PP. &. Raymond—Fauna of the Chazy Limestone. There are three well-marked zones in this division, as fol- lows Zone 8,4, the Glaphurus pustulatus zone, is found at the base of Division 3, at Valeour Island, Chazy, Cooperville, and Isle La Motte. The characteristic fossils are: Glaphurus pustulatus, [llenus globosus, L. erastusi, Isotelus harrisi, Remopleurides canadensis, Pliomerops canadensis, Amphili- chas minganensis, Pseudospherexochus vulcanus, Camarote- chia plena, Conocardium beecheri sp. nov., Lucania sulcatina, and several cephalopods. Zone 3,, the Camarotechia major zone, stands between 3, and 3, and its faunule is a transition between the two. Camarotwchia becomes more abundant and better developed, and fossils, while numerous, become fewer in species. The best development is at Valcour Island. The characteristic fossils are: Camarotechia plena, C. major, Hebertella costalis, Matlocystites emmonsi, Malocystites sp., Palwocystites sp., Lllenus globosus, Pliomerops canadensis, Bucania sulcatina, Raphistoma stamineum, and Lsotelus obtusum. Zone 3,. The Modiolopsis fabaformis zone. In this zone, Camarotechia plena is abundant, almost to the exclusion of other species. The faunule extends to the top of the formation at Chazy, Grand Isle, and Valeour Island. The characteristic fossils are: Camarotechia plena and Modiolopsis fabaformis. Section at Chazy, New York. The section at Chazy has a thickness of 732 feet, but the base of the formation is not shown. Division 1.—The rocks carrying the fauna of Division 1 are well exposed in the ridges south of the village, near Tracy Brook. The thickness is 365 feet, and judging from the fauna at the base, at least 150 feet of strata are missing. SFHebertella exfoliata is very abundant, especially below the horizon of Scalites angulatus. The latter zone is 217 feet above the base of the exposed section, and is zone 1, of the generalized sec- tion. The most common fossils are: Scalites angulatus, Buca- nia sulcatina, Raphistoma immaturum, RR. stamineum, and Thaleops ovata. Higher up in the section, 275 feet above the base, is the zone of Lophospira subabbreviata, about 35 feet in thickness. This is zone 1, of the generalized section. The gastropods are very abundant in the three localities at Chazy where this zone is exposed. Division 2.—The strata of this division are about 195 feet in thickness, and are dark blue, impure nodular limestones, usually full of fossils which are frequently silicified. Stroma- ~ P. EF. Raymond— Fauna of the Chazy Limestone. 359 tocerium, Hospongia varians, Leafinesquina champlainensis, Plesiomys platys, Maclurites magna, Pliomerops canadensis, and several cephalopods are common. Division 3.—The Camarotechia plena division is not very well developed along the line of-the section at Chazy. The thickness is 156 feet, but.a large part of the strata is covered with soil. At the base are about 25 feet of grey dolomite with almost no fossils. The remainder of the rock, as far as exposed, is an impure shaly limestone, abounding in Camarotachia plena. Zones 3,4, 3,, and 8, can not be distinguished just at Chazy village, probably because the strata are so poorly ex- posed. About 3 miles southeast of this point, however, in a field near the lake shore, fine outcrops of zone 3, occur, and -here Glaphurus pustulatus, Amphilichas minganensis, Llle- nus globosus, and the cephalopods are common. Section at Valeour Island. On Valeour Island, the whole of the Chazy is exposed, with a thickness of 890 feet. In one section along the south end, almost the entire thickness is shown, while nearly all the miss- ing parts may be seen in other sections on the east and north sides of the island. Division 1.—The strata of this division are well exposed on the south end. The thickness is 314 feet. At the base is a zone of sandstone and shale in which Lingula brainerdz is the common fossil. Other fossils are rare, /sotelus harris: and a species of Hecyliopterus being the only ones thus far found. Above this zone is that of Orthis acutiplicata, 10 feet in thickness. The Scalites angulatus zone is not exposed on Valcour Island, the rocks usually containing it being absent at the peb- ble beach on the south end of the island. The Lophospira subabbreviata zone is not well developed, but may be indicated by a fauna found on the middle of the west side. Division 2.—The strata of this division are 406 feet in thickness and are usually compact, dark blue and grey lime- stones. The fossils are frequently coarsely silicified, but are almost always difficult to extract. At the base, zone 2,; the Malocystites murchisoni zone, is well developed, and as the fossils weather out in this locality, some 40 species have been listed. While the rocks of this division usually afford poor collec- tions, yet in favorable localities they are found to be extremely rich in interesting species. Thus, one locality on the east side of the island has yielded 60 species of fossils, among them such rare trilobites as Asaphus marginalis, [sotelus ? bearsi, and Lemopleurides canadensis, and many species of pelecypods. 360 PRP. LL Raymond—FHauna of the Chazy Limestone. Division 3.—This is especially well developed on Valeour Island. Zone 3, is exposed in two or three localities on the east side. Zone 3, is best developed about Cystid Point, the southeast point of Valcour Island, and zone 3, is exposed both east and west of Black River Point on the north end. The division is 172 feet in thickness and carries Camarotechia plena throughout. The faunules given for zones 3,, 3,, and 3,., are those found on Valeour Island. Crown Point Section.* The section at Crown Point is 305 feet in thickness. At the base is a zone 25 feet thick in which the strata are sandstone and shale, and the only fossil is Lengula brainerdi. The remaining 280 feet are impure blue and grey limestone, usually very fossiliterous. Division 1 is absent. Division 2.—The fauna characteristic of this division is found all through the section at Crown Point. The character- istic fossils—Maclurites magna, lajinesquina champlainensis, Plesiomys platys, and Leperditia limatula—are very abun- dant, and the whole expression of the fauna is that of the mid- dle part of the section at Valecour Island and elsewhere. Brainerd and Seely assign the lower 48 feet to their Division A, and the upper 57 feet to Division ©, but faunally the whole section belongs together. Cumarotechia plena is absent, as are also the other fossils characteristic of Division 3. The upper 8 feet of the section are a coarse limy sandstone, with Plesiomys platys, Camarelia varians, Raphistoma stamin- eum, and Lsotelus harrisi in a layer a foot thick at the top. Orwell, Vermont. A short distance northeast of Orweil village is the most southern exposure of the Chazy. At that place there are about 60 feet of strata, the fauna of which indicates that they belong to Division 2. Another locality near by shows sand- stone and shale at the base of the formation. North of the International Boundary the various divisions can not be followed in the published lists, but this is due to the fact that no sections have been made in that region. ‘The lists published by Billings, Logan, and Aimi, of the Canadian Sur- vey, however, do show that fossils characteristic of all three divisions are found in that region. The Champlain Valley fauna of the Chazy, which will be uestajumsed as the typical one, is found as far north as Joliette, 35 miles north of Montreal and * For detailed description of this section, see Bull. Amer. Pal., vol. iii, No. 14, 1902. P. EL. Raymond—Fauna of the Chazy Limestone. 361 85 miles north of Chazy. To the west it is found as far as Hawkesbury, 75 miles northwest of Chazy and 55 miles west ot Montreal. Mingan Istanps Recon. The fauna of the Chazy at Mingan Islands is very closely related to that of the typical Chazy of the Champlain Valley, as is shown by the following list of species common to the two regions :— Bolboporites americanus. Orthoceras bilineatum. Phylloporina incepta. O. multicameratum. Columnaria ? ? parva. Pleisoceras jason. Rafinesquina incrassata. Pliomerops canadensis. Camarotcechia orientalis. Lllenus globosus. Camarella longirostris. Loharpes antiquatus. O.. varians. Ottawa VALLEY REGION. The Chazy deposits of this region have been described in detail by Logan,* Ells,t and Ami.{t The formation is not more than 200 feet in thickness, usually less, and is divided into two parts, the lower including shales and sandstones, and the upper, limestones. It outcrops in a narrow belt extending along the north and south sides of the Ottawa River, from Hawkesbury west to Arnprior, and is again exposed south of Ottawa, whence another narrow belt runs to Cornwall, where - it again turns northward. West of Arnprior there are a few outliers ot the same formation. One large one occurs at Allu- mette Island, north of Pembroke, and another 10 or 15 miles south of this and west of Renfrew. The coarse character of the sediments at the base of the for- mation in this region poits to very shallow water and shore conditions and a probable erosion interval between the.end of Beekmantown time and the deposition of the strata of Chazy age. The writer has studied the rocks of this area chiefly in the vicinity of Ottawa and Aylmer, and the fauna there represented seems to consist of about 25 species, only 7 of which occur in typical Chazy deposits. The fauna of the sandstone at the Aylmer region is quite different from that found in the over- lying limestone, and for that reason a list is here given of the species found in each. An asterisk denotes that the species is found also in the typical Chazy :— * Geology of Canada, 1863. + Rept. Geol. Survey of Canada, 1899. ¢ Ibidem ; also Trans. Roy. Soc. Canada, vol. ii, 1896, vol. vi, 1900, and various other papers. 362 P. b. Raymond—Lfauna of the Chazy Limestone. Sandstone. Limestone. Lingula lyelli. Lingula lyelli. * Camarotechia plena. * Camarotvchia plena. * CO. orientalis. * Rafinesquina alternata. Flebertella imperator. Modiolopsis breviuscula. Modiolopsis breviuscula. MM? parviuscula, M, parviuscula. M., sowteri sp. nov. Ctenodonta parvidens sp. nov. Whitelia canadensis sp. nov. * Archinaceila ? deformata. *Raphistoma striatum. *R. stamineum. Raphistoma stamineum. Lophospira billingsi sp. nov. Orthoceras allumettensis. Bathyurus angelint. Bathyurus angelini. Beyrichia clavigera. Leperditia amygdalina. B. clavigera clavifracta. *L. canadensis. Leperditelia labellosa. Primitia sp. LIsochilina ottawa. Lsochilina sp. LI, amiana. Primitia logani. It may be seen from the above parallel lists that there are only 6 species common to the sandstone and limestone divisions of this formation. In the limestones the ostracods are exceed- ingly abundant, often making up entire layers of the rock. The two divisions are intimately connected by very well- defined species, however, and none of the forms pass on into the overlying Lowville limestone. In the Ottawa Valley, the most noticeable feature of the faunas is the absence of the cystids, Bryozoa, and Hydrozoa so common in the typical Chazy. The large number of species of ostracods and their great abundance are in marked contrast to the three or four species found in the Champlain Valley. This difference in the lithology and fauna has led the writer to suggest the name Aylmer* formation for these deposits in the Ottawa Valley. SUMMARY ON THE Lake CuHaAampiLaIn, Mincan IsLanps AND OrrawA VALLEY REGIONS. In the Lake Champlain region occurs the fullest develop- ment of both the strata and the fauna of the Chazy period, and three divisions based upon faunal differences may be recog- nized. The fauna of the Chazy at Mingan Islands, while only partly known, shows that the typical Chazy is also found in that rezion. West of Hawkesbury, Canada, a decided change * Ann, Carnegie Mus., vol. iii, p. 380, 1905. P. EF. Raymond—Ffauna of the Chazy Limestone. 363 in the fauna is seen at L’Orignal, only 16 miles from Hawkes- bury. Here is found a section less than 200.feet in thickness, with sandstone at the base and limestone in the upper por tion. The fauna changes abruptly, several species occurring there which are unknown further east. The typical Chazy fossils found here are: Camarotechia plena, Raphistoma stamineum, and Malocystites murchisont. From this locality west to Allumette Island, a distance of 115 miles, the same succession of strata may be found, and about the same fauna. All through the Ottawa Valley the Chazy is represented by a formation which is sandstone at the base and limestone above. In its most western exposures, the limestones are absent and only the sandstone remains. The base of the Chazy is always a sandstone, but this does not carry the same fauna in all regions, nor does the zone Which rests upon it always have the same fauna. In the Lake Champlain region, the sandstone always contains Lingula brainerdi ; in the Ottawa Valley, it carries a modified Cama- rotechia plena fauna. At the type sections, Lingula brainerds is at the base of the formation, while the Cumarotuchia plena fauna appears 700 feet above. Since the fullest development of limestone deposits of this age is in the region of Chazy and Valcour Island, New York, that must be the locality in which the Chazy sea persisted longest. From the evidence outlined above, it would seem that this sea was a shallow one, invading south and west over a slowly sinking land. Since ‘the Chazy fauna is apparently developed less directly from the Beekmantown of the Lake Champlain area than “from that of Newfoundland, and since there are many European types introduced into the Chazy, it seems. probable that this sea was open to the east. If the sea were thus invading upon the land, the sandstone would represent shore conditions. This is undoubtedly the ease, for the sandstone in both the Champlain and Ottawa valleys frequently presents evidences of shore origin in cross bedding, ripple marks, and worm burrows. If the sea were invading southward in the region now occu- pied by the Champlain Valley, the sandstone should be younger and younger in age as it is traced from north to south. That this 1s actually the case is shown by the faunas, for at Valcour Island all the strata of the Hebertella caf oliata division, 300 feet in thickness, were deposited before the Maclurites magna fauna became prominent, while at Crown Point this second fauna follows immediately upon the basal sandstone. During the greater part of Chazy time, the transgression is southward, but later the shore began to move westward also. The region of the Ottawa Valley was then invaded, and the 364 P. L. Raymond—fauna of the Chazy Limestone. sandstone brought with it a part of the Camarotechia plena fauna. The date of this invasion to the west can be rather closely approximated. Camarotechia plena, Raphistoma staminewm, and Malocystites murchisoni are found in the middle of the section at L’Orignal. At Valcour Island these species occur together in zone A,,, 775 feet above the base, thus showing that the formation in the Ottawa Valley repre- sents the very latest part of Chazy time. Ulrich and Schuchert, in their paper on Paleozoic Seas and Barriers,* bring out this idea of a Chazy sea invading westward and southward. They state: “ With the earlier part of this subsidence [the Chazy invasion], the Atlantic invaded the continent westward. ... The typical Chazy formation =. bears evidence in its members of having encroached south- ward and westward in the arms, the latest beds . . . extend- ing farthest south and west.” THe Crosinc PERIOD oF CHaAazy TIME. In the preceding pages an effort has been made to show that m northeastern New York and in the Ottawa Valley, the Chazy sea invaded over a land surface of Beekmantown rocks, and that the base of the Chazy is a tangential sandstone; also that the invasion was first southward, covering the region of the Champlain Valley, and later westward along the locality of the present Ottawa Valley.t Of the former extent of the formation throughout the St. Lawrence Valley or elsewhere, there is at present little evi- dence. Since the sea did not attain the region of Aylmer until very late Chazy time, it is probable that the formation never extended much further west than the known outcrops in that region (Allumette Island, ete.). From a study of the stratigraphy and faunas it becomes evident that the upper portion of the Chazy is not represented in the region south of Valcour Island. Hither these beds were never deposited there or they were eroded before the Lowville was laid down. The evidence is not of such a character as to - prove definitely which did occur, but for reasons given below it seems more probable that the upper beds were deposited south of Valcour and later eroded. ‘These reasons are as follows :— * Rept. N. Y. State Pal., 1902, p. 639. [+ By these terms, Champlain Valley and Ottawa Valley, the writer does not intend to convey the impression that the Chazy deposits were laid down in narrow arms of the sea, or that the topography was then anything like that of the present time. It should be remembered that strata of post-Chazy age are involved in the Green Mountain uplift, and that there are indications that the Adirondacks did not exist in Ordovician time. | P. E. Raymond—Fauna of the Chazy Limestone. 365 First. All through the Champlain Valley, the Chazy is capped by a bed of sandstone 2 feet in thickness, and this may be interpreted as the invading base of the Lowville formation. From this it would follow that a period of erosion existed between the Chazy ana Lowville formations. Second. If the upper beds were never deposited south of Valeour, the Chazy sea after advancing slowly to the south to some point below Orwell, Vermont, must have then retreated to the northward. Such a recedence could have been caused only by an elevation south of Orwell, for there is no general retreat of the Chazy sea at this time, which is proved by the fact that at a still later period the sea advanced westward beyond Ottawa. . That there was then no uplift in the south is shown by the fact that the Lowville sea invades from the south- west.* On the other hypothesis, which seems more probable, the sea would have invaded southward to the region of Orwell and after depositing there the final, or Camarotechia plena, beds vanished from the area of Lake Champlain. During the latter part of Chazy time or after its close, the Stones River (Lowville) sea was invading from south to north and there was a land interval in the Champlain region, during which time some of the Chazy and Beekmantown beds were removed along the barrier region between Orwell and the Mohawk Valley. Third. By taking the rate of decrease in thickness (11°25 feet per mile) of the Hebertella exfoliata division between Chazy and Valcour Island, to compute the probable southern extent of that division, it is seen that it would have reached only 26°6 miles south from Valcour Island. Therefore, at the same rate of decrease the base of the Crown Point section is 461 feet higher than the base of the Valcour Island section. That this rate of decrease can not be used, is shown by the fact that Division 1 at Isle La Motte is only 225 feet thick, which is less than at Valcour Island, while Isle La Motte is as far north as is Chazy. The only reliable data for an estimate of this character are the facts that there are 300 feet of the beds of Division 1 at Valcour Island and nothing at Crown Point. This is a thinning out of 7:3 feet per mile, which, on the other hand, is probably too small. On this basis, the bottom of the Crown Point section is at least 300 feet above the base of the Valcour Island section and the base of the Orwell section is at least 424 feet above it. If this minimum estimate of the height of the base of the Crown Point section above that of the Valcour Island section is accepted as a work- ing basis, it will be seen that the former lacks the upper 285 feet of the formation. This is a gradient of 6°95 feet per mile * See Ulrich and Schuchert. 3866 P. £. Raymond—Ffauna of the Chazy Limestone. to the top of the beds at Valeour Island. Taking the base of the Orwell section at 424 feet, the upper 407 feet are lacking. The thinning in the 17 miles from Crown Point to Orwell is 122 feet, or 7-1 feet per mile, while the gradient to the top of - the Chazy at Valcour Island is 7-01 feet per mile. The close correspondence of these gradients and the small gradient of 7 feet per mile for 58 miles are significant, and seemingly indi- cate a base-leveled surface of this land durimg the Chazy- Lowville interval. REPRESENTATION oF CHAzZY TIME IN OTHER REGIONS. The Chazy was formerly identified by various geologists as covering a large area, but more recently it has been held that ~ while certain formations may have been laid down during Chazy time, the typical rocks and fossils of this period are restricted to the region of the Champlain and Ottawa valleys and the islands in the Gulf of St. Lawrence. The St. Peter's Sandstone. One of the formations which has long been correlated in time with the Chazy is the St. Peter’s sandstone, which in lowa, Minnesota, and parts of [llinois underlies the lowest member of the Mohawkian series. The fauna* of this forma- tion is meagre and is contained in a few layers near the top. It is made up chiefly of Mollusca, all closely allied to Trenton forms. None of the species are found in the Chazy; hence no new light is thrown on the correlation by the later studies of the Chazy fauna. On lithological grounds, James has cor- related it with the Chazy of the Ottawa Valley, but there are no species common to the two formations. From the close relationship of its fauna to that of the Mohawkian (Trenton) it seems probable that the St. Peter’s was deposited during Stones River time. Stones River Group. In the Columbia, Tennessee, folio of the U. 8. Geologie Atlas, Ulrich has stated that the lower part of the Stones River group, including the Lebanon, Ridley, Pierce, and Mur- freesboro limestones, is to be correlated in time with the Chazy of New York State. This statement is evidently based mainly on stratigraphic grounds, as Ulrich and Schuchert+ have held that the Low- * F. W. Sardeson, Bull. Minnesota Acad. Sci., vol. iv, No. 1, pt. 1, p. 64, 1896. + Paleozoic Seas and Barriers, Rept. N. Y. State Pal.; Bull. 52, N. Y. State Mus., 1902, p. 633. P. EF. Raymond—fauna of the Chazy Limestone. 367 ville of New York is the northeaster n representative of ‘the extreme top of the Stones River” group. In the Columbia folio referred to shee Ulrich has tabulated the fossils of all the divisions of the Stones River group as developed in the middle Tennessee region. In the Lebanon ’ formation, the upper member of the Stones River group which is there correlated with the Chazy, there are, according to the table, 37 species besides 10 undescribed Bryozoa. Of these 37 species, 7 are Bryozoa and 5 are not specifically identified. This large number of Bryozoa—17 species—at once suggests that the formation containing them is much more closely ‘allied to the Trenton than to the Chazy. Leaving out of account the Bryozoa, which in the Ordovician nearly always have a very restricted vertical range, and the 5 forms not specifically identified, it is found that 17 of the 25 species remaining are _ Black River or Trenton forms. All the brachiopods, 4 of the 5 gastropods, and 2 of the 3 trilobites are species occurring in higher formations. Even if all the described species are included, 53 per cent of the species of the Lebanon formation are Black River or Trenton forms. Below the Lebanon is the Ridley horizon, about 80 feet in thickness. Of the 9 species listed from this formation, 6 are found in the Black River. Below the Ridley is the Pierce limestone with 12 species listed and 20 undescribed bryozoans. Only 11 forms are specitically identified and of these 30 per cent are Black River or Trenton forms. The lowest member of the Stones River group is the Mur- freesboro, which is about 60 feet in thickness and contains 24. species, 21 of which are identified. The fauna is composed principally of Mollusca, of which gastropods of the genera Lophospira and Liospira are particularly numerous. Of the 21 species, 1i are Black River or Trenton forms, so that 52 per cent of the species in this oldest member of the Stones River group belong to the Black River or Trenton. This | analysis may be tabulated thus :— Rid- Murfrees- Black Tren- Lebanon. ley. Pierce. boro. River. ton. Bepanou.s soe. 2 25 Ae 1 ope 4 i idlsy en ol 1 cnngeoe a 1 1 | IVES ena aie cama aa a 4 2 9 il 1 2 Murfreesboro... _-_. 3 4 1 mA 4 7 Of the 58 described species occurring in these 4 subdivisions of the Stones River, the above table shows that, 27, that is, 46 per cent, occur in the Black River and Trenton formations. 868 P. EL. Raymond—Fauna of the Chazy Limestone. Couiparing the large percentage of forms common to the Stones River and to the Black River and Trenton with the low percentage—less than 5 per cent—of forms common to the Chazy and Trenton, it becomes evident that the Stones River and Trenton are faunally much more closely connected than are the Chazy and Trenton. This close relationship of the fauna of the Stones River to that of the Trenton, coupled with the stratigraphy, suggests that the whole Stones River is younger than the Chazy. Kast Tennessee. In east Tennessee the Maclurea limestone was correlated by Safford* with the Chazy or Black-River of New York and “Canada. While a large part of this limestone seems to be of Trenton age, a sono around Lenoirs has aftorded the writer a small aia containing fossils characteristic of Division 2 in the Lake Champlain region. This region needs further study before definite correlations are made. DEscrRIPTION OF NEW SPECIES. BRACHIOPODA. Lingula columba sp. nov. Shell small, oval in outline, gently and uniformly convex. There are no flat slopes and the front is semi-circular in out- line. The posterior end is somewhat triangular, the beaks pointed. The surface is covered by very numerous and promi- nent concentric strie, no radiating lines showing except when the surface is exfoliated. One specimen is 10™ long and 7™™ wide; another is 7™ long and 5™™ wide. Locality.—East side of Valcour Island at Chazy, and on Isle La Motte. Type m Yale University Museum. Canarotechia pristina sp. nov. Shells small, transversely oval to subcircular in outline. Both valves moderately and uniformly convex. The dorsal valve has alow fold and the ventral valve a shallow sinus, which is noticeable only toward the front of the shell. There re 10 to 14 strong rounded plications, 4 on the dorsal fold and 3 on the sinus. ‘The 2 plications in the middle of the fold are smaller than the 2 outside ones and the median plication of the ventral valve is the weakest, which is the direct opposite of the state found in Camar otwchia orientalis. Locality.—Valeour Island and Chazy, New York. The type is in the Carnegie Museum. * Geol. Tennessee, 1869, p. 286. P. FE. Raymond—Ffauna of the Chazy Limestone. 369 Canarotechia major sp. nov. Outline somewhat oval, widest a little in front of the mid- dle. Dorsal valve with 10 to 14 strong angular plications. The ventral valve has 9 to 14. The fold and sinus are hardly defined except by a gentle arch in front, but are outlined on both valves by a pair of very strong plications. The dorsal fold bears 5 plications, the middle one of which is the strongest. The ventral sinus has 4 plications, the 2 largest in the middle. The ventral beak is somewhat incurved. Length of a good specimen 23™"; width 21™™. Locality.—Southeast point of Valcour Island, New York. The type is in the writer’s collection. Strophomena prisca sp. Nov. Shell of medium size, resupinate, nearly as long as wide. Ventral valve convex at the umbo, flat in front to about the middle of the valve and then concave. Dorsal valve flat on the umbo and convex in front. Cardinal area narrow, the wide delthyrinm mostly covered by the deltidium, with a small open- ine for the pedicle at the beak. Muscle area in the ventral valve small, confined to the space under the umbo. Surface marked by fine alternating striz, the prominent ones being very numerous and increasing by implantation. Between each pair of the strong striz are two or three finer ones and the whole surface is crenulated by fine concentrie striz. The dor- sal valve sometimes shows very small concentric wrinkles. One specimen is 15°5"" long and 20™" wide; another 16°” long and 19°5™™ wide. Locality.—All the specimens are from Valcour Island, New York, and are in the writer’s collection. Orthis ignicula sp. nov. Shell transversely oval in outline, usually but little wider than long. Hinge width nearly equal to the greatest width of the shell. Ventral valve strongly convex, the area high and a little incurved. Dorsal valve nearly flat, with a broad depression near the front. Area of dorsal valve rather wide. Cardinal process small. Delthyrium narrow and open. Surface marked by 16 to 25 direct rounded plications which increase by implanta- tion. _ Locality—Found rarely on the southeast corner of Valcour Island, New York. 3870 P. &. Raymond—FHauna of the Chazy Limestone. Orthis acutiplicata sp. nov. Shell small, almost circular in outline. Hinge width not gai equal to the greatest width below. Cardinal area of ventral ralve high and a little retrorse. Joramen narrow and open. Ventral valve strongly convex, highest on the umbo. Dorsal valve convex on the umbo, flattened in front. There is a shallow sinus on this valve, which is narrow at the beak but becomes wider in front. Surface marked by 12 to 15 sharp simple strize separated by spaces wider than the strie. Locality.—South end of Valcour Island. The types are in the writer’s collection. Plesiomys strophomenoides sp. nov. Shell small, ventral valve convex at the umbo, concave in front. Dorsal valve convex, with a narrow sinus on the umbo, but frequently with a slight fold on the front of the shell, in which ease the ventral valve shows a shallow median sinus. Surface marked by numerous fine strie, which-inerease by bifureation and implantation. There are usually 7 or 8 in the space of 2 millimeters on the middle of the front of the shell. The cardinal area of both valves is low. The ventral area has a narrow delthyrium, which at the apex is perforated for the passage of the pedicle. The interior of the ventral valve shows small but strongly marked muscle scars under the umbo. The muscle area is roughly quadrate and contains a pair of strong diductor scars, between which are the narrow adductor attachments. Posterior to the latter 1s a deep pedicle scar. The lateral edges of the diductor scars are bounded by strong — plates, which run back to support the dental lamellae. The interior of the dorsal valve shows a robust, simple cardinal pro- cess and small dental sockets bordered by strong plates which do not greatly diverge. In front of the cardinal process is a low but strong median ridge, on either side cf which are the four sears of the adductor, not, however, deeply impressed. Locality.— Crown Point, Plattsbure, and Valcour Island, New York. The type is from the quarries near the Platts- burg Fair Grounds and is in the Carnegie Museum. febertella cafoliata sp. nov. _ This shell is distinguished from ffebertella costalis by its smaller size, more pronounced dorsal sinus, and by the fact that the stria are always simple instead of bifureating. It differs from 77. borealis in its smaller size, and in the narrow and deep dorsal sinus. Locality.—Common in the lower part of the Chazy at Chazy and Valcour Island; also at Plattsburg, Valeour, and Isle La Motte, New York. “The type is in the Carnegie Museum. Ba s4 P. E. Raymond—Ffauna of the Chazy Limestone. 371 Orthidium lamellosa sp. nov. Ventral valve strongly convex, the area high and curved backward. Delthyrium narrow and open. Along the middle of the valve is a narrow and shallow depression in which there is one plication. The outline of the shell is subquadrate. The greatest length is at the hinge and the cardinal extremities are slightly alate. The dorsal valve has a narrow median sinus, which extends from the beak to the front and usually contains 2 plications. There are commonly about 20 sharp pleations, which are crossed by strong concentric lamelle of growth. An average specimen is 8™" long and 5:5™" wide. Loculity.—Valeour Island, Chazy, and Crown Point, New York. The types are in the Yale University Museum. PELECYPODA. Ctenodonta peracuta sp. nov. Shell small, longer than high, the beak about one-third the length from the posterior margin. Front rather drawn out, as in Otenodonta nasuta Hall. The greatest convexity Is at the umbo, the posterior slope very gradual. Both slopes to the hinge abrupt. but that to the basal margin gentle. One speci- men is 127" long and 9”™ high. This. species may be distin- guished from the succeeding one by its more depressed valves and by the prolongation of the: anterior margin ito a some- what nasute extension. Locality.—Found in some numbers in the trilobite layers at Sloop Bay, Valcour Island, and in the middle of the Crown Point section. The type is in the writer’s collection. Ctenodonta limbata sp. nov. Outline nearly circular, the beak back of tie middle Greatest convexity near the middle of the valve; all slopes steep. The cast shows a few faint lines of growth. Length of largest specimen 10"; height 10™™. ; height ¢y Length 975°"; height 6.257". Length 192526 = height Gone. Shell of medium size, a little smaller than Leperditia fabu- lites, oblong in outline, higher behind than in front. Hinge short, straight. Anterior end regularly rounded. ‘The poste- rior end slopes back almost straight for a short distance, but is broadly rounded on the lower posterior margin. The eye tubercle is small, on some specimens sharp, on others obscure. It is situated in the anterior angle, above and a little in front of the “muscle spot.” The latter is large, circular, and very finely reticulated. Back of the muscle spot is a region of the shell which is covered with fine raised lines radiating from the side of the spot. These lines frequently anastomose, making a very pretty reticulate surface. The muscle spot is raised above the general surface of the carapace on the lower poste- rior side, where these lines originate, but the upper and ante- rior sides are level with the main part of the shell. The right valve overlaps the left valve considerably, espe- clally along the ventral edge, which is abruptly deflected and usually shows a low short ridge right at the keel. The lower margins of the anterior and posterior ends are flanged. The border is very narrow and is marked by small pits, which increase in size ventrally. On one finely preserved specimen the anterior flange shows 8 pits, of which the seventh, counted from the front, is largest, and the eighth is very small. On the posterior flange of the same specimen there are 10 pits, the eighth from the posterior end being the largest, the ninth a little smaller, and the tenth minute. The left valve is not so high in proportion to the length as the right valve, but it is also abruptly deflected ventrally. It shows neither anterior nor posterior flanges and there is a small projection close to the hinge line and parallel to it. Below this is a slight depression. Locality—Common on Valeour Island in certain localities. Rare at Valeour and Chazy, New York. Prinitia latimarginata sp. nov. Carapace small and depressed. Front and posterior margins meet the dorsal margin at angles of little more than 90°. Both ends are broadly rounded, the ventral margin is gently curved. The shell is a little higher at the posterior end than P. EF. Raymond—Fauna of the Chazy Limestone. 381 in front. There is a deep sulcus just in front of the middle, which starts from the dorsal margin and extends half-way down the valve, turning a little forward at the lower end. On well-preserved specimens, in front of this sulcus there is a prominent eye spot, which is sometimes translucent. Often there is another slight depression or sulcus in front of the eye spot. The border is wide, concave, and of nearly uniform width all around from the anterior angle of the dorsal margin to the posterior one. The test is frequently punctate. _ Locality.—Common all through the Chazy limestone at Chazy, Valeour Island, Crown Point, and elsewhere in the Champlain Valley. TRILOBITA. Heliomera subgen. nov. fleliomera sol (Billings). Cheirurus sol Billings, 1865, Paleozoic Fossils of Canada, vol. i, p. 288, fig. 276. Cephalon short, wide, the glabella very large and flattened, the cheeks small. Glabella almost semi-circular, with 3 pairs of long, narrow glabellar furrows, all of which turn backward on their inner ends, each joining the one back of it, and the third pair joining the neck furrow, thus producing a central lobe like that of Amphilichas. This central lobe is of uniform width up to the inner ends of the first pair of glabellar fur- rows, but turns outward in front of that point. Toward the front of this median lobe there is a slight depression, some- what similar to that sometimes seen in Pliomerops canadensis. The first pair of glabellar furrows run’ backward at an angle of about 45°, the second pair at a smaller angle, while the third pair are nearly parallel to the neck furrow. The glabel- lar lobes are narrow and club-shaped. ‘This radiating arrange- ment of the glabellar furrows and lobes probably suggested the specific name. The neck ring is wide, flat, and separated from the glabella by a deep furrow, which extends the whole width of the cephalon. The cheeks are not sufficiently well preserved to be described, but enough of the test remains to show that the outline of the cephalon was the same as in Pseudospherezochus vulcanus. There is a narrow smooth border all around the front of the cephalon, and the surface is covered with fine tubercles. The relations of this species are rather doubtful. From the form of the cephalon it evidently belongs close to Pseudospherexochus, but there has not been seen in species of that genus any tendency to vary in the direction of an isolated central lobe and long isolated glabellar 382° Poe. ie fie ooh ec aes of the Chazy Limestone. furrows. The glabellar Fanceies in the various species of Pseudospher exochus ave usually faint, never deeply impressed as in this species. In this last character and in the presence of the median depression of the glabella, it recalls Pliomerops. The glabella is much larger in proportion to the size of the cephalon in Heliomera sol, however, and it is probable that this form must be regarded as intermediate between the two genera. For trilobites with this type of glabellar structure the subgeneric name //eliomera is suggested. Locality.—From the Raphistoma layers in the upper part of the Lower Chazy, at Chazy, New York. The type is in the Yale University Museum. Paleontological Laboratory, Yale University Museum, June 24, 1905. Benton—Properties of Catgut Musical Strings. 383 Arr. XX XIX.— The Mechanical Properties of Catqut Must- cal Strings ; by J. RK. Brenton, Pa.D. THE experiments here described were made in connection with investigations on the stress-strain relation in elastic solids, earried out at the Geophysical Laboratory of the Carnegie Institution under the direction of Dr. G. F. Becker, to whom the writer desires to make acknowledgment for many valuable suggestions in regard to the work presented in this note. Researches on the stress-strain relation have been made for rubber and for the metals; and it was thought of interest to experiment also on a substance of intermediate properties as regards extension within the elastic limit; for this purpose, catgut, as used for strings of musical instr uments, appeared to be Dest adapted. Owing to its hygroscopic properties and the complicated nature of the after-effects it exhibits, it was found that a precise determination of the deviations from Hooke’s law would involve an amount of labor far greater than was thought to be warranted by the importance of the substance. Such results as were obtained, however, together with general data on the mechanical properties of catgut, seem of sufficient interest to justify the publication of the present note. Tensile Strength.—A piece of catgut -061™ in diameter had an average breaking load of 12° 0 kg., corresponding to a ten- sile strength of 43 ke. per mm* (60,000 Ibs. per sq. in.). A piece ‘038% in diameter broke under 4°5 kg., corresponding to a tensile strength of 41 kg. per mm’*. These figures show that it is nearly as strong as copper wire, and must be classed as one of the strongest organic substances, far exceeding all kinds of wood (less than 20,000 lbs. per sq. in.), leather (5000 Ibs. per sq. in.), and hemp ropes (15,000 lbs. per sq. in.). Catgut musical strings, as furnished on the market, are twisted, aud tend to untwist when subjected to tension, and to twist up again when tension is removed. In order to study their elasticity, this twist must be removed, which is accom- plished by soaking the string in water. If hot water is used the string becomes very soft, and contracts greatly in length. In this condition it behaves very much like rnbber ; it can be stretched to nearly double its unstrained length, and when released it snaps back like a rubber band. It is greatly weakened, however, by being wet; but it regains its strength more or less completely upon drying, as shown in the following table: 384 Benton—Properties of Catgut Musical Strings. Catgut string, 038°" in diameter. Average breaking load before special treatment, 4°5 kg. A ba Treatment. ' Protas tests. in kgs. 2 Soaked 4 hr. in water at 30° C., then tested while wet 2°1 1 13 66 6é 6é ; 66 dried, then | tested 5:0 bo ‘Soaked ‘1 hr. in water at 30° C., dried for five days, then tested | 4°3, 4:4 1 Soaked 5 min. in water at 90° C., then tested less | while wet | than 0°5 4 | é< <6 73 73 66 dried, then tested | 1:0-1°:9 Llongation at Rupture.—A piece 0-062™ in diameter, and 59™ long, stretched to 70°" just before rupture, or 19 per cent of its original length. Another test gave 15 per cent. These figures include whatever stretching was due to untwist- ing. After rupture the pieces were too much frayed out for any determination of their length. Specific gravity.—By weighing a piece of catgut (not treated with water) of known length and. cross section, its et gravity was determined as 1°18 (0°01). All the remaining experiments to be described were made on a violin E-string, 150° long and 0:062™ in mean diameter (the diameter was not quite uniform, varying between 0°059 and 0°65). It was freed from its original torsion on August 20, 1904, by soaking one and a half hours in water of 30° C., and while drying it supported a load of 0°5 kg. The experi- ments were carried on from time to time during the following year, at intervals during the prosecution of other work. | fygroscopic properties.—Upon setting up this string and sighting with a micrometer telescope at a point near its lower end, it was at once seen that the length of the string did not remain constant ; and by observing at intervals and determin- ing the humidity of the air at the same time, it was easily demonstrated that the strmg stretched when the dampness increased and contracted when it decreased. This is in accord- ance with the well-known tendency of violin strings to break in dry weather. When the weather is damp the string has to be tightened to maintain the tension to keep it in tune; with increasing dryness its tension increases until finally it snaps. The actual tension required on a violin E-string to produce the proper pitch of 640 vibrations per second may be computed from the length of string (about 33) and its specific mass Benton—Properties of Catgut Musical Strings. 385 (0:0035 g. per em. of length) by the well-known formula for transverse vibrations of strings, and comes out about 6 keg., or about one-half of the breaking load. During the above experi- ments the cae carried a load of 1:0 kg, and the temperature was 20 to 25° C. The order of magnitude of the changes was 0:0002 of the length of the string for each cm. of mercury of vapor tension. Precise determination of the dependence of length on humidity was made prohibitively difficult by the phenomena described in the followmg paragraph. After-effects.— Whenever any change was produced in the conditions, the catgut did not at once come into equilibrium under the new conditions, but did so only gradually. Thus, —— Changes of Length—-cms,——> in cms. of mercury Absolube Humid ity > a yaa = Days (January, 1905) ——_____» ae 7 me Figure 1.—Length and Humidity as Functions of Time. when the load upon it was changed, it exhibited in marked degree the well-known elastic after-effect, requiring some days to come to sensible equilibrium, during which time the change in length due to change in load increased by about 25 per cent of its final amount. It was found also that the stretch due to humidity tended to continue after a change of humidity ceased ; this could easily be seen in the fluctuations of humidity accom- panying changes of weather; but no facilities for controlling humidity were available, and so no attempt could be made to study these phenomena thoroughly. No doubt some time is required for moisture to penetrate into the interior of the cat- gut. The curves of figure 1 represent change of length of the 386 Lenton—Properties of Catgut Musical Strings. catgut, and humidity (absolute), both as functions of time. Aiter-effects were detected also in connection with change of temperature. The elastic after-effect is of course responsible for the beha- vior of new strings on violins, which get out of tune very rapidly at first, and always in the direction of lower pitch. With greater duration of the tension the after-effect becomes less marked and the material approaches equilibrium under the imposed conditions. But if left for some time, the tendency is always towards lower rather than higher pitch, if temperature and humidity do not 2 change. : To protect the strmg from changes of humid- ity, it was placed inside of a tube of galvanized iron (“ spout-pipe’’), 6 inches in diameter, and 2 meters long, the seam of which was soldered up (A, fig. 2). Near the top and bottom, plate glass windows (6, 6), 74x15™, were fastened into it and made moisture-tight by liberal appli- cation of thick grease (‘“‘mobilubricant ”) be- neath the glass and around its edges. Inside the bottom of the tube a circular trough (9, fig. 2) 4™ wide was soldered against it and filled with engine oil. A hanger was attached to the bottom of the catgut, and hung through the hole in the middle of the trough: {o. samc hanger was fastened an inverted cup (@) of tin, which dipped into the oil of the trough. Thus the catgut was completely protected against changes of moisture in the air, and at the same time could be subjected to any desired tension by placing weights upon the hanger sticking out from the bottom of the tube. The whole arrangement was supported from the ceiling and braced against the floor. Micrometer tele- scopes for reading at top and bottom were sup- ported from the tube itself. The top telescope took account of any sinking of the upper sup- port; the distance between the two marks sighted upon on the string was 138°". The tele- scope at the bottom could be shifted bodily to keep up with stretching of the string; a reflection prism was fastened to it in front of its objective, and reflected into its field of view the image of a steel centimeter scale fastened vertically near by. In this way the shift of the telescope when moved could be measured. Coefficient of Thermal. Expansion.—After the catgut had been in the tube four days, readings were made upon it from Benton—Properties of Catgut Musical Strings. 387 time to time, and were still found to fluctuate. Upon compar- ing these readings with those for temperature, it was obvious that the changes were related to it, as may be seen in figure 3; ——Change of length—cms,—> ee ae Daxs Gee May, 1905) mae FIGURE 5.—Temperature and Length as Functions of Time. eww (C)——> th 2 . c=] cet Change of length ——cms Temperatue=—Degrees centigrade —____> Figure 4.—Dependence of Length on Temperature. 388 Lenton—Properties of Catgut Musical Strings. but again the state of affairs is complicated by after-effects. These were simply neglected, however, and equations were set up for the determination of the coefficient of expansion, using eighteen observations ; solving them by the method of least squares, the coefficient of linear expansion came out —0-000081 per degree centigrade. The observations used are shown graphically in figure 4. As the coefficient of expansion comes out negative and larger in absolute value than for most sub- stances, the question arises whether the data obtained were not = otal Elongation— ems—> Ss Hours after Temoving eRe Sut : FIGURE 0.—Elastic After-effect upon removing 05 kg. Time i> 30 45 60 IT Time Hours, after spplying eee _ Figure 6.—Elastic After-effect upon applying 2°5 kgs. Benton— Properties of Catgut Musical Strings. 389 influenced by humidity. In the enclosure containing the cat- gut, the absolute humidity was constant, but the relateve humidity increased as the temperature fell; and it may be that under such conditions the material tends to absorb moisture and thus increase in length. It is possible, therefore, that in a perfectly dry atmosphere, the behavior of catgut under vary- ing temperature might be quite different. Elasticity.—To get at the true elastic properties of the material it would be necessary, after each change of load, to wait until the disappearance of the after-effect before deter- mining the corresponding length of the string. In strictness, = WA, \ Ally > tio — a 7 nga au cms———_> bowd=— kgs——_» an infinite time would be requisite for this; but practically the following procedure can be adopted, and was employed in these experiments: After each change of load, observations on the length of the string were made at intervals of a few hours for several days, and corrected for thermal expansion. From the data thus obtained, a curve was plotted with times for abscissas and lengths for ordinates ; and from this curve the final length which the string tended to reach with disappear- ing aiter-effect, was estimated. Of course such an estimation involves considerable uncertainty and arbitrariness; but no other course seems available, as long as experiments must be Am. Jour. Sci1.—FourtH Serizs, Vou. XX, No. 119.—NoOvEmMBsER, 1905. 27 390 LBenton— Properties of Catgut Musical Strings. limited to finite time. Two of the curves obtained in this manner are exhibited in figures 5 and 6, together with their asymptotes as estimated. Such curves were taken after each change of load; but it is not thought to be of any interest here to submit more than two of them, or to present tables of the numerical data from which the curves are plotted. The results obtained in this manner are summarized in the accompanying table. After any load had acted for a few days it was removed, and the string left unloaded a few days before the next load was applied. The individual readings were made to thousandths of centimeters; but the estimation of the length to which the string tended at infinite time was carried out only to hundredths. The first column of the table gives the loads, in kilograms, placed upon the hanger, which itself weighed about 0°5 kg.; the third column gives the total elon- gation after disappearance of the after-effect, estimated as explained above; the sixth column gives the after-effect, of change of strain from the first instant (practically about 60 seconds) after applying the load until final equilibrium is reached, expressed as percentage of the total final strain; the other columns require no explanation. & Violin E-string, 0°062™ in diameter. Time Stress | Total Young’s the Mean in kgs. | elonga- | — modulus | After- | load | temper- Load, in kgs. per tion in |Strain.| in kgs. | effect. jacted, | ature, mm?. cms. per mm?. in Ge days. ee \ Capplicd yin ha nO oO are E { 30% 5 24 UP | (removed) { Bee ( 0°62 OEE ene | 33% 5 24 1:0 (applied) 3°31 139 \O:ONOI), 2328 44% 6 24 15 (removed) | 4°97 2°26 |0°0164| 308 26% 5 25 Pal Yapiedyy iia ae Na z 29% 4 28 é 0} (removed) § oe ees Coe 29% 7 29 (applied): alo, WSETich (pees 22% 5 aL ae | (removed) § oe eee | Dee iis 29% 4 28 3°0 (applied) 9°95 4°66 (0:0338| 294 20% 8 27 3°5 (applied) 11°60 Orie WOsOsmar lp icaO 15% 6 28 Youngs Modulus.*—The mean value of Young’s modulus from these experiments comes out 322 kgs. per mm’, or 458,000 * The values of Young’s modulus given in the above table are obtained by direct division of each stress by the corresponding strain. In strictness, Young’s modulus should be determined from the slope of the stress-strain curve at the origin. But in the special case that the stress-strain curve is a straight line, the quotient of stress by strain for any point of the curve gives the same result as the slope at the origin. The data under discussion not being sufficiently regular to determine the true form of the stress-strain curve, it is taken as a straight line within the limits of the experiments ; and this justifies the above method of determining Young’s modulus. Benton—Properties of Catgut Musical Strings. 391 Ibs. per sq. in. _ If observations taken immediately after apply- ing the loads had been used, instead of those after the disap- pearance of the after-effect, we would have about 400 ke. per mm* for Young’s modulus; and this latter figure represents the elastic resistance of the material to a stress applied for a short time, as in longitudinal vibrations of the string. Limit of Elasticity.— As is seen from the table, slight per- manent set appears, though not with great certainty, after applymg 2°5 kg (+0°5 ke. for the hanger). That makes the limit of elasticity about 8 ke. per mm’, corresponding to a strain of 2°7 per cent. Stress-strain Pelation.—The results tabulated above are shown graphically in figure 7. It is clear from it that Hooke’s law is approximately true; but the results obtained are too irregular to furnish ground for any definite statement as to deviations from Hooke’s law. Mey Sewickley, Penn. 392 Flora—FEstimation of Cadmium taken as the Chloride. Art. XL.—The Use of the Rotating Cathode for the Esti- mation of Cadmium taken as the Chloride; by Cuartus P. Fora. [Contributions from the Kent Chemical Laboratory of Yale Univ.—exl.] In a previous paper®* the author has described the use of the rotating cathode for the estimation of cadmium taken as the sulphate. In the present paper a simular study has been made of the behavior of cadmium when taken in the form of the chloride. Some differences are to be expected, since it has been established with some certainty that cadmium chloride, when subjected to electrolysis, forms not only positive cadmium ions and negative chlorine ions, but also complex cadmium- ' chlorine negative ions; and, in addition, the chlorine, when set free, does not recombine with the water, to set free axygen, but exists in the solution in its free state. That there are some very important differences a few qualitative experiments showed, so that the study of the estimation of cadmium when taken in the form of this salt was now undertaken. A solution was made up to a convenient strength, and the standard determined by the mean of several closely agreeing determinations by the carbonate method, which the author had carefully tested and found to be perfectly reliable. This showed 071589 grm.-of cadmium in 30°" of the solution, or 0°0052966 grm. per cubic centimeter. I. In Solutions containing Sulphuric Acid. The procedure with cadmium chloride was the same as with the cadmium sulphate, and the results were very satisfactory. But emphasis is to be laid upon the dilution in this case espe- cially, for it was found that from the more dilute solution it is almost impossible to drive the last traces of cadmium. A dilu- tion of 45° was found to give the most satisfactory results. To this solution ten drops of sulphuric acid of 1:4 dilution were added before electrolysis. The following results were obtained under these conditions : s Cd. Cur’t = N.Dyjoo E.M.F. Time. Cd. fd. Error. No. grm. amp. amp. vts. min. gTm. grm. i Ol L059 10-175 3°0-4°5 6°5-7°8 25 01054 —0°0005 10.1059 2°03 Oe GnOn oO) 7°8 15 0°1058 —0:0001 Il. In Solutions containing Acetates. The acetate method has proved one of the most satisfactory for the estimation of cadmium sulphate, but strangely enough, * This Journal, xx, 268 (1905). Flora—Estimation of Cadmium taken as the Chloride. 393 it was found to be absolutely unfitted for the estimation of eadmium when taken as the chloride. The deposited metal was always spongy, often non-adherent and unfitted for quan- titative work. The sponginess was less marked when no potas- sium sulphate was present, but the metal was still poorly adherent and unweighable. The modifications tried are given for the sake of comparison in the following table. The cur- rent potential throughout was 78 volts, while the dilution, excepting in experiment No. 4, was 45™*. In No. 4 a dilution of 65™* was tried, but it offered no apparent advantage. Cd. NaOC.H;0. K.SO.. Cur’t= N. Dio. Time. No. grm. erm. grm. amp. amp. min. Notes. I. O°1324 15 0'5 1°d fo 3 very Sponoy 2-0-1324 SS: none 1:0 3°0 8 01314 orm. fd. 3. 0°1059 1:0 * O75 225° 20 non-adherent. 4. 0°1059 15 0-5 Weoee eg cael ee adherent 2 2 gem? formalin 0°1059 1°5 none 1:0 3 added, spongy 6. 071059 15 BE 0°75. 2:25 — non-adht., cryst. 7. 0°1059 0°5 0°5 1:0 3°0 = ee - Ili. In Solutions containing Cyanides. The use of a cyanide solution gave results with the chloride of cadmium as satisfactory as were given when the sulphate of cadmium was taken. As in that case, care must be taken to avoid foaming of the solution. The best dilution seemed to be 65°. The time required is a trifle longer than in the estima- tion of cadmium sulphate by this method. The following results were obtained : : No. Cd. KON. NaOH. Cur’t= N. Diop E.M.F. Time. Cd. fd. Error. SE. Ci. Sri.e amp. amp... vis; min: erm, erm. 1. 0°1324 1°5 1:0 + 12 18 350 6©0°13822 —0°0002 2. 0°1324 1°5 1-0 + 12 78 40 0°1317 —0:0007 In experiment No. 2 there was much foaming, and a trace of cadmium remained in solution, the deposition being much retarded. IV. In Solutions containing Pyrophosphates. The different modifications of the pyrophosphate method gave results which were quite satisfactory, and in every respect comparable with the results obtained with this elec- trolyte in the estimation of cadmium sulphate. As was the case with that salt, the use of ammonium hydroxide to dis- solve the precipitate gave the most satisfactory results; while . after that, sulphuric acid seemed to be the most suitable 394 Llora—FEstimation of Cadmium taken as the Chloride. solvent. The deposits obtained from solutions to which were added free phosphoric acid showed a shght tendency toward sponginess. When hydrochloric acid was added, the deposits were good but deposition was slow. The total volume in each ease was 45°"; the amount of sodium pyrophosphate used was 9°5 grm.; while the current potential was 7:8 volts. The following results were obtained: Cd. Curt = .N. Dio: Time; Conde wren: No. grm. Solvent. amp. amp. min. grm. grm. 051324 NEL OF cones 1c: 0°5 13s) 15 0°1327 +0:0003 O:1324 HeSOR 4) 12 dps: 0°75 225 85 0°1328 +0°0004: 0:1342 H,PO, (sp. gr. 1°7), 15 dps. 0°75-1°0 2°25-3'0 30 0°1331 +0°0007 071324 HCl, 1:4, 15 dps. 07-05 21-15 45 071819 —0-0005 me Gi) 1S) te V. In Solutions containing Phosphates. With cadmium chloride, hydrogen disodic phosphate must be used with even more care than with cadmium sulphate, if deposits which are even slightly satisfactory are to be obtained ; and even when used with caution, the tendency to form spongy deposits is so persistent that this method is not to be recom- mended where other methods are available. The following are the solutions tried, the concentration being 45° throughout: Cd. HNa,PO,. H:PO.. Curt= | N. Dio. E.M.F. Time. Cd. fd” Mirror No. grm. orm. (Gp. i i)) vamp: amp. vis. min. owns erm. 0°1059 = 0°25 som? 20-30 60-90 78 15 0082 | EG paee 0°1324 0°25 Deremay 2 0=370 6°0—9'0 7°8 18) 071344 =+0°0020 0°1324 0°20 10 dps 1:0 3°0 78 15 0°1330 = =+0°'0006 0°1324 0°20 6dps 0°25 0°75 7'8 30 6©0'13810 )=60-—- 00014 Hq Co DO ee Numbers 1 and 2 gave spongy deposits ; number 4 showed no color upon testing the solution at the end of the operation with hydrogen sulphide, but this test does not seem to be very sensitive in this solution. Number 3 seems to represent the best conditions. VI. In Solutions containing Oxalates. Several qualitative tests, using conditions identical with those giving the least unsatisfactory deposits with cadmium sulphate, were tried upon. the cadmium chloride, with like unsatisfactory results, so that the work upon the oxalate method was not pursued further. VII. In Solutions containing Urea, ete. A few qualitative tests seemed to indicate that solutions con- taining urea, formaldehyde or acetaldehyde would furnish very satisfactory media for the estimation of cadmium, taken in the form of the chloride, but further experimentation proved these appearances to be deceptive. Under these conditions, No. bo Cd. 0°1324 0°1324 0°1324 071324 1°5 Lees I alg Estimation of Cadmium taken as the Chloride. 395 Flora the cadmium is deposited much more quickly from a solution of its chloride than of the sulphate, and the deposit in the earlier part of the precipitation appears to be very satisfactory. But the chlorine set free apparently has some action upon the organic compound present to produce substances detrimental to the process. With care, however, satisfactory results may be obtained, as may be seen ‘from a study of the following tables: Series A.—UREA. The amount oe urea present, therefore, should be between 15 grm. and 2 grms., and the current potential should not exceed 8 volts, ed of the 12 volts permissible when cad- mium sulphate was used. The test with hydrogen sulphide does not seem to be very delicate in this solution, so that at least 30 minntes should be allowed for each determination. Some writers recommend the testing of the end-point im simi- lar cases by raising the level of the liquid upon the cathode, but this was not proved of much value in this work, as the amount of metal deposited upon the fresh cathode surface from the solutions near the end of the process is imperceptible. A solution to which formaldehyde was added gave the fol- lowing results: SERIES B.—FORMALDEHYDE (FORMALIN). Tot. Gree ren. Curt — NeDi,o. MCE. Lime vol. Gad: td: Error. erm. grm. amp. amp. vts. miny = cm? erm, erm. Notes. spongy, 324 a 1°0 3°0 12 20 60 0°1812 —0°'0012 not all out ( spongy, 0°1324 2 1s) a0 L2 30 60 0°1336 +0:0012 py all fe very 0°1324 3 1°0 She, LO eED 55 6©—0°13842 +0°0018 spongy, all out 0°1324 2 O25 O75 SEIS, 60 0°1324 +0:0000 all out 01324 2 0°5 1:5. 12° 20 60 0°1333 + 0-0009 me SPY: all out 0°1324 1 1-0 3:0 12 20 60 071370 -+0:0046 Very: spongy Paar a 0°25 —0°5- 0° 7515. 78°30 60 01328 +0°0004 good Cotas O50 25—-0°9 0:75-5:78 30 60 0°1329 +0:°0005 fair Tol. Horm: Curt —- N-Dii)- KE. MCR. Time; vol... -Cd. fd: >. Error: grm. em’. amp. amp. bse.» Mi CM: «form, erm. Notes. x0 025-13. 0: la—f£0 78 90 Abe 2 b355 1-0-0009) tair 270 050-2°0 1°50-6°0 11°8 15 45 0°1330 +0:°0006 slt. spgy £-5-. OAD 2°25 78 30 60 0°1324 +0°0000 compact 1°0 3°0 7°8 35 60 0°1325 +0°0001 sé 396 Flora—Lstimation of Cadmium taken as the Chloride. It will be seen from these results that it is better to use a somewhat smaller quantity of formaldehyde than was used with the cadmium sulphate; the current potential should not exceed 8 volts; while the solution should be rather dilute (GOs) These cautions are of even greater iipoeanee when acet- aldehyde is used, if satisfactory results are to be obtained, as the followmg results show : SERIES C.—ACETALDEHYDE (95%). Aldehd. Tot. Cd...(992). Curt’ =] NDiog | EM Dime. vol. Cditd) hiner, Nos Aerie) e emer: amp. NAisig ypu iagubangs (Ghaaky~ area Lia grm. Notes. not all 1. 0:1324 52:0 0:2-0°7 > -0°6=251 7°8 30. 60) So 00"1811 —0°0013 4 out slt. spgy 2, 2071324 ele (02-085 NOVO. 2°25) ods 30- 60 0°1346 +0°0022 spongy : Spon y® 3: O:1324 "(0:5 -0:2-1°0)* 0:6-350 = 7-8 65 60 0:1307 —0-0017 ~ not all out 4, 0°13824 1:0 0°1-0'°75 0°38-2°25 8:0 35 60 0:1828 —0:0001 fair VI. Ln Solutions containing FKormates and Tartrates. Like cadmium sulphate, so also cadmium chloride gave nega- tive results when‘solutions containing im addition “potassium formate in the presence of formic ‘acid were subjected to electrolysis. Moreover, when no formate was added, but formic acid alone, the results were still unsatisfactory. To solutions containing 0713824 orm. of cadmium in the form of the chloride was added 1-5°* of formic acid, the whole diluted to 50°", and electrolysis conducted under potentials of 7-5 and 11°8 volts. In each case the precipitate was spongy and non- adherent, while the solution persistently held traces of cadmium, even after subjecting to the current nearly two hours. Solutions containing tartaric acid behaved in a similar man- ner. In the presence of 3 grm. of tartaric acid, under current tensions of 8 and 12 volts, the precipitated metal peeled from the cathode during revolution, the deposit was spongy, and deposition seemed to be complete at no point of the operation. Chemistry and Physics. 397 SCIENTIFIC INTELLIGENCE. I. CHEMISTRY AND PHYSICS. 1. The Formation of Ozone by Ultra-violet Light.—FiscuHErR and BraEMER, by employing a mercury-vapor lamp with quartz walls, have studied the effect of ultra-violet light upon oxygen. They have found that ozonization takes place if the temperature is not too high, for above 270° ozone is decomposed more rapidly than it is formed. ‘Thorough cooling of the oxygen by means of a water-jacket increased the yield of ozone, while a greater intensity of the light of the lamp also increased the yield to a certain limit and then decreased it, probably on account of the effect of greater heating. Upon doubling the speed of the oxy- gen through the apparatus the total amount of ozone formed was nearly doubled, but the percentage of ozone formed was some- what diminished. The authors believe that their experiments show the correctness of Warburg’s view, that the formation of ozone by the silent electric discharge is due to the ultra-violet light thus formed. Uranium Vanadium Yttrium Zirconium and other rare metals. Offers from producers should be accom- — panied by samples. SYSTEMATIC COLLECTIONS OF TYPICAL SPECIMENS In sets of twenty-five up to fifteen hundred specimens. Prices $5.00 upwards per set, the average price for students’ specimens being about twenty cents. We have supplied the leading institutions for thirty years, having lately completed a single order for over 60,000 specimens. Our material is the accepted standard both as to correct labeling and high quality. . E Free Collection Catalog, containing lists and illustrations of General Mineral Collections, Series of Ores for Prospectors, Sets of Crystals, Series illustrating Hardness and other Physical Characters of Minerals, with Price List of Laboratory Material and Individual Specimens. FOOTE MINERAL OO. Established 1876, by Dr. A. E. Foote. W. M. FOOTE, Manager. DEALERS IN MINERAL SPECIMENS AND COMMERCIAL RARE MINERALS, 1317 Arch Street, Philadelphia. THE AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES.] Art. XLI.— Two New Ceratopsia from the Laramie of Converse County, Wyoming; by J. B. Harcuer. (With Plates XII, XIII.) [From a Monograph on the Ceratopsia by J. B. Hatcher. Published by permission of the Director of the U. 8S. Geological Survey. | Editorial note.—In the course of his extensive study of the Laramie Ceratopsia contained in the U. 8. National Museum and in that of Yale University, Mr. Hatcher discovered two forms which were new to science. These he described in the above mentioned monograph, giving to the first, an undoubted Triceratops, a new specific name, while for the second speci- men, which represents a new genus as well as species, no name was suggested by the author. The duty of naming this form devolves therefore upon the editor. The generic name Dicera- tops is suggested by the lack of a nasal horn, while the specific name hatcherz will serve to commemorate Mr. Hatcher’s work in connection with this remarkable type. | In view of the recent discoveries among these most interest- ing forms, it has been deemed advisable to publish these descrip- tions at the present time without waiting for the publication of the monograph.—Ricuarp S. Lut. Triceratops brevicornus sp. nov. Plate XII, Figures 1 and 2. Type No. 1834, Yale Museum. Char. Specific: Supraorbital horn cores short and stout, not much com- pressed, nearly circular in cross-section. Nasal horn core short and stout with the anterior border vertical instead of being directed upward and forward at an angle of 30 degrees. Vertical and longitudinal diameters of lateral temporal foramen nearly equal. Orbit irregularly elliptical in outline with the longer axis running from above downward and forward, Postfrontal fontanelle open even in old individuals. The type, No. 1834, Yale Museum, of the present species consists of a nearly perfect skull with lower jaw and a com- Am. Jour. Sci1.—FourtH Series, Vou. XX, No. 120.—DEcEMBER, 1905. 414 J. B. Hatcher—Two New Ceratopsia. plete series of presacral vertebrae, together with a number of ribs more or less complete, and portions of the pelvis, includ- ing a portion of the right ilium and a nearly complete pubis. The vertebral series lay in position with the vertebre inter- locked by their zygapophyses from the axis to the last dorsal, though portions of some of the vertebree had weathered away when found. Behind the posterior dorsal, impressions of the centra of the first two sacrals were preserved in the hard sand- stone in which the skeleton was imbedded. Locality.—The skeleton was discovered by Mr. W. H. Utterback, and the exact locality was some three miles above the mouth of Lightning Creek and about one and a half miles south of that stream, in Converse County, Wyoming. The horizon was near the summit of the Laramie, and the specimen was collected by the present writer assisted by Messrs. W. H. Utterback, A. L. Sullins, and T. A. Bostwick. When dis- covered the skeleton lay imbedded in a hard sandstone concre- tion and was much shattered and weathered about the pelvic region. None of the limb bones and no part of the tail were recovered. The Skuit. The extremely rugose nature of the skull together with the closed condition.of the sutures, many of which are almost or entirely obliterated, make it certain that the type of the present species pertained to an old individual. The Cranium.—The chief distinctive features of the cranium are as follows: The supraorbital horn cores are unusually short and stout, especially at the base. ‘They are less compressed and more nearly circular in cross-section than in most other species. The nasal horn is short and very stout with the antero- posterior diameter much exceeding the transverse. Its anterior border is directed upward in a line perpendicular with the longer axis of the skull instead of forward and upward at an angle of about thirty degrees to that axis as in the type of T. prorsus. The lachrymal foramen, as in 7. serratus, lies between the maxillary and the nasal, but in the present species its anterior half is entirely enclosed by the maxillary, that bone sending upward a short process alongside the premaxillary process and forming the anterior one-half of the superior border of the foramen. The orbit is elliptical in outline with the longer diameter inclined backward from the perpendicular at an angle of about ten degrees. The lateral temporal fossa is triangular in outline, its respective borders describing nearly an equilateral: triangle, the fore and aft diameter only slightly exceeding the vertical. The rostral bone is heavy and very deeply excavated beneath. The epijugal is rather J. B. Hatcher—Two New Ceratopsia. 415 acutely pointed and regularly triangular in cross-section. The infratemporal arch, as in 7. serratus, is formed by the quad- rate with overlapping processes from the jugal and squamosal, that from the latter element occupying a slightly more ele- vated position in the type of the present species than in that of Z. serratus. The exoccipital process extends distally beyond the quadrate and projects as a small angular process. There are six exoccipitals, borne wholly on the squamosal, and at least three more between the last of these and the single median one situated at the median parietal region. Though the frill is not sufficiently perfect in this region to determine the number of epoccipitals with accuracy, there cannot be fewer than nineteen. The postfrontal fontanelle is large and circular in outline. The median longitudinal crest of the parie- tals is well defined and bears the usual rugosities. Near the apex the right horn core has been worn into a peculiar form by the aetion of wind, sand and water while it protruded from the sandstone concretion in which it was found prior to its dis- covery. The palatial view shows no characters essentially dif- ferent from those of other species of this genus. In the region of the supraoccipitals and parietals the sutures are so obliter- ated by age and obscured by distortion and crushing that it is quite impossible to determine their nature. The Lower Jaw.—The lower jaws with the predentary were in position and in a splendid state of preservation. The pre- dentary is rather longer than is common. On the superior surface of the mandibular fossa near the anterior end two large foramina pierce the wall and pass upward toward the dental chamber. ‘The splenial is very broad posteriorly and entirely encloses the mandibular fossa, except at the opening of the internal mandibular foramen. The coronoid process is low and stout and superiorly it is produced forward into a broad and somewhat decurved projection. At its greatest expansion the superior border of the splenial covers over for a short dis- tance the series of dental foramina on the inner side of the dentary. The principal characters of the skull are well shown in Plate XII, figures 1 and 2. The Vertebre.—The vertebre will be fully described in that portion of the monograph relating to the osteology of the genus Triceratops. Principal Measurements of Type of T. brevicornus (No. 1834, teateshdemoim On Simi oes oa ie oy ale 1652™™ ECACeS, OTeAGIOF titles. Se a Sr Gite A 1120 Expanse of jugal ------ ema ech SoMa a ae Se 600 416 J. B. Hatcher—Two New Ceratopsia. Expanse of frontal region at anterior border of orbits... 357™™ Greatest: diameter of onoity ch. ie sink ne We ee, eee 168 ~ Least ie Neary men any Lagi, meAN mets | 2D Fore and aft diameter of lateral vey ore foe eae 105 Vertical ia Rn: pee care tek HES Fe" RR ene 85 Distance from posterior border of orbit to posterior border OF Pye ee ee ee ats ee 840 Thickness of postirontal behind orbit_._-._ _-_ 27) 2a Least antero-posterior diameter of horn core immediately abOVe-OTDlG te oo enews ee Ea ee eee Antero-posterior diameter of horn core, six inches above OTD TUE eee a a Yc ay ee ve Ly Transverse diameter of horn core immediately above orbit 140 Transverse diameter of horn core, six inches above orbit 97 Greatest lenothof squamosal::: 42 208 be ee 870 a breadth jot eS cae eG 2 acl ee ee 433 Length of parietals along median line ______._:_..-2 222 712 Distance between squamosal sutures at posterior border Of frull 23 use Be Se ee ee ee eee 874 Distance between squamosal sutures at junction with post- fromtals. ies 2 a MCs eee ee ee eer 330 Distance from anterior border of orbit to posterior border of nasal opening ee ea 228 Distance between orbit and lateral temporal foramen __. 142 Distance between.lateral and supra-temporal foramina... 285 Distance from lateral temporal foramen 1 to poster ior border Of sqQuamMosal 9 See ee ee as ee 705 Distance from occipital condyle to Saute aor margin of CRESG) oe ee eee rea a ee 650 Distance from occipital condyle to interior border of ros- ee eM te sens Ree ge | 8) Distance from posterior border of anterior nares to ante- rior borderof rostral (2. a es | Sere ee 525 Distance from postfrontal foramen to extremity of nasal Ja 0) y (apera dese OSSCREn NR elie NS ES DIL eg 750 Greatest expanse of exoccipital processes aa eA nae pa 550 Distance from inferior border of orbit to bottom of jugal 343 Diameter of occipital condyle -22) 22). 2s. ee 88 Distance from mid-frontal region to apex of supraorbital 1A G) ott Wyse Pau N earn er Malay NRO Ten sat 500 Length of splental a. hoc Sere Soe cee re ee ae 503 es et ppmedenpary, (tree Las era ee 255 Greatest breadth ot predemtary: 2). 28.0 3 ee 142 Combined length of ene ands predenbany 72.5.0 ae aee 681 oh Ss S vartieular ss OU as wea 620 Total length of presacral vertebral series 22. 9220275222 2290 os tf ‘6 dOTSal SCTICS esas 2 ela sa ae ees 1490 J. B. Hatcher—Two New Ceratopsia. ALT Diceratops hatcheri Lull, gen. et sp. nov. Plate XIII, Figures 3 and 4. Mr. Hatcher’s description is as follows : “Char. Generic: Nasal horn core absent. Squamosal bones pierced by large fenestre, while smaller ones penetrate the parietals. The inferior border of the squamosal lacks a quadrate notch. Type No. 2412, U.S. National Museum. “Char. Specific: Supraorbital horn cores short, robust, and nearly circular in cross-section at base, erect and but slightly curved. Orbits project in front of the horns, the frontal region lying between the horns being concave. Exoccipital processes slender and widely expanded. “ The type, No. 2412, of the U.S. National Museum, con- sists of a skull without the lower jaw. ‘The posterior portion of the frill is somewhat weathered but the specimen appears to have suffered comparatively little from crushing. “ Locality: The specimen was found in a hard sandstone concretion about three miles southwest of the mouth of Light- ning Creek, Converse County, Wyoming. When found the concretion in which the shell was imbedded had entirely weath- ered out of the surrounding sandstone and stood at an altitude of five or six feet above the ground, firmly attached beneath to another concretion. The skull stood on its nose with the frill pointing upward. “ The Skull: The chief distinctive features of the skull are as follows: ‘The supraorbital horn cores are comparatively short, robust, and nearly circular in cross-section at the base instead of compressed, as in most other species. They rise more directly upward than in other species and are only shghtly eurved. The orbits also oceupy a position more anterior than that seen in other species; the anterior borders of the horn cores rise from about the middle of the superior borders of the orbits so that the orbits project well in front of the horns. The frontal region between the orbits is concave. The exoccipital processes are rather slender and widely ex- panded. “The nasals terminate anteriorly in a rounded rugosity not developed into anything approaching a nasal horn and resem- bling that of the type of Zrzceratops obtusus. The rostral bone is small and firmly codsified with the premaxillaries. The lat- ter are elongate but not deep. The maxillaries are massive and the lachrymal foramen is elongate and below and com- siderably forward of the orbit. The jugal is broad distally and firmly codssified with the epijugal. The lateral temporal fossa is nearly as deep vertically as longitudinally. The squa- 418 J.B. Hatcher—Two New Ceratopsia. mosal is elongate, and just posterior tothe quadrate groove it is pierced by. a large fenestra. The antero-inferior angle is little produced and ‘there is no quadrate notch, the inferior border in this region describing widely an open concavity. The parietals are broad and thin and, on either side of the median line about 235"™ in front of the posterior border, there is an elongated fenestra with a longitudinal diameter of 150™ and a greatest transverse diameter of 52™™. This fenestra is com- pletely enclosed on the right side, but on the left the parietal is injured in this region. In the drawings it has been restored from the right side. The supra-temporal fossa is elongate. There is a single median postfrontal fontanelle as in 77icera- tops, but posteriorly this gives origin to two deep channels, one on either side. These run backward along the surface of the parietal and terminate in two small circular fontanelles, condi- tions very similar to those which obtain in Zorosaurus. Measurements of the type. ‘* Distance from anterior end of rostral to posterior of squa- TOS ec ee oe lec ean ee 1990n Distance from anterior end of rostral to anterior of orbits 845 s ‘“‘ inferior border of orbit to lower end of jugal 363 ay ‘¢ posterior border of nasal opening to ex- tremity:of bed (205 ae ee ee ee 614 Distance from posterior border of orbit to posterior sur- faee-o£ horn core ss j= Ae See ee ee 175 Distance between anterior bos ders: of orbits 3-233 a= 340 Circumference of supraorbital horn cores at base........ 610 ee a es 6) mM. AbOVe” OEDIty as se 340 Vertical diameter of torbits © toys 32S es Se 165 Antero; posterior ‘diameter of orbits’) 22.432. eee 12522 [ Vote.—This genus is most nearly allied to Zriceratops and is distinguished therefrom mainly by the much smaller rostral bone ; by the absence of a nasal horn, which in all species save i obtusus is fairly well developed ; “by the very erect, short, robust, supraorbital horn cores which seem to take their origin much further back with relation to the orbit ; by the concavity of the frontal region between the orbits and by the peculiar form of the postfrontal fontanelle. The general proportions of the skull resemble Triceratops rather than the contemporary genus Zoro- saurus, in which the great frill so preponderates over the compara- tively abbreviated facial region. ‘The parietals resemble those of Triceratops except for the presence of the small fenestre on either side of the median line. | The squamosals differ from those of Z’riceratops in the con- formation of the lower border, which lacks the quadrate notch, and in the presence of the unique fenestre. Plate XII. Am. Jour. Sci., Vol. XX, 1905. \ A \\ teenth natural size. . nus Hatcher, one-six UtCO1 ceratops br ri Am. Jour. Sci., Vol. XX, 1905. Plate XIII. Diceratops hatcheri Lull, one-sixteenth natural size. J. B. Hatcher—Two New Ceratopsia. 419 Aside from the general proportions of the skull, Diceratops and Torosaurus differ in the presence in the former of separately ossi- fied epoccipital bones around the margin of the frill. These ossicles are apparently entirely lacking in Zorosaurus. The two genera agree in the possession of parietal fenestre though these are evidently not homogenous. They also agree in the form of the postfrontal fontanelle. While I believe Diceratops to be a valid genus, I am not inclined to lay the stress upon the parietal and squamosal fenestrz which Hatcher does, as they may possibly be pathologic. Those of the squamosal bones, which are found in no other form among Ceratopsia, are not of the same size, while only one is known in the parietals for the sufficient reason that the bone is broken away on the left side where the fenestra would come if present, and it is quite possible that it may never have existed. There is preserved in the Museum at Yale University a Clao- saurus scapula with a clean cut foramen through it with perfectly healed edges. This foramen is not present in the other scapula from the same individual and Professor Marsh used to say that the perforation was caused by a Zriceratops horn. This certainly seems suggestive of the manner in which the Diceratops fenes- tre may have arisen. Ricwarp 8S. Lut. Amherst, Mass. | DESCRIPTION OF PLATES. PLATE XE Skull of the type specimen of Triceratops brevicornus Hatcher. No. 1834, Yale University Museum. One-sixteenth natural size. FicurReE 1.—Lateral view. ang, angular; art, articular; cp, coronoid process; D,dentary ; ep, epoccipital ; ju, jugal; Uf, lachrymal foramen; maz, maxillary; no, nasal opening; nh, nasal horn core; o, orbit; pa, parietal ; pd, predentary; pmax, premaxillary ; qu, quadrate; r, rostral bone; sang, surangular; sq, squamosal ; soh, supraorbital horn core. FIGURE 2.—Palatal view. dc, dental channel ; exo, exoccipital; ju, jugal ; mex, maxillary; pa, parietal; pal, palatine; pmax, premaxillary; pt, pterygoid ; gu, quadrate; r, rostral bone; sg, squamosal; BO, basi- occipital ; C, occipital condyle. PLATE XPT. Type skull of Diceratops hatcheri’ Lull. No. 2412, U. S. National Museum. One-sixteenth natural size. Figure 1.—Lateral view. ep, epoccipital ; [f, lachrymal foramen ; mt, max- illary teeth; mx, maxillary; n, nasal; NO, nasal opening; o, orbit; pa, parietal; pmax, premaxillary; qu, quadrate; r, rostral bone; SF, squamosal fenestra; soh, supraorbital horn core. Ficure 2.—Dorsal view. ep, epoccipital; lf, lachrymal foramen; n, nasal opening ; 0, orbit; pa, parietal; paf, parietal fenestra; pff, postfrontal fontanelle ; r, rostral bone; S#, squamosal fenestra; sg, squamosal: soh, supraorbital horn core. 420 Lull—festoration of the Horned Dinosaur Diceratops. ART. XLII— Restoration of the Horned Dinosaur Dicora- tops; by Ricuarp $. Loi. (With Plate XIV.) THE new genus and ae described by Hatcher in the pre- ceding articlerepresents perhaps the most bizarre and grotesque form among all the race of horned dinosaurs, and the author has attempted an interpretation for the purpose of empha- sizing the features wherein this animal differed from any of its allies. Diceratops comes from the Laramie of Converse County, Wyo- ming, and while contemporaneous with Triceratops and Toro- saurus it is probably as late in geological time as any of the species of either genus, and may be said to represent the cul- mination of at least one phylum of the Ceratopsia. Diceratops differs from Torosaurus in the proportions of the skull, for in the latter genus the frill is relatively huge as contrasted with the abbreviated facial region. In this Diceratops and Tricera- tops agree, and it is quite evident that there is a genetic rela- tionship between these genera, while Torosaurus represents a totally distinct phylum. Perhaps the most notable point of distinction between Tri- ceratops and Diceratops is the presence of a fairly well devel- oped nasal horn in the former while in the latter genus it is lacking, a feature which in the author’s mane represents the culmination of specialization. The earliest known Ceratopsia are the J udith River types, characterized by an incomplete frill, by rudimentary horns above the eyes, and by a very well developed, generally erect or backwardly curved nasal horn. The supraorbital horns are progressive structures while the nasal horn is retrogressive, and during the lapse of time between the Judith River and Laramie periods, when the marine Bear- paw shales and Fox Hills sandstones were laid down, the Cera- topsia underwent a remarkable though unrecorded ‘evolution, _for when they again come into view in the Laramie the arma- ment is reversed, in that the great temporal horns are by far the larger and more efficient weapons, and the diminishing nasal horn, while supplementing the others in the various spe- cles of Triceratops and Torosaurus, is vestigial in the form under discussion. This change of armament was necessarily accompanied by a change in “the method of attack, for while the Judith River types probably used the one horn much as the rhinoceros does, with an upward thrust, Triceratops seems to have charged with lowered head, the small forwardly directed nasal and the larger { " 4 % ; 4 b y bs Lulli— Restoration of the Horned Dinosaur Dicerutops. \ 421 supraorbital horns meeting the enemy at the same moment of impact. The frill now becomes of greater protective value instead of affording leverage merely for the muscles of the neck. Diceratops exhibits the extreme of development of this style of warfare, for the supraorbital horns are the sole aggressive weapons while the widely expanded frill served admirably to withstand the shock of the adversary’s horns. We have here a precise analogy with the knight of old tilting with his spear and shield. The skull of Diceratops shows the horns to be very erect, much more so than in Triceratops, so that the head would have to be carried much lower in charging than in the latter genus and the horns through relatively short are extremely powerful. I have indicated a callosity, the last vestige of a horn, over the nasals, for they still remain very highly arched and evidently bore some of the impact of the adversary’s blow. The eyes were set in deep thick-rimmed sockets which look directly out- ward, evidently limiting the forward range of vision, but afford- ing ample protection to these highly necessary organs. If one will turn to Hatcher’s figure of the Diceratops skull (Plate XIII, figures 1 and 2), he will notice in the frill several apertures which Hatcher has called “fenestra.” Two of these are through the squamosal portion of the frill, one on either side, and one through the parietal.* They are irreoular in size and in position, and while the Judith River types and Torosaurus among the Laramie forms have parietal fenestre, they are large and symmetrical, and there is no instance of squa- mosal fenestree in any known genus of Ceratopsia. If the author’s conception of the final function of the frill is correct, there would be no reason for the development of apertures through it, which would only tend to weaken it and mar its usefulness. It seems vastly more probable that these are ‘old dints of deep wounds” received in combat. None of them, not even the great one on the left, were necessarily fatal, as they all seem to be through the free portion of the frill, and, while the bone was destroyed, the horny or leathery integ- ument may have grown again over the gap as indicated in the model. The edge of the apertures are healed, showing that the animal lived for some time after the injuries were received. I have represented the gape of the mouth with much less * Mr. C. W. Gilmore, who prepared the type specimen of Diceratops, is by no means sure of the “‘ parietal fenestra.” There was no bone adhering to the matrix at that point so he left the opening through the frill for want of evidence to the contrary. The bone forming the margin of the left squamosal aperture is decidedly pathologic. 422 Lull—fRestoration of the Horned Dinosaur Diceratops. backward extent than in other restorations of Ceratopsia. Here we cannot be guided by the form of the mouth in existing reptiles, for none living have the same feeding habits as these dinosaurs. Here the mouth may properly be divided into an anterior prehensile portion, the turtle-like beak, and a posterior masticating portion, the dental armature. In herbivorous mammals the gape only includes the prehensile and never the masticating portion, because of the necessity of muscular cheeks to retain the food in the mouth. The Ceratopsia had a dental apparatus which chopped the food into short lengths, and the pieces, falling outside of the lower jaw, would have been lost had the gape extended backward beyond the beginning of the tooth series. Massachusetts Agricultural College, Amherst. y Am. Jour. Sct., Vol. XX, 1905. Plate XIV. Restoration of Diceratops hatcheri Lull, from a model by the author. The upper figure is that of the front view of the model with the muzzle somewhat depressed. C. R. Keyes—Triassic System in New Mexico. 4238 Art. XLIII.— Triassic System in New Mexico ; by CHARLES R. Keyes. Tue “ Red Beds” of the Southwest, from central Kansas to the Grand Canyon, have long defied every attempt to deter- ‘mine their geological age, and to satisfactorily settle even the larger problems connected with their stratigraphy. In Kansas, in Oklahoma, in Texas, and on through New Mexico and Arizona to Utah, these formations have for more than half a century remained a puzzle. Those who have had to give some attention to the Red Beds have, in the absence of abundant characteristic fossils, considered the entire sequence either Tri- assic in age or (so-called Permian) Carboniferous. Since the making of extensive examinations of the Red Beds formations over broad areas in New Mexico and the adjoining states during the past few years, it has been found that there are a number of important general features that have either not received the attention they deserve, or have escaped notice altogether. When two years ago I made the statement* con- cerning the Kansas section, that after seeing at close range the Red ‘Beds of New Mexico sufficient data had been obtained to clearly demonstrate that their stratigraphy could not be unrav- eled on the basis of the Kansas scheme, the separation of the Red Beds into their component parts was then beginning to resolve itself into a satisfactory reality. The Red Beds do not form the homogeneous succession that they have been generally regarded as doing. Lithologically they are broadly divisible into two easily distinguishable parts. There is a large portion of the entire section composed of heavy argillaceous shales and clayey sandstones usually of deep red colors, rather uniform throughout, with much gyp- sum intercalated and disseminated, and with saline shales abounding. The upper part consists of light, sandy shales chiefly, with some heavy sandstones; the colors, while prevail- ingly reds, are quite varied; gypsum and saline shales are pres- ent only sparingly. The plane separating the two parts of the Red Beds section, as thus defined, is, when once recognized, a conspicuous one. In eastern New Mexico, in the Canadian and Pecos valleys, around the northern and western margins of the Llano Esta- cado, there is at the base of the upper one of the two terranes a well marked conglomerate that has been widely traced. Un- conformable relationships exist between this and the strata beneath. In western Texas, Drake+ and Cummins have also well established these facts. * American Geologist, vol. xxxii, pp. 218-228, 1903. + Texas Geol. Sur. , Third Ann. Rept., p. 227, 1892. > im 494 CO. R. Keyes—Triassie System in New Mexico. In western New Mexico, in the Zuni uplift, there exist, as was first shown by Dutton,* similar conditions, except that the evidences of unconformities have not as yet been noted, and in fact no attempt has yet been made to look carefully for them. Between the two lithologically different parts of the Zuni section of the Red Beds there also exists an important. conglomerate which the author just mentioned correlates with Powell’s Shinarump conglomerate of the Grand Canyon, and which is considered the base of the Triassic of that district. According to all available data, derived from the biologic contents, which at best are rather meager, the stratigraphic relationships, and lithologic characters, there is a lower portion of the Red Beds belonging to the so-called Permian (Carbon- iferous) and an upper portion which appears to be Triassic in. age. So: great difficulty which has been encountered in the econ- sideration of the Red Beds in the southwestern United States has been the existence of a great erosion interval during Early Cretaceous times when the Red Beds suffered severely from planing off during the period when they constituted part of a vast land area. This fact has only lately been fuily appre- ciated,t and its full significance grasped. The three general sections of western, central and eastern New Mexico may. be paralleled as in the subjoined table: Z Bs aS Ae aed bom Set Oa Sm tial a aie ey bn) ah Sealant aetp ee Emm ro lee See a Hed er el eee eon GENERAL Ren BEps Sections In NEw MEXICO. Western Section. Central Section. Eastern Section. Dakota sandstones - - Dakota ss. Dakota sandstones_- Wanting cess =e. Wanting Comanche sandstones 300 Zuni shales ._.- ._-- 1200 Wanting Pyramid shales __--- 100 Wingate sandstones 800 Wanting Amarillo sandstones 200 Shinarump shales -. 1500 Wanting Endee shales__.-..- 300 Moencopie shales... 500 Wanting Cimarron shales._.- 1000 Madera limestones . Madera li. Not exposed .-__-__- The geographic distribution of the Triassic beds presents some special points of interest. East of the Rio Grande the Carboniferous part of the Red Beds probably greatly predom- inates over the Triassic portion. West of that stream the lat- ter no doubt has very much the larger section. Owing to exten- sive erosion that took place over the Red Beds district, at least throughout much of what is now New Mexico, before the deposition of the Dakota sandstone, a large portion of the Tri- assic portion must have been removed. It may be that part of this erosion took place just prior to Triassic times, as the conglomerate bed 500 feet above the base of the Red Beds *U.S. Geol. Sur. , Sixth Ann. Rept., p. 185, 1886. + This Journal (4), vol. xviii, pp. 360-362, 1904. C. R. Keyes—Triassic System.in New Mexico. 425 section in the Zuni region and in the middle of the section in eastern New Mexico would indicate. — In the Canadian valley, at the eastern border of New Mexico, the sediments of the Triassic system are well represented at the top of the Red Beds section. Farther westward, where the Rio Pecos cuts the Glorietta escarpment, Newber ry distin- guished both Triassic and Permian (Cimarron) plant remains. Around the entire escarpment of the Llano Estacado, or Staked Plains, in eastern New Mexico and western Texas, embracing an area of over 50,000 square miles, the Triassic beds are more or less well exposed. The New Mexico portion of this belt is 300 miles long. The greater part of the Red Beds section seen in the Canadian and Pecos valleys is of Tri- assic age. Only in the bottom of these valleys is the Carbon- iferous part of the Red Beds found. It now seems quite likely that within the boundaries of Kansas none of the Red Beds section can be considered as being of Triassic age. Early Cretaceous erosion, which bevelled off the Red Beds, appears to have removed the Triassic strata alto- gether east of the New Mexico line and north of the Canadian river. The youngest layers of the Early Cretaceous (Comanche series) in overlapping northward on the old, even, erosion-sur- face, now appear to rest, in southern Kansas, on the lower part only of the Red Beds. West of the Rio Grande, in north-central New Mexico, along the Chama river, at the locality known as Abiquiu, N ewberry and Cope regarded a very thick Triassic section to be repre- sented. Around the Zuni mountains is an important belt of Triassic strata, which according to Dutton are more than 3500 feet in thickness. As detailed mapping of the region goes on, the beds which have been considered as belonging to the Triassic system will be found to have a very much wider geographic distribution than is at present known, and many new localities will doubt- less be discovered in which these strata are well represented. In eastern New Mexico the basal plane of the Triassic appears to be well established at the bottom of a well-marked conglom- eratic sandstone which separates the lower, dark red, clayey Red Beds from the upper, light reddish, sandy portions. At the base of this conglomerate there are abundant evidences of unconformable relationships between the two parts of the section. These relationships are well displayed along the northern magnificent escarpment of the Llano Estacado, which forms the south side of the Canadian valley. Drake,* who has traced the formation along the entire length of this great wall, and * Texas Geol. Sur., Third Ann. Rept., p. 229, 1892. 426 CO. Le. Keyes— Triassic System in New Mexico. who has examined its details rather carefully, says regarding the character of this unconformity: ‘The slight difference in dip, and sudden change in lithological character of the Triassic beds from the Permian, point conclusively to a break in the sedi- mentation of the two formations. At some localities the Tri- assic beds are overlain by Cretaceous, but generally by Tertiary material. The Cretaceous escarpments or buttes resting on the Triassic beds are often two hundred feet thick, and mostly limestone. The denuding forces that for an immense length of time were cutting these Cretaceous rocks back towards their present limits must have carried away a great deal of the Tri- assic before it was covered by Tertiary. The strata thus enclosed between two uneconformable beds must of necessity vary in thickness, and so we find it varying from a few feet to: nearly four hundred feet. Even in localities close together the beds vary considerably in thickness. The average, however, will probably reach two hundred feet.” Of the appearance of the two formations a short distance east of the New Mexico line the same writer * observes: “The contact between the Dockum beds and the underlying Permian is clearly marked. Both the color and lithological char- acteristics of the two formations bear a striking contrast. The Permian is a bright red argillaceous sand, slightly shaly, though sometimes massive, is characteristic for stratification planes, and below the top forty feet is interstratified with massive and fibrous gypsum, the gypsum becoming more abun- dant toward the base of the section exposed. The Dockum beds, arenaceous clays, in contact are a yellowish purple or a yellowish red, sometimes decidedly yellowish. The bedding is usually uniform and lacks the stratification planes so character- istic or the Permian. The contrast between the formations along their contact is so great that the contact may be located as far as the eye can see stratification planes in the freshly eroded outcropping bed, or as far as it can distinguish sharply contrasting colors.” The upper limiting horizon of the Triassic section is well defined. In the east, around the Llano Estacado in the Cana- dian and Pecos valleys, the superjacent formations are beds of the Comanche series of the Early Cretaceous. A little farther to the westward the massive Dakota sandstone of the Mid- Cretaceous age is the capping member. West of the Rio Grande there comes in between the Wingate division of the Triassic Red Beds and the undoubted Dakota sandstone a series of red and white shaly sandstones having a thickness of 1,200 feet, the exact age of which is at present not definitely determined. This formation is thought to belong to the Tri- * Ibid., p. 241. O. R. Keyes—Triassic System in New Mexico. 427 assic system. It is not impossible that it is Jurassic or Creta- ceous in age. It is the beginning of a great formation which extends a long distance to the northwestward into Arizona, western Colorado and Utah and which has been regarded as representing the Jurassic period. East of the Rio Grande a very marked plane of unconform- ity separates the Dakota sandstone from the Red Beds beneath. This break in sedimentation represents a profound erosion period, to which more detailed reference is made in another lace The stratigraphic extent of the Triassic strata in eastern New Mexico embraces about 500 feet of the general geological section. In the west the vertical measurement is very much greater. Dutton places it at 3,500 feet. As regards correlation of the Triassic formations, that por- tion of the Red Beds which has been regarded as of Triassic age may be compared, on the one hand, with the cognate beds of the Texas section, and on the other with the enormous thicknesses of Triassic strata in Arizona, Utah and Colorado. The standard section of the Triassic in New Mexico should be considered typically developed in the northwestern portion of the region, where the section is most complete and most extensive. In comparing this sequence with the Texas, Okla- homa and Kansas sections of the Red Beds there are presented some difficulties of an unusual kind. The planing off of the folded Paleozoics including the Red Beds in great part, during the erosion interval which existed in the eastern New Mexico region just before the deposition of the Dakota sandstone, . removed a very large portion of the formation. | As at present understood, the general relationships of the Carboniferous part of the Red Beds, the Triassic Red Beds, and the associated formations are best indicated by diagram as given below. LARAMIE > OE TT ERTIARYSE COLORA DAKOTA SANDSTONES _TAIASS IS ah HE = —— Li. oe: Fic. 1.—Relationships of the Triassic Formations in the southern Rocky Mountains. The cross-section traverses New Mexico in a nearly east and west direction, passing through the Cerro Tucumcari and the Zuni mountains. ; Owing to mountain-making movements which took place in the region in the latest Carboniferous or in Early Cretaceous 498 0. R. Keyes—Triassic System in New Mexico. times, or possibly in both, the Paleozoic formations and Tri- assic beds were bowed up, somewhat folded and faulted and then eroded off as a land surface. When Mid-Cretaceous (basal part of “ Upper”) beds were laid down, they were deposited largely on this old land surface worn out on the bevelled edges of the older formations. In the east there are shown marked unconformable relation- ships not only between the Cimarron Red Beds and the Tri- assic Red Beds, but between the latter and the Comanche series, between the last mentioned and the Dakota sandstone series, and between all of these and the Tertiary formations. In central New Mexico, the Dakota series rests directly on the Madera limestones of the Carboniferous. The Red Beds of both the Cimarron series and the series of the Triassic have been entirely removed through Early Cretaceous erosion. The Comanche series, which had “been constantly encroaching upon the old land area from the beginning to the end of its period of deposition, did not reach this far. In consequence the Mid- Cretaceous sandstones (Dakota) were deposited directly upon the Carboniferous limestones (Madera). In the west the sequence was very much as it was in the east, except that the Early Cretaceous appears to be entirely missing, the Triassic section very much thicker, and the Cim- arron eens very much reduced. It is a singular fact that the tripartite character of the Tri- assic sections in the west has a triple counterpart in the east. No direct connection between the two sections has been actu- ally traced in this field, for in central New Mexico a wide gap exists. Comparing the eastern section with the sections oF the adjoining portions of Texas, this agreement is very close. The entire Triassic section is there called the Dockum beds. As already stated, it is not believed that any portion of the Red Beds of Kansas are represented by the Triassic formations of New Mexico. In the Zuni uplift, where the Triassic beds are so well dis- plaved, they come up from beneath the vast field of Cretaceous sandstones. The sequence between the so-called Permian Red Beds and the Dakota sandstones of the Cretaceous is very thick. The data upon which the geoiogical age has been determined have been already given. Dutton, who a quarter of a century ago had perhaps given the subject more attention than anyone else, was unable to satisfactorily separate the two parts. He says that the “ Triassic system of New Mexico cannot be corre- lated so easily with its cognate beds in southern Utah and the Grand Canyon district as the Carboniferous and Permian. In the former region it has yielded but few fossils, while in the C. R. Keyes—Triassic System in New Mexico. 429 latter it has yielded none at all. We have here as well as there only an arbitrary provisional horizon for its base, and we are if possible still more uncertain where to assign its summit. The paleontological doctors disagree, and who therefore shall decide? It all hinges upon the question whether the Jurassic system has any representatives in this region. If not, then the summit of the Trias can be established at once. But if the upper portion of the enormous series of sandstones and gypsum beds which lies between the Shinarump conglomerate and the lower Cretaceous sandstone is Jurassic, the problem must wait for a solution.’’* Of the Zuni section of the Triassic system it may be that only the lower portion is represented east of the Rio Grande. The upper part of this section has been regarded as belonging to the Jurassic age; but until fuller data are obtained it does not appear advisable to recognize the Jurassic system in this part of the country. For the present, at least, all of this part of the sequence will be considered as a portion of the Triassic succession. New Mexico School of Mines, Socorro, New Mexico. * U.S. Geol. Sur., 6th Ann. Rept., p. 135, 1886. Am. Jour. Sci.—Fourty Series, VoL. XX, Ne. 120.—DrEcemBeEr, 1905. 30 430 G. Re. Wreland— Upper Cretaceous Turtles. Art. XLIV.—Structure of the Upper Cretaceous Turtles of New Jersey :* Agomphus ; by G. R. Wreranp. THE genus Agomphus was first proposed by Cope for the reception of Leidy’s Hmys jirmus and Adocus petrosus and Adocus turgidus,t all of which are based on very fragmentary and scanty remains from the Upper Cretaceous mar! beds of New Jersey, indicating a genus of heavy shelled turtles next related to Adocus. ‘Two of these original types, A. petrosus and A. tur- gidus, are now conserved in the Cope Collections in the Amer- ican Museum of Natural History, where the writer has been extended the courtesy of seeing them, together with the allied Adocus pectoralis Cope. An additional type from the Tertiary of Georgia, Amphiemys oxysternum,t is no doubt correctly referred to Agomphus, but has not been accessible. Since the brief descriptions unaccompanied by figures were given by Cope, the only addition to the very meager knowledge of Agomphus was made by Baur,$ who briefly noted in addi- tion to the close relationship to Adocws and inclusion in the Adocidee as next related to the existing Central American Dermatemydidee, the peculiar costiform processes and the interesting fact that Agomphus includes forms with relatively the heaviest carapace and plastron known. These latter facts were doubtless based on the specimens of the Marsh OCollec- tion obtained about the same time as the Leidy and Cope material, but never formally described or further mentioned although now found to make possible a complete description of the structure of the carapace and plastron, and to include at least two new species and a topotype as follows: Agomphus tardus Wieland (sp. nov.). (Figures 1-7.) By far the best specimen of the Marsh collection referable to the genus Agomphus is that numbered 774 (Accession No. 323), and now made the type of the new species A. tardus. This fine fossil was obtained from the Pemberton marl pits at Bir- mingham, Burlington County, New Jersey, in 1869. It is of especial interest as affording the structural characters of the * The first paper of this series, on Adocus, Osteopygis, and Propleura, appeared in this Journal, vol. xvii (pp. 112-182, pl. I-IX), Feb. 1904, and the second on Lytoioma, in vol. xviii (pp. 183-196, pl. V-VIII), Sept., 1904. + The description of these forms under the generic name Emys appears on pages 125-8 of Cope’s Synopsis of the Extinct Batrachia, Reptilia and Aves of North America. Philadelphia, August, 1869.—Agomphus in Supplt, 1871. +t On a New Species of Adocide from the Tertiary of Georgia; by EH. D. Cope. Proc. American Phil. Soc., vol. xvii, July, 1877, pp. 82-4. § Notes on some little known American Tortoises (on pp. 429 and 430), Proc. Acad. of Natural Sciences, Philadelphia, 1891 (pp. 411-430). G. R. Wieland— Upper Cretaceous Turtles. 431 shell of another genus of a well represented Upper Cretaceous to Tertiary family, the Adocidee, and as being relatively the heaviest and most massive turtle shell yet discovered. Although originally a perfect fossil with suturally united carapace and plastron, only thirteen complete and five incomplete elements of the carapace, together with the hyo- and hypoplastron, have escaped the accidents of discovery and collection. Of the imper- fect parts but four are diagnostic as to form, whence the recov- ered elements that are wholly determinative virtually number 1 FIGURE 1.—Agomphus tardus Wieland (sp. nov.). Carapace and plastron of type specimen* with the missing portions restored in the estimated natural size and position. Actual length of carapace 33°™. Elements present indicated in the succeeding figures 2-5. but nineteen, or exactly one-third of the original fifty-seven elements of which the carapace and plastron was composed. These recovered elements of grayish to dark, marl green color, are however perfectly fossilized, uncrushed, disarticulated, and without crumbling or breaking of the sutural faces. Moreover they are by a rare and noteworthy chance so distributed as to clearly outline the missing elements and make possible a res- toration by the Museum preparateur, Mr. Gibb, and the writer, which it is confidently believed by both will be found essen- * Elements present : nuchai (incomplete), 2d and 5th neurals, left 1st and 2d pleurals, right 4th and 5th pleurals (incomplete), right 6th and 7th pleurals, left 2d marginal, left 5th and 6th marginals (incomplete), left 10th and 11th marginals, ight 8th-11th marginals, the left hyo- and the right hypoplastron. 432 G. R. Wieland— Upper Cretaceous Turtles. tially correct as to form and size whenever a complete indi- vidual of this species is fortunately discovered. ) Hornshields.—A medium-sized nuchal and twelve pairs of marginals with the inner or marginalo-costal suture not rising onto the pleurals, as in Adocus, but traversing the marginal plates throughout (except in the single species A. masculinus, where the penult and final or pygal shields respec- tively overlap the 8th pleural and pygal plate. Plastron.—Of medium size, without fontanelles and very heavy with the strong bridge suture extending from the pos- terior end of the 3d to the anterior end of the 8th marginal. Entoplastron large, of sub-isosceles triangular to rhombic out- line. LEpiplastral border rounded; anal region acuminate in every known species—not rounded as in Adocus. Agomphus is held to be distinct from the earlier proposed genus Adocus mainly because of the position of the marginalo- costal suture on the marginals, the very characteristic form of the plastron, and the enormous thickness of shell. Although some of the imperfectly known species may prove to intervene and bridge these gaps—not large when taken singly,—it appears at present that they uniformly separate an Agomphid series of closely. related, mostly small turtles ranging from the Upper Cretaceous into the Eocene. Therefore, as both Adocus and Agomphus are numerous in species, it would seem to be much the better policy to retain the latter genus so long as not definitely proven to merge into the former. The species respectively assigned to these two closely related genera of the Adocidee therefore are: 1. Adocus (Emys) beatus (Leidy) Cope. 2. fe agilus Cope. 3. «< pravus (Leidy) Cope. 4, S syntheticus Cope. bi. es punctatus Marsh. 6. ATES (Eimys) turgidus (Leidy) Cope. ie << dfirmus (Leidy) Cope. 8. i (Amphiemys) oxysternum one) Hay. 9. petrosus Cope. 10. a (Adocus) pectoralis (Cope) Wieland. itd e tardus Wieland. ~ 2. masculinus Wieland. — 444 G. LR. Wieland— Upper Cretaceous Turtles. From this numerous assemblage of species we naturally come to ask how turtles with such thick shells as the Agom- phids, the more naturally ascribed to land forms, came to be so intimately associated with Osteopygis, Lytoloma, and the various other semi-marine to marine turtles and other forms which teem in all the Agomphus localities in the Upper Cre- taceous marl beds of New Jersey. Being mostly small turtles the heavy specialized shells would mainly serve as a protection from the other larger and more powerful reptiles, which swarmed along and into the bays and estuaries of the New Jersey Cretaceous shore line, so that a salt water littora! habitat is not precluded. But while no specimens of the Marsh or other collections illustrating limb or cranial structure have yet been referred to Agomphus, it would seem that at least some of the species of the genus dwelt back from the shore line along the streams, on the more or less sandy river, ox-bow, or delta banks, and doubtless in the vast numbers paralleling the Ormocan Podocnenws, the easy prey of the jaguar, and once far more abundant on lower river courses than now. From such locations many shells might be carried forward to the shore front in flood time or in the course of estuarial change. Also, if congregating in any considerable numbers on the more nearly forest-free river banks, or on dune slopes, at egg-laying time, many individuals might then be either preyed on by other animals, or swept shoreward. It is a fact of some slight bearing on such a conjecture that while the Adocidee are much more numerous than the other Testudinates of the marl beds, nearly all the limb bones recovered pertain to the semi-marine to marine Osteopygid and Lytoloman series. Moreover the abundance of the fossils of the marl beds is probably not generally understood, since almost no specimens have been secured in the past twenty-five years. Only a very few per cent of the specimens uncovered in the marl pits, when excavation was actively carried on thirty years ago, ever made their way into the museums; and these were all from restricted areas, although these fossils were for the greater part abundant everywhere in the several fossiliferous horizons of the entire New Jersey mar! belt. Yale Museum, New Haven, Conn. Hf. D. Campbell—Cambro-Ordovician Limestones. 445 Arr. XLV.—The Cambro-Ordovician Limestones of the Mid- dle Portion of the Valley of Virginia ; by H. D. Camrsett. Neirger the Knox dolomite nor the Shenandoah lime- stone, if used as the name of a geologic formation, should be made to include all of the Cambrian and Ordovician limestones of the Valley of Virginia from Tennessee to Maryland. . M. R. Campbell* makes the Shenandoah limestone of south- west Virginia comprise not only the Knox dolomite but at least 1500 feet of Cambrian strata beneath it. He also describes two formations of limestone above the Shenandoah and recog- nizes 500 feet of subjacent variegated shale and impure lime- stone. At the border between Tennessee and Virginiat he makes the Shenandoah limestone include not only the Knox dolomite but five other formations, remarking that the six merge into one formation which prevails along the eastern side of the Appalachian valley at least as far as Pennsylvania. It is not the purpose of this article to discuss the correlation of the Knox dolomite and the Shenandoah limestone, which can be satisfactorily accomplished only after several additional sections across the valley of Virginia have been described in detail and the fossils from fixed horizons have been com- pared. This introduction is offered as an explanation for using the ‘following entirely new names for the formations recognizable in the limestones of the portion of the Appalachian valley near Lexington.and the Natural Bridge, Virginia. Section of the Valley Limestones near Lexington, Virginia. Period. Name of formation. Thickness in feet. Baiosicnn Liberty Hall limestone | 1000+ Murat limestone 100-150 Natural Bridge limestone 3500+ akan Buena Vista shale 600-900 Sherwood limestone 1600-1800 Sherwood limestone.—In the bluff of James River at Sher- wood, Va. and for more than twelve miles to the southwest, the lower part of this formation consists of several hundred feet of white crystalline dolomite. This dolomite is overlaid by heavy beds of hght blue and gray magnesian limestone with oceasional beds of shale and shaly limestone. It was just beneath or at the very base of the Sherwood limestone that *U.S. Geol. Surv., Geol. Atlas of U. S., Folio No. 26, 1896. + Ibid., Folio No. 59. Am. Jour. Sci.—FourtTs SERIES, Vou. XX, No. 120.— DECEMBER, 1905. 31 446 HH. D. a Cambro-Ordovician Limestones. C.D; Walcott found Lower Cambrian fossils. The forma- tion is superjacent to the quartzites and shales of the Baleony Falls section. Buena Vista shale.—Bright variegated shale is conspicuous in the bluffs of James River between Sherwood and Buchanan, and along the road between Sherwood and Natural Bridge. Red bands predominate, but green, yellow, and brown colors are common. Mottled blue limestone beds alternate with the shale in the lower part, and it passes by a succession of shale and limestone beds into the superjacent limestone. In this formation C. D. Walcott+ found a Ptychoparia closely related to species from the Middle Cambrian beds of Tennessee. The formation is from 600 to 900 feet thick. It receives its name from Buena Vista, Va., where it is well developed. Natural Bridge limestone.—The formation consists princi- pally of heavy-bedded gray and light blue magnesian limestones with thin siliceous laminege as a conspicuous feature, especially upon weathered surfaces. Beds of white and pinkish dolomite occur now and then. Calcareous sandstone strata from a few inches to eight feet thick are occasionally prominent. Black chert occurs in nodules more or less throughout the formation, but heavy beds of chert are usually very conspicuous near the top. Specimens of Lingulepis and Obolus were discovered by C. D. Walcott in this formation two miles below Buffalo Mills on Buffalo Creek in June 1898, thus establishing by fos- sils the age of part of this limestone as Cambrian. Fossils — from 300 or 400 feet below the top of this formation make the age of its upper beds Beekmantown (Calciferous).¢ On account of the difficulty of determining the geologic structure in this section, the total thickness of the formation has not been accurately determined, but measurements which were made in a continuous series where there was no indi¢a- tion of folding or faulting indicate a thickness of over 3500 feet. The Natural Bridge and its canyon display part of this limestone, and hence the name. Murat limestone. —Superjacent to the heavy chert beds of the Natural Bridge limestone occurs a massive gray crystalline limestone containing bryozoa and other fossils in abundance. About 125 feet of it are well exposed along Buffalo Creek at Murat, Va., whence the formation takes its name. Its lower portion often contains chert nodules. The deep red clay soil resulting from the Murat limestone is conspicuous in contrast with the gray cherty soil from the top of the Natural Bridge formation. *This Journal, July 1892, p. 58. + Ibid., p. 52. sd Site A Bassler, Bull. U.-S. Geol. Surv. No. 248, 1900, p. 310. H. D. Campbell—Cambro-Ordovician Limestones. 447 Liberty Hall limestone.—In describing a section through this region in 1879, J. L. Campbell* used the name Lexington limestone for this formation, but inasmuch as the same name is given to certain Silurian rocks in Kentucky,ft it has been rechristened Liberty Hall hmestone from the name of an old historic ruin which is constructed on and of this rock, and which has been standing for more than a century and is as well known in this region as Lexington itself. The Liberty. Hall limestone is usually a succession of rather evenly banded beds of fine-grained, dark blue limestone and darker, more argillaceous limestone which weathers shaly. As we ascend into the formation calcareous shale predominates and limestone beds are less frequent. In this region the forma- tion has been much fractured and folded, and sometimes appears massive with innumerable veins of infiltratior of cal- eite filling the crevices. Again it appears shaly after long exposure to weather. Brachiopods and trilobites of Mohawkian age are especially abundant in the lower beds. From the top of the Murat through the lmestone and calcareous shale, so long as it carries vonspicuous limestone beds, the Liberty Hall limestone is about 1000 feet thick. Then follows about 600 feet of shale and slabby sandstone to the bottom of the first bed of quartzite above the Valley limestones. The thick beds of shale above the limestones, both northeast and southwest of the section here considered, give rise to another problem of correlation. Washington and Lee University, Lexington, Virginia, October 31, 1905. * This Journal, xviii, 1879, p. 29. + U.S. Geol. Surv. Geol. Atlas of U. S., Folio No. 46, 1898. 448 C. Barus—Ilons and Nuclet in Dustfree Arr. Art. XLV 1.—felations of Lons and Nuclei in Dust-free Air; by Cart Barus. 1. ly the following table I shall give typical results of the nucleation computed from the coronas observed in a glass fog — chamber, in the presence or absence of external radiation, when the saturated dust-free air contained is suddenly cooled by par- tial exhaustion of successively increasing magnitude. The amount of exhaustion (with which the supersaturation goes in parallel) may be conveniently specified in terms of the drop in pressure, 6p, between the outside and the inside of the given moderately efficient fog-chamber. Since the barometer was nearly normal, the corresponding volume increase, etc., may be readily derived. TABLE I.—Typical results of the ionized and colloidal nucleation of dust-free air, energized (or not) by weak and strong radiation. Fog chamber about 50 long, 15°™ in diameter; walls of glass 38°" thick, ends 1™ thick- Piping of one inch gas pipe. Barometer about normal. JD, distance between walls of fog chamber and anticathode or sealed alumintm tube with weak radium. X-rays X-rays Radium X-rays ‘X-rays from from from from from D=» end |D=600™ end | D=0:™ side |D=100™ end |D=50™ side Op n x 10-2. op n x 10-3. Op nx 10-8, Op n x 10-3, Op n x 10-3. 2A 6) 19 wo) 19 2 18 0 18 0 23 oO 20 2 20 2 19 1 19 2 25 7 oe 8 21 10 20 20 PAU Mirena) 430) O7 5 22 20 22 28 Pad By ail *98 28 18 23 28 23 38 22 50 22 57 29 Ad 24 30 24 45 24 75 24 93 30 a9 26 36 25 50 26 95 25 110 32 ae 28 37 26 52 28 110 27 138 34 87 30 39 28 53 30 124 29 145 36 100 34 4] 30 d5 39 13h) 32 155 ae see cat iat 35 56 sh he 390 160 In computing the (fleeting) nucleation one is left in doubt whether the nuclei (ions) are restored to the air more quickly than they can be removed by exhaustion; or whether the reverse is true. I have assumed the former to be the case and call this nucleation (number per cubic em.) n. If the nuclei are removed more quickly than they are reproduced, it will be necessary to multiply m by the corresponding volume increase, and I shall call this value VV. In the present experimentst V * Probably reduced by the presence of persistent nuclei in small number. + At high values of dp both n and N become untrustworthy as absolute values; but they suffice very well to indicate the relations. C. Barus—Ions and Nuclei in Dustfree Arr. 449 is usually much larger than m. In the cases of persistent nucleation due to the X-rays or other causes, JV is obviously to be taken; but here from the low values of 6p which suffice for condensation, the difference is not so important. 2. To vary the intensity of radiation, the anti-cathode of the X-ray tube or the radium tube (of thin aluminum, hermetically sealed, holding 10 mg. of weak radium —-10,000 * — within), is placed at a distance, /, from the outside of the fog-chamber. This was.a horizontal cylinder of glass, 50° long and 15°™ in diameter, with the end toward the bulb 1™ thick and the side wall 3°" thick. When J is measured from the end, persistent nucleation is not usually producible* because of the thickness of the glass to be penetrated. When 7 is measured from the sides, however, persistent nucleation just begins at ) = 50™ and increases at a rapidly accelerated rate for smaller distances. Hence the ionization corresponding to D = 50™ is a transi- tional value at which fleeting nuclei or ions merge into persis- tent nuclei. : ; 3. It the data are constructed graphically, it appears that all the curves are eventually intersected by the curve for dust- free non-energized air. In the latter we may recognize a region of ions (say from dp = 21™ to 27™) and a region of colloidal nuclei for larger values of 69; but the whole phe- nomenon is continuous. The effect of radiation as seen in the other curves is therefore to decrease the efficient nucleation of TABLE IT.—Persistent (large) nuclei(N, number per cm*) produced by intense X-radiation, in dust-free air. dp —18™, being. decidedly below the fog- limit. D measured from side of glass fog chamber (wall °3°™ thick) to anticathode. Aluminum screen inserted. p= 12 20 30 40 50° MES VO> 2 = 140 56 10 1 1 dust-free air more noticeably when the radiation is weaker and the supersaturation higher. These results may be tried directly for instance, with the radium tube at different distances, ); for a fixed pressure difference, dp. Thus at 9 = 41™ the nuclea- tion passes throngh a minimum at D = 25™ when D increases from 0 to 50™. The effect of radiation is then virtu- ally an aggregation of the colloidal nuclei of dust-free air. If the effect of ionization were merely to mask the presence of the smaller colloidal nuclei, the same effect should occur at intense ionization. Here, however, nuclei larger as well as indefinitely smaller than the mean ionic gradation are produced like the latter in continually greater numbers as the radia- tion increases. The case is rather one in which relatively *T have since succeeded in producing persistent nucleation through thin tin plate. 450 C. Barus—Ions and Nucler in Dust-free Air. small as well as large colloidal air nuclei are all successively aggregated into larger particles. In other words, as the radia- tion increases the whole air curve is bodily shitted more and more and with change of form to the left into smaller super- saturations, until eventually the persistent nuclei due to X-rays actually appear, and require scarcely any supersaturation to condense water vapor. 4. Using an apparatus with a larger exhaust cock and ex- haust pipes and therefore more rapid exhaustion, data were obtained of which the following table contains examples. TaBLeE III.—Nucleation of dust-free air, energized or not. New apparatus, 116" stop cock, connections of inch gas pipe. Bar. 76°23. Walls of fog chamber ‘38° thick. Three horizontal partitions of wet cloth within chamber. Op. s n x 10-°, Op. 8. mx ee Sept.i3: 44 "7-7. 180.) | Sept.3 10:6, 216 ee if AD 726) 195 204 35 TG 160 CANAD 5) 43 OS ice mignon 137 ee eG ae? wa 94°3 "7 2) X-rays, D=50°™ from side. te 5) “1 48:9. ig 190 25°3 1°4 Ley 43°6 §8:°0 216 26°1 2°6 6°2 BE 6 BO 240 27°0 3°6 17 30°8 49° 258 2 6) 5°0 46 26°6 = +92 229 28°6 5°4 60 22°80 21856 LS % 29°3 6°3 92 21s) Miiee 120 30°5 7°0 125 DO cS alles 100 35°0 Cg 160 19°5 ovl 39 RO 5 136 | 18°6 Tal 1 : ny 18°3 0? 0? II. Radium tube added on side. Barometer 76:05 (mean values) xe 43°6 4°4 43 X-rays, D = 12° from side. 35°0 5°4 67 18°9 202 207 30°8 5°6 69 20°0 Dei 54 DOA) 576 64 210.0 Tae ae DYDD I BSG. 60 O30 Tte73. a ie 223 5°6 57 | DA bons 195 ian) 5°6 55 298 Ors 290 20-0 4°1 20 sare 1 One 320 18°8 0 0 43°7 ||10°3 306 19°0 0 0 52°2 §9-4 240 19°0 9) gl The feature of these data is their steadiness in the lapse of time. The curves have naturally moved bodily into lower *G' B P corona steadily repeated, but not quite clear. | ¢+wreg. pl Cad ofl 2b Sw P corona. — | w y g. PO Oe tte yog’. C. Barus—Ions and Nuclei in Dust-free Air. 451 supersaturations and the asymptotes (or maxima) are in every ease reached and much higher in value. The range of super- saturation within which the condensations begin and are nearly completed is reduced so that the curves are usually steeper. In case of the new curve for non-energized air (and to the same extent in the others) the relative absence of nuclei in the region of ions is a distinguishing peculiarity. Investigated by the coronal methods, the curves rise, as it were, abruptly from the abscissa, and there is a rise of fog limit. ' ! / / / ° / ’ ¢ oe / ] [7 3% ’ 4 2 9 ald S 4 > Al) 4] r, G) Z eg / P 5 20 22 j 5 is) Fic. 1.—Charts for Table IIT, showing the coronal apertures (angular diameter being s/30) in cases of different supersaturation (pressure drop on exhaustion dp) in cases of non-energized dust-free air, and of dust-free air energized by radium and the X-rays from different distances, D. The dotted curve corresponds to less rapid exhaustion (Table I). Its intersection with the corresponding curve drawn in fullshould be noticed. It indicates the presence of a group of larger nuclei present in the former case and absent in the latter. 5. In connection with the present data, different suggestions made in the earlier paper and diverging from the more usual explanations, may be recalled: The effect of radiation if not too strong, has been shown to be virtually an aggregation of the colloidal nuclei of dust-free air. It seems probable A52, C. Barus—Ions and Nuclec in Dust-free Avr. that the ions or fleeting nuclei are such loose aggregates built up out of colloidal nuclei, because evidence of the pres- ence of colloidal nuclei absent at the low ionizations (exposure to weak radiation) is manifest at the high ionizations (exposure to intense radiation). If the radiation is very strong all sizes are represented, showing that the aggregates are virtually built up out of continually smaller colloidal nuclei probably closely approaching the molecular sizes, while at the same i : Ve Fic. 2.—Charts for Table III showing the nucleation 1 in terms of the supersaturation (pressure drop dp), for dust-free non-energized air, and for dust-free air energized by radium (10,000 x , 10 mg.) from different distances D. The dotted curves show. the corresponding cases (Dust-free air, 4 ; Radium, R; X-rays, X.) for less rapid exhaustion (Table I). The vertical spaces represent 20,000 nuclei each. C. Barus—Tons and Nuclei in Dust-free Arr. 453 time the existing nuclei are further aggregated into larger systems. Within the fog-chamber it 1s probable that the radiations whether undulatory or corpuscular, is at any point the same in all directions, for the nuclei in any given case are largely produced by secondary radiation. Henee it follows, qualitatively at least, that the inside of the fog-chamber is an ideal Lesage medium. One may argue, therefore, that a corresponding tendency for the preéxisting colloidal nuclei of dust-free air to aggregate into ions or larger bodies, should be manifest. Again the ions, conditioned by the presence of radiation, must fall apart when the radiation is withdrawn, and this is the case. One may infer also that the nucleating effect produced by negative corpuscles would be different from that corresponding to the positive residuals. Let the kinetic ionization pressure be supposed to mcrease as the square of the velocity of the corpuscles and as their density of distribution; then if the ionization becomes very intense it is possible that the pressure becomes strong enough to produce permanent union of the loose aggregates, or that the fleeting nuclei eventually become persistent, as is the case. If the ionized field is intensely produced by corpuscles issu- ing from within the body itself, as for instance, in combustion, or ignition, etc., one may expect that large nuclei as well as fleeting nuclei should simultaneously appear; or that nuclea- tion, passing through a transitional stage from fleeting to per- sistent as the electrification is more intense, should be the invariable concomitant of ionization. It is probable that the expulsion of corpuscles takes place whenever persistent nuclei are produced. Thus in the case of the X-rays, the generation of persistent nuclei occurs at an accelerated rate with time for a fixed radiation. If the radiation is cut off, nuclei are spon- taneously generated (secondary generation) for some time after. Arguments to the same effect would follow for hight pres- sure for wave lengths small enough to be easily scattered. Thus persistent and fieeting nuclei as a simple continu- ous phenomenon are produced by the X-rays (Table II), in the manner identical with the case of ultra-violet light. Simi- larly nuclei grow large in size as the ignition, the potential differences, etc., are larger. Finally, although the colloidal nucleation of dust-free air may be conceived to be aggregated both by undulatory and by corpuscular pressure, it is only in the latter case that the nuclei can be influenced by an electric field because the cor- puscles are themselves actuated. This distinction in fact exists between nuclei otherwise quite identical fleeting or persistent, but produced in one case by ultra-violet light and in the other by the X-rays or the action of radium. Brown University, Providence, R. I. 454 Flora—Cadmium by Means of the Rotating Cathode. Art. XLVII.— Additional Notes upon the Estumation of Cad- moium by Means of the Rotating Cathode, and Summary ; by Cuaries P. Frora. [Contributions from the Kent Chemical Laboratory of Yale Univ.—cexli. ] I. The Behavior of Cadmium Nitrate. SIncE cadmium is not readily precipitated by the electric current from solutions containing even small amounts of free nitrate acid, it was to be expected that cadmium nitrate would prove to be little fitted for estimation by electrolysis, since the action of the current would produce nitric acid. This, in general, was the result of my experiments upon the esti- mation of cadmium in the form of the nitrate upon the rotat- The deposits obtained from solutions containing sulphuric acid, the phosphates, pyrophosphates, urea or formal- dehyde, were satisfactory, but the time necessary for complete - deposition is so prolonged that these solutions are compara- tively valueless for the estimation of cadmium taken as the nitrate, since it would be easier and more trustworthy to trans- form the salt to the sulphate by evaporation with sulphuric With solutions containing acetic acid the metal was not precipitated, except in a narrow ring at the surface of the liquid. The behavior of solutions contain- ing formic acid, tartaric acid, acetaldehyde and formaldehyde was similar, but less pronounced. ‘The only solution from which I was able to obtain satisfactory results in the estimation of cadmium nitrate was a solution containing potassium cyanide. This solution was prepared by adding to the solution of ead- mium nitrate, which had been standardized by the precipitation and ignition of the carbonate, the desired amount of sodium hydroxide, and then redissolving the precipitated hydroxide in The time needed for com- plete deposition is somewhat longer than that required where the chloride and sulphate of cadmium were taken, but the deposit was bright and very satisfactory. Care must be used to avoid the use of too large an amount of potassium cyanide. The following table shows the results obtained : ing cathode. acid before electrolyzing. an excess of potassium cyanide. No. Cd. KCN. grm. germ. 1.020 920. wales 2-0 O92 0 as ne) B OO73: 2057 NaOd. erm. 0°5 0'5 0°5 Cur’t = N.D100. E.M.F. Time. vol. vts. amp. 4°0 3°0 iad amp. 12°0 9:0 Mas min. Tot. Cd. fd. cm, germ. 60 0°0933 60 0°0924 60 0:1072 Error. germ. + 0°0013 + 0:0004 —0°0001 Il. The Behavior of Solutions containing Free Nitric Acid. If free nitric acid be added to a solution containing a salt of cadmium, the precipitation. of the cadmium by the electric cur- Flora Cadmium by Means of the Rotating Cathode. 455 rent will be retarded, and even prevented altogether if the nitric acid be present in sufficient amount. Upon this behavior have been based methods for the separation of copper, bismuth and mercury from cadmium.* ‘Tests were made by me to determine the amount of free nitric acid necessary to prevent deposition of the cadmium, and it was found that 2°" of nitric acid of 1:4 dilution im 50° of solution (approximately 1 per cent of free acid) will absolutely prevent the precipitation of the cadmium upon the cathode (current, 38 amperes ; E.M.F., 7°5 volts). If less nitric acid was used, traces of cadmium were deposited upon the cathode. Summary of Results obtained in the Estimation of Cadmium by means of the Rotating Cathode. The results of the work described in this and the previous papers upon the estimation of cadmium by means of, the rota- ting cathode may be briefly summarized as follows: Under the conditions used, cadmium taken in the form of the sulphate may be very accurately and satisfactorily estimated by deposi- tion from solutions containing sulphuric acid, sodium acetate and acetic acid, or potassium cyanide; but little less satis- factorily from solutions containing urea, formaldehyde or acetaldehyde; and also with proper precautions, from solutions containing pyrophosphates, phosphates, tartaric acid or formic acid. From solutions containing oxalates or oxalic acid, ammonium tartrate, or potassium formate, however, I was unable to obtain satisfactory deposits. When taken as the chloride, cadmium does not permit such a wide range of condi- tions. Nevertheless, from solutions of the chloride containing sulphurie acid or potassium cyanide, or the pyrophosphates, the metal is deposited in a form comparable with that obtained when cadmium sulphate is taken. Solutions of the chloride of cadmium to which is added hydrogen disodic phosphate gave less desirable results; while solutions containing urea, formaldehyde or acetaldehyde gave deposits free from spongi- ness only after careful regulation of the conditions. In addition to the solutions containing the oxalates, oxalic acid, the for- mates and the tartrates, negative results were given in the case of the chloride by solutions containing the acetates, formic acid, and tartaric acid. The nitrate of cadmium is ill-fitted for electrolytic estimation, the cyanide solution being the only one from which satisfactory results were obtained. From solutions containing one per cent or more of free nitric acid, the cadmium is not deposited by the enrrent. * Edgar F. Smith, Am. Ch. J. ii, 42 (1880) ; Smith and Mayer, J. Ch. Soc., lxiv, ii, 496 (1893); Kammerer, J.. Am. Ch. Soc., xxv, 94 (1908); Riidorff, Z. angw. Ch. (1894), 388. 456 3 Flora— Estimation of Cadmium as the Ovide. Arr. XLVIUI-—The Estimation of Cadmium as the Oxide ; by Cuartes P. Frora. [Contributions from the Kent Chemical Laboratory of Yale Univ.—exlii.] Capmium may be simply and accurately estimated by con- verting the carbonate to the oxide by ignition. The oxide of cadmium may be subjected to very intense heat without loss from volatilization ; but in the presence of any carbonaceous matter 1t is very easily reduced to the metal, which is quite volatile at high temperatures. Consequently, this method has always been subject to more or less change of error where: paper filters have been used. The usual course has been to wash thoroughly and then dry the precipitated carbonate, which is then removed as completely as possible from the - filter: the latter then being burned separately. Even here there has always been avery appreciable loss, to avoid which various more or less complicated modes of treatment have been offered. As a type of these, we may take that of Max Mus- pratt,” who, after noting that high results were given by the ignition of the nitrate formed by dissolving the precipitated carbonate in nitric acid on account of included sulphate, pro- ceeded as follows: The precipitated carbonate was washed and dried, and as completely as possible scraped free from the filter paper, and then converted to the oxide by gentle igni- tion. This carbonate was entirely free from suiphate. The filter paper was treated with nitric acid and the resulting solu- tion and rinsings brought into a large porcelain crucible and evaporated to dryness. ‘The dry nitrate was gently heated and the weight of the oxide obtained added to that of the mass of the precipitate. Even after this tedious procedure, Muspratt is obliged to suggest that the results will be more satisfactory if the oxide obtained by the ignition of the paper and the residues upon it be calculated as Cd,O rather than CdO. Evi- dently the method would give satisfactory results if this redue- ing action of the filter could be avoided, and in former papers from this laboratory,+ it has been shown that this may be accomplished by the use of asbestos filters in a Gooch crucible. There is then no danger of loss from reduction, and the earbo- nate method is simplified and placed among the good analyti- cal methods. Recently, however, Miller and ‘Paget have found that “the carbonate method is the most troublesome and the least satisfactory ;’ but these investigators did not use the asbestos filter. * J. Soc. Ch. Ind.; xiii, 211 (1894). + Browning, this Journal (3), xlvi, 280 (1898) ; Browning and Jones, ibid. (4), ii, 269 (1896). + Ch. News, Ixxxiv, 312 (1901). Flera—Estimation of Cadmium as the Oxide. ADT The work of the writer upon the carbonate method fully substantiates the previous work from this laboratory. For the determinations given, a solution of cadmium sulphate was used, whose standard was accurately given by the average of a large number of closely agreeing electrolytic determinations. Por- tions of this solution were accurately measured from a burette and diluted to 800% with hot water. A 10 per cent solution of potassium carbonate was then added, drop by drop, with constant stirring until no further precipitation took place. The whole was then boiled for about fifteen minutes, when the precipitate became granular and quickly settled. It was then filtered upon an asbestos mat in a Gooch crucible, which had previously been ignited and weighed, and was then carefully washed with hot water. In several cases, washing by decanta- tion was used. The precipitate was then dried and ignited over a Bunsen burner, first gently, then at full red heat until a constant weight was obtained, care being taken to avoid the reducing action of any unburned gas from the burner. The following results were obtained : CdO taken. CdO found. Error. No. of exp. erm. germ. erm. A 0°1277 O25 —0°0002 = OZ 0°1280 + 0°00038 3. OT 27 0°1272 —0°0005 4 Ors Gou 0°1391 —0'0008 5 0°1399 0°1399 + 0:0000 6. 0°1703 0°1700 —0°0003 Qs 0°1703 071700 —0°0003 8. 0°2129 0°2128 —0°0001 9. -0°2129 0°2128 —0'0001 10. 0°2554 0°2554 + 0°0000 The method is simple in execution, and the above results prove its accuracy. Some of the older manuals also give as a method for the estimation of cadmium that of igniting to the oxide the pre- cipitated hydroxide obtained by adding a solution of sodium or potassium hydroxide to the solution containing the salt of cad- mium. Follenius has published* some results obtained with the use of an asbestos filter, and it was decided to try this method in comparison with the carbonate method. As in the experiments with the carbonate, portions of the solution of cadmium sulphate were carefully measured off from a burette, diluted to about 300°’, and heated to boiling. A 10 per cent solution of potassium hydroxide was then added drop by drop and the whole boiled for about fifteen minutes. Upon cool- ing, the precipitate quickly settled in a semi-granular state, *Z. anal. Ch., xiii, 284 (1874). 458 C. Schuchert—Mounted Skeleton of Triceratops prorsus. and was best filtered and washed by decantation. The results were lower than when the cadmium was precipitated as the carbonate, as is shown by the following table : CdO taken. CdO found. Error, No. of exp. germ. gTm. grm. 1; Om ia7 0-127 +0°0000 2: OLR 207 0:1270 —0:0007 3. 0°1277 0°1260 —0°0017 4, Oly —0'1286 + 0:0009 5. 0°1362 0°1350 —0°0012 6. 0°1899 0°1389 — 0°0010 he 0°1708 0:1697 —0:0006 8. 01703 0°16938 —0:0010 De 017038 0°1699 —0°0004 10. 0°1788 0°1802 +0°0014 1a 0°2129 0°2139 +0°0010 ee 0:2129 0°2128 —0°0001 While the figures show that fair results may be obtained by the hydroxide method, it can be compared with the carbonate method neither for accuracy nor convenience: the precipitate does not attain the same granular form as that of the carbo- nate; it is hard to filter, difficult to wash, and can be removed completely from the beaker in which precipitation takes place only with the utmost difficulty. : Ann, XLIX — The Mounted Skeleton of Triceratops prorsus in the U.S. National Museum; by C. Scuucuurtr. (With Plate XV.) Nore.—At various times articles on Triceratops by the late Professor Marsh have been printed in this Journal, and as the U. 8. National Museum is the first institution to possess a mounted skeleton of this, the largest-headed Dinosaur, it is deemed advisable to complete the records by reproducing here the illustration recently published in the Proceedings of that Museum.* Mr. G. W. Gilmore did the mounting, and from his article the following extracts are taken : Among the vertebrate fossils included in that part of the Marsh collection, now preserved in the United States National Museum, are the remains of several individuals pertaining to the large Cretaceous dinosaur, Triceratops. All of this material, which comes from the Laramie division of the Cre- — taceous, was collected by or under the supervision of the late Mr. J. B. Hatcher in the northeastern part of Converse County, Wyoming, a locality made historic by the researches * Article 1426, vol. xxix, 1905, pp. 483-485, 2 plates. Plate XV. Aime Jouneocl., Vols XA 905: C. Schuchert— Mounted Skeleton of Triceratops prorsus. 459 of this enthusiastic student. From this one region he col- lected the remains of more than forty individuals of the Ceratopsia, a record that has never been equaled. From the tip of the beak to the end of the tail the skeleton as restored is 19 feet 8 inches in length. The skull, which is 6 feet long, equals nearly one-third of this length.- At the highest point (the top of the sacrum) it is 8 feet 2 inches above the base. The mounted skeleton presents several features which would otherwise be lost to the observer if seen in the disarticulated condition. The short body cavity, the deep thorax, the massive limbs, and the turtle-like flexure of the anterior extremities are characters only appreciated in the mounted skeleton. The position of the fore limbs in the present mount appears rather remarkable for an animal of such robust proportions, but a study of the articulating surtaces of the several parts ‘precludes an upright mammalian type of limb, as was represented by Marsh in the original restoration. Moreover, a straightened form of leg would so elevate the anterior portion of the body as to have made it a physical impossibility for the animal to reach the ground with its MEAG cis 2 In constructing these parts we have followed -Marsh’s drawing, assisted somewhat bv fore-foot material kindly loaned by Dr. H. F. Osborn, of the American Museum ot Natural History, New York City. / The nasal horn of the skull used in the present skeleton appears to be missing, and on account of the unsatisfactory evidence as to whether the horn is wholly or‘only partly gone, it was decided not to attempt a restoration at this time. This will account for the absence of one of the important features upon which the name of the animal is based, Z7rzceratops meaning three-horn face, in allusion to the presence of the two large horns above the eyes and the third smaller horn on the nose. It may be of interest to mention here that Prof. O. C. Marsh used this skeleton (No. 4842), supplemented by other remains now preserved in the collections of the Yale Museum, for.the basis of his restoration of Z7riceratops prorsus, published as Plate LX XI in the Dinosaurs of North America. . . . A com- parison of the above restoration by Marsh with the mounted skeleton [see Plate XV] shows several differences in points of structure, due chiefly to the better understanding of these extinct forms. The most striking dissimilarity is in the shortening of the trunk by a reduction of the number of presacral vertebree. . . . Mr. Hatcher determined, from a well- preserved vertebral column in the Yale Museum, the number of presacrals as twenty-one, this being six less than ascribed to the animal by Marsh. 460 - Seientific Intelligence. SCIENTIFIC INTELLIGENCE, I. CHEMISTRY AND PHYSICS. 1. A New Formation of Diamond.—In a. lecture delivered before the British Association at Kimberly, South Africa, Sept. 5, 1905, Str Wittiam CrookeEs stated that he had found what were, in all probability, microscopic diamonds in residues obtained by Sir Andrew Noble in exploding cordite in closed steel cylin- ders. Crookes had calculated the theoretical melting point of carbon as 4400° absolute, and the melting pressure as 16°6 atmos- pheres; hence he concluded that the conditions of the cordite explosions, where a. pressure of 8000 atmospheres and a tempera- ture reaching in all probability 5400° absolute, would be favor- able for the formation of diamonds. Upon examining the resi- dues from such explosions, octahedral crystals were found which had high index of refraction, the proper cleavage, and the absence of birefringence of the diamond, and, although their other properties have not yet been determined, the chemical ordeal to which they were subjected in the treatment of the material leads to the belief that they must be diamonds.— Chem. News, xcii, 148. Fc Woeee 2, A New Compound of fron.—Orro Hauser has prepared a curious ammonium-ferrous-ferric basic carbonate, evidently a triple salt, to which he gives the formula Fe’,NH,Fe’”O(CO,),. 2H,0. It may be prepared, in the form of a crystalline precipitate, as follows : Ammonium-ferrous sulphate is dissolved in five parts of water, then a solution of commercial ammonium carbonate in about five parts of water is added until the precipitate which forms at first has redissolved. The quickly filtered liquid is then placed in a loosely closed bottle where only a small surface of the solution is exposed to the air. The liquid now becomes brown very rapidly from the top downward without any separa- tion of ferric hydroxide, and in about half an hour a slightly greenish precipitate settles to the bottom, and after about two days the iron is completely removed from the solution in the form of the new compound. The substance has a light green color when fresh, but it rapidly darkens upon exposure to the air. It dissolves readily in acids with effervescence ; it gives a black ferrous-ferric oxide with alkalies, and at the same time evolves ammonia.— Berichte, xxxviul, 2707. H. L. W. 3. Nitrosyl Fluoride.—Ru¥F¥F and Stivper, by the action of nitrosyl chloride upon silver fluoride have prepared this substance, NOF. It is a colorless gas which melts at about —134° and boils at —56°. In its chemical activity it resembles free fluorine Chemistry and Physics. ‘461 as well as nitryl fluoride, NO,F, which has been prepared by Moisson, but from the latter it differs in its density, in reacting readily with iodine to form iodine pentafluoride, and in reacting with water to form nitrous acid instead of nitric acid.—Zeztschr. Anorgan. Chem., xlvii, 190. : H. L. W. 4. The Atomic Weight of Stronttum.—About ten years ago T. W. Ricnarps determined the atomic weight of strontium by means of a comparison of the bromide with silver. He now pub- lishes the results of a comparison of the chloride with silver, made several years ago, but not published at the time on account of a discrepancy in the results due to an unknown cause. This discrepancy has been explained by the recent revision of the rela- tion between silver and chlorine made by Richards and Wells, and it is found that the averages of the two series of determina- tions agree with remarkable closeness when the correction in the atomic weight of chlorine is made; thus: From SrBr,, Sr = 87°663 From SrCl,, Sr = 87°661 These results are based on silver as 1077930, but Richards remarks that this number is probably not exact in comparison with oxygen as 16, and that the result will require modification when the true atomic weight of silver has been determined.-——Zettschr. Anorgan. Chem., xlvii, 145. H. LW. 5. Qualitative Analysis ; by E. H. 8. Barney and H. P. Capy. 8vo, pp. 278. Philadelphia, 1905, P. Blakiston’s Son & Co.—This book is interesting in being based upon the applica- tion of the theory of electrolytic dissociation and the law of mass action; but, however important these principles may be con- sidered in connection with teaching the subject, it appears that the extent to which they are carried in this case often causes a loss of clearness and conciseness as far as qualitative analysis is concerned. it may be mentioned that the book is not as large as its pages would indicate, for nearly half of them are left blank for use in keeping notes. H. L. W. 6. Charging Liffect of Réntgen Rays.—The ionizing effect of these rays has been apparently fully proved, and there is a satis- factory agreement in the results of observers. This is not, how- ever, true in regard to the question whether the rays give various bodies upon which they impinge electric charges. The subject has been investigated by Kart Hain, whose results support the contention of Prot. J. J. Thomson, that the rays give a positive charge to bodies. His results are summed up as follows: (1) All bodies upon which the rays directly impinge are charged positively. (2) Very thin sheets of metals are charged more strongly than thick sheets, and the difference is greater the shorter the exposure to the rays. Am. Jour. Scr.-—Fourts Srrizs, Vou. XX, No. 120.—DrcrembBer, 1905. 32 462 Scientific Intelligence. (3) The influence of the character of the metal surface is negh- oible. 7 ‘ (4) The potential of the charged plate is dependent upon : (a) The capacity with which the plate is connected ;—the quantity of electricity, that is, the product of capacity and poten- tial is smaller for the greater potential, that is for smaller capac- ity. If we assume that this is due to the conductivity of the air, one can assume that the quantity of electricity is constant. (0) The time of exposure to the rays. The potential increases with the time of exposure up to 20 sec. and then remains con- stant. (c) On the nature of the rays. Hard rays exert a stronger influence than weak rays. (d) On the nature of the metals; the potential is greater the greater the atomic weight, and the more the metal is electronega- tive. The influence of atomic weight is more notable with the hard rays ; the position of the metals in the electromotive series has greater effect in the case of the weak rays. (e) On the surrounding gas. The potential is greater in air than in CO,. (5) Secondary rays tend to neutralize the charges. This phe- nomenon explains the discordant results obtained by various observers.—Ann. der Physik, No. 11, 1905, pp. 140-171. 5. 7. 7. Emission of Negative Corpuscles by the Alkali Metals.— Elster and Geitel discovered that even the light emitted by a glass rod heated to a dull red heat was sufficient to cause rubidium to emit corpuscles. Professor J. J. THomson shows that rubidium and the liquid alloy of sodium and potassium give out corpuscles in the dark, This result leads Professor Thomson to speculate upon probable differences of temperature between the interiors of bodies and their surfaces arising from the explo- sion of atoms.—Phil. Mag., Nov. 1905, pp. 584-590. ovis 8. A New Method of showing the Presence of Neon, Krypton, and Xenon.—S. VALENTINER and R. Scumripr depart somewhat from the method of Dewar, by which, using the singular occlu- sion power of charcoal at low temperatures, Dewar showed the presence of neon, hydrogen and helium. Valentiner and Schmidt exhaust the spectrum tube and connected apparatus ; then admit a large quantity of argon, submit this argon to the occlusion action of charcoal at the temperature of liquid air. By this process neon is left in the spectrum tube ; and the quantity can be suitably increased by varying the pressure and amount of exhaustion. Suitable modifications of this method of employing argon as a basis and the occluding power of charcoal at different low temperatures enabled the authors to show the presence of krypton and xenon.—Ann. der Physik, No. 11, 1905, pp. 187-197. In 9. The Mechanical Properties of Catgut Musical Strings ; a Correction by J. R. Benton (communicated).—-I have to correct Geology and Mineralogy. . 463 an error in my article on the Mechanical Properties of Catgut Musical Strings, which appeared in the last issue of this Journal. On page 384, under the heading ‘‘ Hygroscopic Properties,” some observations are discussed which appear to show that the string in question increased in length with increasing humidity ; although, as mentioned there, its behavior was much cemplicated by after-effects. It is well known that the catgut strings of musical instruments are affected by changes of humidity: but they tend to contract with increasing humidity, and not to stretch, as stated in the article. The string on which I made observa- tions showed just the opposite behavior ; but it was under dif- ferent conditions from the strings in musical instruments, In the first place, it was under far less tension ; in the second place, it was free from any torsion; consequently any lateral swelling of its fibers which might occur would have no tendency to shorten it, while such swelling would tend to shorten a twisted string. IJ. Gkotogy AND MINERALOGY. 1. Lowa Geological Survey, Volume XV. Annual Report, 1904, with accompanying papers. Krank A. WILpER, State Geologist; T. EH. Savage, Assistant Geologist. Pp. vill, 560, 7 plates, 51 figures and 10 geological maps. Des Moines, 1905.— ' The Geological Survey of Iowa, under the charge of Professor Samuel Calvin, has enjoyed an excellent reputation for the thor- ough and systematic work which it has accomplished since its inauguration. Professor Calvin has now found it necessary to resign from the position of chief responsibility, and the place has been filled by Professor Frank A. Wilder, under whose auspices the present volume has been published. This volume gives promise that the future work done for the State will be carried forward on the same hnes and with the same excellent results that have characterized its predecessors. The volume contains, in addition to the administrative report, a chapter on the Mineral Production in 1904, by 8S. W. Beyer ; another on Cement and Cement Material, by E. C. Eckel and H. F’. Bain ; and then a series of chapters discussing in detail the geology of a number of counties accompanied by geological maps. These special reports include the following: On the Geology of Benton County, by T. E. Savage; of Emmet, Palo Alto and Pocahontas Counties, by Thomas H. Macbride ; of Jasper County, by Ira A. Williams ; of Clinton County, by Jon Andreas Udden; of Fayette County, by T. E. Savage. It is stated that the field work for 1905 is being carried on preéminently along economic lines, the earlier stratigraphic work having laid the necessary foundation. In the report of Mr. Beyer alluded to above, it is shown that the value of the mineral productions of the State in 1904 was 464 Scientific Intelligence. $15,000,000, of which the chief items are coal, making two-thirds of the whole, and clay more than_a fifth; others are building stone, gypsum, lead and sand-lime brick. 2. Summary Report of the Geological Survey Department of Canada, for the Calendar Year 1904. Roserr Bry, Acting Deputy Head and Director. Pp. xxxviil, 392, with seven geo- logical maps. Ottawa, 1905. — This volume gives a concise account of the work accomplished by the Canadian Survey dur- ing 1904. The total number of parties engaged was twenty- eight, and their labors extended over the entire area of the country, extending not only from the Atlantic to the Pacific but also into the Arctic. In general, the work carried on, as with other surveys at the present time, was largely on the economic side. As an illustration of what may be accomplished in this way by careful geological work, the Director mentions the recent discovery of a seam of coal, 10 feet thick, at a depth of 2,340 feet, near Pettigrew, in Cumberland County, Nova Scotia. The bore-hole was sunk through an unproductive covering at the sug- gestion of Mr. Hugh Fletcher of the Survey, as the result of his knowledge of the minute structural geology of that district. This successful result opens the prospect of finding numerous coal seams through an area of fifty miles in length and thirty in breadth. This discovery is given as an illustration of the very important economic results that follow accurate topographical and geological work. . Of the special reports given in the volume, two are devoted to the Kluane and Duncan Creek mining districts, in Yukon Terri- tory, others to the different coal-basins of British Columbia and so on. An interesting account is also given by Commander A. P. Low of the expedition to Hudson Bay and northward in 1903-4 by the 8. 8S. Neptune. Among other points may be mentioned a detailed statement of the phenomena accompanying the fall of the meteorite at Shelburne, Ontario, on August 13, 1904. 3. Glaciation of Southwestern New Zealand. — HK. C. ANDREWS, of the Department of Mines, Sydney, New South Wales, has written on “Some interesting facts concerning the glaciation of Southwestern New Zealand” (Trans. Austral. Assoc. Adv. Sci- ence, 1904, 189-205, 8 plates), in which he sets forth with much clearness the evidence of intense glacial erosion in the district of the fiords about Milford sound. Hanging lateral valleys, par- tially or totally truncated spurs, and the resulting rectilinear cliffs or over-steepened valley sides, with lakes and over-deepened fiords along the valley courses all occur in abundance. ‘These peculiar features are compared with those developed by normal erosion in the highlands of northeastern Australia (“ New Eng- land”), and the conclusion is reached that as normal erosion can- not possibly account for both, glacial erosion must be responsible for the peculiar features that occur where glaciers are independ- ently shown to have existed. In a supplementary note, Andrews ee —e Geology and Mineralogy. 465 well says: “The author feels confident that the glacial explana- tion is most convincing to students of geography, who .. . have not either lived in or even seen any region of former or present intense glaciation. Only to such workers does the whole series of novel perceptions presented during a first glimpse at a former strongly glaciated region come with the startling force of a reve- lation.” W. M. D. 4, Mastodon-Reste aus dem interandinen Hochland von Bo- livia ; von J. F. Pomercxs. Palaeontographica, Bd. 52, 1905, pp. 17-56, 2 pls.—Dr. Pompeckj, during his travels in Bolivia in 1902, collected near Ulloma and Calacoto a number of mastodon jaws and teeth here described and discussed in great detail. These were found at an elevation of 3800 meters above sea level, and belong to Mastodon bolivianus and M. humboldti. The belief is held by some that these mastodons lived at a time when the mountains had a far lower altitude than now, but Pompeckj holds quite the contrary opinion. He states: “During Diluvial time, or at least during that portion of it when the fauna containing Mastodon bolivianus existed, the high Bolivian plains at an elevation of about 3800-4000 meters prob- ably had the character of a steppe, similar to that of to-day, but with a greater rainfall and therefore with a richer growth of grass and bushes than at present. “‘ Neither the geological structure of the Bolivian highland nor its Diluvial fauna compels the conclusion of decided Diluvial or Postdiluvial elevation of the Andes.” C8. 5. Description of New Rodents and Discussion of the Origin of Daemonelix ; by O. A. PetERson. Mem. Carnegie Museum, 11, 1905, pp. 139-191, text figures and pls. 17-21.—The part of this paper of greatest interest in general paleontology relates to the interpretation of the so-called ‘“ Devil’s corkscrews,” so well and fully described by Professor Barbour. The general explana- tion has been that these gigantic screw casts are the fossilized and infiltrated roots of water plants. However, the suggestion has also been made that they represent the burrows of some fos- sorial rodent. Last year Mr. Peterson made a careful search for vertebrate fossils in the Daemonelix beds as exposed in the adjoining counties of Sioux in Nebraska and Converse in Wyoming. He states that one is always sure to find rodent remains in a locality where Daemonelix is found in great numbers, and he was rewarded in his exploration by securing a number of good skele- tons of the beaver-like Steneofiber within the spiral of Daemone- liz or its so-called rhizome. This led him next to study the tunnels of the living prairie-dog so common throughout the semi- arid region of the West. He did this by making a mixture of plaster, water, and sand, and pouring this into and filling the tunnel. Later this filling was dug out, and two of these casts are illustrated in the paper here reviewed; they certainly suggest 466 Scientific Intelligence. Daemonelix. The rhizomes of the gigantic corkscrew were found. to be either simple or several times branched ; some of them ended in an enlargement, but none of them showed small Daemonelix spirals emerging from them, as has been stated by other authors. The skeletons within Daemonelix are usually scattered “and quite often only the head is found crowded close to the wall, or inside of the rim of the compact mass of roots.” The best skeleton was found near the end of one of the rhizomes, Some of the latter attained a length of fifteen feet. In regard to the plant material found within the spirals, Mr. O. E. Jennings states: “‘ The vegetable tissues are apparently sim- ply the remains of a mesh of roots such as is sometimes found clogging a tile drain or sewer. . . . Enough was evident, how- ever, to plainly indicate that nearly all the roots were those of angiosperms, the cells discerned being quite typical.” The evidence thus far presented is decidedly more in favor of — Daemonelix being the cast of fossorial rodent burrows than the roots of some gigantic aquatic plant. CS 6. Heonomic Geology of the Bingham Mining g District, Utah; by Joun Mason Boutwe tx; with A Section on Areal Geology, by ArrHuR Keira and An Introduction on General Geology, by SAMUEL FRANKLIN Emmons. U. 8. G. 8. Professional Paper, No. 38, 396 pp., 49 pls., 10 figs. in text. —This paper, which almost approaches a monograph in size and scope, is a valuable contribution to the literature of economic geology as well as being a detailed description of an important and interesting min- ing district. Bingham has been a producing camp since about 1870. In the early days of the district the lead-silver deposits were worked, the carbonate ores being first mined and then later the sulphides. Some gold mining has also been carried on, both placer and vein deposits. In 1896 large bodies of low-grade copper ore were first seriously exploited and since then Bingham has steadily risen in importance as a copper producer. ‘The pro- duction of copper from the district for the year 1902 was nearly 15,000,000 lbs. The Bingham district is situated on the east side of the Oquirrh Mountains about fifteen miles south of Great Salt Lake. The rocks of the section are made up chiefly of quartzites, sandstones and limestones of Upper Carboniferous age, with intrusive bodies of monzonite and monzonite porphyry and extrusive flows of andesite. ‘There is one broad open flexure of the rocks in the district, a synclinal fold which pitches toward the northwest. Besides this many. smaller folds are found. ‘The rocks are also extensively faulted. The ores occur in vein, bedded and disseminated deposits. The vein deposits are chiefly those of argentiferous lead ore which fills fissures that traverse all of the rock types. The bedded deposits are of copper ore and are found in the limestones, while the dis- seminated copper ore is restricted to the monzonitic intrusives. cso ee aa! AE Neeps scully Geology and Mineralogy. 467 The most important ore bodies are those of the copper-bearing sulphide deposits which occur in massive marbleized limestones along particular beds in the vicinity of the intrusives. Mr. Boutwell sums up the geological history of the district as follows: “‘ Between Carboniferous and late Tertiary time monzo- nitic intrusives invaded sediments in the Bingham area, meta- morphosed them and introduced metallic elements which replaced marbleized limestone with pyritous copper sulphides. After the superficial portions of the intrusives had cooled to at least partial rigidity they and the inclosing sediments were rent by northeast- southwest fissures. “Heated aqueous solutions from the deeper unconsolidated portions of the magma then ascended these channels, altered their walls, and introduced additional metallic elements. At this time more pyritous copper sulphide may have been added to that formed earlier in the limestone in connection with contact meta- morphism. Monzonite, including its original metallic constitu- ents, was altered ; copper, gold and sulphur were probably added, and auriferous copper sulphides were formed. The silver-lead ore was deposited in the fissures, mainly by filling, partly by replacement. “Since this period of mineralization these original sulphide ores have been altered by surface waters, in their upper portions, into carbonates and oxides, and relatively enriched in their under- lying portions.” WwW. E. F. 7. Economic Geology, a Semi- Quarterly Journal devoted to Geology as applied to Mining and Allied Industries. Volume I, Number 1. (Published by the Economic Geology Publishing Company, Lancaster, Pa.)—The appearance of the first number of this new journal is an event of unusual interest and importance. Economie geology has only within the last quarter century estab- lished its place as a distinct and important department of geo- logical science. In Germany the Zeitschrift fur praktische Geo- logie was the result of this movement among German geologists and it has done much to place this branch of geology on a firm basis both at home and abroad. It is only recently, however, that in the English-speaking world economic geology has begun to occupy its rightful position, and this new journal has been established on account of this fact and with the hope that by its efforts a still larger recognition may be given the subject. Until now the American geologist who had interested himself in the problems of ore deposits had no field for the publication of the results of his investigations outside of the channels of the United States Geological Survey, except in various technical or semi- technical journals devoted to mining. It is, therefore, with a distinct sense of congratulation that we find provided here a proper place for the printing of such papers. The editor outlines the scope and oftice of the journal in his first editorial as follows: “The chief purpose of ‘ Economic Geol- 468 Scientific Intelligence. ogy’ will be to furnish its readers with articles of a scientific character. These will deal with the application of the broad principles of geology to mineral deposits of economic value, with the scientific description of such deposits and particularly with the chemical, physical and structural problems bearing upon their genesis. With the engineering and commercial aspects of min- ing this journal will not be directly concerned, as these subjects find ample representation in the technical mining journals.” The editor of “Economic Geology” is Prof. J. D. Irving of Lehigh University, and the associate editors; Mr. W. Lindgren of Washington, Prof. J. F. Kemp of Columbia University, Mr. F. L. Ransome of Washington, Prof. H. Ries of Cornell Uni- versity, Mr. M. R. Campbell of Washington and Prof. C. K. Leith of the University of Wisconsin. The magazine in its mechanical make-up has evidently been. modeled after the Journal of Geology. The paper, type, and general appearance are all excellent. The first number embraces 100 pages, of which about three-fourths are given to articles, whose titles ure as follows: The Present Standing of Apphed Geology, by F. L. Ransome; Secondary Enrichment in Ore- Deposits of Copper, by J. F. Kemp; Hypothesis to Account for the Transformation of Vegetable Matter into the Different Vari- eties of Coal, by M. R. Campbell; Ore-Deposition and Deep Min- ing, by W. Lindor en ; Genesis of the Lake Superior Iron Ores, by C.K. Leith ; The Chemistry of Ore-Deposition—Precipitation of Copper by Nataral Silicates, by E. C. Sullivan. There are also sections devoted to the informal discussion of topics relating to economic geology, to reviews and to scientific notes and news. W. E. F. 8. Minerals in Rock Sections ; the practical methods of identi- Sying Minerals in Rock Sections by means of the Microscope ; by Lea MclI. Luqurr. Revised Edition. 147 pp. New York, 1905 (D. Van Nostrand Co.).—The additions and changes intro- duced in the revised edition of this useful volume (see vol. vii, 319) are numerous and such as to materially increase its value for the practical worker with the microscope. III. MiIsceELLANEOUS SCIENTIFIC INTELLIGENCE. 1. National Academy of Sciences.—The autumn meeting of the National Academy was held at New Haven, Conn., on November 14 and 15. The following list contains the titles of papers read : JOHN TROWBRIDGE: Slow movements of electrical discharges. E. B. Witson: Sex-determinations and the chromosomes. L. B. MENDEL: Studies on the chemical physiology of devclopment and growth. W. M. Davis: The Dwyka glacial conglomerate of South Africa. B. B. Bottwoop: The disintegration products of thorium as indicated by the proportions of lead and helium in minerals. Miscellaneous Intelligence. 469 A. Hatz: Relation of the true anomalies in a parabola and a very eccen- tric ellipse having the same perihelion distance. S. L. PENFIELD: On a new mineral from Borax Lake, California. F. E. Beaca: On errors of excentricity and collimation in the human eye. ©. S. Prerrce: The relation of betweenness and Royce’s O-collections. L. P. WHEELER: Some problems i in metallic reflection. FRANZ Boas: On Pearson’s formulas of skew distribution of variates, A. AGassiz: On the variation in the spines of sea urchins. W. H. Brewer: Further observations on sedimentation. H. A. Bumsteap: The effect of Rontgen rays on certain metals. Recent publications of the Academy are as follows: Memoirs, Vol. [X.—Monograph of the Bombycine Moths of North America, including their Transformations and Origin of the Larval Markings and Armature. Part II], Family Cerato- campide, Subfamily Ceratocampine; by ALpHEus SPRING PackarRp. 149 pp., 61 plates, in part colored. Vol. X, No. 1.—The Absolute Value of the Acceleration of Gravity determined by the Ring-Pendulum Method ; by Cuarizs E. MENDENHALL. Pp. 1-23, 3 plates. No. 2.—Claytonia Gronoy, a Morphological and Anatomical Study; by THEopoRE Hotm. Pp. 25-37, 2 plates. No. 3.—A Research upon the Action of Alcohol upon the Cir- culation; by Horatio C. Woop and Danirt M. Hoyr. Pp. 39-68, 3 plates. 2. The Geological Society of America.—The eighteenth win- ter meeting of the Geological Society will be held at Ottawa, Dec. 27-29, in the House of Commons Building’; this is by invi- tation of the Logan Club of the Geological Survey of Canada. President R. Pumpelly will preside. The Cordilleran Section of the Society will meet at Berkeley, Cal., Dec. 29, 30. 3. A Laboratory Guide in Bacteriology ; by Pau G. Heine- MANN. 143 pp. 1905 (The University of Chicago Press).—This little manual of 145 pages contains clear and concise directions for a thorough course of laboratory work in the subject, includ- ing the preparation of culture and staining media and the col- lection, isolation, and method of studying the different groups of bacteria. The course, as outlined, is that pursued by the medical students of the University of Chicago. There are descriptions and illustrations of practically every piece of apparatus used in the laboratory. Between each two pages is a blank sheet for notes and additions to the text. W. RB. C. 4. British Tunicata ; by AtpER and Hancock, edited by the Secretary of the Ray Society. Vol. I. Ray Society, 1905. Pp. 146, with 20 plates.—This long delayed monograph, that was begun in 1855, has now made its appearance, more than thirty years after the death of both the authors. The entire work will be completed in three volumes and will contain descriptions and colored illustrations of all the British tunicates known up to the year 1873. The present volume contains a general account of the anatomy, physiology and relationships of the class Tunicata, together with extended specific descriptions of the thirty indige- 470 Scientific Intelligence. nous species of the genus Ascidia. Nearly all these forms are illustrated by beautiful colored drawings by the authors. There are also many anatomical figures. W. R. C. 5. Catalogus Mammalium tam viventium quam fossilium a Doctore K.-L. Trovgssart. Quinquennale Supplementum (1899- 1904) Fasciculus IV. Pp. vil, 753-929, Berlin, 1905 (R. Fried- lander & Sohn).—This part completes the Supplement begun in 1904 (this Journal, xviii, 95) and gives the volume contents and index. It includes the Cetacea, “Edentata, Marsupialia, Allo- theria, Monotremata. 6. Car negie Institution of. Washington. — Recent publica- tions, not before announced, are as follows: No. 9.—The Collected Mathematical Works of GrorcE Wittam Hirt. Volume one. Pp. xviii, 363. With an intro- duction by M. H. Poincaré and a portrait (frontispiece). No. 35.—Investigations of Infra-red Spectra. Part I, Infra- red Absorption Spectra ; Part IL, Infra-red Emission Spectra ; by Winuras W. Costenrz, 331 pp., 152 figures, 3 folded plates. No. 36.—Studies in Spermatogenesis, with especial reference to the ‘‘ Accessory CHromosoure *; by N. M. STEVENS) 303 ppeeg plates. No. 37.—Sexual Reproduction and the Organization of the Nucleus in certain Mildews; by R. A. Harper. 104 pp., 7 plates. No. 41.—Traditions of the Caddo; collected under the auspices of the Carnegie Institution of Washington; by Grorce A. Dor- SEY. 136 pp. 7. A Handbook of the Trees of California ; by Attcre East- woop, Curator of the Department of Botany. Occasional Papers of the California Academy of Sciences, IX. 86 pp., 57 plates. San Francisco, 1905.—The. scope of this work will be seen from the following statement quoted from the preface: ‘The aim has ~ been to prepare a work which, while giving all the information necessary for the identification of the different trees of our val- Jeys and mountains, will be so brief and concise that the entire matter can be put into a book that can be carried into the field.” The description of species are quite brief, but are well supple- mented by a series of 57 excellent plates. OBITUARY. Professor Drwirr Bristoi. Brace, head of the Department of Physics in the University of Nebraska and author of numerous physical papers, died at his home in Lincoln, Nebraska, on Octo- ber 2, in his forty-seventh year. Professor RatpH Copetanp, Astronomer Royal of Scotland and Professor of Astronomy in the University of Edinburgh, died on October 27, at the age of sixty-eight years. Dr. W. von Bezoup, Professor of Physics and Meteorology at the University of Berlin and Director of the German Meteoro- logical Institute, died early in October, in his sixty-ninth year. ENDEX TO.NOLUME XX;* A Academy, National, New Haven meeting, 468; publications, 469. Aldrich, J. M., catalogue of No. American diptera, 77. Allen E. T., thermal properties of feldspars, 72. Alpen, die, im Eiszeitalter, and Briickner, 407. American Museum Journal, 77 Arizona, petrography Mts., Guild, 3153. Ashley, R. H. estimation of sulphites by iodine, 15. B Bacteriology, Heinemann, 469. Bailey, E sis, 461. Barus, C., groups of efficient nuclei in dust-free air; 297 ; ions and nuclei in dust-free air, 448. Baskerville, C., phosphorescence of zine suiphide, 93; action of radium emanations on minerals, 95. Becker, rock cleavage, 407. Beiknap Mts., N. H., son and Washington, 344. Benton, J. R., properties of catgut musical strings, 583, 462. Bieres, Fabrication, ' Lévy, 168. Blondlot’s Emission pesante, 400. Bolivia, mastodon remains, Pom- peckj, 465. Boltwood, B. B., ium in radio-active minerals, 50 ; radio-active waters, Hot Springs, Ark., 128; disintegration products of radio-active elements, 253 ; pro- duction of radium from uranium, 239. BOTANY. Croomia pauciflora, Holm, 50. Cyperacez, studies in the, Holm, No. XXIV, 301. Trees of California, Eastwood, 470, Moreau British Museum, Catalogue of Birds’ | Eggs, Oates, 412. * This Index contains the general heads, Penck of Tucson H. S., Qualitative Analy- | geology, Pirs- | and | radium and uran- | Bronson, H. L., effect of tempera- ture on rate of decay of radium deposit, 60. Brown, T. C., Lower Tertiary fauna of Chappaquiddick Is. , 229. Briickner, E., die Alpen im Eiszeit- alter, 407. | Bush, K. J. tubicolous annelids from Pacific, 75 C Cady, H. P., Qualitative Analysis, 461. |Campbell, H. D., Cambro-Ordovi- cian limestones of Virginia, 445. Canada geol. survey, 1904, 464. Canadian Rockies, glacial studies, Scherzer, 80. | Cape Colony, Geology, Rogers, 163. Carnegie Institution, publications, 80, 411, 470. |Carolinas, tin deposits of, Sterritt and Pratt, 75. Cathode, rotating, for estimation of cadmium, Flora, 268, 392, 455. Chemical Arithmetic, Wells, 399. |CHEMISTRY. Actinium, 319. | Aluminium, iodometric determin- ation, Moody, 181. Bromides, typical hydrous, Kreider, 97 gases from, Debierne, Cadmium, estimation of, as sul- phate, Flora, 268; as chloride, Flora, 392; as oxide, Flora, 457; by rotating cathode, Flora, 455. Ferric sulphate solutions, hydroly- sis of, Recoura, 320. Gold, separation from platinum metals, Jannasch and von Mayer, 320). Helium from radium emanation, Himstedt and Meyer, 399. Hydrogen, liquid, and air calorim- eters, Dewar, 152; nascent, dif- fusion through iron, Winkleman, 400. Iron, new compound, Hauser, 460. BOTANY, CHEMISTRY (incl. chem. physics), GEOLOGY, MINERALS, OBITUARY, RocKS, ZOOLOGY, and under each the titles of Articles referring thereto are mentioned. 472 CHEMISTRY continued. Neon and helium in the air, Ram- say, 65. — krypton and xenon, new method of detecting, Valentiner and Schmidt, 462. Nickel, new reagent for, Tschugaeff, 397. Nitrosyl fluoride, Ruff and Stauber, 460. Ozone, formation by ultra-violet light, Fischer and Braemer, 397. Precipitates, handling of, Gooch, Hie Radium. See Radium. Solution, new heavy, Duboin, 519. Strontium, atomic weight, Rich- ards, 461. Sugar, determination with Fehling’s solution, 320. Sulphites, estimation Ashley, 18. Thorium, radio-activity of, Sackur, 5 by iodine, Zine sulphide, phosphorescence, Baskerville and Lockhart, 95. Chemistry, Engineering, Stillman, 98 Inorganic, Gooch and Walker, 66. Physiological, Long, 399. Qualitative Analysis, Bailey and Cady, 461. Cincinnati. See Observatory. Cleland, H. F., ral bridges, 119. Conn, H. W., protozoa of Connecti- cut, 76. Connecticut, clays of, Loughlin, 408. Crinoids, Helderbergian, of New York, Talbot, 17. Cumings E. R., Fenestella, 169. Cushing, H. P., geology of Little Falls, N. Y., 156. formation of natu- development of D Dalyo wk as, granites, 189. Darton, N. H., age of Monument Creek formation, 178; geology and water resources of Central Great Plains, 70. Davison C., Recent Earthquakes, 163. Day, A. L., thermal properties of feldspars, 72. secondary origin of INDEX. Dewar, J., liquid hydrogen and air calorimeters, 152; thermo-electric junction for determining ues tures, 153. E Earthquakes, Recent, Davison, 163. Eastwood, A., Trees of California, 470. Electricity, side discharge of, Trow- bridge, 57. Electrolytic dissociation Talbot and Blanchard, 398. theory, F Fairchild, H. L., a fallacy, 164. Feldspars, see MINERALS. FitzGerald-Lorentz effect, Morley and Miller, 67. Flora, C. P., as sulphate, 268; do. as chloride, 392 ; do. as oxide, 457; do. by rota- ting cathode, 455. Forestry, report for Minnesota, 1904, 167. Fossils, see GEOLOGY. G ice erosion theory, Gardiner, J. phy of the Maldives and Lacea- dives, Vol. ii, pt. iv, 77. Gas mixtures, Spectroscopic Analysis, Lilienfeld, 67. Geikie, J., Structural and Field Geol- ogy, 408 GEOLOGICAL REPORTS and SURVEYS Canada, 1904, 464. Cape of Good Hope, 1904, 406. Indiana, 29th annual report, 322. Towa, Vol. xv, 1904, 463. Louisiana, bulletins, 328. New Jersey, 1904, 828. United States, 69, 402. Geological Society of America, Ottawa meeting, 469. GEOLOGY. Baptanodon, osteology, Gilmore, 403. Bingham mining district, Utah, geology, Boutwell, Keith and Emmons, 466. Cambrian faunas of India, cott, 404, 405. Wal- estimation of cadmium — G., Fauna and Geogra-. INDEX. GEOLOGY continued. Cambrian, Lower, fauna of, Portu- gal, Delgado, 159. Ceratopsia, two new, Wyoming, Hatcher, 413. Chazy limestone, fauna of, Ray- ‘mond, 393. Clays of Connecticut, Loughlin, 408. Daemonelix, origin of, Peterson, 465. Diceratops, restoration, Lull, 420. Faults, overthrust, in central New York, Schneider, 308. Fenestella, development, Cumings, 169. Fossil Invertebrates, Catalogue, U.S. Nat. Museum, Schuchert, 405. Geologic map of Tully Quadrangle, Clarke and Luther, 158. Geology of Bingham mining dis- trict, Utah, Boutwell, 466 : of Central Great Plains, Darton, 70; of Little Falls, N. Y, Cush- ing, 106; of Littleton. New Hampshire, Hitchcock, 161 ; Watkins and Elmira quadrangles, Clarke and Luther, 157. Glacial conglomerate of Africa, Mellor, 107. — studies, Canadian Rockies, Scher- zer, 80. Glaciation of the Green Mts., Hitch- cock, 166; of New Zealand, An- drews, 464, South Glacier, Delavan lobe of Lake Mich- | igan, Alden, 409. Graptolites of New York, Ruede- mann, 406. Helderbergian crinoids York, Talbot, ive Ice erosion theory, fallacy of, Fair- child, 164. Limestones, Cambro-Ordovician of Virginia, Campbell, 445. Martinez group, paleontology, | Weaver, 199. Mastodon remains from Bolivia, 465. of New Mesozoic plants from Korea, Yabe, | Monument Creek formation, age of, Darton, 178. Natural bridges, land, 119. Paraphorhynchus, pod, Weller, 160. new. brachio- Peat bogs of Switzerland, Frith and | Schroter, 162. of | | Hobbs, W. H., formation, Cle- 473 GEOLOGY continued. Rock cleavage, Leith, 406, Becker, 407. Sympterura minveri, ophiurid, Bather, 160. Tertiary, Lower, fauna of Chappa- quiddick Island, Brown,, 229. Devonian Thalattosauria, California, Mer- riam, 161. Toxochelys, new Niobrara, Wie- land, 325 Triassic system in New Mexico, Keyes, 428. Triceratops prorsus, mounted skel- eton, Schuchert, 458. Turtles, marine, new, Wieland, 320; Upper Cretaceous, Wie- land, 430. aonides origin of No. American, White, 160. Vaileys, hanging, Russell, 165. Geology, Economic, Journal of, 467. —of Cape Colony, Rogers, 163. —Structural and Field, Geikie, 408. Gilmore, C. W., osteology of Bap- tanodon, 403. Glaciation, Glaciers, GEOL- OGY see ‘Gooch, F. A., handling of precipi- tates for solution and reprecipita- tion, 11, Inorganic Chemistry, 66. Graton, L. C., purpurite, new min- eral, 146. Green Mts., glaciation, Hitchcock, 166. Guild, F. W., petrography of the Tucson Mts., Arizona, 313. H Harrington, B. J., apparatus for vapor densities, 220. Harvard College, see Observatory. Hatcher, J. B., two new Ceratopsia, Wyoming, 413. | Heinemann, P. G., Laboratory Guide in Bacteriology, 469. Hidden, W. E., cassiterite, a new cleavage, 410. Hitchcock, C. H., glaciation of the Green Mts., 166. origin of channels around Manhattan Island, 71. Hofmeister, F., Beitrage zur chem. Physiologie, 168. Holm T., Croomia pauciflora, 50; studies in the Cyperacee, XXIV, Carices from N. W. America, 301. 474 Hough, R. H., mechanical equiva- lent of heat vaporization of water, 81. Howorth, H. H., Ice or Water, 166. I Ice or Water, Howorth, 166. Iddings, J. P, optical study of the feldspars, 72. India, Cambrian faunas of, Walcott, 404, 405: Indiana, geol. survey, 322. Ions and nuclei in dust-free air, Barus, 448. Iowa, geol. survey, 1904, 463. J Jamieson, G. S., tychite, 217. Japan, magnetic survey, Tanakadate, 167. K Kayser, H., Handbuch der Spectro- scopie, 69. Keyes, C. R., Triassic system in New Mexico, 428. Kilimandjaro to Meru, Africa, Uhlig, 78. Korea, Mesozoic plants from, Yabe, 406. Kreider, D. A., iodine titration vol- tameter, 1 Kreider, J. L., typical hydrous bro- mides, 97 L Laboratoire del’ Usine, Houllevigue, 168. Lakes, A., Geology of Western Ore Deposits, 409. Landolt-Bornstein Physikalisch- Chemische Tabellen, 401. Leith, J. B., rock cleavage, 406. Littleton, N. H., geology, Hitch- cock, 161. Lockhart, L. B., action of radium emanations on minerals, 95; phos- phorescence of zine sulphide, 93. Long, J. H., Physiological Chemis- try, 399. Loughlin, G. F., Clays of Connecti- cut, 408. Louisiana geol. survey, bulletins, 320. Lull, R. S., restoration of Dicera- tops, 420. Luquer, L. Mcl., Minerals in Rock Sections, 468. INDEX. M Magnetic survey of Japan, Tanaka- date, 167. Maldives and Laccadives, Fauna and Geography, Gardiner, 77. Manhattan Island, origin of chan- nels around, Hobbs, 71. Manila, ethnological survey, publi- cations, 167. Mastodon remains from Bolivia, Pompeckj, 465. Mellor, E. T., glacial conglomerate of South Africa, 107. Merriam, J. C., Thallatosauria from California, 161. Metals, emission of negative corpus- cles by alkali, Elster and Geitel, 461. MINERALS. Beckelite, 323. Bowmanite, 324. Cassiterite, new cleavage, 410. Diamond, formation, Crookes, 460. Enargite, 280. Feldspars, thermal properties, Day and Allen, 72; optical study, Iddings, 72. Hematite, Na , 289. Hutchinson- ite, 324. Lengenbachite, 524, Luzonite, 277. Marrite, 324. Northupite, 217. Platinum in black sands, Day, 410. Purpurite, 146. Quartz, San Diego Co., 125. Riebeckite, genesis, 133. Smithite, 824. Sylvanite, 282. Tin deposits of the Carolinas, “Pratt and Sterrett, 75. Trech- mannite, 324. Tychite, 217. Wolframite, Boulder, Colo., 281. Minerals, action of radium emana- tions on, Baskerville and Lockhart, 95. —in Rock Sections, Luquer, 468. — optical character of birefracting, Cal., Wright, 285. — radio-active, Rutherford and Bolt- wood 5a; Strutt, 68. Montana, Agricultural College, Science Studies, 78. — petrography of Central, Pirsson, D0. Moody, S. E., iodometric determin- ation of aluminium, 181. Murgoci, G. M. , tiebeckite and rie- beckite rocks, 133. Musical strings, properties of, Ben- ton, 388, 462. INDEX. N Negritos of Zambale, Reed, 167. New Hampshire, geology of, Pirs- son and Washington, 344. New Jersey geol. survey, 1904, 323. New Mexico, Triassic system, Keyes, 423. New York, graptolites of, Ruede- mann, 406 — Helderbergian crinoids, Talbot, 17. — overthrust faults in central, P. F. Schneider, 308. New Zealand, glaciation, Andrews, 464. Nobel prizes in 1902, 167. Nuclei, efficient, in dust-free air, Barus, 297. O OBITUARY. Von Bezold, W., 470; Brace, D., 470 ; Buckton, G. B., 412. Copeland, R., 470. Errara, L., 412. Reclus, E., 412. Von Richthofen, F., 412. Observatory, Cincinnati, 411; Har- vard College, 411; Mt. . Wilson, Cal., 80; West Hendon House, 80; Yerkes, 411. Optical character of birefracting minerals, Wright, 285. Ore deposits, Geology of western, Lakes, 409. EP Pacific, tubicolous annelids from, | Bush, 75. Paleontologia Universalis, 406. Penck A., die Alpen im LEiszeitalter, 407. Penfield, S. L., tychite, 217. Peterson, origin of Daemonelix, 465. Physikalisch-chemische Tabellen, Landolt-Bornstein, 401. Physiologie, Beitrige zur chem. Hofmeister, 168. Pirsson, L. V., petrographic pro- vince of Central Montana, 35; | geology of New Hampshire, 344. Portugal, Lower Cambrian fauna, Delgado, 159. Pumpelly, R., Explorations in Turk- estan, 245. Q Quartz apparatus for laboratory pur- | poses, Mylius and Meusser, 66. — permeability to gases, Berthelot, 66. 475 R Radio-active elements, disintegra- tion products, Boltwood, 253. —minerals, Rutherford and Bolt- wood, 55; Strutt, 68. —waters, Hot Springs, Ark., Bolt- wood, 128. Radio-activity, absence of excited, due to temporary exposure to y- rays, 68; of thorium, Sackur, 695. Radium, action of emanations on minerals, Baskerville and Lock- hart, 99. —effect of temperature on rate of decay of deposit from, Bronson, 60. —emanations, helium from, Him- stedt and Meyer, 399. — production from uranium, Bolt- wood, 239. —slow transformation Rutherford, 321. —and uranium in radio-active miner- als, Rutherford and Bol*wood, 55. Ramsay, neon and helium in air, 69. Raymond, P. E., fauna of Chazy limestone, 353. products, ROCKS. Andesites, Arizona, Guild. 314. Basalt, Arizona, Guild, 316. Belknap Mts., N. H., Pirsson and Washington, 544. Granites, secondary origin of, Daly, 185. Petrographiec province of Central Montana, Pirsson, 35. Quartz-basalt, Arizona, Guild, 318. Rhyolite, Arizona, Guild, 513. Riebeckite rocks, genesis, Murgoci, 135. Rock cleavage, Leith, 406; Becker, AQT. Rogers, A. W., Geology of Cape Colony, 163. Rontgen rays, Hahn, 461. Rutherford, E., radium and uranium in radio-active minerals, 55); slow transformation products of radium, del, charging effects, S Shaller, W. T., purpurite, a new mineral, 146. . Schneider, H., Soils and Fertilizers, 398. | Schneider, P. T., overthrust faults in central New York, 308. : 476 INDEX. Schuchert, C., Catalogue of Fossil Invertebrates, U. S. Museum, 405 ; mounted skeleton of Triceratops prorsus in U. 8. Nat. Museum, 408. Sedgwick, A., text-book of Zoology, 76 Smithsonian Institution, annual re- port, 412. Soils and Fertilizers, Schneider, 398. South Africa, glacial conglomerate, Mellor, 107. Spectral lines, influence of character of excitation upon structure, 68. Spectroscopie, Handbuch, Kayser, 69 Stillman, T. B., Engineering Chem- istry, 398. Switzerland, peat bogs of, Frth and Schrotter, 162. T Talbot, M., revision of the New York Helderbergian Crinoids, 17. Taschenberg, O., Bibliotheca Zoo- logica, IT, 412. Temperatures, thermo-electric junc- tion for determining, Dewar, 153. Trouessart, E. L., Catalogus Mam- malium, 470. Trowbridge, J., side discharge of electricity, 57. Tully quadrangle, geologic map, Clarke and Luther, 158. Turkestan, Explorations in, 245. U Uhlig, C., Kilimandjaro to Meru, Africa, 78. United States, geol. survey, 69, 402. National Museum, Report for 1908, 167. — — Catalogue of Fossil Inverte- brates, Schuchert, 405. Utah, geology of Bingham mining district, Boutwell, Keith and Km- mons, 466. V Vapor-densities, apparatus for deter- mining, Harrington, 225 Virginia, Cambro-Ordovician lime- stones, Campbell, 445. Voltaic element, normal, 68. Voltameter, iodine titration, Kreider, W Walcott, C. D., Cambrian faunas of India, 404, 405. beset C. F., Inorganic Chemistry, Walther, J., Geology, 161. Waring, G. A., quartz from San Diego Co., California, 120. Water, heat vaporization, mechan- ical equivalent, Hough, 81. Watkins and Elmira quadrangles, geology, Clarke and Luther, 157. Washington, H. S., geology of Bel- knap Mts, N. H., 844. | Wells, H. L., Text-book of Chem- ical Arithmetic, 399. Wieland, G. R., new Niobrara Toxo- chelys, 325; Upper Cretaceous tur- tles, 480. Wright, F. E., optical character of birefracting minerals, 285. Wyoming, new Ceratopsia, Hatcher, en restoration of Diceratops, Lull, ¥ Yerkes Observatory, report, 411. Z Zoologica, Bibliotheca, II, Taschen- berg, 412. Zoology, text-book, Sedgwick, 76. ZOOLOGY— Annelids, tubicolous, from the Pa- cific, Bush, 79. Birds’ Eggs, catalogue, British Mu- seum, Oates, 412. Cold Spring Harbor Monographs, (ish Diptera, No. American, catalogue, Aidrich, 77. Fauna of Maldives and Laccadives, Gardiner, 77. Goat, Rocky Mt., Grant, 77. Insects, Habits and Instincts, Fabre, 77. Mammalium, Catalogus, Troues- sart, 470. Protozoa of Connecticut, Conn. 76. Tunicata, British, Alder and Han- cock, 469. MERELY A REMINDER THIS PAGE is occupied by us merely as a reminder that the only concern in America en can supply you with SPECIMENS IN ALL DEPARTMENTS OF NATURAL HISTORY (Except Botany and Entomology) IS Ward’s Natural Science Establishment To our old friends, the leading museums, colleges, educators and private collectors of the world, we are continually offering new things, some of which will be announced at the bottom of this page, from time to time. We endeavor to always remember you with our latest circulars; if you have failed to receive these recently, a line from you will bring them by return mail. To those who have not formerly dealt with us—our establishment was founded in 1862, and incorporated in 1890 as a stock company with a paid-up capital of $125,000. We occupy a frontage of 250 feet facing the University of Rochester, and are known from the Yukon to the Ganges as the largest institution in the world dealing in Natural History Specimens. For over forty years we have made it our sole business to collect, prepare and sell these, individually or in collections. Quality rather than extreme cheapness is our aim, and we have spared no expense to maintain a high standard and a ‘standing in scientific circles. We pay no commissions, but deal direct with our customers, and sell at list prices only. We offer school collections as low as $5 and have made one cabinet costing over $100,000 (Field Columbian Museum, Chicago), and seventeen others ranging from $10,000 to $70,000. In numerous instances we have built a large public museum complete at one stroke. Our catalogues, over twenty in number, are not mere price-lists, but are valuable as reference works, and have even been used as text-books in academies and colleges. A small charge is made for these, except to our regular customers or teachers intending to purchase; a list will be sent upon request. We also issue free circulars in all departments, and shall be glad to place your address on our mailing list. OUR DEPARTMENTS. Mineralogy (Minerals, Rocks, Crystal Models, Meteorites, etc.). Geology (Phenomenal Series, Relief Maps, Geological Models). Palaeontology (Fossils, Casts of Celebrated Fossils, Wall Charts, etc.). Archaeology and Ethnology (Specimens, Models, Casts of Monuments). Inveriebrates (Shells, Echinoderms, Corals, etc.; Biological Supplies). Zoology (Mounted Skins and Skeletons, Custom Work in Taxidermy). Ga” Anatomy (Human Skeletons, and Anatomical Models of all inds). SPECIAL ANNOUNCEMENT FOR THIS MONTH. Out of Print.—Our well-known “‘CATALOGUE OF MOLLUSCA’”’ will shortly be out of print, less than too copies remaining. This elaborate book of 150 pages and 200 illustrations will not be reissued. Send forty cents for a copy before it is too late. Ward’s Natural Science Establishment, 76—104 COLLEGE AVENUE, ROCHESTER, N. Y. CONTENTS. Page, ART. rou. New Ceratopsia from the Laramie of Converse County, Wyoming; by J. B. Harcuer. (With Plates: XT (X11) esi Ava” XLIL—Restoration of the Horned Dinosaur Diceratops ; by Rrcwarp 8..LuLu. (With Plate XIV.) >> 1a) seen XLIIi.—Triassic System in New Mexico; by Cuartus R. KEyESe. O02 So ee ee 423 XLIV.—Structure of the Upper Cretaceous Turtles of New Jersey: Agomphus; by G. ‘R. Winianp.__. 2 ee 430 XLV.—The Cambro-Ordovician Limestones of the Middle > Portion of the Valley of Virginia; by H. D. Campprtn 145. XLVI.—Relations of Ions and Nuclei in Dust-free Air; by OARL- BARBUS 20 eo ee ee 448 XLVII.—Additional Notes upon the Estimation of Cad- mium by Means of the Rotating Cathode, and Summary; by Cuarenns PP). Prora:. 0/22 ee 2 oe ee 454 CHARLES P: FrORA =< = 2: OL. BE 2S eee XLIX.—The Mounted Skeleton of Triceratops prorsus in the U.S. National Museum; by C. ScaucuHenrt. (Wak : Plate: XV.) \s.2 G2 Se 458 SCIENTIFIC INTELLIGENCE. Chemistry and Physics.—A New Formation of Diamond, Sir W. Croomae A New Compound of Iron, Orro Hauser: Nitrosyl Fiuoride, Rurr and STAUBER, 460.—The Atomic Weight of Strontium, a: “We RICHARDS : Qualitative Analysis, E. H. 8. Barney and H, P. Capy : Charging Effect of Réntgen Rays, Kart Haun, 461.—Emission of Negative Corpuscles by the Alkali Metals, J. J. THOMSON : A New Method of giowing the Presence of Neon, Krypton, and Xenon, S. VaLEentTiner ahd R. Scumipr: The Mechanical Properties of Catgut Musical Strings, J. R. Benton, 462. Geology and Mineralogy.—lowa Geological Survey, Volume XV, 463.—Sum- mary Report of the Geological Survey Department of Canada, R. BELL: Glaciation of Southwestern New Zealand, E. C. ANDREWS, 464. —Mastodon- Reste aus dem interandinen Hochland von Bolivia, Lk POMPECKG : Description of New Rodents and Discussion of the Origin of Daemonelix, O. A. PETERSON, 465.—Economic Geology of the Bingham Mining District, Utah, 466. —Economic Geology, 467.—Minerals in Rock Sections, L. MclI. LvQueR, 468. Miscellaneous Scientific Intelligence.—National Academy of Sciences, 468.— The Geological Society of America: A Laboratory Guide in Bacteriology, Paun G. Hetvemann: British Tunicata, ALDER and Hancock, 469.—Cata- logus Mammalium tam viventium quam fossilium, HE. L. TROUESSART: Carnegie Institution of Washington: A Handbook of the Trees of Cali- fornia, A, Eastwoop, 470. Obituary.—Professor DEwirt BRistoL BRacE, Professor RALPH ‘COPELAND, Professor D. W. von Brzoxp, 470. INDEX TO Vou. XX, 471. TEN-VOLUME INDEX. An extra number, containing: a full Index to volumes XI to XX of the Fourth Series, will be ready in December or January. Sent only to those who __ Whe specially order th PE ee order it. Price one dollar. XLVIIL—The Estimation of Cadmium as the Oxide; by | Cy 9% Md saan 3) } ad ENDO Pew ot a . i Hi, aa is a q . - on - : ne ou | : - ee em ee ham a na ‘WA Q 3 9088 01298 571