Ho iD ie pat Nas ii tle thn Hn Pe ath a Oras NG- mate AR Hp A nt Nl Mee Mt tte KN ote Lh 4 a 2 Ace Darr Manat Seabee HA ad Reel Datla # «l= tallah! «Hata tint alee =~ eMail Mn Og RD A ihe t= “rl Ppprieeetrerinere se aranar ire pone ee tetany Maha! 8 -ttigibo Balt 1 uh Menten has Bee GBS esha Inga Moin stat ein teeth) hel patina i eave nae th ne Of 6 tome law Pacts Alon ig Henn Mantes ~~ - wo Rae 24 De eo tn” te Ry tines Hnihn me ae he eta Mee le ath ete ate rae ee ead Pte ee It he a i i MAL He ae tah ’ a Ceara (Se ee Seat A hit gh aA Mn Petal est eos Fa the eG tT 8 ete te i it te A owe tan Rs meee. Hn eel ie te ithaithn Mn AataihCaitnMn Ate te ath i tig Mell aA Re, = nnn Sw Ani Pan =r ag FACT Daa NA tll AMIR lm te Me Ord Na Ph ty Pn Ns MAL he MA ors SA Ma acta Mn ie Maia Mert Aimache HS ank th Nor 61% ee ee ee ipaaantet a neh Som Neth als ninmtn Bh Ao Rares nine Walt RED tbl Bae i inMe iat Me Bhan Ma ~~ Din - =e" nah, “me nit Parca TTR thn tm esomt Rnd Re eA At Ratha te Rtg et Raia fete eRe iAB Fini? Maite! Tn a Rag aN pacha Retell, Sollonthw teeth Ta th Big he Pattee tag B Mahan a ein fe Bh Poon nme er eae Safe i Ag Aj "Re a mn Pt A Tet e Me Moe x ce aT cM i ct heen eh Pin Ae Me Be te ee Reine em Mt CAN rt a ta ata aM ain oe APSARA ah Ra tnt WMT M Ae tind to Pa ARAM ALY ee eS ee Ee ete ee eg Ab aan lite Oe Aen tn Canin Mei Pan tae Ue then athe erates eh Naat ihe Ut me Binsin ree er ee ee ee ett Alla lt MN thine the PR Ree MT TO il, MMM Rai en Tare Ate ee ray Rte Beit ALAM nate l Poo Nie nih tna atta set ae AAA DBAS R AS 8 RNAS ee ee ee ee cipher a Degli tem ce Nth My Dts Mt i tlh lp tht (a i ht Ma Min te Pern a Ce ee ee ea St athe in naan stir aM nN HSI Natt B De 6 ete eat eh Hei te Me eh Na Tine eile ae are ethan, ato Bre - We eh Ree etek th aie haan Naas i afin Bn sO Cen Pella ta Aholha Seta Watt od Ab Faire tn" se parr are ea ee eee eee ee Pan ew BaP th gs ng Ph Mathes ee Picts he Fin Ae eh tet Rc TAL Ri es Meat malay rt RAIN An reine Real tatM AU MAD tar Yo Me PME Tt Netw he leat Phe AN th ith AN Ae Bint he hy Sema Aen inthe Fh tht inf VAL A Uric a hte an in rasan acheter ain Bala Feb me te gale Mortage Oe 1a he Gi thee Daathnth Mee att ike eer Hae Mathes Peete A Pe eee aaa ha = Mlle ie a ir te Neti Ne A eh ton etl Bice Me ha Patni tM ee Nem Mt Net ete ue Mle eR Ne hee en Raat lhe Rollin MD ail RAs Me Aas Malet in a Mle ig Mo Be Sn Pte Maint Ne ha tg ne hak APD, Dateline Rathea™ a? Liter Hatee Rae maarithe RatMeM, Rai yfrutinha tn ta en- epereeenee ar en ne sn Ne Min hs hn Gan ta A> he Bas mt i A 2 a Tatts Plt fe trate stefan Meier LR Neto tN A Rin AAA Fl Pape ng: Lan tne the SNL hn ON ttn, ea Tehibe aBRNAM Sitm Ahead MAA rela nage Bat AE Ying Bans te Mant OMe pe eee ero See nee ALA the as ia thins MD ain Anthea goat i ¥— inn = neta 6) oT eMetbetat othe! sa. Whe fo rhe than te a he ln nee the en Artin en at me a tng Pat oh lh he “et Mes tn Tan Aenea Ram Sh ee te Mle Me A ay he eR Mah. GO Se Athan fg ae ee eee RN he Fe ome AMPED AM Nace Nae Aas Me Ni Ni IN te AN Ot Me aia Alt ae Mt = ee ee rarer wre era arn see Sgt inline Reha Bad 3 Ament teeta at atatamata? A~ 7 ed dh, righ Retin NB ig A Mien “GH alt i a lt holes digest, A Maan Nn Re Pn ting Hm ee ed RA A a tt Me te saat Retr arate aie Vacre INTRA a eon Mae PIR Nene Re tinct HE i Ratton a a ee ethan Bn Pas mn NAAM Marte Bey is ea Nhe oot bata Rvttattng ttn =a torn tatsaDonet« * te tee? P= eg tetas! ne ae BPs 18 te Nn Ratan Phy Fete Ra ae He aati tie tnt ae he i et Me ee Boerne nt Ae te Ne Te Re EN Mo tte © ee ee ae ee Nl, a tte ie, eh Gene ta, ee ee eg en Pane We ee a o> BA err ae ne ee ee er er oe i a er ee ee eee ete te nem oy Ap te nT tis at Deat al Aalto Mende IM, ee eed “aaah Phat ra dtclivat Ae. ghee Aa attty At Ay Pc A Malan, Fs Sore ee ee Wh Thee teh in Menara ha tailor Tha Daring Heth has Aa mM li thm fh ha Ta, es nha rin Pn te ns sete ewan ne aw Fe tottwt fin Param bo Tete! Ta me Naoto: hao Ren Ma a ht tle hella A Mia My, Mf lite ae PN eal Me Meten tnhn. Sef lie Be Pe We re ne ee ede amet a oe PA: TN HH Sc A Rat tO te ch Rett ne A HA Saat} Saale Bg ee ee Oe Bo SP ae TA at: eRe lla ig th he ele ain Nant on tee le a8 wike hue! ot a Pee Se ‘pete Biba, De A pee witli Ap Naw Fina ie ew 8 Me hth Mae ng Wah -NILF th BA ALTA et etait he et de PMc dee SL OL Fh Pe Ne th alge SM hg te ot Mt ia Tine eee Maes the a ath ata tee AM BERR nw AT ee 9 Oh ee Fa hg ei ln Ph Ot Ma Mn tg Mm me that NeoPaint: hiiten ow ey ee Fe a a ve er! we ee ens a pre ee ee ee Aelita Me Mert RPh Daath Lang tine ae Mele van Samet Ne Puc g Rotate baMaroliag Qn fs M2” atte Bn gate tee a Ball neta lam pte! ome stig. a9’ he ne a Ne et le Sn Pe was The! ee Se Bab AR Mab A'S NedleAetanel A Mane a aN PE IR A cert They shen oP hath RANI AN grt ARAN TR Aa a AL Toa Att I Bs Nets AN IRIN et lh Mag tg ih Phe gia ated diy) 4h * 1) dt a healthy Ratt ee: 7 Aen Meh We aa Ra ay ath M0 nt a lied te a Matton oR HM hile Hm oe A ee a ee ed i an: te ee te etl then A Meme he ke anh PG Weiner Bie He tat a RR Meg en a wine Eade tad eS Meta Rei RAM Ve A TE DM Pst hn Rae tte MAHI lll Pattie agi Rema Me Ste ag Me AM + ata ee ey ln Me aD eg eee Me athe ta Peet ol af a tet ste — ati a tan.” Me Ahn he tiadg In” Oe eee eee ~~ ee Pai te ts > Le Paani he aM, OR Ry. UM Rta Rit PA athe Maney 7 Homo ag IAL ED iin eis Hotes ~My Mime ela PM As? nla ig thn My Migtten ga Maken Me ~* PL te tly i Bn an Htimnen on. De TE Mant 0 ee Mat oat Mea! Batter saabnc hs oa Py Mla cae i Pee! Orth dite | Aan Mn Pont ne et ee ee et et te Te ner RMR le tee ete Rm ne ln, eal: Pn SS ee A eal RAN SU Fic Dinl he eR: Gayle ee ee eA EK, «Earn, Be testy ae ta "nln Mn nd ar eeerer eS wre Ed Ag Ree Be a ON Retin ity te Matera et. Viton tae Me nn BaP ah ut “wo cine “na a Ni “te et Be tll Hy he he in ig, Pn eI Rag ss a tm Serer ae WIC Se ee RP. Pate than Rete i Mp Dotan te re ee ree Ca “de AR AA AN op tn Rh EM ng NA + Brn sa Recta eg tat, Nein Hh pathic May hi Mette bar Aay Se ee MM Reale Te Al Mes gt hens neg tanta net = mis ra ratlnts G2 * tet to Pear arene ss The tae atte Palette Meet, mn Nae tae SIRS Ne an A tk it beth Merete No en Se cel tal enc am gm eae Nah Piet PA si Mi TRAM esa Ac tea AAs War hes Ne eee ee ar ae Ribrtriiirard Ian ee Sethe r= SN ha eet le Fae eta a tothe Paes sa A er Ran ome EA Dig emg art Ran gh Manin Te a Re tin thin" Pam Min at ee en eee dP ct tm tt Magara eta M nite AN AIO i Ae NNR A fle Ee it PA ase NC a be Mon Min leet te» A A Pg ALAN Mag DI Ee Rly Mem gto nn Mat a aA Pe ig a Ai A aaah Raga PM ng Sel tn ai + wt Me eA dee We en iano WW ee ee Ss Me Ne ta ee TL Tee Nae te te ha tee a ee ER, tt eS A a oe ee AME Ain ese Rag = ey Om a et oe ee ee AA Rie PP sn aA My De ee Be AO A eta ee rel Me Set ate Ae Oe Re te te oe he é a ee ey ee ret ag Ml A ae ag Ms gh i NB Bane eh Ms Te de leo EW Bele tote he ee oh ee ee le re ee ee ee ee PO ee ear, er AH ey er te de a AR LR 8 He ae ee ee aH” be th ee Maen X ee ee ee es . ee ne ee i Ne ig Me ie Mt ee a ee ee Peck tth Prager ats aM ih Ty Ramat eM er ee li sh AN A le Re TM Ath “ate Ae ie Ne, alta a me Re aT Antes Pets A oe wee ee 2a” fev alae ts eden. ery eee DDT Bee he ME te Pi rg ay Ht, Rea rt OR Fm te te * i a a Ret ti he 1 ee eR, eh hg eM eM ine RTL tn Hine oe Me Ny ett it a ltormie (teeem A yA ee AT Peet eesubeante: ie Maat Dem tne fats hath. Oe TeaRa tale tetintinns AewiMe tertgteGs ste = Se Oe ~ =e ten eek AL ee ei Fe em ~ Rg ee LS ett MA Re tenet ee ee ee Bee ee eR A ee a et te eT a ba tele Tel on De tee Aen en ~ Seen ho aes Ae Wann Mg ng ha Ags en egy te Mow tn Pee am ee . a ee ae eee ee a a a ee rere a ee ears ae oe —— + Fabel, Ramee OT Ba Nt fe eB hn CPR ete het ee ed agente ete tae Ain Mee A oe tha orate: Bim San Be a we a ag ipa im Matai Me Mn ae Mars er re eee a lity pasa eal etn Res yen Ae Ran it 8 mt ARMIN PGND RIB Noe SAT Si aM Mth Bh in eet em Hom Ra oe a ame ete Sacha grt Se eM oo ve thm tte Set nae Pi ti AOA = ~ ~ ~~ we wade e® - a al en re td ih Spey Pe Sh gM yet el Rete, he Ry ds on me el om iMe!S . Ao ~~» re eee er Orme So ee Od teh Te kA oie at eee tk ay Mp em A ont te ete eh ke Re whe ras - ~~—s = es ed Mae Me ee Ma te ee NN Tay = tent Peat te Tae PI AN lM, on Ne eM ae oe a Et ee ee Yt an een ee ee ee ee Se eee, ea a * io pul » = “ote - ¢ seis me Pete ee KI Dee 6 oy Re Pe ee ee fe lg tt ate I Rte ea ee a ning Dag dee SM ne, EN es ay ea ne tat eer Re pe aah ag He she Se 7 whet tebe te fee im ante me Me ee a ee ae on gg ll Mr ee ee A tl i MEE te FM Gel oe et ee eM Sem 2 TL Tate. ae ee OP ae eo a ees Ree ARAMA Bhi” he 7 i ws Se, ee a te mL ee mF “othe he A, A. Nt. te hee i ee n = 4 * ». oe ere me \ i er a ee Ne = he Pot er ee ee ae Rae ete ee De - MMs = aoe 7 - Age + -" ee er Pe We Pentti Sl renee ow we ee ee = ke ’ hs Peete Pe Va Do a? ‘ wee —s ee ners 7 Tee aN at egy, tee Peat, Ae in Mean tT nie het athe om Sg ee eA em a “oe , ~ te CAD Mowett et ’ ea Mg Pele Big thar, “ mete ie ate “are 7 othe AAR Bate ee tia teMe ta cte = ota SR be ee oe A, eee aw Sepa ee DR eat et te ee o s -. - A,.42 a ~* - ok WHO oe ” b . tata ev te wait are eee ee ee ee = n me Mele aca QAP Be He Mee Bert ee OT Oe ~ cee het yp soften Se TURE Saba Ome ee Scat, s A — -_~- 2 ~~ “ se ae _ tigi a4 ne a a) Ay ba! — ae x a: te “a a“. Perens Pe Pe me ~ _ > - .* Lae ~™ oe! ~ - ” . - ; : tom Ni ' 4 .* i a 4 *. ~ 4 ot Ob ete * om , at Pal . = . - ~ * » a * ma fas ‘ re pe A Nat A Bat CS te er eee! eta 9 ae” Or tages oo oe tf oe - 18 « +N 4 “ ua D “ a s 4 ee ak et Mm PT Pte ea OH ee -_ ~ ~ a * ~~ r™ - re, 7 : : = ~ ‘ _ os - 7 . a a n o- . . . =) f , ie = = ‘ -. t ~ en 2 i = -‘ ‘a ~ : ~ 4 - r ~ 2 > . . = ’ = A a . , e ee oe , . . 7 nae i 7 . % 2 tet 8 my 7 ae . “ » ¥ - ’ — . -". male ‘* r a =} Pa hs ge RY '. tig? v6 x. - eres A = a. wm 4 a i @. . ; eyewares pt tA Bate etd eR fy Re! a eee 7 a ‘ * ‘ -* i= . F ra . * a ed ew patents Hien tn nat ge 2 oa wn patient —At ait A R a | dd me Gre ot TE aE “ “ ay o, we + te * ea. : ee ae ina aioe AG epraest. > i ihn 00 ny enn = emp OS enw S ieee te +o ee pee Re Pee Nabe oe ee pee, i ye a ee Py Rt oP ee Re mee “Swe eee ly, D5, 75 THE AMERICAN | JOURNAL OF SCLENCE. Epirorn: EDWARD S. DANA. ASSOCIATE EDITORS Proressors GEO. L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or Camprince, Proressors A. E. VERRILL, HENRY S. WILLIAMS anpb L. V. PIRSSON, or New HAVEN. PROFESSOR GEORGE F. BARKER, oF PHILADELPHIA, PRoFEssor JOSEPH S. AMES, or Ba.timore, Mr. J. S. DILLER, oF WasHrINncTon. FOURTH SERIES. VOL. XII-[WHOLE NUMBER, CLXIL] WITH NINE PLATES. NEM. -ELAVEN, CONNECTICUT. 19:0-1- A WWSTITUT, Liss 2) ‘dy THE TUTTLE, MOREHOUSE & TAYLOR COMPANY, NEW HAVEN, CONN. Arr. I.—Geology of the Shonkin Sag and Palisade Butte CONTENTS TO VOLUME XII. Number 67. Page Laccoliths in the Highwood Mountains of Montana; by pve nix and-By V: Presson’ = 22 02 2022 2 1 II.—Manganese Ore Deposits of the Queluz (Lafayette) Dis- trict, Minas Geraes, Brazil; by O. A. Derpy___..___-- 18 IiI.— Bituminous Deposits situated at the South and East of Gardengs, Cuba; by H: EK. Prckwam).. ._ i. 2.5422. 22- 33 TV.—Mineralogical Notes, No. 2; by A. F. Rogers ._-_-.-- 42 V.—New Solution for the Copper Voltameter; by W. K. SISINDI IE) 2 29982 eee rene he ee aL oes er eee ee: |!) VI.—Thermo-magnetic and Galvano-magnetic Effects in Welurium: by M.-G. Lioyp .--. 222.222 2.J2 57 VIl.—Additions to the Avifauna of the Bermudas with diagnoses of two new Subspecies; by A. H. Verrinn._. 64 VILI.—Induced Alternating Current Discharge studied with Reference to its Spectrum and especially the Ultra-Violet Spectrum; by A. W. Wriaut and EH. 8. Downs.-_-...- 66 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Iguition temperature of Phosphorus, EypMAN: Composi- tion of -‘Caro’s Acid,” BAEYER and VILLIGER, 73.—Vitrified Quartz, SHEN- STONE, 74.—Influence of Magnetism upon Supe: aturated Solutions, DE HEMp- TINNE: Ammonium Cyanate, WALKER and Woon: Selenium in Sulphuric Acid, JOUVE, 75.—Qualitative Chemical Analysis: Viscosity of Argon, SCHULTZE: Conductivity in Hydrogen and Carbonic Acid Gas, TowNSEND and KirKBY: E.M.F. of Clark- and Weston-Cell, JAEGER and LINDECK, 76.—Electrical Flow in Gases, STARK: Treatise on Electro-Magnetic Phenomena, LYONS, 77. Geology and Natural History—Kocene Deposits of Maryland, and Systematic Paleontology, 77.—Annual Report Geol. Survey Arkansas, 1892, 78.—Summary Report Geol. Survey Canada, 1900: Revision of Genera and Species of Cana- dian Paleozoic Corals, LAMBE, 79.—Amygdaloidal Melaphyre in Mass., Burr, 80.—Explosive Activity of Vesuvius, MatrEucct: Der Vulcan Etinde in Kam- erun und seine Gesteine, Escu, 81.—S6ndre Helgeland, Voet, 82.—Classification of Igneous Rocks, FEpEROw, 83.—Boletin del Instituto Geologico de Mexico: Brief Notices of recently described Minerals, 84.—Iron Meteorites, 86.—Milch- saft und Schleimsaft der Pflanzen. MoLiscu, 87.—Remarkable Instance of Death of Fishes at Bermuda, VERRILL, 88. Miscellaneous Scientific Intelligence—Atlas Meteorologico de la Republica Argen- tina, 88.—British Museum Catalogues: Select Bibliography of Chemistry, 1492-1897: Publications Bureau of American Ethnology: Field Columbian Museum: Scientia, 89.—Zoologisches Addressbuch: Publications Earthquake Investigating Committee ; Publications U. 8. Naval Observatory, 90. iv CONTENTS. Number 68. Page Art. [X.—Experiments on High Electrical Resistance, Part Il; by O..N. Roop - 2.00.2 25S ee a4. X.—Mineralogical Notes; by A. J. Moses .-..-.---.-_._-- 98 XI.—Motion of Compressible Fluids; by J. W. ‘Divi 107 XII.—Action of Sodium Thiosulphate on Solutions of Metallic Salts at High Temperatures and Pressures; by J. T. NortTon, JB... 2. i222) een Sc i ee XIII.—Secondary Undulations Shown by Recording Tide- gauges ; by A, W. Durr 22-2... 7.2.0 XIV.—Mathematical Notes to Rival Theories of Cosmogony ; by Ow FISHER #_ 2-2-0 2.2. Us ee) 2 oS rrr XV.—Studies of HKocene Mammalia in the Marsh Collection, Peabody Museum; by J. L. Worrman.-.._---------- 143 XV1.—Electromagnetic Effects of Moving Charged Spheres ; by Hi. P. Apams ..-- 02. Sk oe 155 XVII.—The Nadir of Tomer e and Allied Problems; by J. DEWAR 2. 252 see bo elated eh er SCIENTIFIC INTELLIGENCE. Miscellaneous Scientific Intelligence—Magnetic effect of Electrical Convection, H. PenpDER, 173.—American Association for the Advancement of Science, 174, Obituary.—DR. JOSEPH LECONTE: PROFESSOR PETER GUHTRIE TAIT, 174. CONTENTS. Vv Number 69. Page Art. XVIII.—Discharge-current from a Surface of large mutauute oye de be. ALM Ye oe. ee ig he oe 175 XIX.—Octahedrite and Brookite, from Brindletown, North Marelina by MeeneTROBINSON 2222 2850 e022. 2. 180 XX.—Behavior of Small closed Cylinders in Organ-pipes ; RETA Via Sunes os ee aoa 185 XXIJI.—Cesium-Tellurinm Fluoride; by H. L. Wetts and cl MS NAY GUTLAIATSY 25 es SS Ma ae we eR a 190 XXII.—Double Chlorides of Cesium and Thorium; by H. Meevormo ands. Me OWiILLIs. 2. f.5 22 22 tees 191 XXIII.—Studies of Eocene Mammalia in the Marsh Collec- tion, Peabody Museum ; by J. L. Worrman _.--_.--- 193 XXIV.—Separation of the Least Volatile Gases of Atmos- pheric Air, and their Spectra; by G. D. Liveine and 2. LOTSA TRS eee DERE rie aa oy ei 207 XXV.—KEstimation of Calcium, Strontium, and Barium, as Bieeoxaiates, by ©. A. bP RTERS 2 2 eS 7.2 LS ee 216 XXVI.—Calaverite; by 8. L. Penvretp and W. E. Forp 225 SCIENTIFIC INTELLIGENCE. Miscellaneous Scientific Intelligence—Temporary Set, C. BARUS: Some new rock-types, 247. Obituary—JOsePH LECONTE, 248. vl CONTENTS. Number 70. Page Art. XX VII.—Galvanometers of High Sensibility; by C. E. Munpenuatr and ©. W. Warner __.___.---.-.. 249 XXVIII.—Method of Locating Nodes and Loops of Sound in the Open Air with Applications; by B. Davis. _... 263 XXIX.—Anatomy of the Fruit of Cocos nucifera ; by A. L. WINTON .. 22 S205 ee te eo ae Soe XXX.—Studies of Eocene Mammalia in the Marsh Collee- tion, Peabody Museum; by J. L. Wortman. With Plates I-IV :. 2025.) seo 2 oes) oo 2 a XX XI.—Crinoid from the Hamilton of Charlestown, Indiana 5 by E. Woop. WithoPlate’V (222 2222.22 ee 297 XX XII.—Estimation of Czsiam and Rubidium as the Acid Sulphates, and of Potassium and Sodium as the Pyrosul- phates; by P. HW. BROWNING 27) 122.) 2225535 301 XX XIII.—Time Values of Provincial Carboniferous Ter- tranes’; by CuK. Kayes 20525 1s 2 305 XXXIV.—Spectra of Hydrogenand some of its Compounds ; by J. TRowseriper. . With Plate VI-°___. 2... = ae 310 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Induced Radio-activity produced by Salts of Radium, P. CurRIE and A. DEBIERNE, 319—New Method of Quantitative Analysis, R. W. THATCHER, 320.—Kuropium, a New Element, DeEMARGAY: Research Papers from the Kent Chemical Laboratory of Yale University, F. A, Gooou, 321.— Magnetic Effect of Electrical Convection, H. A. WILSON: Effect of Amalgamated Gases on Resistance, W. ROLLINS, 322. Miscellaneous Scientific Intelligence—American Association, 323.—British Associa- tion: Catalogue of the African Plants collected by Dr. F. Welwitsch in 1853-61: Vol. II, Part II, Cryptogamia: Leitfaden der Wetterkunde, gemeinverstandlich bearbeitet, R. BORNSTEIN, 324.—Studien tiber die Narkose, EK. OvERTON: Ueber Harmonie und Complication, V. GoLpscumMipt: Nature’s Miracles, Familiar Talks on Science, -E. GRAY, 325. Obituary CHARLES ANTHONY ScHott: BARON ADOLF ERIK NORDENSKIOLD: Dr. WILHELM ScHUR: BARON DE LACAZE-DUTHIERS, 326, CONTENTS. — Vil Namie Cl: Page Art. XXX V.—Effect of Temperature and of Moisture on the Emanation of Phosphorus, and on a Distinction in ° the Behavior of Nuclei and of Ions; by C. Barus__--- 327 XXX VI.—Determination of the Heat of Dissociation and of Combustion of Acetylene, Ethylene and Methane; by MMIC MR VUDNCHIIr 8.52 oer igh ene nN Ge 4 < — am ioe) Hs fl ~ Ss > — qo 0 OS = or (6) Calcite with a new form from Colorado.—Some calcite erystals from Seguache Co., Colorado, obtained from Geo. L. English & Co., are differ ent in habit from ae described from this countr y (fizure 3). The forms observed are *x:, (49°41°90°8), UW: (24°8°39- 7), 7 (LOIL), m (4041), ¢ (0112), ¢ ¢ (0001). Ki, New for calcite, is the dominant form. The faces are bright and give excellent images except e and ¢, which are striated or roughened. These were identified by their position, the for- mer truncating the polar edge of 7. The following measure- ments made with the reflection goniometer identify “the forms. Limits. Average. Cale. 65° 28’—-65° 30’ 65° 29’ 65° 29' 53° 48'-58° 49’ 53° 49’: 53° 484 fe 30-13" 3) 13° 304’ 13° 29’ 88° 48’-88° 49” 88° 481’ 88° 484’ 71 Baek ieee Wh Vad UB bag 31° 10’ 31° 104° —~I ~ k:" (49°41:90-8) . (49 90-41°)8) a #:* (49°41°90°8) 4 (90°41:49°)8) pe ONE (24°8 '32°7) » (4041) ay Wi * (24.8 -32:7) ~ (24:32:8°7) A, ME .@2011) ~ (4041) SHS Shen Ww ww oOo 44 Rogers—Mineralogical Notes. There is no doubt as to the symbol given for «: being the correct one though it is complicated. The images were excellent and the crystal was recentered several times. The nearest known form is (6°5:11:1), for which the calculated angles are 65° 386’ and 58° 40’ instead of 65° 29’ and 58° 484’ respectively. (c) New form on Calcite from Prizington, England.—The small acicular calcite crystals from Frizington, Cumberland, - England, seem not to have been described. The forms identified are as follows: m (1010), a (1120), *y (5°5°10°1), v (2131) v (1011), e (0112), and a negative scale- nohedron probably B (2-8: 10°3) and an undetermined dihex- agonal prism. The pyramid of the second order y, new for ealcite, is the dominant form. The following measurements were made with a No. 2 Fuess goniometer with diminishing. eye-piece. The measurements made over adjacent polar edges of the pyramid show conclusively that it is not a scalenohe- dron. Limits. AV. Cale. Von H Sio LO x (U0r5 51) Oe ooo mtaoy 59° 34! 59° 344 4 XY (85 101) 2 (oo: 100) 4 59° 34'—59° 35’ 59° 348! (d) Calcite Crystals from Hudora, Kansas.—In eavities of the brachiopod fossil Hnteletes hemiplicata, which oceurs so abundantly in limestone at Eudora, Douglas Co., Kansas, are found small colorless calcite er ystals. The observed forms are o (5164), r (1011), ¢ (0112), 2 (0445), M (1010), ¢ (2134), m (1010). o is ordinarily the dominant form, though sometimes 7 is. The faces are somewhat striated and otherwise imperfect. Consequently the measurements are not all good, but they serve to identify the forms without question. | Average. — Cale. Tey SABE ON (G18) 14° 19’ Ve a g Ria (5164) A (615s? (eon, 77° 54 Ci) EP XKOVL2) X04 4a) aGe LO! lore Livi (OLAS) oA LOnO)e a2 SUS e3 OF 519 as Mp ME (1010) A. (4041) 2 1 aly 4 ey be nie (2134)0~ (O112),.4 AN SO 20° 58’ (7) Calerte Crystals from Kansas City, Mo.—Dull opaque calcite crystals from 1 to 3°" in length, occurring in a limestone at Kansas City, Mo., present some peculiar features. They are singly ter minated and are always attached at the terminated end. The other end is hollowed out and occupied by parallel growths. ‘They would scarcely be called crystals were not the rough faces seen to be definitely related to the cleavage planes. Tke forms are g (0552) and n° (5051), rare rhombohedrons for Leogers—Mineralogical Notes. 45 ealcite. g consists of smooth narrow faces truncating the polar edges of n° ‘The faces of the latter are very rough, channeled parallel to their intersection edge with g, and pitted. The angle g ~ 7 (cleavage face) 0552 ~ 1011 was measured with the following results, the contact goniometer being used. “Average. Calculated. Crystal 1 8 measurements Gia D5: GPs 380 = 2 6 ve Giese DD: 67 30’ The form m° was identified by the fact that the g face trun- cates its polar edges, as measurements were not possible on account of the rough surface. 9, A New Form on Galena. Several crystals of galena associated with sphalerite and ealcite from an unknown locality are thought to be worthy of brief mention here. The forms present are the tetragonal 4 5 6 crystals have a tetrahedral aspect due to the unequal develop- ment of the octahedral faces (fig. 4). p (13°1:1) is a new form for galena and is not mentioned as occurring on any isometric mineral in the 6th edition of Dana’s System. It is striated parallel to its intersection edge with the octahedron, but good . measurements were obtained, as the following table shows. ; Limits. Average. Cale. pps 811A 13821.1..54 8° 44’— 8° 48’ 8° 46’ 8° 46’ pe Ole ele Ike YS 4829 48-039) 485-80) . 48° 314’ 3. Pyrite Crystals from Weehawken, N. J. Occurring in the calcite veins of the trap near Weehawken, N.J., are found small crystals of pyrite with the following forms o (111), e (210), JZ (432), the latter as narrow faces trun- cating the edge e/o. Figure 5 gives an idea of their habit. The markings, equilateral triangles on o and isosceles triangles on e, are almost identical with the etch figures, and like them indicate 46 Logers—Mineralogical Notes. the symmetry of the crystal. The diploid is sometimes lack- ing and oceasionally the pyritohedron and octahedron are in equal development, the crystals then simulating the regular icosahedron of geometry. On account of striations good images were not obtained, and in the case of the diplcid the reflections had to be relied upon. The measurements are as rien: Average. Calculated. REO CABO Ah ay HSS Ie! 15° 184! CGO Ook eee 39° 14¢ 39° 14! ON Re M210 Re 0ers e587 0) 58 ate 4. A Topaz Crystal of Unusual Habit from Pike's Peak. A topaz crystal from Pike’s Peak, Col., measuring (25 x 20 x 15") differs in habit from most of the described Te See figure 6. The forms present are ¢ (O01), 6 (010), m (110), } (120), J (021), y (O41), and a small undetermined pyramid of the unit series. ; and y are the dominant forms and the erys- tal is lengthened in the direction of the d-axis. m and / are vertically striated and the other faces are dull. The forms were identified with the contact goniometer, 5. A new locality for Leadhillite. The rare mineral leadhillite has been reported from but three localities* in this country. The writer here records the mineral from a new locality, namely the Cerro Gordo Mines, Inyo Co., Cal. It occurs as small imperfect crystals of a pale sea- oreen tint on a specimen associated with linarite and caled- onite. The crystals are short prismatic or thick tabular, of hexagonal aspect, much resembling those from Granby,t Missouri. The following forms were observed: e¢ (001), m (110), @ (100), and a narrow face in the vertical zone which could not be identified on account of its small size. The erys- tals were not suitable for measurement. There is perfect cleavage parallel to ¢ and the cleavage plates have a high pearly lustre. The mineral is soluble in EAT, with efferves- cence leaving a white residue. 6. Linarite Crystals from California. No linarite crystals have been described from the United States though it is known to occur at three localities.{ The erys- * Newberry Dist., Spartanburg Dist., N. C., by Shepard, Dana Min. 6th edit, p. 922, 1892. Schultz Gold Mine, Arizona, by Penfield, ibid. Granby, Mo., Pirsson and Wells, this Journal, vol. xlviii, p. 219-226, 1894. + Pirsson aud Wells, loc. cit. t+ Cerro Gordo mines, Inyo Co., Cal., Dana’s System, 6th edition, p. 925, 1892. Stevenson Bennett mine, near Las Cruces, New Mexico, Farrington, Publications Field Columbian Museum, Geol. Series, vol. i, no. 7, p. 225, 1900. Galena, Kansas, Rogers, Kansas Univ. Quarterly, vol. ix A, p. 165, 1900. —= Rogers—Mineralogical Notes. 44 tals from Cerro Gordo mines, Inyo Co., Cal., present the following forms: ¢ (O01), @ (100), m (100), i (010), w (201), s (101). ~The crystals are very thin, being flattened parallel to the hemi-orthodome s (101) and are elongated in the direction of the b-axis. Good measurements were not obtained on account of narrow faces and multiple images, but they are sufficient to identify the forms as given above. This mineral also occurs at the Alice mine, near Butte City, Mont., and at the Daly mine, Park City, Utah. 7. Caledonite Crystals from Montana. Caledonite has been reported from the same three local- ities as linarite. Its occurrence at Mine la Motte, Mo., needs confirmation. The crystals examined are from the Alice mine, near Butte City, Montana. They resemble in habit those deseribed by Farrington* from New Mexico. The forms are ¢ (001), 6 (010), m (110), ¢ (011), f (021), s (223), ¢ (221, besides several undetermined vicinal forms between f and ¢ and between ¢ and m. 8. Highly modified Barite Crystals from Kansas City, Mo. In cavities in limestone at Kansas City, Mo., are often found small crystals of barite. They are usually simple combinations of the common forms, but occasionally highly modified ones are found. One such measuring about 1$™™ (fig. 7) was studied with the following results. The crystals are tabular in habit, being flattened parallel to the basal pinacoid. The forms are as follows: ¢ (001), 6 (010), a (100), m (110), Z (140), » (820), ¢ (104), d (102), 0 (911), F (118), v (112), 2 111), y (122). Not all the faces gave good images, but the forms are identified without doubt. Average. Calculated. GA Get LOO. AgNO *t3 Bile Sil 83 Pee! VA ROOK 15 OS aay DNs ve Core OO LR 02%. 23 Bs sly 38° 514! Pete o AO0T KV OL 43 5D Ee 52> 437 Woes ALLO A Td as 95% 39" Dee Auk arene Moon OO NS AGe Or AG 6. Wawa tse O01); 3 B4° 45! 84° 43’ Cet DOE ae VIO. ! 98 SiGe Dil Sta l Were Overs 1 LOK v3 BO tel 3 nH Nw 320 ~.110 8 On 2 0) 10° 39° pine. Ob ~ V40. 6 FS AC): DO Mas * Farrington, loc. cit. 48 Logers— Mineralogical Notes. 9. Celestite Crystals from Salina Co., Kansas. The writer is indebted to Dr. J. W. Beede for some erystals of celestite obtained at a quarry east of Mentor, Salina Co., -—I co Kansas. As far as can be learned no celestite crystals have been described from Kansas, though it is a common mineral at several localities in the state. The crystals, one of which is represented by fig. 8, are of the usual tabular habit elongated in the direction of the a-axis. The following forms were identified: c¢ (001), 6 (010), mm (110), m (120), 1 (104), ad (102), 0 (O11), 7 (113), and o (22)a setae, a prominent form on the e¢rystals, is mentioned by Hintze.* m and n appear as narrow faces and did not give images. They were identified by their zonal relations and by app ate measurements. The following are the results : Average. Calculated. CR AE L00V AC OAT as DOS Wf ype ON) PPR STR ATO OR 8 Wena 17° «42 CK oC 001E DX” 201 Oe 5200 Con of OOM MBs a DOr 34° 463’. Con SS 001 ee OA BG. 38° 264 SRS NOL aes 38° 114’ Bie” In conclusion the writer wishes to thank Prof. A. J. Moses for suggestions in the preparation of this paper. * Hintze, Zeitschr. f. Kryst., vol. xi, 232) 11885. Mineralogical Laboratory, Columbia University, New York, January, 1901. Shepard—New Solution for the Copper Voltameter. 49 Art. V.—A New Solution for the Copper Voltameter ; by Wiirram K. SHeparp. THE importance of the convenient and accurate method of measuring electric currents by the voltameter depends very largely upon the construction of a voltameter which admits of high current densities. The silver voltameter gives results of a high degree of accuracy, but is not conveniently used except for small cur- rents ; the copper voltameter, however, from the cheapness of its materials, finds employment in the laboratory for the measurement of comparatively large currents. As ordinarily used, the copper voltameter is subject tu irregularities due to the solvent effect of the solution in some eases and the oxidation of the copper in others. The volta- meter solution recently investigated by the writer commends itself by the facts that these irregularities are avoided and that approximately twice the current density permissible with solu- tions hitherto in use may be employed without loss of accuracy, thus allowing large currents to be measured without incon- veniently increasing the size of the electrolytic cell. Many different solutions for the voltameter have been recom- mended by various experimenters. Gray,* using a copper sul- phate solution of density 1:15 to 1:18 with one per cent free sulphuric acid, found that the highest current density desirable was ‘02 amperes per square centimeter of cathode surface. Vannit observed that a copper sulphate solution of density 1:12 with one per cent free sulphuric acid dissolved copper and therefore gave indications always too small. With an increase of the current density the deduced electro-chemical equivalent of the copper increased. A solution neutralized with copper hydrate, on the contrary, yielded results in constant excess, probably on account of the oxidization of the copper. By adding to a liter of the normal solution one half a gram of a solution which contained one per cent sulphuric acid, he found neither a gain nor a loss in the weight of a piece of cop- per immersed in it., With this solution he successfully used a am ; - , but no experiments seem to cm have been made to determine the limit. Oettel { obtained smaller deposits of copper from copper sulphate solutions than Faraday’s law demands resulting from * Phil. Mag., xxii, p. 289, 1886. + Wied. Annal., x, p. 214, 1891 and Phil. Mag., xxv, p. 179, 1888. ¢ Chemische Zeitung, p. 543, 1893. Am. Jour. Sci1.—FourtyH Srrizs, Vou. XII, No. 67.—Juty, 1901. 4 ; current density of 011 50 Shepard—New Solution for the Copper Voltameter. the formation of oxygen compounds such as HSO,. By the addition of aleohol these losses were avoided. He advised as the best solution : 150 gms. CuSO, 50.) TESS Os 50 * Aleohol 1000 “ H,O Beach* investigated the use of copper nitrate in the volta- meter. His solution after neutralization had a density of about 1°53 and contained one drop of a saturated solution of NH,Cl to 100ce. of the nitrate. The NH,Cl was added to prevent the oxidization and consequent excess of deposit which occurs when a neutral solution is used alone. He found that he could use a much higher current density with this solution than with the copper sulphate solutions commonly used. But as copper nitrate is relatively expensive in comparison with the sulphate, it is not so well suited for general laboratory use. It seemed to the writer of interest to attempt the use of a neutral solution of the sulphate in the same manner, with the aim of increasing the allowable current density. The solution was prepared as follows: A saturated solution of copper sul- phate was boiled fora short time to expel the air and then kept at 100° C. for about an hour in contact with metallic copper in order to neutralize the solution. The density was then about 1°20. A very small amount of NH,Cl was added, about ‘05 of one per cent. Two voltameters were used in the investigation. They were connected in series in order to see how the different solutions would agree when using exactly the same current. Each voltameter consisted of three anode and two cathode plates, the cathode plates being placed alternately between the anodes at distances of about 2 centimeters. The plates were 10°5x19cm., of which a working cathode surface of about 600em* was employed. The voltameters were used in battery jars 20cm. high and 15cm. in diameter. The method used in the preparation of the cathode plates was that given by Gray.t| The edges and corners were well rounded and the plates polished with sand paper. They were next placed in water slightly acidulated with sulphuric acid for a short time to remove any trace of oxide, rinsed in clean water, dried with filter paper and then over a gas flame. After cop- per had been deposited, the plates were not exposed to the air for a time longer than absolutely necessary before the copper sulphate solution had been completely washed off. ‘The reason * This Journal, xlvi, p: ‘81; Wegs7 + Phil. Mag., xxii, 1886. Shepard—New Solution for the Copper Voltameter. 51 for this is that a copper plate oxidizes rapidly when wet with a solution of neutral copper sulphate and exposed to the air. This was accomplished by removing the voltameter plates as a whole from the solution and dipping at once into water con- taining a few drops of sulphuric acid. They were then rinsed in clean water, dried with filter paper and over a flame. Gray’s solution was first tried for comparison. Very good deposits were obtained using a current density up to about -03 ampéres _ cm? one per cent in the two voltameters. Oettel’s solution gave fine deposits for low current densities. When, however, the current density was increased the plates blackened badly and copper was found in the form of powder on the bottom of the jars. The highest current density per- amperes ; but the deposits would sometimes differ as'much as missible with this solution was :02 oy, Then the solution with NH,Cl prepared as above was investi- gated in the same manner. For low current densities it gave as good results as had been obtained with either Gray’s or Oettel’s solution, but, in contradistinction to these, the results continued to be satisfactor y when these limits were widely surpassed. With a current of 20 amperes the density was increased, by raising the plates so as to lessen the immersed surfaces to a value even more than ‘05 a Notwith- standing this very material increase in the demands upon the voltameter, the deposits still remained perfectly satisfactory. With 30 amperes and a current density of -045 a the solution still gave results differing by less than one ee cent. The voltameters were then filled, one with Gray’s or Oettel’s solution and the other with the new solution. When employed for low current densities the solutions in all cases gave deposits agreeing within the limits of error imposed by the methods employed. With regard to the constancy of the apparent electro- chemical equivalent when a high current density was used with this solution, no exact statement can be made, as the abso- lute measurement of the current was not attempted. In the first part of this investigation a Weston ampére-meter was included in the circuit, the error of which the writer has reason to believe was less than one per cent. Assuming the indica- tions of this instrument to give the true values of the current, the electro-chemical equivalent obtained from the new solution was fairly constant. 52 Shepard—LNew Solution for the Copper Voltameter. The results are not given, as different values of the current being used they included the errors of the instrument. Having now performed these preliminary experiments show- ing that the solution could be employed for large current densities without loss of accuracy and having obtained a fair idea of the working of the solution, experiments were begun with the aim of eliminating the error of the instrument. Another Weston ampére-meter was obtained which read only to 15 ampéres, each ampere division being divided into ten parts. So that with this instrument the current could be measured with an error in the reading of less than th of an ampere. The instrument was not assumed to be correct, but after the experiments had been finished the results were aver- aged, and taking the electro-chemical equivalent of copper to be :000829 ee the true value of the current at the division © used on the instrument was obtained. Only one point on the ampere-meter was used, namely, the one which gave the nearest value to 15 amperes. Thus with the same division and changing the current density by raising the immersed surfaces of the voltameter plates, the error of the instrument was eliminated. From what has just been said, it will be plain that the results given below are not presented as determinations of the electro-chemical equivalent of copper; the experimental data are reduced to that form simply for convenience in comparing them. The current was obtained from a storage battery of 50 cells, and was kept constant by means of a copper sulphate resistance which could be varied at will by varying the distance between the copper plates. An auxiliary circuit was arranged whose resistance was equal to that of the two voltameters. The current was thrown by means of a switch first through the auxiliary circuit to the ampere-meter and resistances included in the circuit. These were allowed to heat up until everything became constant and the resistance adjusted so as to give the desired reading of the ampere-meter. Then the current was thrown through the voltameters and assumed the desired value instantly. The duration of an experiment was 20 minutes, so that a deposit of about 6 grams was obtained. This deposit showed no oxidization even up to the highest current density used amp. em? which was greater than ‘07 The results are arranged in series acaordine to their respec- tive current densities, as follows: Shepard—New Solution for the Copper Voltameter. 58 SERIES I. : Current Density = 02 — ; No. Equivalent x 107. | Soe en ae SW Seeere he 3287°2 Deets ee Lk) eee ee 3295°0 Rs ete Sk SS See een ate 3288°4 AD re Dols Sl nd ee ONO 5) cd ee em gpa ye PS thet on ee 3292°8 On ak Eee Ua ok ek eee oe Se S202 EE OA OES ANF eee ee 3286°8 OE vt eee Sree CaLe PA ea eo 3288°4 OS Ni Sa pe Se Repent eek Ae Se Pe 3293°9 ieee ahicsae eerie wine So SI FA PO a it lesa sete ees eres PAO Sache Sees 3287°2 ND ere eee eee es Se oie Pete 3293°4 The mean of these values is 3290°7 with a probable error of +2°4 for a single observation. SERIES II. Current Density °03 maa em No. Equivalent x 107. LL eee age eee ee Sy tte een ee 3288°4 ee re ne no eee 3291°7 Bi IN ie a ie aa RO PRE ped eS OAL S| Raver ee Stee NS ee oe Se 3291°7 eee ae oe Sea Ue See ee 3292°8 SENS Sol ee NRE BOC ont LON s fa 3288°'3 Tear NR seg Se N 3295°5 eee SIA 8 Saas See 3295°9 PRE ie © Ps) se REE ROL A J AO) fc cee ae pm cE sae oe eae A be 3287°2 Tee ie bare Sen re ES RS Se 3288°4 Pe eh a a oe se Sees ee BORE 9 The mean of these values is 83291°7 with a probable error of +2°3 for a single observation. SERIES [IT. Current Density -04 eo cm No. Equivalent x 107. 1 pp ie eee ee er eee Sone Aen 3287°8 AS a EE PBL SAP. ES, UC d294°4 7 a ee NG Pie O82 Ian, ea ae 3287°2 Lb at Es Sa ei ites BL, SES ext aa 3284°5 E08 Si rae pi eS a A 3283°9 Tt ee aig ree Se eee 3283°4 (Lelie 6s Se 2 eng a 8 NN ae 3296°7 Sp Re et eee gs es, Ss 3295°0 7 sn aria a PRs fe ee 3292°2 BG eee ote ane SS 3295°5 | SSS (ea a Sy ee 3296°1 [2 EM aN eee eee 3296°7 The mean of these values is 3291:1 with a probable error of +3°6 for a single observation. 54 Shepard—New Solution Jor the Copper Voltameter. SERIES IV. Current Density °05 seat cm No. Equivalent x 107. Wek ea ete Ute ah ae eer ee 3286°2 Ds) au Ae GER NG tha AOU eS ge a 3286°2 Dal habbo aa Sey yee 3290'5 i A De eee ree Mae tt aS EON CS 3293°4 Sea aeee Dn ee eye men EN am. ol. Seiad: CEPR aOR aren RR vari ey La AOR Saat 3286°8 MH hl IE Vee Lag De RD LN 38297°2 Beate eee aie ce ronnie eres eres 3292°8 Oi a ie tee Sn Ramee sasecec Tt A Nuel 3285°5 Th O seth eat > eee Ee Paneer ecn pase files 3287°2 a i ok ae eae ae ne ee aes Oe OES 1 Ae hai rae cae OTS ENON) adr MEN, Se Re 3293°4 The mean of this series is 8289°5 with a probable error of +2°6 for a single observation. SERIES V. Current Density :06 ie cm No. Equivalent x 107. De Os ene en Seno eer Dai ee) SP TRL EN Mi Ca eas eng 3294°4 3 fe eRe AME ate ALON Ge WR ae 32 Sie PTT g A AG Oy See Se ae OLR Re a WR ty rl 3290°0 Fp a OR Me Men ey Ca I ey SE GN ECT Malt 3292°8 CO alpen gear a tiaras or AGM A ia ip ica EA oeal 3290°5 (IS cpa CN UA NG ah Lh aE 3291°7 Seer Nice aCe weeny Ul eamena i pre rsteeaetea NO WO 8 peepee traaias tie at Ft RE Pa Sa eR 3283'°9 BD) ST i ea as fg re Np 3287°2 be Reape ee eee MN NI a ca BH} 7) 8 Wee ae Macey ay eee sy RNY SN aS) Let The mean of these values is 3290°0 with a probable error of 2°2 for a single observation. Shepard—New Solution for the Copper Voltameter. 55 SERIES VI. Current Density ‘07 peal cm No. _ Equivalent x 107. ae a ae Ee eee ON Ly: FSO A RUN et MUR uD Ne ok PF Pe Be EP 3288°9 3 Es Sy ae Sn e221 Cag Foon aap 329 1°h YS eaee eet OR Mg ee Sen OR (lens ah aun aa 3286°2 Deis shite SNL Seah ae, Aiea i ae 3290°5 (6 ae agg ioe nae ele pia tees Pagid nao 3290°0 (Seen RUE! Selim rk | ceokeeel | pint P asate BW. Sane (OP) i) a al Ke tp, a i eee ea SOY oS iS ert ee in A ORR era ele EONS 3290°5 TOP SSRRSIN rea eS ies ee OS TPE iecd Vou, Sa 2S Ae er Ba an ie eam Omer IES Res 72) 5) 0k [Dee eee MRS IPE Lh See eMac eae eL as 3282°8 RSS le A i SO se A re 3291°7 eee ate er us tae Meru ky Ate) (oe MOPED Pe eee ee SIME yl NT NS eS Te 3292°2 LUG) gyi See ha UE IS Rar Ra era ES 3288°9 The mean of these values is 3289-0 with a probable error of +1:9 for a single observation. The results are arranged for comparison in the following table : Series. Current Density. Mean. Prob. Error of Mean. ie (02 mine 3290°7 +0°7 II °03 3291°7 ae OPT Il "04 | Sy ASD at se 0) IV °05 3289°5 ae OW V "06 3290-0 aE OAD VI 07 : 3289°0 +0°6 These results show no dependence upon the current density. In fact, series VI, which differs most from the others, could be slightly increased for the following reason. The day after this series was taken the watch which had been used to measure the time was rated with a clock and found to be gaining at the rate of 2 seconds an hour when lying at the side of the ampere- meter. This was doubtless caused by the magnetic field of the ampere-meter, as the watch at other times was very correct. If a connection were applied for this to series IV it would increase it by about one part in 1800,so that it would become 8290°6, which is in still closer agreement. During series III, IV, V, which were taken after this fact had been discovered, the watch was rated and proper connections were applied. But as the watch had not been rated during series I, II and VI, the corrections cannot be applied to them. 56 Shepard—New Solution for the Copper Voltameter. It is also proper to observe that although the current density was widely varied there is no loss of accuracy, as is shown by the probable errors of a single observation in the different series. _ amp em* found to heat a little during the first experiment. Results which were obtained during the rise in temperature of the solution were always slightly greater than those obtained after the solution had become warm and at a uniform temperature throughout. The unequal temperature of the solution may be avoided by stirring or by making a preliminary experiment to warm up the solution. The best way was found to heat the solution by means of a flame to a temperature of about 35 or 40° C. before making an experiment. The deposits were then found to be very satisfactory even up to a current density of — over ‘07 —= and the resulting values of the equivalent to agree closely with those obtaied with the low current densities. The results given in series III, [V, V and VI were obtained with the solution at a temperature from 35° to 40° C., while series I and II were obtained with the solution at the tempera- ture of the room, about 20°C. Wesee therefore that the temperature from 20° to 40° C. has practically no influence upon the results given by the solution; but that any znequal- ity in the temperature of the solution will cause slight errors and therefore should be carefully avoided. The same solution was used throughout the second part of this investigation. Perhaps fresh solutions would have given more consistent results, but certainly no large error is intro- duced by the repeated use of the solution. We may collect the results with this solution as follows: The weight of copper deposited is practically independent of the temperature between 20° and 40° C. The results are independent of the current density within the limits given. The solution may be used a large number of times. The solution may be used with a current density two or three times as great as that permissible with others commonly in use, so that to measure a given current a considerably smaller voltameter may be employed than with the older solutions. Current-measuring instruments may be calibrated by its means with an error of only about one-tenth of one per cent. Sheffield Scientific School, New Haven, Conn. For current densities of 05 or greater the solution was Lloyd—Magnetic Effects in Tellurium. 57 \ Art. VI.—-The Thermo-magnetic and Galvano-magnetic Effects in Tellurium ; by Morton Gituens Luoyp, Ph.D. THE investigations here described were begun with the idea of measuring all of the thermo-magnetic and galvano-magnetic effects in the same specimen of a non-metal, tellurium being the substance chosen. It has heretofore been investigated for only two of the transversal effects. Satisfactory numerical values of the magnitude of most of these effects have not yet been obtained, owing to the difficulties involved in their meas- urement, but as the work must be interrupted for some time, it is thought best to publish the qualitative results at once. If a conducting plate carrying an electric current be placed in a magnetic field perpendicular to the lines of force, the fol- lowing galvano-magnetic effects have been observed. 1. A difference of electric potential is produced between the edges of the plate, perpendicular to the direction of the cur- rent and of the magnetic field. (Hall effect.) 2. A difference of temperature is produced between the same points. 3. A change of resistance occurs. (Longitudinal Hall effect.) 4, A difference of temperature is produced in the direction of the current.* If a flow of heat replace the primary electric current, four analogous thermo-magnetic effects are observed. . 1. A difference of electric potential between the edges of the plate. 2. Rotation of the isothermal lines or a difference of tem- perature between the same points. 3. A change of conductivity for heat. 4, A difference of potential in the direction of the heat- flow. (Longitudinal thermo-magnetic effect.) The transversal effects are reversed in direction when either the magnetic field, or the primary current of heat or electricity, is reversed in direction; but the longitudinal effects are inde- pendent of the direction of the field. | It has already been proposedt to call the galvano-magnetic temperature-difference, the thermo-magnetic temperature-dif- ference, and the thermo-magnetic potential-difference by the respective names, Ettingshausen effect, Leduc effect and Nernst effect. Most of these effects were first observed in bismuth, and some of them only in bismuth. The change of conduc- * Nernst, Wied. Ann., xxxi, p. 784, 1887. + Thesis: The Transversal Thermo-magnetic Effect in Bismuth, M. G. Lloyd, Philadelphia, 1900; Beiblatter, 24, p. 1014. 58 Lloyd—Thermo-magnetic and Galvano-magnetic tivity for heat was first announced by Righi* and shortly afterward by Ledue.t . The Leduc effect has been observed by Ledue,t Righi,t and van Everdingen.§ The Nernst effect, discovered by von Ettingshausen and Nernst, has been observed by van Everdingen, Yamaguchi,4] Moreau** and the writer, these observations covering a number of metals. The Ettingshausen effect was observed by its discoverer t+ in bis- muth, antimony and tellnrium, and was included in van Ever- dingen’s observations on bismuth. The longitudinal thermo- magnetic effect has been observed in bismuth by von Ettings- hausen and Nernst, van Everdingen,tt and Lownds.§§ Chemically pure tellurium was obtained from Eimer and Amend, and cast into the form of a plate, 6-7™ long, 25™ wide and 0:1 thick, with two lugs projecting from the middle of the lateral edges. Six thermo-electric couples of copper” and German silver wire made electric connection with the plate; one to each of the lugs, the other four along the center line of the plate, separated by distances of 15, 138, 18™™. These couples served to determine the temperature at the six points and thus the temperature-gradient, and in addition could be used to determine the difference of potential between any two points. The ends of the tellurium plate were pressed by springs against projections from two brass tubes, through which steam and cold water, respectively, could be passed. The whole was mounted on a wooden frame, and suspended between the poles of a powerful electro-magnet. The pole- faces were 6°3™ square and 1°3™ apart, and the magnetic field between them practically uniform. The strength of field was determined from the resistance of a calibrated bismuth spiral which was also attached to the wooden frame. When sending an electric current through the plate, cold water flowed throngh both brass tubes, to which the lead-wires were attached. Dzif- ferences of potential were measured by comparison with a standard Clark cell, using the potentiometer. A very sensitive astatic reflecting Thomson galvanometer was used as a current- indicator. : The electromotive force, E, in microvolts, of a thermo- electric circuit of copper and German silver, one junction of * Rend. Acc. Lincei, iii, p. 481, 1887; Beiblatter, 11, p. 670. + Comptes Rendus, civ, p. 1783, 1887. t Rend. Acc. Lincei, iii, p. 6, 1887. § Proc. Royal Acad. Sci. of Amsterdam, i, p. 72, 1898; Comm. from Leiden, No. 42. || Anz. d. kais, Akad, Wien, xiii, p. 114, 1886; Wied. Ann., xxix, p. 343, 4 Ann. d. Physik, i, p. 214, 1900. ** Journal de Physique ix, p. 497, 1900. ++ Wied. Ann., xxxi, p. 737, 1887. t+ Ann. d. Physik, iv, p. 776, 1901. S§ Leiden Comm., No. 48; Proc. R. A. 8. A., Mar, 25, 1899, Liffects in Tellurium. 59 which is at 25°C. and the other at T°, has been found by Dr. H. C. Richards to be E = 15-20 (T — 25) + 0°0272 (T — 25)’. One of the couples used in this work was tested between 0° and 100° and found to agree with the formula, which has been used in determining temperatures. The plate was insulated and protected from air-currents as well as possible by the use of mica, asbestos, cotton, ete. Hitingshausen effect.—If the primary current flow from right to left, and the lines of magnetic force are directed away from the observer, the upper edge of the plate is cooled and the lower warmed. This is in the same direction as in bis- muth. The effect was measured by passing current through the plate between the two tubes, and observing the tempera- tures at the edges of the plate for both directions of the field, giving time in each case for temperatures to become constant. A series of readings was taken, reversing field each time, in order to eliminate any error due to the gradual change in tem- perature of the whole plate. Readings in zero field were not used in the calculations. They showed, however, that the effect was not symmetrical, i. e., the change for one direction of the field was not the same as for the opposite direction of the field; starting with zero field in both cases. The coefii- cient of the Ettingshausen effect, P, is defined by the equation T = PHS where T is the difference in temperature produced, by a change of field H, between the edges of a plate of thickness ¢ when traversed by a current C. Temperatures are measured in Centigrade degrees; other quantities in C.G.8. electro- magnetic units. The values of P observed ranged between 0:00014 and 0-00029 for fields up to 5500 gausses, at a tempera- ture of about 65°. The high resistance of tellurium causes a correspondingly large generation of heat and the temperature of the interior is considerably higher than that of the sur- roundings. ‘The flow of heat from the center of the plate to the edges (cooled by water flowing through the tubes) will produce a Leduc effect which may interfere with the measure- ments of the Ettingshausen effect. In this case the measure- ments were made at the center of the plate, where the tem- perature-gradient may be assumed zero. Hali effect.—It the primary current flow from right to left, and the lines of force are directed away from the observer, the upper edge is at the lower potential. This direction coincides wita the results of other observers and is usually considered as 60 Lloyd—Thermo-magnetic and Galvano-magnetic positive. It is in the opposite direction in bismuth. To determine the Hall effect, the difference of potential was meas- ured between the copper wires connected to opposite edges of the plate. This P.D. includes the thermo-effect due to the two points of contact not being at the same temperature. If the temperature remained constant, the thermo-effect could be eliminated by taking readings for both directions of the field. But the temperatures are not constant, the Ettingshausen effect producing a change whenever the field is reversed, and this must be allowed for. It was found in this case to affect the second significant figure by one unit only and hence is neglected. The coeflicient of the Hall effect, R, is defined by the equation E=RH. In a field of strength 4,000 gausses, the value of R found was 430 C.G.S. at a temperature of 65°. Change of resistance—The resistance was determined by the difference of potential of two points between which a current was flowing. The current and P.D. were measured with Weston ammeter and voltmeter. The accuracy of the observa- tions would serve to determine a change in resistance of one- half.of one per cent. No change was found in fields up to 5,500 gausses. Other observations* on tellurium show an increase in resistance less than this. | Leduc efect.—If the heat flow be from right to left and the lines of force are directed away from the observer, the upper edge of the plate is cooled. The direction in bismuth is the reverse of this. The coefficient of the Ledue effect, S, is defined by the equation | where 6 is the breadth of the plate and et the longitudinal temperature-gradient. Values of S between 0:000002 and 0:000005 were obtained at temperatures between 30° and 38° in fields of strengths from 2,500 to 5,200 gausses. The coefficient appears to increase with the temperature. Nernst effect.—lf the heat flow be from right to left, and the lines of magnetic force are directed away from the observer, the upper edge is at lower potential. In bismuth the direction is the same. The coefficient of Nernst effect, Q, is defined by the equation dr = QOH: * Goldhammer, Wied. Ann., xxxi, p. 360, 1887. Lifects in Tellurium. 61 The difference of potential between the copper leads connected to opposite edges of the plate is the resultant of the Nernst effect and the thermo-electric effect at the two junctions. Since the two junctions are in general at slightly different temperatures, they do not balance one another. If their tem- peratures were constant, the thermo-electric effect could be eliminated by taking readings for both directions of the mag- netic field; for in one case the Nernst effect would be added to it and in the other case subtracted. Owing to the Leduc effect, the temperatures do not remain constant, but are altered by reversing the field. | Let E=Nernst effect; V and V’=the observed P.D.’s for two directions of the field; e=thermo-electric power for cop- per and tellurium; ¢, and ¢,=the temperatures of the two junctions for one direction of the field, 7,’ and 7,’ for the other ; T=Ledue effect=4 (¢,—7,/+7,/—7,'). Then ef, —t,) +E V'=e(t,’ —t,')—E whence Tea | 2 R= ede. It is thus seen that the observed values of V must be cor- rected for the Ledue effect. In bismuth this correction was found negligible ; in tellurium it is not so, owing to the enor- mous thermo-electrie power of copper and tellurium, but assumes a magnitude of the same order as the Nernst effect. An accurate determination of the Nernst effect requires an accurate knowledge not only of one value of this thermo- electric power, but also of its variation with temperature and with strength of field. The values of ¢ observed are slightly over 600 microvolts, whereas the value usually given is about 500. Observations of the Nernst effect give a value of Q=0°36 at 33° in a field of about 3,000 gausses. As in the ease of the Hall effect the constant is larger than has been observed in any metal. Change of heat-conductivity.—The conductivity is lessened in the magnetic field. The relative conductivities were calcu- lated from observations of temperature at four points along the center line of the plate, assuming the thickness to be uni- form and the surface-conductivity to be constant. Neither of these conditions is fulfilled, and hence the results are only approximate. The ratio of conductivity in a field of strength 4700 gausses to that in zero field was found to be 0°9. Longitudinal thermo-magnetic effect.—The warm end of the 62 Lloyd—Thermo-magnetic and Galvano-magnetic plate is at the higher potential. This direction is the same as Lownds* has found in bismuth at ordinary temperatures. Two suggestions have been madet which might explain the — origin of the longitudinal effects. One is, that the thermo- electric-power is altered in a magnetic field. Lownds has demonstrated that the effect observed by him cannot be so explained. The other suggestion is, that owing to a change in heat-conductivity, the temperatures at the junctions are changed, thus changing the E.M.F. in circuit. This explana- tion is plausible where the effect was measured by compen- sating the E.M.F. in zero field, and observing the change produced by the field. Lownds does not mention whether he made any test of this possibility, but in one of his experiments at least (p. 779) it is evident that the E.M.F. could not have had this origin. = The longitudinal effect has been observed in bismuth also by. van Everdingen,t and found to vary with the strength of field according to the same law as the variation of resistance. In my own experiments the temperatures at the junctions were observed in the given field as well as in zero-tield, and allowance made for the change. Theory§ indicates that longi- tudinal effects should be proportional to the square of the strength of field, and independent of its direction. I have noticed that in every case the effect was greater for one direc- tion than for the other. The mean value for a field strength of 2750 gausses, mean temp. = 39°, difference of temp. = 19°, was 1050 microvolts. If we regard the transversal differences of temperature and potential as due to the transfer of heat and electricity respect- ively across the plate, we notice that in tellurium the transfer of electricity is always in the same direction as the transfer of heat. That is, the edge which is raised in potential is also raised in temperature. Moreover, to produce the same effects with a current of electricity as with a current of heat, the direction of flow must be the same for both. For example, the edge which is warmed by the Leduc effect has its potential increased by the simultaneous Nernst effect. And to warm the same edge by the Ettingshausen effect as by the Ledue effect, the primary currents of electricity and heat must have the same direction. In bismuth, the reverse is the ease, the heat and electricity being transferred in opposite directions. Thus the edge which is warmed by the Leduc effect has its potential lowered by the simultaneous Nernst effect. And to warm that same edge by * Anme d. Phys, 1v, p.. 106; 1901. + Grimaldi, Nuov. Cim. (3), xxii, ». 5, 1887; Goldhammer, J. ce. ¢ Leiden Comm., No. 48, 1899. § Drude, Ann. d. Physik, iti, p. 377, 1900. Liffects in Tellurium. 63 the Ettingshausen effect would require a primary electric cur- rent in the direction opposite to the heat flow producing the Ledue effect. These facts are of especial interest in connection with Liebenow’s* theory of thermo-electric currents, in which he arrives at the conclusion that in metals heat always flows with the negative electricity, whereas in non-metals it accom- panies the positive electricity. If this relation between the directions be a general one, a knowledge of the direction of any one of the four transversal effects would be sufficient to determine the direction of the other three. Of the eleven elements examined by Nernstt+ for the Nernst effect, six (Bi, Sb, Ni, Fe, Zn, Pb) bear out this relation with the Hall effect, if we regard antimony as a non-metal. Of the other five, three (Cu, Ag, Sn) showed such a small effect that they may be regarded as doubtful. The two notable exceptions are car- bon and cobalt. It has been already noted by several writerst that cobalt occupies an anomalous position as regards its Hall- constant, and it may be that the Ettingshausen effect is large enough in this metal to entirely vitiate the measurements which have been made of its Hall-effect, even to changing the sign. In this regard it is to be noted that the thermo-electric- power is large between cobalt and copper, the metal usually used for lead-wires. Or the measurement of Nernst effect may have been equally influenced by Ledue effect. Moreau (J. ¢.) however, found an opposite direction for the Nernst effect in cobalt, which would make it agree with the general rule. The direction of the Ettingshausen effect in antimony is in har- mony with the other effects. This idea comes in conflict, however, with the ‘“ Electronen- theorie” of Drude,§ which requires that the Ettingshausen effect be in the same direction in all metals. For the above relation to hold, the Hall-effect must also be in the same direc- tion in all metals, which we know is not the fact. In the only three substances yet examined, bismuth, tellurium and anti- mony, the direction of the Ettingshausen effect has been found the same. _ I summarize the constants for the four transversal effects. lB sages t Ettingshausen.---. P= 0-°0002 5500 65° JIE STS Se at Q= 0:36 3000 Bor | ie | ea Rene a ee ea R = 480: 4000 a ES eA == 2,3 es Sr S = 0:000004 5200 30-38 * Wied Ann., Ixviii, p. 316, 1899. + Wied. Ann., xxxi, p. 760, 1887. ¢ Von Ettingshausen u. Nernst, Wiener Berichte, xciv, p. 560, 1886; J. C., Beattie, Proc. Royal Soc. Edinburgh, xx, p. 481, 1895; Drude, Ann. d. Phys., iil, p. 392, 1900. § Ann. d. Physik, iii, p. 369, 1900. Laboratory of Physics, University of Pennsylvania, June 1, 1901. 64 Verrill—Additions to the Avifauna of the Bermudas Art. VIl.—Additions to the Avifauna of the Bermudas with diagnoses of two new Subspecies; by A. Hyatt VERRILL. DuRING a recent collecting trip to the Bermudas, from March 10th to May 9th, 16 species of birds were observed that appear not to have been previously recorded : Phaéton wthereus. Red-billed Tropic-bird. Several were seen on Harrington Sound in April. Larus glaucus. Glaucous Gull. A large flock remained some time. Seen about the islets in Harrington Sound early in March. Were regarded as something new by the inhab- itants. : * Melanerpes Carolinus. Red-bellied Woodpecker, Chab. - Seen April 8th, on a Pride of India tree. Passer montanus. European Tree-sparrow. Locally com- mon in Paget Parish. Naturalized; resident. Probably intro- duced with the English Sparrow. Carduelis carduelis. European Goldfinch. Abundant on the southern and eastern parts of the islands, especially about Hungry Bay. Accidentally introduced about 1885, from a wreck. Previously recorded by Reid as an escaped cage-bird. Spinus tristis. American Goldfinch. Resident. Not un- common. Intentionally introduced about 1896, near Hungry Bay. | Spizella monticola. ‘Tree-sparrow. A flock was seen sev- eral times at Hungry Bay during the latter part of March. Sitta Carolinensis. White-breasted Nuthatch. Seen April 14th to 30th, on cedars at Harrington House. Dendroica Pennsylvanica. Chestnut-sided Warbler. Dendroica striata. Black-poll Warbler. Dendroica Blackburnie. Blackburnian Warbler. The last three were seen in flocks of other migrants, March 12th to 15th, at Hamilton. Saxicola wnanthe. Wheatear. Introduced recently near St. Georges. Appears to be perfectly naturalized. Previously recorded by Reid as a rare migrant. Mimus polyglottus. Mocking Bird. Resident. Introduced about 1892, at Bailey Bay. Not uncommon at Walsingham and Paynter’s Vale. Appears to be now naturalized. The following four species were identified from the local collection in the Public Library at Hamilton: Orchard Oriole ; Thrasher or Brown Thrush ; Blue Jay; Red-shouldered Hawk. The abundant resident Ground Dove proves to be the Bahama subspecies (Columbigallina passerina Bahamensis). It always has a black bill. with diagnoses of two new Subspecies. 65 The resident Bluebird is decidedly larger and brighter col- ored than the true szalzs. It is a new subspecies. Sialia sialis Bermudensis A. H. Verrill. Blue of the upper parts of male brilliant purplish azure, a little brighter on rump; chin light blue. Breast, sides and flanks deep purplish-cinnamon, much darker and richer than in North American specimens. Female more brownish with brighter rnmp and back than in the true sza/zs. Edge of wing, at carpal joint, distinctly pure white. Length, 6°75 to 7-5 inches; wing, 4 to 4°25; tail, 2°75 to 3°25. Nest usually built in crevices and holes of cliffs; eggs usually pure white, rarely tinged with greenish-blue. The resident cardinal bird of Bermuda also differs as a sub- species, from the American forms: Cardinalis cardinalis Somers A. H. Verrill. Adult male: Lower parts brilliant orange-vermillion, brighter and more orange than in C. cardinalis. Upper parts are also ~ clearer and brighter, deep lake-red, with scarcely any gray on tips of feathers. Vermillion of checks and crest brighter and clearly defined. Bill deep scarlet. Female lighter than in cardinalis, especially below ; breast buffy yellow; belly almost pure white; upper parts clear ashy gray; crest and ear-coverts strongly tinged with red; wings and tail nearly as in the male. Length, 8°75 inches; wing, 3°75; tail, 4°75 ; culmen, 0°80. Am. Jour. Sc1.—Fourta Series, Vou. XII, No. 67.—Juty, 1901. 5 66 Wright and Downs—Spectrum of the Art. VIIl.—Zhe Induced Alternating Current Discharge studied with Leference to its Spectrum and especially the Ultra- Violet Spectrum; by A. W. Wricur and E, 8. Downs IN some experiments made in this laboratory a few years ago, in photographing the spectrum of an induced alternating current dischar ge, produced under special conditions between copper terminals, a a plate was obtained which contained a very great number of fines. The primary purpose of this investi- gation was to locate these lines and if possible discover their origin, and secondly to study this peculiar form of discharge when produced on a very large scale. Spottiswoode* was among the first to give an account of this mode of exciting an induction coil by the direct application of an alternating machine, without the intervention of a contact breaker or the use of acondenser. In the Pr oceedings of the Royal Societyt he points out some peculiarities of the discharge which he noted. In our experiments the induction coil was used without a contact breaker or condenser. It had four hundred and twelve turns of wire about 38™" in diameter for its primary, and a secondary of about 25°5™ in length, consisting of about fifty thousand turns of wire 0°37" in diameter. In the core of the primary, which was 385°5™ long and 8°5™ in diameter, there were nearly a thousand soft iron wires carefully annealed and insulated with shellac. As a result of having such large wire for its secondary, the current was very large, and the potential as determined by the length of the spark was in the neighborhood of 120,000 volts. The current was supplied to the coil by a Siemens alternating current dynamo of about five horse power which was driven by a gas engine, and whose magnets were excited by a small dynamo using approximately three horse power. With the terminals arranged horizontally the discharge was exceedingly intense and vigorous, being accompanied with a loud singing noise. It consisted of a light nebulous flame of a whitish color, with a slight tinge of yellow or green, rising gradually from the terminals and~ meeting at the center, where the color changes to a reddish. The flame appeared to the eye to be continuous, but, when observed in a rotating mirror or photographed on a swiftly moving plate, the alternate dis- charges were plainly discriminated. About and close to the terminals the bluish-purple flame, caused by the nitrogen of the air was plainly seen. The spark would leap across an interval of about 3:5, and when the discharge was once started, the terminals could be drawn apart for as much as 20™ * Phil. Mag., Nov. 1879. Pav iol. xx’ pla go. Induced Alternating Current Discharge. 67 or 25™ without extinguishing the flame, presenting a striking and beautiful spectacle. It may be of interest here, in view of the peculiar character of the discharge, to give the results of some experiments which were made a few years ago in this laboratory to determine the time and mode of its formation. A camera was arranged so that the discharge, just as it was forming, could be photo- graphed upon a rapidly moving plate. At the same time the image of a spark from a tuning ‘fork whose point dipped in mer- cury was adjusted so as to fall upon the plate side by side with that of the discharge. The photographs show that the spark passes directly across at first. As each spark has a tendency to follow its predecessor and the air becomes heated causing an upward current, the path of the discharge rises upward little by little until it reaches its permanent condition. From the period of the tuning fork it was calculated that about one- tenth of a second was required for this. In the Chemical News* Crookes has an article on this flame in which he states that it consists chiefly of nitrogen burning with the formation of nitrous and nitric acids, and that it shows no lines, but that the spectrum is faint and continuous. It is more probable though that the conditions here are similar to those in a tube from which a portion of the air has been exhausted. Owing to the heat, the air within the aureola becomes rarified and partially conducting. The luminosity is caused by the air being heated and electrified. Stratifications can be seen and photographs show them distinctly somewhat as they are shown in a vacuum tube. An examination of the spectrum with terminals of several different metals, made by means of an ordinary prism spectro- scope, shows in ‘all cases, and as the most prominent feature, a continuons background which is very bright in the red, yellow, and green, but gradually decreases in intensity until it fades away In the violet. The yellow sodium lines are always pres- ent, due to particles in the air or impurities of the electrodes. Generally a few of the stronger lines, which are due to the metal of the electrodes and are ordinarily seen in its flame spectrum, appear and others which come out only occasionally. With zine terminals and also with those of aluminum, the flame is yellow- ish in color but has a peculiar bluish-green core which is very intense, extending in streaks throughout the flame. The dis- charge from electrodes of soft iron which are small in diameter presents a beautiful spectacle, inasmuch as the iron, because of its large resistance, becomes highly incandescent, and bursts into a brilliant combustion, throwing off luminous particles in all directions. A banded ‘structure. from about wave length * June 17, 1892. 68 Wright and Downs—Spectruin of the 4247 to about wave length 4225 came out which appeared to be the same in the ease of all the metals employed. When the coatings of a large Leyden jar were connected to the ter- minals of the secondary, the discharge took the form of intensely brilliant sparks following one another in such rapid succession and such energetic detonation as to cause a jagged and almost deafening roar. In general the spectra of the metals employed in the induction flame are less complete and less developed than their flame spectra, owing to the fact that they are obscured by the rather intense continuous spectrum. For a more careful study of the spectrum than the one pre- viously described, which was made with the naked eye, it was necessary to have recourse to photography. To obtain the photograph of this discharge a Rowland concave grating havy- ing a radius of 21°5 feet was employed. The upper half of the photographie plate was exposed to the spectrum of the sun’s rays for a comparison and the lower half to that of the flame between the terminals of the metal to be studied. The plate upon which was the spectrum which it was the primary purpose of this investigation to locate, was stained with erythrosine to make it more sensitive to the rays at the red end of the spectrum. This staining also appears to increase the sensitiveness of the plate for ultra-violet rays.. These lines are in the region where the red of the first spec- trum and the ultra-violet of the second overlap, and they begin abruptly at wave length 6127°32 in the red and extend down to 6269-09, or at 3063°66 in the ultra-violet, extending down to 318455. Inasmuch as the two spectra overlap, it was un- certain to which spectrum they belonged. The first’ thing that suggested itself was that these might be due to copper from the electrodes. Very few copper lines so far down in the red are given in the very admirable tables of the British Association Report.* This fact renders it impos- sible to make a comparison of any significance with the lines given in this region. The number of copper lines given in the ultra-violet was much larger, but still so much smaller than the number of lines upon the plate that no comparison of any value could be made although there were a few cases of approx- imate coincidence. With a view to ascertain to which spectrum the lines belonged various devices were employed. At first plates stained with cyanine were tried, and some preliminary experiments made in obtaining the solar spectrum to get the time of exposure. It was found that an exposure of eighteen minutes gave a good spectrum extending into the red far beyond the situation of the lines in question. Several plates were then exposed to the. * 1884, p. 384. Induced Alternating Current Discharge. — 69 discharge for an hour and a half to secure the lines, but, as no lines were obtained, and none of the ultra-violet lines of the solar spectrum were visible, it was suspected that the staining with cyanine, while it made the plate more sensitive to the red rays, diminished the sensibility to the ultra-violet, and there- fore, that probably the lines sought were in the ultra-violet. Plates stained with erythrosine were next tried. It was found that an exposure of five minutes gave the ultra-violet spectrum of the sun’s rays together with the lines in the red, in the neighborhood of the lines sought for. To obtain the lines several exposures to the discharge of an hour and a half were made and a few feeble lines obtained. The time of exposure was then increased to twochcurs. Upon develop- ment the lines upon the first plate tried were sharp and dis- tinct. The wave lengths were calculated as accurately as possi- ble by means of the comparison solar spectrum and Rowland’s charts. This is first done on the supposition that they are in the lower part of the first spectrum and the result divided by two to get their trne wave length in the ultra-violet of the second spectrum, since it was now regarded as almost certain that they were ultra-violet lines owing to the fact that we did not obtain them with plates stained with cyanine, which makes them more sensitive to the red rays, and on the contrary did obtain them when the plates were stained with erythrosine, which increases their sensitiveness to the ultra-violet, as was proven by exposing a dry unstained plate to the sun’s rays when, although the ultra-violet lines were visible, they were exceedingly faint. The wave lengths as calculated by means of Rowland’s charts of the solar spectrum and the comparison on the plate are as follows: 3063°66. 3086°48 3105'71 3063°83 3087°40 3106°02 3064°30 3089°80 3106°55 3065:04 3089°94 3109°46 3065°17 3090°24 3110°30 306627 3091°35 3112°17 3068°00 3091-48 311343 3068°31 3092°45 3114°S85 306911 3092°84 3115°49 3070°02 3070°52 3094°70 ST 87 3070°83 3095°41 3118°03 3072°05 3096'16 3122°70 3078°47 3096'87 3124°14 3080:06 309863 3130°46 3080:27 3099 53 3184°55 3081°65 3101°20 3083°35 3102°13 3085°25 3102°35 70 Wright and Downs—Spectrum of the In the next experiment platinum electrodes were used in order to determine whether the lines were due to copper or to the gases through which the discharge passed. This plate upon . development showed exactly the same series of sharp lines. Hence they must be due to some one of the gases, or the oxides of them, which occupy the intra-polar space. To settle the location of the lines conclusively the camera was adjusted so as to include the region in the first spectrum in which they were supposed to be. An ordinary dry plate without staining was used and an exposure of an hour given it. After the development of the plate the lines sought were seen sharp and clear, which proved conclusively that they were in the ultra-violet of the first spectrum. The work of ascertaining to what one of the intra-polar gases, or oxides of them, the lines belonged, was next taken up. A careful comparison of the recorded lines of the hydrogen spec- tra did not reveal any coincidence. In the Report*-of the British Association for the Advancement of Science some lines due to water vapor are given, and a comparison with the lines studied in this investigation showed that a large number of them were apparently in exact coincidence. These lines were discovered by W. Hugginst and -also by G. D. Liveing and J. Dewart at almost the same time. Hug- gins exposed a photographic plate to the flame of hydrogen burning in air for one minute and a half and was very much surprised upon developing it to find such a strong group of lines. He traced the lines in this group from wave lengths 3062 to about 3290, the upper limit of which is about the same as that of the lines studied here and the lower limit consider- ably below. Upon placing a spirit lamp before the slit of the spectroscope the spectrum is essentially the sarhe, but, as it is less intense, only the strongest lines are seen. In their article Liveing and Dewar give the results of the study of some lines which they had obtained and noted in the spectrum of coal gas burning inoxygen when investigating the spectrum of “Compounds of Carbon with Hydrogen and Nitrogen.” The group of lines which they obtained extends from wave lengths 3062 to about 3210. The same spectrum was given by the electric spark taken, without condenser, in moist hydrogen, oxygen, nitrogen, and carbonic acid gas, but it disappeared if the gas and apparatus were thoroughly dried. Hence they were led to the conclusion that the spectrum was that of water. In conclusion they were not prepared to guar- antee that oxides of nitrogen from traces of air might not have something to do with some parts of the spectrum here observed. * 1886, p. 169. + Proc. Roy. Soe., vol. XXX, p. 576. ¢ Proc. Roy. Soc, vol. xxx, p. 580. Induced Alternating Current Discharge. tL In a second* communication to the Royal Society, “ On the Spectrum of Water,” Liveing and Dewar state that the spec- trum which they had figured in the article just mentioned did not by any means exhaust the spectra of flames observed by them, but that it was as much as they were able at that time to trace to water as its cause. In a third} article they give the results of a very searching investigation of the spectrum of the oxy-hydrogen flame. By making long exposures they obtained photographs of the oxy-hydrogen flame, showing closely set lines from wave lengths 2268 to 4100, with traces of lines beyond those limits. The whole spectrum appeared to consist of a rythmical series of lines, the strongest of these series being the one first described. Liveing and Dewar refer to an articlet by M. Deslandres in which he states ‘that the first band of the water spectrum (i. e. the group beginning at wave length about 3063) includes a series of rays which reproduce, line for line, at the same dis- tance and with the same relative intensities, the band A of the solar spectrum; and that the second band (i. e. the group beginning at a wave length about 2811) includes a series cor- ‘responding to B, and that in the third a may be found to be reproduced. lLiveing and Dewar were not able to make out such an exact correspondence between the lines of the water spectrum and those of A, B, and a as M. Deslandres’ words seem to imply, but nevertheless the similarity of the grouping was very remarkable. To study further the question whether these lines were due to water vapor, or to the effect of water vapor upon the gases occupying the intra-polar space, a plate was exposed to the discharge when it took place under special conditions, namely that the upper half of the plate should be exposed to the dis- charge when it took place in ordinary air and the lower half of the plate to it when it took place in air containing a large amount of moisture. [or the purpose of supplying a large amount of water vapor to the air about the discharge during the exposure of the lower half of the plate, water was boiled ina glass flask, containing a cork stopper through which passed a glass tube so shaped that by its means the issuing jet of steam could be easily directed into the flame. In this experiment the upper half of the plate was exposed for forty-five minutes to the discharge in ordinary air, and the lower half for the same length of time when steam was sup- plied to it. In the latter case the time during which the light actually fell upon the slit of the spectroscope was not really as * Proc. Roy. Soc., vol: xxxiii, p. 274. + Trans. Roy. Soc., vol. elxxix (A), p. 27. ¢t Comptes Rendus, vol. ¢, p. 854. 72 Wright and Downs—Spectrum of the great as in the first case owing to the fact that the flame was somewhat disturbed by the jet of steam. Nevertheless, when the plate was developed, the lines upon the lower half came out. much more strongly than upon the upper, the only difference being in the intensity, which shows that if the lines are not directly due to water vapor, the latter greatly facilitates the development of lines due to the several gases of the air, or combinations of these gases, probably both. Tables of wave lengths of the different constituents of the atmosphere were next examined to see if any of the lines given coincided with the lines studied. In a note* regarding a very careful search which he made for the line spectrum of hydrogen in the oxy-hydrogen flame, Liveing states that he failed to find the slightest trace of any one of the hydrogen lines. As oxygen is a very large constituent of the air, it was very important to make a comparsion of the lines due to it, which were catalogued, with these lines. There were found, how- ever, to be no coincidences, which indicates that none of the lines are due to oxygen. Just before the completion of this work an article by G. Berndt appeared.t While studying the spectra of several metals, Berndt had noticed in all of them a banded spectrum extending from wave lengths 5100 to 2000, a portion of which covers the region in which the banded structure previously described was located. He concluded from his observations that the banded structure in the visible portion of the spectrum of different metals is dependent upon the presence of oxygen and also suggested that the banded spectrum in the ultra-violet is due to air. Atthe close of his article Berndt publishes a list of nitrogen lines obtained by an induction discharge in nitrogen free from oxygen. Upon comparing them with the lines studied in this investigation, a number of apparent coincidences were seen, making it probable that some of these lines at least were due to nitrogen. The last experiments made were a continuation of the attempt to ascertain if possible to which one of the intra-polar gases the lines belonged. As the solar light in this region is very feeble, the light from a voltaic carbon are was passed through bulbs containing nitrogen tetroxide, and an attempt made to secure the absorption lines due to it. The lower half of the photographic plate was exposed to the light which passed through the nitrogen tetroxide and the upper half to the light from the discharge alone to see if any of the lines coincided with the absorption lines due to the nitrogen tetrox- ide. Although several exposures of nearly two hours were * Phil Mao: t vol. xxxivy p: oie + Drude’s Annalen der Physik, No. 4, p. 788, 1901. Induced Alternating Current Discharge. 73 made to get the absorption spectrum, the lines obtained were so feeble and so few in number that it was impossible to make a comparison of any significance in determining the coincidences with the lines investigated. After considering the large amount of work done in the past in studying this strongest group of the so-called water spectrum an@ also the work done in this investigation, it seems hardly possible that the lines are due to water vapor alone, but rather that they are due to the various constituent gases of the air and combinations of them which are facilitated by the presence of water vapor. On the whole, owing to the small amount of water vapor in the air, the experiments seem to point to the conclusion that the lines of this group are due for the most part to the nitrogen of the air and its oxides. Sloane Physical Laboratory, June 11, 1901. Sco NE Pe Le. EN DEE EkG-E NC E. I. CHEMISTRY AND PHYSICS. 1. The Ignition-temperature of Phosphorus.—The temperatures given in chemical literature as the point at which phosphorus takes fire in the air or in oxygen vary from 38'1 to 75°. The amount of moisture present in the air or oxygen probably has a great effect upon this temperature, as it is well known that phos- phorus may be distilled in perfectly dry oxygen without ignition. F. H. Eypman, JR. has now made a careful series of experiments with air, oxygen, and air diluted with an equal volume of carbon dioxide. These gases were used in a moist condition, in fact they were made to bubble through molten phosphorus which was under water in a test-tube. The test-tube, which contained a thermometer, was placed in a flask of water, and the latter was gradually heated until the phosphorus was ignited by the gas which bubbled through. With air the observed temperatures of ignition varied from 44°9 to 45:4°, with an average of 45°; with oxygen the mean of four closely agreeing results was 45:2; while with the mixture of air and carbon dioxide two experiments gave 45° in each case. The interesting conclusion is reached that the ignition-point of phosphorus does not vary with the concentration of the oxygen. ‘The author intends to extend his experiments by using dry air and oxygen.—fecueil Trav. Chim. Pays- Bas, xix, 401. H. 1. Ww. 2. The Composition of ‘‘Caro’s Acid.”—Baxnyer and VILLIGER have made an investigation of this powerfully oxidizing acid, which is prepared either by treating a persulphate with con- centrated sulphuric acid, by the electrolysis of rather concen- i4 Scientific Intelligence. trated sulphuric acid, or by the action of concentrated sulphuric acid on hydrogen peroxide. These chemists were unable to obtain the acid in a pure state or to prepare any salts of it, but — by studying the behavior of hydrogen peroxide, Caro’s acid, and persulphuric acid they found that the first substance could be determined and destroyed in the presence of the others by means of potassium permanganate, and that Caro’s acid liberates iodine immediately from acidified potassium iodide solution, while with persulphuric acid this liberation is very slow. By applying these facts they were able to determine the ratio of the oxygen to the sulphuric acid liberated by the acid under investigation, and the conclusion was reached that it has the composition represented by the formula H,SO,. The probable structure of the acid is ieee Ore The change from persulphuric acid to Caro’s acid, which takes place gradually in solution, is as follows : H,S,0,+H,0 = H,SO,+-H,SO, Caro’s acid is then gradually changed in solution to sulphuric acid and hydrogen peroxide according to the equation, H,80,+H,0 = H,S0,+H,0,. In the presence of strong sulphuric acid the last equation is reversed to a certain extent, but the product still contains much unchanged hydrogen peroxide.— Berichte, xxxiv, 853. 4H. L. W. 3. Vitrified Quartz.—In a recent discourse delivered before the members of the Royal Institution in London, W. A. SHENSTONE exhibited the production of tubes and bulbs of fused silica. The first stage of the process consists in getting quartz into a condi- tion in which it will not fly to pieces when heated. ‘This is done by heating it in small fragments to about 1000° and throwing it quickly into cold water. It then becomes white and enamel-like, and after the treatment has been repeated it may be thrust sud- denly into the hottest part of an oxyhydrogen flame without splintering to the slightest extent. Silica becomes hot enough to be worked only above the melting-point of platinum, and it is only the hottest part of the oxyhydrogen flame, just beyond the inner blue cone, where the temperature is sufficient for the pur- pose. To produce tubes and other vessels from the enamel-like silica, two fragments of it held in platinum forceps are heated and pressed together until they adhere, then other lumps are added, one at a time, until a rough rod is produced. This rod is afterwards reheated and drawn out into a finer rod about 1™™ in diameter. Some of these fine rods are then bound round a stout platinum wire, while soft, and heated until their sides adhere. The rough tube thus made is reheated, drawn out and, after being closed at one end, a bulb is blown in it. By applying thin rings of silica to the small bulb, heating until the silica begins to spread and blowing out, it is increased in size, and by drawing out such bulbs long tubes may be prepared. When a silica tube has been produced it may be worked in the flame as easily, though not as Chemistry and Physics. 75 inexpensively, as glass. It may be thickened by adding fresh rings of silica, and all kinds of joints can be easily made. In one respect silica is easier to work than glass. It never breaks when suddenly thrust into the flame, and the finished apparatus needs no annealing. For this work the eyes must be protected by very dark glasses. Vitrified quartz is harder than feldspar, but less hard than chalcedony. When cut with a file it breaks like glass. Its con- ducting power for heat is about equal to that of glass. Its den- sity, 2°21, is much lower than that of quartz, 2°66. It has been found that the coefficient of expansion by heat is only ;4 as great as that of platinum, and much smaller than that of any similar substance that has hitherto been studied. The substance is shown to be remarkably transparent to ultra-violet rays, hence its application to spectroscopic work will probably be important. Perhaps the most remarkable property of vitrified silica is its resistance to sudden changes of temperature. Water may be dropped upon a white hot rod of the substance, or a similarly heated tube of it may be plunged into cold water or even into liquid air without injury. It is suggested that the material will be useful in the preparation of thermometers, for instance, those in which tin is used in the place of mercury, and also for the bulbs of air-thermometers. The interesting observation has already been made that when oxygen and nitrogen (air) are heated above the melting-point of platinum in a silica tube, nitrogen peroxide is produced.— Chem. News, 1xxxiii, 205. H. L. W. 4. Influence of Magnetism upon Supersaturated Solutions.— Attempts by A. pE HemprinneE to produce the crystallization of supersaturated solutions and supercooled fusions by exposure to a strong magnetic field were in every case failures. By indirect means it was shown that crystalline substances in saturated solu- tion showed a distinct orientation in this field; hence it is con- cluded that crystallization in these cases does not depend upon the orientation of the molecules.—Zeitschr. physikal. Chem., MEV. 223. H. L. W. 5. The Preparation and Properties of Ammonium Cyanate.— This salt was not obtained as a solid in a pure condition by Liebig and Wobler, on account of its transformation into urea. WALKER and Woop have recently succeeded in preparing this famous salt by mixing solutions of cyanic acid and ether at —17°, and also by mixing the gases diluted by air. It is a colorless salt which is readily soluble in water, and this solution gives the reactions for the cyanate and ammonium ions. Upon heating it fuses at 80°, but then becomes solid again, being changed to urea.— Jour. Chem. Soc., \xxvii, 21. HH, Tj. W. 6. Detection of Selenium in Sulphuric Acid.—It has been found by Jouve that crude acetylene gas furnishes a very delicate test for selenium in sulphuric acid by precipitating it in the ele- mentary condition. Hestates that a red coloration can be observed when the acid contains only one part of selenium in 100,000, while 76 Scientific Intelligence the test with sulphurous acid is only about one-tenth as delicate, The sensitiveness of the reaction is increased, as far as the rapid- ity of the coloration is concerned, if the acetylene contains hydro- chlorie acid fumes. — Chem. News, |xxxiii, 248. H. Ls We 7. Qualitative Chemical Analysis, Organic and Inorganic ; by F. Mottwo PeERKIN. 8vo, pp. villi, 266. London, 1901 (Longmans, Green & Co.).—This is an elementary text-book that is unusually full in respect to the explanations and chemical equa- tions that are given, and good judgment has been used in the selection of the analytical methods that are employed. The book presents a novel feature in including a treatment not only of the more common organic acids, but also a variety of other organic substances, such as aldehydes, alcohols, acetone, glycerol, several sugars, starch, a few bases and glucosides, and about a dozen of the more important alkaloids. H. L. W. 8. Viscosity :of Argon.—Lord Rayleigh obtained the value 1:21 for the frictional coefficient of argon at ordinary tempera- ture. HuGo Scuutrze, thinking that the apparatus employed by Lord Rayleigh was not suitable for obtaining extreme accuracy, has made new determinations. Extraordinary care was taken in cleaning the glass of the apparatus and in purifying the argon. It was possible to force the argon back and forth through the glass tubes, and to measure the time of flow with accuracy. ‘The gas was also heated in a well devised calorimeter and a much improved manometer of water in combination with mereury was employed. The relative value obtained by the author is a little larger than that obtained by Rayleigh. Tables are given of the relative values of the coefficient with reference to air. The tem- peratures varied from 15° ©, to 183° C.— Ann. der Physik, No. 4, pp. 140-165. JeeEs 9. Conductivity produced in Hydrogen and Carbonic Acid Gas by the Motion of Negatively Charged Ions.—This subject has been studied by Professor TowNnseNnD and Mr. P. J. Kirksey by the aid of the Rontgen rays, which were allowed to fall on the gases contained between two parallel plates—m aintained at various differences of potential. The authors conclude that a negative ion makes 11°5 collisions per centimeter in hydrogen at 1™™ pressure, and 29 collisions per centimeter in carbonic acid gas at the same pressure. It was found that the mean free path of an ion is longer than the mean free path of a molecule in the following ratios: 4°8: 1 in hydrogen. 4-6: ] in carbonic acid gas. 4°3: 1 im air. — Phil. Mag., June, 1901, pp. 630-642. Te hs 10. Electromotive Force of the Clark- and the Weston -Cell.— An exhaustive examination of the constancy of the electromo- tive force of these elements has been made at the Reichsanstalt, by W. Jarcer and 8. R. Linpecx. This result of their work shows that the criticisms of Cohen, in regard to an instability of Geology and Natural History. { cadmium sulphate and cadmium amalgam in the proportions employed in the cells used at the Reichsanstalt, are not well founded, at least when the thirteen and twelve per cent amalgam is employed.—Ann. der Physik, No. 5, pp. 1-50. ars 11. Electrical Flow in Gases.—L. STarK gives a theoretical discussion of this subject in which he shows under what condi- tions Ohm’s law does not apply to electrical discharges in Plucker tables. The streaming and the phenomena of stratification depend upon the rate of arrival and departure of ions—also upon the amount of ionization and mobilization. Ohm’s law holds for that portion of the discharge in which the ionization remains con- stant. In general there are only portions where this is true. In the neighborhood of the electrodes on account of the transport of free ions by the streaming the ionization is lowered. This happens especially at the cathode. In metallic conductors the force lines coincide with the stream lines. This is not true in the case of electrical discharges in gases. The author discusses also the inner electromotive force developed in electrical discharges. He also takes issue with Professor J. J. Thomson theory of stratifica- tion.— Ann. der Physik, pp. 89-112, No. 5, 1901. a, GE. 12. A Treatise on EHlectro-Magnetic Phenomena and on the Compass and its Deviations aboard Ship ; mathematical, theoreti- cal, and practical ; by Commander T. A. Lyons, U.S. Navy. Volume I, pp. xv and 556. New York, 1901 (John Wiley & Sons). — The. author has undertaken the preparation of an exhaustive treatise upon the use of the magnetic needle on ship- board. This is a subject of the greatest importance as well as of very general interest, and one which it is difficult to find ade- quately treated in any single volume. We have before us the discussion of the physical subjects which form the basis of the practical discussion which is to follow. In order to bring out clearly the nature of: the periodic changes that take place in the earth’s magnetism, as well as the complex relations of magnetism and electricity in general to the ether, the author has begun with the foundation subjects and treats them in a familiar and popular way, so as to be intelligible to one whose studies have not been definitely along this line. For example, he treats of wave motion generally, from the sound wave to those of Hertz, and in the part of the volume following, discusses not only the magnetism of the earth, and the methods of determining magnetic elements, but also the phenomena of magnets and magnetic fields in general. The second volume, which is to contain the discussion proper, will be awaited with interest. Il. GroLocy AND NATURAL HISTORY. 1. The Eocene Deposits of Maryland, by Wiut1am Buttock + Crark and GrorcE Curtis Martin, pp. 1-92; and Systematic Paleontology, by Messrs. Casz, Eastman, Utricu, CLarx, Martin, VaucuHan, Bace, Hoxtick. Pp. 93-331, Plates t-txiv. Mary- 78 Scientifie Intelligence. land Geological Survey. Baltimore, 1901.—The authors of the first paper have drawn the distinction between stratigraphic and paleontologic units; but as they state, ‘‘the former are desig-- nated as formations and members, the latter as stages and sub- stages. As their limits are the same the same name is employed for each.”” Names are given in the following table: Group. Formations or stages. Members or substages. ( W oodstock ; Nanjemoy bb Potapaco Pamunkey. | Aquia. Paspotansa \ Piscataway. The several members or substages are still further divided into zones from which the faunules are obtained which are listed. The zones are apparently discriminated on a lithologic basis . and grouped together on a paleontologic basis into members. The details are carefully worked out and the results are well arranged. The only criticism the writer would suggest is that the discrimi- nation between formation and fauna is not carried far enough to exhibit the full value of the fossils in correlation. If the combination of species, called substages, is supposed to be restricted to the formational divisions, called members, then the paleontological statistics may serve to identify the members; but if the species instead of stopping their existence migrated at the end of each member, the true value of the species in corre- lation is not reached by applying the same name to the forma- tional members and to the faunal substages. The paleontological part has been (as it should be) assigned to specialists. ‘The whole fully keeps up the reputation these reports have held from the beginning of the new administration. H, 8. W. 2. Annual Report of the Geological Survey of Arkansas for 1892. Vol. V.}The Zine and Lead Region of North Arkansas ; by Joun C. Branner, State Geologist. Pp. 1-395, plates 1-38, and figs. 1-92. Little Rock, 1901.—An appropriation made by . the Arkansas legislature in 1899 provided for the publication of this volume, the investigation for which was completed before the official termination of the Survey in 1893. The chief part of the volume is concerning the zine and lead deposits of northern Arkansas. These ores were originally deposited in sedimentary beds of the Ordovician age, and by infiltrating waters were carried through fissures to form secondary deposits in cavities in rocks of later age. The Paleozoic faunas of northern Arkansas, Chapter vil, pp. 268-362, by Henry 8. Williams, is here published for the first time. The report was prepared during the progress of the Sur- vey, its detention having been caused by lack of appropriations for its publication. The report contains a classification of the Paleozoic formations of the northern part of the State, with lists of the contained faunas, and discussion of several complex strati- Geology and Natural History. — 79 graphic problems, one of which concerns the geological horizon of the manganese deposits. H. S. W. 3. Summary Report on the operations of the Geological Sur- vey of Canada for the year 1900; by Gore M. Dawson, Deputy and Director. Pp. 1-203. Ottawa, 1901.—This last report of the late director, issued after his death, exhibits the customary energy and progress of the Survey in developing the geological resources of the Dominion. The commercial value of mica and molybdenite has led to their special examination by experts. The mica (phlogopite) of Que- bee and Ontario is of special value for electrical purposes; but the most satisfactory market is shown to be the United States and not Great Britain. Molybdenite samples subjected to criti- cal test show that cobbing and hand-picking may be carried on with the ordinary ores profitably, but when the ore is of low grade, rolling, screening and jigging are not economically suc- cessful. Salt is reported from the ‘‘ Medina formation” in a well near St. Grégoire, Beauce Co., Quebec. The Crows Nest coal field-is estimated to contain 22 billion tons of possibly workable coal. The coal is of Cretaceous age, and has excellent coking qualities with small percentage of ash. A similar Cretaceous coal area has been discovered between the 55th and 57th parallels of latitude in British Columbia; and anthracite has been found in the region of the head waters of the Skeena and Stikine Rivers. The fuel near Forty Mile Creek on the Yukon is Tertiary lignite, but samples of anthracite are reported from a locality west of Lake Marsh. Study of the formations of the Brockville map-sheet (Ontario) has revealed the fact that fossils referred to in earlier reports as obtained from beds of Potsdam age, are found in what are known as transition beds between the sandstone and the lime- stone. The best specimens were obtained ‘from weathered sur- faces of a siliceous limestone which represents the base of the Calciferous formation.” Investigations in the Cambro-Silurian slate region of York and Carleton Counties, New Brunswick, have developed the fact that the fossiliferous Silurian rocks overlie discordantly the Cambro- Silurian slates; and Cambrian fossils have been found in black shales near Benton, Carlton Co., belonging apparently to the lower series. Nevertheless a satisfactory solution of the difficult stratigraphical problem is not yet reached. The Arisaig formations have received special study with the result of their reduction to four divisions, to which the names Stonehouse, Moydart, McAdam and Arisaig formations are applied. H..S. W. 4, A revision of the genera and species of Canadian Paleozoic Corals—the Madreporaria aporosa and the Madreporariarugosa ; by Lawrence M. Lampe. Geological Survey of Canada. Con- tributions to Canadian Paleontology, Vol. 1V, Part II, pp. 97-197, 80 | Scientific Lntelligence. plates vi-xvin. Ottawa, 1901.—The paleontologic and literary part of this report leave nothing to be desired, and the drawings are admirable portrayals of the characteristics of the species drawn by the author and thus having all the value of scientific definition. A few notes regarding correlation are of importance. It is suggested by Dr. Whiteaves that “ American stringoceph- alus zone” be applied to the Manitoba horizon to avoid confusion with the European zone of that genus (p. 104). Attention is called (p. 124) to Billings’ identification of the Gaspé limestone, No. 8 of Indian Cove, Gaspé, “as nearly of the age of the Oriskany sandstone.” The application of ‘“ Lower Helderberg ” to the fauna of L’Anse a4 la Barbe, Baie des Chaleurs, Que., is commented on; the opinion of Dr. Whiteaves cited that the limestone seems to be most nearly equivalent to the Guelph formation of Ontario, Ohio and Wisconsin (see foot-note, p. 129), and the remark is made that the term Lower Helderberg has been used in part I of this volume, “in a sense as comprehensive as that evidently implied by the term in the ‘Geology of Canada.’” This comprehensive use of the term should be con- sidered in reading the review of the writer on the first volume. (See this Journal, ix, p. 155, 1900.) H. S. W. 5. The structural relations of the Amygdaloidal Melaphyre in Brookline, Newton and Brighton, Mass. ; by Henry T. Burr. Bull. Mus. Comp. Zool. at Harvard College, Vol. 38, Geol. series, Vol. V, No. 2.—This paper once more brings to our notice the dispute regarding the structural relations of the melaphyres of the Boston Basin and the associated sedimentary rocks. Mr. Burr concludes, from observations made in the field, that, for the area studied, the beds of melaphyre are strictly intrusive in character; and in so doing he takes very decided and pointed exception to the views of Prof. W. O. Crosby, who, as is well knuwn, considers the main beds, at least, to be unquestionably contemporaneous flows. The facts upon which Mr. Burr bases his main conclusions are: ‘1. The conglomerate associated with the melaphyre con- tains no fragments of it. 2. The contacts, wherever found, are igneous in character. 3. The melaphyre is seen in contact with sediments varying from the coarsest of the conglomerate to the finest of the slate. 4. The distribution of the melaphyre showsit to be discordant with the structure of the sediments under any interpretation of the latter that has been offered.” In the May number of the American Geologist, Prof. Crosby has taken occa- sion to call into question some of Mr. Burr’s statements and con- clusions and to present again the arguments in favor of the con- temporaneous character of the melaphyre. He calls attention to the fact that, contrary to Mr. Burr’s statement, fragments of melaphyre are found in the overlying conglomerate; and shows, we think, very clearly, that, so far as the character of the con- tacts and the other relations to the associated sedimentaries are concerned, the melaphyres are as well considered contemporane- ous as intrusive sheets. He points out that the main melaphyre Geology and Natural History. 81 beds have all the characteristics of lava flows, and raises the very pertinent question why, if, as is claimed, the melaphyre and trap dikes of the region are petrographically hardly distinguishable rocks, we find ahuge dike of melaphyre amygdaloidal and scoria- ceous in texture, while a three-foot dike of trap is a dense, holo- crystalline rock. Attention is also called to the significant pres- ence of well defined beds of tuft, a fact entirely overlooked by Mr. Burr. In the light of these and other facts brought out by Prof. Crosby it would seem, even to the casual reader, that the weight of evidence is overwhelmingly in favor of a contemporane- ous origin and that Mr. Burr’s paper is thus the unintentional cause of strengthening an exactly opposite conclusion from the one given in the paper itself. In view of the fact that Prof. Crosby had, some twenty years ago, fully described the “new interpretation” of the structural relations of the Chestnut Hill slates to the northern conglomerate, now presented by Mr. Burr, the word “new” seems, to say the least, rather superfluous. Cc. H. W. 6. Sul Periodo di forte Attivita esplosiva offerto net mest Aprile-Maggio 1900 dal Vesuvio; par R. V. Matrevucci. Boll della Soc. Sism. Ital., Vol. vi, 8°, 110 pp., 6 pl., 1901.—This volume gives a detailed account of the phenomena and products attend- ing the period of explosive activity of Vesuvius at the time men- tioned. The author begins with a general description of the phenomena of the explosions, discusses their frequency and de- scribes their effects on the crater, etc. He then discusses the material ejected, the powder and sands, the lapilli, scoria, bombs and solid blocks, each of which is made the object of particular study. An immense quantity of these blocks were ejected, a photograph being given of one containing twelve cubic meters of material. Chemical determinations show that the ejected matter was somewhat lower in silica than previous effusions of the vol- cano. Some calculations of the force involved in projecting the large block mentioned above are also given, showing that it reached an altitude of about 300 meters. The work closes with a discussion of volcanic explosions, both former ones of Vesuvius and of other volcanoes as well. The plates are very beautiful and interesting heliotype repro- ductions of instantaneous photographs, showing the form and development of the smoke cloud from the great vapor puffs in different stages and the swarm of projected blocks in mid-air. The work as a whole is an excellent contribution to the study of volcanic outbreaks in general and to the history of Vesuvius in particular. ja 7. Der Vulcan Etinde in Kamerun und seine Gesteine ; von E. Escu. Sitzber. d. k. preuss. Akad. d. Wiss. zu Berlin, phys. mat. Classe 1901, x11 and xvii, pp. 41, figs. 22.—Etinde is one of the smaller volcanos of the Cameroon Mts., a volcanic group cov- ering some 150 square kilometers, close on the coast of western Central Africa, and whose highest peak attains an altitude of JULY, 1901. Am. Jour. Sci1.—FourtH Series, Vou XII, No. 67. 6 82 Serentific Intelligence. 13,700 feet. The volcano is greatly eroded and is one of the oldest of the group. Its age cannot be exactly told, but the tuffs and ashes of the group cover sediments of Upper Cretaceous age and contain plant remains representing the present flora. a The chief interest lies in the igneous rocks which form the lavas of the mountain. They are of highly alkaline types, and among them are represented leucitite with phenocrysts ot nephe- lite, hauynite and augite, occasional pyrite and perofskite in a eroundmass of leucite, nephelite and augite, with some apatite SiO, Al,O3 Fe,03 FeO MnO MgO CaO Na,O K.O H,O P20; TiO; SO; Cl CO. 1/46°48 19°00 4:74 2°30 tr. 2°49 4°35 846 6°78 3:31 “15 1:22. 19) .08) 362-599 901 2/40°10 15:27 10:13 1°85 -08 459 12:08 4:78 3°34 2:93 -87 3464 — — -23=— 99-89 3/39°97 17:30 7:41-3:05 :-09' 3°82 1053 5:14 3:56 4:11 “84 3:34 0b) eos onao 4|39°30 13°66 7:42 4:45 -08 4:46 11°37 5°78 1:44 4°53 -85 3°62 2.1% “48 sib== 99-%6 5/39°37 16°50 2°28 7-97 “06 448 1022 4:73 3°38 4:77 13 B31 2:49.09) ot OF 6/38°39 12°64 7°40 6:15 ‘02 6:46 14°17 4:35 2-44 1:62 116 444 cq sae eS aouean 1/40°16 (7'32 7-25 4:00 -08 4:43 11°78 5°99 3.78 1:18 °71 3°21 == ——=1b==10038 1, Leucitite; 2, Leucite nephelinite; 3, Leucite nephelinite, including ZrO,.=:20; 4, Hauynophyre; 5, Hauynophyre; 6, Nephelinite. of a special variety; 7, Normal nephelinite, including ZrO,="35. All analyzed by Dr. Max Dittrich. and perofskite. Leucite composes about one-half the rock. Leu- cite nephelinite also occurs, composed of augite, apatite and ores in a groundmass of augite, leucite and nephelite, the latter pre- dominating. Hauynophyres occur containing phenocrysts of hauynite, augite and ore grains in a groundmass much like those just mentioned. A nephelinite of a special variety in which the nephelite in the groundmass is not fully individualized is also described. Phenocrysts of augite, hauynite, apatite and titanite le in a clear colorless base with microlites of augites and ore grains. ‘The colorless base is held to be partly individualized nephelite. Nephelinites of normal type are also described and the twinning and optical properties of the larger individuals is made a special study. The accompanying table of analyses will be of interest to petrographers. tae 8. Sdéndre Helgeland af J. H. L. Voet. (Norges Geologiske Underségelse No. 29, 1900, Kristiania, 8°, 180 pp., | map, 21 figs.) —This work is devoted to an account of the geology of Helge- land, the southern part of the district of Nordland in northern Norway. With the exception of a small area of Archean the region is composed of metamorphic schists and limestones, prob- ably of Cambrian-Silurian age. The author divides the metamor- phosed sediments into mica schist, marble, and a later gneiss. In addition there are extensive areas of eruptive rocks consisting mainly of soda granite and gabbro. Sometimes these are strongly stretched and squeezed and are then held to have been injected previous to the final dynamic foldings which gave birth to the mountain chains; sometimes they are in normal condition and are then held to be of contemporaneous or later age. Geology and Natural History. 83 A careful study of the morphology of the region is given by the author, the quaternary geology by him and Rexsrap, and the work is concluded with an account of the silver ore veins of Svenningdalen. A résumé in German is given at the end. ‘The work through- out is full of the suggestiveness in ideas and thoroughness in details that are characteristic of the author. Livi Bs 9. Classification of Iyneous Rocks.—E. von FEDEROV, in an article published in the Journal of the Mineralogical Society of St. Petersburg in 1900, page 395, gives a method for the classifi- cation of igneous rocks in which he endeavors to show that it is possible to divide all of the eruptive rocks into types and classes by assigning to them simple symbols like those used in crystal- lography. In considering the general chemical Eom poios of rocks in the order of the oxides, 1, R,O; 2, RO; 3, R,O,; 4, RO,, it be- comes evident that all of the ordinary rock- -making minerals may be expressed by symbols of four numbers. Thus, quartz (0001), corundum, hematite, etc. (0010), spinel, magnetite (0120), olivine (0201), diopside and actinolite (0101), garnet (0323), biotite (1223), phlogopite (1313), anorthite (0122), orthoclase and albite (1013), leucite and aegirite (1012), nephelite (8089). The reckoning of the different compositions is based on a reg- ular tetrahedron whose summits are the points (1000), (0100), (0010), (0001); the center of the facesare the points (0111), (1011), (1101), and (1110); the center of the edges (1100), (1010), (1001), (0110), (0101), (0011), whilst the center of the tetrahedron is 1111 Usioe these points the tetrahedron can be divided into 24 sphenoids, to whose centers there correspond 24 typical composi- tions. Of these fundamental types of chemical composition only four are represented, those having the symbols (1423), No. 1, (1324) No. 11, (1234) No. 11, and (2134) No.1v. No. 1 is repre- sented by the periodotites, No. 2 by rocks of the gadbbro group, No. 3 by rocks of the diorite and diabase groups, and No. 4 by’ rocks of the granite and syenite groups. The division of the types into classes is done by dividing the sphenoids into 24 sphenoids of the third order. By the aid of the principles of zonal crystallography the author explains the process to be followed to obtain the chemical composition when the indices corresponding to the third order are given, and how to determine if this composition is a typical one (represented by an interior point of the sphenoid), one intermediate between the classes, one of transition between the great types, or an extreme rock, The author shows how these may be determined by graphic methods and makes a number of applications of his system. ‘The main body of the text is in Russian, but a short résumé in French permits the salient features of the article to be understood by those who are not versed in that language. i, Vi Bs : Scientific Intelligence. CD ho 10. Boletin del Instituto Geologico de Mexico, Num. 18, Las Rhyolitas de Mexico. Primera parte, por EKzEquiEL ORDENEz, Mexico, 1900, 4°, 75 pp., 1 map.—This bulletin is the first of a series to be issued on the eruptive rocks of Mexico. ‘The aim of the series is to make known the great number of varieties of rocks, their geographical distribution and field relations, their physiographic features, and economic importance as building ~ material and soil. It is also expected to bring about uniformity in the petrographic nomenclature of Mexico. The first part of the bulletin is devoted to historical notes, principal subdivisions of the rhyolite family, macroscopic characters of the rhyolites and _ geographic distribution of the rhyolites in Mexico. ‘The second and greater part is devoted to the description of various localities under the following four heads: (1) The “ Bufas,” (2) The Nava- jas, (3) Other regions of the Central Plateau, (4) The western Sierra Madre. The term “ bufa,” for many years applied by the miners in the region about Guanajuato and Zacatecas to curiously ~ shaped masses of rock crowning the summits of prominent points, is adopted by the writer to designate these isolated erosion rem- nants of rock of a rhyolitic nature. The bulletin closes with a chapter upon the age of the rhyolites and a résumé in French. Points of general interest are: the great distribution of the rhyolites, surpassed only by the andesites, in a belt reaching southward from the northern boundary of the country through the western Sierra Madre into the Central Plateau to the vicinity of the city of Mexico; the greater part of them were erupted from the end of the Miocene to the middle of the Pliocene. The succession of Tertiary eruptives, based on the evidence thus far presented, is: (1) Granites-Granulites, (2) Diorites-Diabases, (3) Andesites-Dacites, (4) Rhyolites, (5) Dacites-Andesites, (6) Basalts-Basaltic andesites. There are two modes of occurrence, namely, in the western Sierra Madre from fissures parallel to lines of relief, and. in the Central Plateau from local vents. The intimate connection between the geologic structure and the erup- tive rocks of the western United States and the mountains of Mexico is also of interest. HH.) Be ll. Brief notices of some recently described minerals.—Concu- irE. This name has been given by Agnes Kelly to the calcium carbonate forming a considerable part (often associated with cal- cite) of the calcareous secretions of molluscan shells. ‘This form, unlike aragonite to which it has been referred (G. Rose), is uni- axial and optically negative, as is calcite, and yet it differs from the latter mineral in some other characters. The most conspicu- ous difference is in the specific gravity, which is 2°87 for conchite, and 2°71 for calcite. It shows no twinning and has higher refrac- tive indices; thus « is 1°524 for conchite and 1°486 for calcite. Conchite is an unstable compound changing into calcite at 300° to 310°. Various natural incrustations (Karlsbad, Yellowstone, etc.) were found to consist of conchite; here also belongs the Flos Ferri.—Min. Mag., xii, 363. Geology and Natural History. 85 SELIGMANNITE. A new species described by Baumhauer from the dolomite of the Binnenthal in Switzerland; it is named after the mineralogist, G. Seligmann of Coblenz. The material was too scanty to allow of analysis, but its close identity with bourn- onite in complex crystalline form and the similarity in color and luster and brittleness to sartorite (scleroclase), have led the author to the conclusion that it is a lead sulpharsenite isomorphous with the former species.— Ler. Ak. Berlin, p. 110, 1991. HussaxitE. Of —ga 18 zero, Fo = 0. When the capacity of the tube is reached, the conjugate flows meet at the critical section with equal velocities, densi- ties, and pressures, and are, consequently, there indistinguish- able. On either side of this section, which of the two possible flows is actual, is determined by conditions existing on that side. Hence it is possible for the actual flow to have a veloc- ity less than sound on one side and greater on the other. That this is not a natural case appears as follows: On either side of J. W. Davis-—Motion of Compressible Fluids. 1138 the critical section the density in one of the flows is every- where considerably greater than in the other, and, therefore, the mass of fluid is considerably different in the two flows. Now suppose the quantity, M, of fluid passing any section in unit time increases from zero to the capacity of the critical section. Through all this range the velocity of the fluid everywhere in the tube remains on one side of the velocity of sound until at the limit it may still remain everywhere on that side, or may pass from one side to the other at the critical section. Hence one solution at the limit is continuous with all the previous solutions, and the other is isolated. The iso- lated case cannot originate from a flow that increases to the capacity of the tube, for this would involve at the final instant in the total mass of the fluid an instantaneous change of con- siderable magnitude. Neither can it exist in a tube that returns into itself and has only one section of least area or of greatest potential. : The behavior of the conjugate flows in any space may be summed up as follows: If in either flow the velocity of the fluid at any point is less than the velocity of sound at that point, then, in general, the velocity of the fluid is everywhere less than the velocity of sound, or may reach the latter as a limit. If the velocity is greater than that of sound at one place, it is greater at all places, or may diminish to the velocity of sound as a limit. One conjugate velocity is everywhere greater than the other conjugate velocity, or may diminish to equivalence with the latter as a limit where each becomes equal to the velocity of sound. The places of greatest velocity and least density in one case are the places of least velocity and greatest density in the other. In the approach to a gorge or place of higher potential the more rapid flow is retarded and the other is accelerated; the recession from a gorge or place of higher potential produces inverse effects. Under very special conditions the velocity of the fluid is not limited by the velocity of sound. There is an inferior limit, p,, to the density of a compres- sible liquid, whereas the density in the pressure-density formula, (3), of that liquid, and, therefore, in equation (4), is continn- ous from zero to infinity. Hence the foregoing conclusions have a limited application to liquids. Suppose the line p = p, to be drawn in figure 1. If now the actual motion is such that both intersections of (4) and (2) always occur on the posi- tive side of the line p =p, for every elemental tube of flow, there is also a conjugate motion; otherwise, there is no conju- gate motion, and the conclusions of the preceding paragraph are only partially applicable. 114 JS. W. Davis—Motion of Compressible Fluids. Nore. On page 177, volume x of the Memoirs of the Man- chester Literary and Philosophical Society, 1887, in a para- — graph of a paper entitled ‘“* On the Flow of Gases,” read before the Society November 17, 1885, Professor Osborne Reynolds notes that, when a gas is flowing through a channel of variable section, the pressure at every point of the channel may have in general either of two unequal values; whereas, if the fluid were incompressible, the pressure could have but one value. This is the only published notice the present writer has seen of the dual values that occur in problems relating to the motion of compressible fluids. The apparent governing effect of the velocity of sound upon the behavior of compressible fluids, is set forth in an interesting manner by Professor Reynolds in the paper just mentioned, and by Hugoniot in a communication published in: Comptes Rendus, page 1178, volume cili, 1886, in the case considered by these authors of a flow through an orifice or minimum section. The present writer has recently been perplexed by meeting with the dual values in several problems. Professor Horace Lamb of Man- chester, England, who courteously examined one of the prob- lems, and rendered the solution more definite by employing the method of discussing the intersections of the curves (2), (4), as followed in this paper, expressed the decided opinion that only those values which are continuous with the observed or assumed values at a given point belong to the problem in hand. Impressed by this statement, the writer set about inquiring into the meaning of the values that appear, only to be disregarded, with the result, arrived at in the foregoing paper, that they belong to a separate motion related to the actual motion, as the branches of a hyperbola are related, by having the same equations and the same parameters. Norton— Action of Sodium Thiosulphate, ete. 115 Arr. XII.—The Action of Sodiwm Thiosulphate on Solu- tions of Metallic Salts at High Temperatures and Pres- sures; by Joun T. Norton, JR. [Contributions from the Kent Chemical Laboratory of Yale University—C. ] THE use of sodium thiosulphate as a substitute for hydrogen sulphide in effecting precipitations and its application in ‘the case of arsenic, antimony, copper and platinum was suggested by Himly* before the middle of the present century. Thirteen years later Vohl}+ and Slater,{ independently, drew attention to this use of sodium thiosulphate and extended the investiga- tion to salts of tin, mercury, silver, gold, lead, bismuth and cadmium. Slater in addition studied the action of sodium thiosulphate upon chromic acid, molybdates, ferrous and ferric terrocyanides, ferric sulphocyanide and potassium perman- ganate. Following out these lines, the precipitation of copper together with arsenic and antimony by treating with sodium thiosulphate the hot solution containing sulphuric acid and the separation of these elements from tin, zine, iron, nickel, cobalt and manganese has been advocated by Westmoreland§ ; and quite recently Faktor| has studied the action of sodium thio- sulphate upon neutral salts of several of the elements men- tioned, as well as the modifying influence of ammonium chloride and other salts upon the course of the reaction. Subsequently to the work of Himly, Vohl, and Slater, Chan- cel*| developed his well-known method for the precipitation of aluminum as the hydroxide and its separation from salts of iron by boiling with sodium thiosulphate the nearly neutral solu- tion, containing the salts of aluminum and iron, at suitable dilution; and upon an extension of the principle of Chancel’s separation of aluminum from iron, Stromeyer** founded his well-known processes for the separation of titanium and zir- conium from iron. The latter process appears to be fairly trustworthy; but of Chancel’s method, although it has met with wide acceptance, it was shown by Wolcott Gibbs very soon after its announcementtt that it fails to bring about complete separation of alumina within a reasonable period of boiling, and this result has been confirmed by Zimmerman,{t who has shown that the boilmg must be continued fifteen hours in order to complete the precipitation of the alumina. * Ann. Chem. (Liebig), xliii, 150. + Ann. Chem. (Liebig), xcvi, 237. t Chemical Gazette, 1855, p. 369. § Jour. Soe. Chem. Ind., v., 51. || Centralblatt, 1900, ii, 20, 67, 239, 594. [ Compt. rend., xlvi, 987. ** Ann. Chem. (Liebig), cxiii, 127. ++ Zeitschr. anal. Chem,, iii, 389. tt Inaug. Diss., Berlin, 1887. 116 Norton—Action of Sodium Thiosulphate on It was shown by Dr. Gibbs that when the treatment of salts of aluminum by thiosulphate was carried on in sealed tubes under pressure at 120° C., the precipitation of alumina was complete, and further that the precipitation of sulphides of nickel, cobalt and iron, though partial under ordinary atmospheric pressure, was made complete by heating in sealed tubes to 120°-140° C. In repeating the experiments of Dr. Gibbs qualitatively and extending them, I have made use of the well-known Pfungst tube to secure the necessary pressure. In each experiment a test tube containing the mixture of an excess of sodium thio- sulphate with the salt whose action was studied was placed within the Pfungst tube containing some water, the cover of the latter was set in place and firmly bolted upon a washer of lead, and the whole was submitted to temperatures varying - from 140° to 200° C. for an hour by immersing in a bath of parafiine. After cooling, the test tube was taken ont, the pre- cipitate was filtered off, and the filtrate tested by appropriate reagents to determine the completeness of precipitation. The following table records the details of these experiments : Action oF Na,S,O, ON SALTS UNDER PRESSURE. Salts used. Precipitates. Degree of Precipitation. Sulphides. NisO, NiS +8 Complete. CoSO, CoS +8 «“ FeCl, FeS +8 “ ZnSO ZnS +8 HP PbO,(C,H,0), PbS +S ce He(NO,), Hes +8 “ AgNO, Ags +5 * CusO, CuS,(Cu,S), + 8 “s CaSO, CaS + 8 «“ KSbC.H,O, Sb,S, + § “ Bi(NO,), Bi,S, + 8 «“ Hydroxides. NH,Al(SO,),12H,O A1O,H, + 8 od K,Cr,0, CrO,H, +8 y K,ZrF, ZrO, = S “ K,TiF, TiO,H, +8 “ Th(NO,), ThO,H, +8 “ Elements. SeO, Se +8 | TeO, Te +58 sf Solutions of Metallic Salts. 117 Salts used. Precipitates. Degree of Precipitation. Sulphides. MnsSO, MnS +8 Partial. AuCl, Au,S +5 (NH,),MoO, MoS,(?) + 8, Red liquid “ Hydroxides. BeCl, BeO,H, Hi } Undetermined. EN EL), UO, Black a Brtcl, Gray, reddish brown liquid ue CeCl, White, yellow liquid ie CaCl, 66 6é eG 4 SrCl, 6s 66 6¢ 4 6¢ BaCl, 66 66 6¢ 66 MgsoO, Trace NH,VO, Brown liquid ¢ I], KAsO, —— fe A perusal of this table brings to light several interesting facts. It appears that salts of nickel, cobalt, iron, zinc, lead, mercury, silver, copper, cadmium, antimony and bismuth are completely precipitated as sulphides by sodium thiosulphate under the prevailing conditions of temperature and pressure. In the case of manganese precipitation is only partial, and arsenic does not seem to be precipitated from an arsenate without the addition of acid. Tin, curiously enough, is not thrown down as the sulphide froma stannous salt, but gives a dirty white precipi- tate of uncertain composition. Salts of aluminum, chromium, titanium, zirconium and thorium are completely precipitated as the hydroxides ; but in the case of beryllium, which one would expect to act similarly, the precipitation as the hydrox- ide is incomplete. Salts of selenium and tellurium are re- duced, aud the elements are precipitated. The precipitates obtained with barium, strontium and calcium were white in a bright yellow liquid, but no study was made of the constitution of precipitate or liquid. In the case of magnesium there was no precipitate. Salts of molybdenum, vanadium and ura- nium gave dark-colored liquids. Thallium yielded a white spongy mass which on compression was reduced to a very small bulk without disintegrating. Salts of gold and platinum gave slight dark precipitates, presumably sulphides, surrounded by dark-colored liquids. The apparatus used in these experiments and described above is easily handled and answers sufficiently well for quali- tative purposes. But, obviously, the introduction into pre- cipitates of foreign matter caused by the action of water on 118 Norton—Action of Sodium Thiosulphate on the glass of the test-tube and porcelain lining of the Pfungst tube precludes the possibility of an exact quantitative study of the reactions involved. For the subsequent experiments, there- fore, conducted upon the same general lines, a digester with an interior cylindrical cavity of about 12°" in depth by 5°" in diameter, and provided with a pressure gauge, was employed. As a container for the solutions to be tested use was made of a platinum cylinder, 4° in diameter and 10™ deep, provided with a loose cover. With this apparatus the following quantitative experiments, which deal with those ele- ments which are precipitated as hydroxides, namely, aluminum, beryllium, chromium, zirconium and titanium, were made. In each case a weighed quantity of the salt taken for the experiment was dissolved in 50% of water in the platinum ves-~ sel, and to this a known amount of sodium thiosulphate was added. The vessel was placed in the digester, and the latter was heated by a Bunsen burner in the customary way until the required pressure was shown on the gauge. The apparatus was then cooled and the platinum vessel removed from the digester. The precipitate was filtered off on ashless paper, ignited, and weighed. Experiments with a Salt of Aluminum. In a series of experiments made according to the method of Chancel, the results of which are shown in Table II, the solu- _ tion in water of a weighed portion of pure ammonium alum was treated with an excess of sodium thiosulphate and boiled vigorously for periods varying from ten minutes to half an hour. TABLE II. ee Oe Amount of Na2S2Q03. Al.O; found. Error. orms. erms. orms. erms. 1. 0537 Large excess 0471 0066 — Mi 0537 oS ge 0397 OL O0== 3. "1083 Ne uh "0931 "0152 — 4, eS 7 5 grms. ‘0979 0158 — Be "1139 2 ee - "1002 °01387— These results substantiate the observations of Gibbs* and of Zimmerman,t+ and show clearly that the boiling of solutions of the aluminum salt and sodium thiosulphate for a reasonable time does not effect the complete precipitation of aluminum as the hydroxide. Table III shows the result of submitting solutions of am- monium alum treated with varying quantities of sodium thio- * Loc. cit. : j Moe mclts Solutions of Metatlie Salts. 119 _ sulphate to a pressure of 20 atmospheres in the digester. It usually required about 40 minutes to raise the pressure to the limit set; but this limit once reached, the digester was allowed to cool slowly. The duration of an experiment was about two hours. TABLE III. Alum taken Amount of as Al.Os3. Na2S203 used. Al,O; found. Error. orms. erms. germs. orms. 0565 5 | 0633 "0068 + lio. 10 1154 °0022+ °11538 3) °1186 0033 + AA28 3 HH ROS Bene °0001+4+ "1126 3 mine? 0016 + sR IPAS) 2 °1120 0008 — "1136 2 “1121 0015 — E28 io "1136 0008 + "1124 2°5 SUL ON "0003 + 1134 OAS) °1133 0001 — This table shows that sodium thiosulphate precipitates alumi- num completely as the hydroxide when pressure is employed. The high results seen im some of the experiments appear to be due to the difficulty of removing by ignition the large amounts of sulphur formed in the action, as well as to the salts mechani- cally included in the precipitate. The amounts of sulphur and contaminating salts present depend upon the amount of thio- sulphate taken ; therefore this should be as small as possible, 2 to 3 germs. being. sufficient to precipitate all the alumina in a gram of alum. When the amount of thiosulphate is reason- ably restricted the weights of alumina accord fairly well with the theory. Experiments with a Salt of Chromium. Up to the time of the completion of this work nothing appears to have been done upon the quantitative precipitation of chromium as the hydroxide by means of sodium thiosul- phate. Slater* and Roset make mention of the action of sodium thiosulphate upon chromic acid, bichromates and neutral chro- mates, but give no quantitative data. Recently, however, F. Faktort has studied the action of sodium thiosulphate on chromium compounds. This investigator has found that if aqueous solutions of potassium bichromate and sodium thio- sulphate are boiled together, a brown precipitate of hydrated Cr,O,, CrO, separates out and the liquid turns yellow owing to 2 Loer eit + Traite de Chimie Analytique, vol. i, p. 479. t Zeitschr. anal. Chem., 1900, xxxix, 345. 120 Norton—Action of Sodium Thiosulphate on the formation of normal chromate. A solution of potassium . chromate is unaffected by boiling with thiosulphate, but in. presence of ammonium or of magnesium chloride the chro- mium is separated rapidly and completely in the same form as with the bichromate, and after continued boiling with an excess of thiosulphate all the chromium present is precipitated. Faktor also found that a solution of chromic chloride is com- pletely decomposed by continued boiling with thiosulphate, ehromic hydroxide and sulphur being precipitated. In the experiments shown in Table IV a weighed quantity of pure potassium bichromate was dissolved in water, a known amount of sodium thiosulphate added, and the whole submit- ted to a pressume of 20 atmospheres in the digester. After cooling, the precipitate was filtered off on an ashless paper,. ignited and weighed as Cr,Q,. TABLE IV. K.Cr.0, taken Amount of as CroQOa. Na.S.0«. Cr.O; found. Error. eTms. erms. orms. erms. A "1330 3 °1341 ‘OOLL + 2. °1330 2.5 "1326 ‘0004 — 3. “1322 2°5 "1318 “0004 + 4, "1303 Z °1303 - "0000 — . ASO 2 °1310 0009 + 6. °1320 2 S22 0002 + The results of these experiments are very satisfactory and show that under pressure sodium thiosulphate precipitates chromium rapidly and completely as the hydroxide. It is advisable to use as small a quantity of thiosulphate as possible in order to prevent the presence of much free sulphur in the precipitate. Experiments with a Salt of Beryllium. In experiments dealing with beryllium’ the salt used was the chloride, a certain amount of which was dissolved in water diluted to a liter and the amount of beryllium present deter- mined by precipitating with ammonia and weighing as the oxide. Measured quantities of this solution were drawn from a burette as required. When a solution of a salt of beryllium and sodium thiosulphate are merely boiled together, nearly all the beryllium remains in solution. It was expected that the use of pressure would throw out all the beryllium, but, curiously enough, when solutions of beryllium chloride and sodium thio- sulphate were submitted in the digester to pressures ranging from 10 to 80 atmospheres, only a partial precipitation of the hydroxide took place. Solutions of Metallic Salts. 121 Experiments with Salts of Zirconium. To prepare a standard solution of the salt of zirconium it was found to be most convenient to heat the double fluoride of potassium and zirconium with sulphuric acid, evaporate to dryness in platinum, dissolve the zirconium sulphate remain- ing in water and enough sulphuric acid to prevent the precipi- tation of the basic salt, and dilute to standard volume. Meas- ured portions of the solution were taken from a burette as required for the experiments. The presence, however, of so large an amount of sulphuric acid as was necessary to keep the zirconium salt in solution tends to decompose sodium thiosul- phate so rapidly that it was found necessary to nearly neutralize the solution with ammonium carbonate before adding the sodium thiosulphate. The solution of zirconium sulphate was standardized by precipitating with ammonia aud weighing as the oxide. In experiment 1 of Table V, the solutions of zirconium sul- phate and sodium thiosulphate were boiled together for a few minutes and then the precipitate filtered off, ignited and weighed as the oxide. In experiments 2-5 inclusive, similar solutions were submitted to a pressure of 20 atmospheres in the digester. TABLE V. ZrO, taken. Na.S.03 taken. ZrO, found. Error. grms. erms. grms. germs. ie "0658 ; 3 "0651 "0007 — 2. "0658 3 0676 0016+ 3. "0666 2; ‘0670 "0004 — 4, "0641 2 0648 ‘0007+ 5. "0641 2 "0645 "0004 +- ‘These results clearly show that sodium thiosulphate precipi- tates zirconium completely as the hydroxide either with or without the aid of pressure. Experiments with a Salt of Titanium. The solution of the salt of titanium was obtained by treating the double fluoride of potassium and titanium with sulphuric acid, evaporating to dryness, and dissolving the residue m sul. phuric acid and water. The solution was standardized by pre- cipitating the titaninm hydroxide with ammonia and then add- ing an excess of acid as recommended by Gooch.*: This method of procedure avoids the tendency to excessive weight observed when the titanium hydroxide is precipitated by ammonia in presence of salts of the alkalies. * Amer. Chem. Jour,, vii, 285. 122 Norton—Action of Sodium Thiosulphate, ete. In the following table is shown the effect of treating a solu- tion of titanium sulphate with sodium thiosulphate. Experi- ment 1 was conducted by merely boiling a solution of the reagents named above, filtering off the precipitate and weigh- ing as the oxide. In experiments 2 and 8 the solution of titanium sulphate and sodium thiosulphate was submitted to a pressure of 20 atmospheres in the digester. TABLE VI. TiO. taken. Na.S.0 taken. TiO. taken. Error. erms. erms. germs. grms. ie "0240 2 02am (0003 — oD, 0240 2 0240 -0000- 3. "0240 2 "0240 ‘0000+ These results show that titanium is completely precipitated by sodium thiosulphate either with or without the aid of pressure. To recapitulate: J have shown that sodium thiosulphate will completely precipitate aluminum, chromium, zirconium and titanium as the hydroxides with the aid of high tem- perature and pressure. Beryllium is only partially precipitated under similar conditions. Mere boiling for a reasonable time will net precipitate aluminum and chromium, but it is suffi- cient in the case of zirconium and titanium. In conciusion I wish to thank Prof. F. A. Gooch for his kind advice and assistance. A. W. Duff—Secondary Undulations, ete. 128 Arr. XIII.— Secondary Undulations Shown by Lecording Tide-gauges ; by A. W. Durr. Ow recording tide-gauges at many ports rapid oscillations are observed crossing and recrossing the main tidal record: Mr. W. Bell Dawson* in a short article seems to have been the first in recent years to recall attention to these “ secondary undu- lations,’ and well describes them as standing “in much the same relation to the main tidal wave as a higher octave would to a low musical note when their undulations are recorded graphically.” Others have given occasional attention to them, especially Mr. F. Napier Denison,t whose two brief papers . suffer somewhat from a lack of detailed evidence for the “ ten ehief points deduced” and too exclusive a concern with the meteorological aspects of the question. No one seems to have attempted to give a full account of the phenomena for even a hmited region. The best field for the study of the question is apparently the Eastern coast of Canada bounded by the Bay of Fundy, the Atlantic Ocean and the Gulf of St. Lawrence, where, because of the great extent and complexity of the tides, numerous recording tide-gauges have been established in the interest of navigation. Four years ago [ publishedt some observations made at St. John, N. B. The evidence advanced, depending on a calcula- tion of the period of undulation from the dimensions of the basins in tne neighborhood, seemed to point strongly to the undulations being ‘of the nature of free vibrations of a par- tially confined body of water under the force of gravity. Two years ago I took advantage of a visit to Ottawa, Canada, to examine (with the permission of the authorities of the Tidal Survey of the Dominion of Canada, to whom my thanks are due for their courtesy) the tidal records’ from the Canadian tidal stations and made extensive notes therefrom. Other occupations interfered with the pursuit of the subject, but as no one has since then taken it up, I have thought it well to publish a brief outline of the information eathered together with an examination of its bearing on certain proposed ‘expla- nations. Materials studied. In order to make clear the extent of the data employed in what follows, it may be stated that the tidal records froin twenty stations were examined in detail, these records covering * Trans. Roy. Soc. Can., i, 1895-96. + Proc. Can. Inst., Jan. 16, 1897; April 23, 1898. t This Journal, iii, 406, 1897. Am. Jour. Sci1.—Fourts Series, Vou. XII, No. 68.— mm 90 m. Farther Point St. Lawrence River 21 m. SO ea) AKO oa S.W. Point Anticosti to Gaspé 19 m. 30nd. 160 Mm. Forteau Bay Strait of Belle Isle 16 m. 20 cumin a7 in The above calculated periods can only be considered as roughly approximate. They suffice, however, to show that, except in the first four cases, the periods cannot be those of binodal seiches, and if they are seiche periods at all the seiches must be multinodal (e. g., at S.W. Pt. and Forteau Bay trinodal, at Farther Pt. quinque/nodal). Local peculiari- ties, for example variation in the depth of the basin, would, 132 A. W. Duff—Secondary Undulations of course, tend to set up partial vibrations (as would ine- qualities in the thickness of a cord.) The above comparison then, while not at variance with the seiche theory, affords only partial confirmation of it. To other cases no calculation can be applied. Digby is ona basin *— of irregular form which may have somewhat regular vibrations, though theory is at present inadequate to calculate the period of such irregular bodies of water. The same is true of Halifax, with the peculiarity that the basin is one marked out by a series of banks (Roseway, Le Have, Emerald, &e.) which, while not coming to the surface, bound an irregular bay, attaining at places a depth of 300 fathoms. Pictou and (Juaco are also on partially enclosed bays (reckoning the Quaco “ledges” as part of the boundary of the latter). At Grind- — stone Island (in the middle of the Gulf) and St. Peter’s Bay (on the outer or Gulf side of P. E. I.) the undulations are ~ always or nearly always irregular, as might be expected of points near the middle of such an irregular body of water as the Gulf of St. Lawrence. The case of St. Paul's Island (at which there are periods of — 4-7 and 12°9 m.) might be thought at first inexplicable, as the island is in Cabot strait between prominent headlands on elther side, with no suggestion on an ordinary map of a channel or basin of oscillation. But a glance at the nautical chart (of 1891) shows that, while the main chan- nel has a great depth (approaching 300 fathoms), on both sides of this channel, there is an abrupt change of depth and gradient and the contours of, say, 100 fathoms on opposite sides of this channel run roughly parallel for a distance greater than the width of the strait. Moreover, on the Cape Breton side and adjacent to St. Paul’s Island, there is a de- pressed basin, only part of the boundary of which reaches the surface. Thus we have clearly the conditions for second- ary undulations according to the “seiche”’ theory. The case of Yarmouth presents a striking peculiarity. The seiche theory gives a very satisfactory account of the periods at Yarmouth; but why should the period at high tide be regu- larly twice as great as at low tide? The mere change in depth of the channel can have no such striking effect. The only other factor that changes with the state of the tide is the direction of the t¢dal current. I find myself unable to ex- plain satisfactorily why the reversal of the current should have such a marked effect. A thorough examination and elu- cidation of this peculiar case might lead to a more satisfactory explanation of the whole subject than has yet been offered. It might also be noted that at St. Paul’s Island the unusual elevation of water level at both high and low tide, on the Shown by Recording Tide-qauges. 138 oceasions on which the period shifts from 4:7 m to 12°9m, _ would also mean some marked change in the currents through Cabot Strait. Does thevarying depth of the Bay of Fundy affect the period at St. John ? As stated in an earlier paper (already referred to), a conclu- sive test of the seiche theory would be the following. The period of oscillation of a body of water varies inversely as the square root of the mean depth. Now the mean depth of the Bay of Fundy at low water is something over 200 ft., and at high water (on the average) nearly 20 ft. greater. The width of the bay at St. John being about 40 miles, it can be calculated that the period of secondary undulation at St. John should, according to the seiche theory, be about 1:8 m. greater at low water than at high water. To test this I selected 106 tracings of particularly well- marked secondary undulations at St. John. These I measured for period with the greatest care. Then I sorted them out into high tide undulations and low tide undulations, finding 49 of the former and 57 of the latter. he mean period for high tide was 40°5 m. and for low tide 41:7—a difference of 1-2 m. Now this was wholly confirmatory of the seiche theory. That the difference 1:2 should come out less than the cal- culated difference 18 is readily understood from the fact that the undulations nearly always continued before and after high and low tide, so that the effective difference of depth would be less than the mean range of the tide. To test the probability of the above differences being mere chance, I calculated “the probable error” of each series, and found it to be the same—O0-4 m. Now the mean of each series only differs from the mean of the whole by 0°6 m. From this it is a comparatively simple problem to calculate what is the probability that by mere chance the mean period at high tide was 0°6 m. higher than its real value and the mean value fur low tide the same amount lower. The probability is 4, or the chances are 12 to 1 against the differ- ence being due to the mere chance distribution of somewhat rough determinations of what is really the same quantity. Thus this test, while very strongly in favor of the seiche theory, is not conclusive. Connections between undulations at different places. Taking first points on the Atlantic and the adjacent Bay of Fundy, it is shown by a comparison of the records at Halifax, Yarmouth and St. John, that marked oscillations at one are 134 A. W. Duff—Secondary Undulations usually accompanied by marked oscillations at the other. Con- spicuous examples are, Jan. 13th, 1898; Nov. 26 and 27, 1898; Dec. 28, 1898 ; Dee. Bi, 1898 ; Jan. 1 and 2, 1390: &e. The exceptionally strong undulations on these dates coincided with severe storms from the south on the Atlantic. But this — connection is not quite invariable, e.g.,on Aug. 14, 1898, strong oscillations of 15 inches at Yarmouth were not accompa- nied by any exceptional oscillations at St. John or Halifax. Turning now to a comparison of stations on the Atlantic and stations on the Gulf of St. Lawrence, it is found that in general no such connection exists. For instance, in my note book, opposite each of the cases of marked oscillations at St. John, Halifax, and Yarmouth, just enumerated, I have a note of the state of affairs at St. Paul’s Island, South West-Point’ and Forteau Bay, and the notes, with two exceptions, are ““no oscillations,” “ particularly slight,’ ‘less than normal,” ‘ only normal.” In the two exceptional cases the note runs ‘ above normal,” and these two notes refer to St. Paul’s Island, which, being in Cabot Strait, belongs as much to the Atlantic as to the Gulf. The lack of connection here noted seems a point of such importance that I will illustrate it by another example. Pictou on the Gulf and Halifax on the Atlantic are on opposite sides of Nova Scotia and only about 100 miles apart, a much shorter distance than separates St. John and Halifax. The large tide gauge at Pictou was only maintained during the summer of 1896. I bave compared the records from it with those from Halifax. In the following from my note-book the first note refers to Pictou, the second to Halifax:—July 3, “marked” —‘normal”; July 17, ‘nothing ”—“strong”; July 25, “nothing”’—“ strong”; Sept. 19, “nothing whatever” -— “strong”; Sept. 20, “feeble”—“ strong”; Sept. 24, “ weak ” . Herons ps 5, Sept. 28, “strong ”’—“ normal ” : Nov. 1, “feeble” “str prone ; Nov. 11, “nothing ”—‘‘strong ” ; Nov. 14, “ noth- ing” « strong, » These were, I believe, the only cases in serene the undulations at either place were decidedly above normal and therefore suitable for founding such a comparison on. Thus there is no correspondence between the occurrences of secondary undulation at Halifax and Pictou. Now, according to the theory of atmospheric billows, two points so comparatively close together as Halifax and Pictou should show a close correspondence of times of strong undula- tions ; for it cannot be conceived that “atmospheric waves of from 15 to 30 kilometers long’”’ whose “average velocity is 60 miles per hour in summer”™* could raise strong secondary * Denison, Proc. Can. Inst, Jan. 16, 1897, p. 30; Feb. 6, 1897, p. 58. Shown by Recording Tide-gauges. 135 undulations at Halifax and fail to affect Pictou and vice versa. This discrepancy is still more marked when it is considered - that Pictou is almost exactly N.E. from Halifax, and it is on the poleward current, moving N.E., that Mr. Denison relies for the production of the atmospheric billows in question. (It is to be understood here that I am not discussing the existence of atmospheric billows but the effects attributed to them.) On the other hand, the close correspondence between Atlan- tic (including Bay of Fundy) ports, in contrast with the lack of correspondence between Atlantic and Gulf ports, is strongly in favor of the view that the disturbance at any port is usually due to the transmission by water of distant disturbances of the general body of water caused by storms. Secondary Undulations and barometric records. Any connection between atmospheric disturbance and second- ary undulation should, of course, be shown by a comparison of tidal and barometric records. I have carefully compared the tracings of secondary undulation at St. John for the sum- | mer of 1896 with tracings of the St. John barograms for the same date. These latter I owe to the kindness of Mr. Bell Dawson. They were made as carefully as possible with a hard, sharp-pointed pencil. The dates are June 8, 9, 10, 21, 22, 23; July 8, 25; Sept. 1, 5, 6, 7, 18, 19, 20, 22, 28—the only occa- sions, I believe, on which marked secondary undulations occurred at St. John during the summer of 1896. A careful comparison of both series of records shows that (1) disturb- ances of atmospheric pressure usually occur about the times of secondary undulations, and, in general, the stronger the atmos- pheric disturbances the stronger the water undulations, but (2) the atmospheric disturbances rarely if ever begin and end at the same time as the water undulations; they usually precede but sometimes seem to follow the latter, and the secondary undulations may continue for several hours (e. g. 12 hours on Sept. 20) after the barometric disturbances have ceased ; (3) in no case is there any suggestion of regular periodicity in the barometric disturbances such as there is in the water undula- tions. The utmost therefore that it seems to me possible to deduce is that secondary undulations usually accompany storms near by or at a distance, but that the perzod of the oscillations at any place is wholly unaccounted for by atmospheric conditions. (This is also in accord with an extensive series of comparisons that I made, of the Atlantic Pilot Charts and the Canadian Weather Maps on the one hand and the records of secondary undulations on the other—into the details of which it seems superfiuous to enter.) 136 A. W. Duff—Secondary Undulations Mr. Denison, however, states as the first of the ten “chief points deduced ” by him (but without any indication of the details of the evidence), “that the undulations are due to the direct action of atmospheric waves upon the surface of the water at stations [italics mine] and not to ground swells, due to distant storms, or ‘seiche’ movements, as found upon lakes during atmospheric disturbanees.” Now with the for- mer or positive part of this conelusion the evidence I haye adduced above does not seem at all compatible; the latter or negative part, denying any other agency, can only mean the sufficiency of the theory of atmospheric waves to account for the period as well as the origin of the secondary undu- lations, which again is wholly at variance with the fact that the barometric disturbances show no periodicity and the. water indulations continue regular and periodic after the baro- metric record has become quite smooth. ‘The second point enumerated by Mr. Denison is that “there is a marked rela- tive correspondence in amplitude between the barometric and water undulations.” This I also have found to be the ease, provided it be not understood to imply that the two kinds of undulations always (or even usually) occur or continue at the same time. (I would also take some exception to the word marked. ’”’) From the above statements in Mr. Denison’s paper it has been inferred by some readers that he found regular atmos- pherie oscillations coinciding with and accounting for the water undulations both as regards origin and period. Mr. Denison does not directly say so and this cannot be his mean- ing. There is absolutely no evidence for it in the cases I have referred to above, covering a whole summer of St. John records. How in fact is it conceivable that atmospherie bil- lows many miles in length and moving with a velocity of from 60 to 112 miles per hour, should always have a periodicity of half a minute at the mouth of the St. John River—41 m. at St. John Harbour (less than a mile away)—124 m. at Quaco (20 miles away)—69 m. at Yarmouth at high tide, but 30 m. at low tide—20 m. at Halifax on some occasions, but 33 m. on other occasions and occasionally probably both—47 m. at St. Paul’s Island—29 m. at Picton, ete. ? Effect of Contraction of Bay. Reference must be made to another point in the “ atmos- pheric billows” theory. To account for such slight variations of pressure causing such large variations of water level, it 1s necessary to suppose that, as in the case of ordinary tides, “minute undulations as they move farther into the bay become magnified as they reach. narrower and shallower portions.” Shown by Recording Tide-gauges. 137 Yet, if we consider the Bay of Fundy, exactly the opposite is true. Yarmouth is at the mouth of the bay and there the oscillations sometimes reach 5 ft.; St. John is half way up the bay and there the oscillations never exceed 2 ft.; while at the head of the bay, where, as is well known, the rise of the tide is enormous, Hopewell Cape, Parrsboro and Windsor show no secondary undulations. Again Carleton, Q., is at the head of Chaleurs Bay, but the undulations there are particularly shght. At Cape Tormentine, which is in the narrowest part of Northumberland Strait, there are no secondary undulations and the same is true of Charlottetown, while they are very frequent at Pictou and Souris near the entrance of the Strait. On the other hand, it is to be noted that, according to the seiche theory, secondary undulations would not be expected in shallow basins such as those at the head of the Bay of Fundy; since, if oscillations of the whole body of water were started, they would rapidly die away in such shallow water. On the whole, the theory of “atmospheric billows,’ while ingenious and attractive, seems to me wholly untenable as regards the percod of undulation and very improbable as regards the orzgin of such undulations (I am not of course referring to the agency of atmospheric disturbances in general), whereas the “seiche” theory offers the only tenable explana- tion hitherto advanced, but leaves points still unexplained. A general view of the vibrations of large bodies of water. It may be well at this stage to make explicit a general view implied in the preceding discussion. It seems probable that a large body of water, such as the Gulf of St. Lawrence or the Bay of Fundy, when disturbed breaks up into smaller areas of oscillation, determined by irregularities in the level of the bed; these minor basins oscillate to a considerable extent independ- ently, at rates determined by their respective dimensions and the depth to which they are filled, the oscillations being at times in the fundamental mode, but at other times in some higher mode, but with a tendency to the establishment of some particular mcde that gives a characteristic period to the basin. Frequently, owing to the co-existence of different modes of vibration, and owing to invasions from the vibrations of neigh- boring basins, the vibrations of each minor basin become more or less irregular ; but such irregularities are comparatively rare in a part of the main basin with only shght irregularities of bottom and bounded on opposite sides by nearly parallel coasts (as is the case of the Bay of Fundy at St. John, where the records show a more constant period of undulation than at any other tidal station on the eastern coast of Canada). 138 A. W. Duff—Secondary OUndulations Another possible source of periodic fluctuations under certain circumstances. Can any other explanation be found for the definite period that characterizes the regular undulations at each place? I can think of but one other. Most people who have, from the shore, watched the progress of a ship have noticed the regular series of waves thrown off from the bow. If the ship be at anchor in a rapid current, a similar series of waves are produced, reaching any point on the shore at equal intervals. In general it may be stated that “a line of pressure athwart a stream flowing with velocity ¢ pro- duces a disturbance consisting of a train of waves of length 2arc*/g lying on the down-stream side.”* The effect of inequali- ties in the bed would be similar. Would an island, shoal, bank - or headland, past which a current is setting, cause a similar series of waves of a length and period commensurate with the size of the obstruction? For example, St. Paul’s Island divides two opposite currents between the Atlantic and the Gulf of St. Lawrence; at Yarmouth the direction of the tidal current is reversed with the tide and the Lurcher shoal is a prominent obstruction. | I do not attempt to consider this suggestion in detail. It seems to me a much less likely explanation than the theory of stationary undulations or seiches; but no adequate attempt, ir a brief space, to test its applicability would be possible, even if data as regards currents, etc., were to be had.t Summary of preceding. sy, J AOUNS. 1. Secondary undulations occur at most of the places where observations have been made, the undulations at any place being sometimes irregular but at other times of a regular periodic nature. | 2. When the undulations are regular, the period of undula- tion has a distinct and characteristic value for each place, varying, from place to place, from less than a minute to over an hour. 3. At some places at least two systems of regular periodic undulations of different periods are found; they usually occur * Lamb’s Hydrodynamics, § 228, where illustrations of such waves from experi- mental papers by W. and R. EK. Froude and theoretical papers by Lord Kelvin and Lord Rayleigh are given. + Moreover, in making this somewhat vague suggestion, 1 am aware that the formula quoted above for the length of such waves is independent of the size of the obstruction: but this does not preclude the possibility that hydrodynamical theory, applied to the case of very large obstructions, might be able to account for all the phenomena. Shown by Recording Tide-gauges. 139 at different times, sometimes they occur together; at one sta- tion the period at high tide is entirely distinct from that at low ‘tide. 4. Undulations usually occur simultaneously at stations on the Atlantic Ocean (including the Bay of Fundy), but there is apparently no connection between this occurrence at stations on the Atlantic and stations on the Gulf of St. Lawrence. 5. Barometric records show no similar periodic oscillations, but they do show disturbances at. or about the times of marked secondary undulations. 6. As the head of a bay is approached secondary undulations do not become more marked and frequent, but apparently less. 7. According to the best available evidence the period at St. John is less the greater the depth to which the tide fills the Bay of Fundy. B; DeEpvcTIONS. 1. The theory of atmospheric billows does not and cannot explain the characteristic local perzod of secondary undulations. 2. As an explanation of the origin of secondary undulations the theory of atmospheric billows seems at variance with cer- tain facts of fundamental importance to the theory (see 4 and 6 under A, above). 3. The seiche theory, which is fully established in the case of lakes, has a high inherent probability and is not at variance With any ascertained facts and in certain respects receives strong confirmation (see 2, first part of 3, 4, 6, and 7, under A, above). - 4. Even if the “seiche” theory be correct, it still remains to be explained why the oscillations at certain places are usually binodal, at others trinodal, etc., and why in one striking case (Yarmouth) the oscillations are always uninodal at high tide and binodal at low tide. 5. The whole subject is apparently much more complex than has hitherto been supposed and it may yet be found that no single explanation will apply to all cases. Am. Jour. Scr.—Fourtu Series, Vou. XII, No. 68.—Aveust, 1901. 10 140 = Fisher— Notes to Rival Theories of Cosmogony. Arr. XIV.—Mathematical Notes to Rival Theories of Cos- mogony ; by O. FISHER. THE following mathematical notes are supplementary to the discussion of “ Rival Theories of Cosmogony” in the June number of this Journal (pp. 414—422).* Let— 7 = distance of any point from the earth’s center. a = the earth’s mean radius in feet. log a = 7°3201961. ga = 2°4605 im radians = 1407 58’ 35”: s = the density of the surface rock (2°75). ) gs = the weight of a cubic foot of the same, viz: 171°875 pounds. p = the density at any depth. p = the pressure in pounds per square foot at the depth where the density is p. = the coefticient of compressibility at that depth measured by atmospheres. J = Joule’s equivalent, 772. o = specific heat of rock. ie) (1) Laplace’s law of density is, p = Os a Lr where gq is the density at the center, 10°74, and gw = 2:4605 in radians = 140° 58’ 35”. (2) The relation upon which the following calculations depend is one which I have proved in the Appendix to my “Physics of the Harth’s Crust” (p. 32), viz: gs a (OS) 5 2s? 1 — ga cot ga = B(p’—s’), suppose, P= which, reduced to numbers, gives for the relation between the pressure measured in pounds per square foot and the density, p = 58855 X 10°{ p?—(2°75)’}. Hence we see what is the value of the constant 6, which is implied in Laplace’s law of density, viz: 5°8855 x 10’ in the above units. ips : —s Also since p= be ore = ee ; p DoT mae * Attention is called to the following erratum: On p. 421, for 20 miles read 40 miles. Fisher—WNotes to Rival Theories of Cosmogony. 141 we get for the compressibility in pounds per square foot, 1-699 x 10° 27% p which reduced to atmospheric pressure is Ss eo s-c0eo tne ee (3) The above relation between the pressure and the density renders it .easy to calculate the work of compression, and thence the temperature ; v for work = — f pdov, Vo where p = B (p’— 8’). v s And since ——e Oye ae ae dv = — 0,8 ene p AE ao Ue and work = v,s X 5°8855 X 10 (p’—s’) —> s p 2 = ,8 X 5°8855 X 10°(¢ ing 28) p : 2 Sex eee p work J X specific heat X mass Mallet’s value for the specific heat of rock is 0°199. The mass will be s X v,, which will divide out, and thus B C= or * o p : And the temperature = temperature Bah. = This will give the temperature so long as we can reckon upon the specific heat remaining constant, which probably will not be the case for the higher temperatures, and certainly not so if the substance becomes liquid or gaseous, when the latent heats will disappear, but it will be sufficient to show what enormous temperatures such condensation would produce, even if the specific heat should be as great as that of water. ; B * The log of the coefficient 5s is 5°5833156. 142 Lisher—WNotes to Rival Theories of Cosmogony. (4) The lava issued from Etna at 1932° Fah. We can find the depth at which that temperature would have been pro- duced by compression of surface rock. We have temperature Fah. = iss s)s Jovi Applying this to the present case, we get (p — s)" = 5:043 x 10° = 2m, suppose, whence p—s=m+ V/(s > my =s. Since the surd is greater than m and (p — s) must be pe tive, we must use the upper sign; we then obtain , p—s=—0120: This will be the increase of density to the depth we the temperature is 1932° Fah. or p = 2°87. We then require the value of 7 which will make @ It appears on trial that a depth a—~7 of about 40 miles satis- fies this condition. It follows that the depth at which the temperature 1932° Fah. would be produced would be about 40 miles. sin qg7r ~. 9g 7. TABLE. Denes Condensation gee Banas by Pressure in eS requisite to Laplace’s atmospheres. 0 radius. miles. Law. reduce surface rock to density. 1°0 0 2°75 0°000 =) 07000 0'9 *400 3°88 Del S< 0 0°29 0°8 800 5°03 505 > 10k 0°45 0°7 1200 onli 8°460 x 10° 0°55 0°6 1600 (P25 12344 se 0g 0 62 0°5 2000 8°23 16°867 < 10’ 0°67 0°4 2400 9°09 20°698 X 10° 0°70 0°3 2800 9°80 94-370) < 107 0°72 0°2 3200 10°32 272030 10: 0°73 O'l 3600 10°64 29°052 x 10’ 0-741 0°0 4000 10°74 30°061 x 10° 0°744 Wortman—Studies of Eocene Mammalia, ete. 143 Arr. XV.— Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum; by J. L. WortMAN. [Continued from vol. xi, p. 450.] Family Viverravide. Miacide Cope (in part), Tertiary Vertebrata, 1884; Miacide Scott, Jour. Acad. Nat. Sci., Phila, 1886, vol. ix, p. 169; Viverravide Wortman and Matthew, Bull. Amer. Mus. Nat. Hist., 1899, p. 136. A FAmity of small or medium-sized Carnassidents ancestral to the viverrines, ranging in time, as far as at present known, from the Torrejon to the Bridger Eocene, having pentedactyle limbs with ununited scaphoid, lunar, and centrale of the carpus, a civet-like perforation of the transverse process of the atlas for the passage of the vertebral artery, a well-developed anterior basal cusp upon the superior sectorial, no antero- posterior femoral curvature and no postero-internal cingular cusp upon the superior molars. Blades of the superior sec- torial separated by a deep slit-lke notch, and molar formula either 3 as in the primitive Canide, or 3 as in the Viverride. In defining the foregoing family, it is necessary to distin- guish it from the contemporary Canidee on the one hand and its successors, the Viverridee, on the other. As regards the former, while in all probability they have been derived from a common source, yet they have departed sufiiciently in the direction of their subsequent and final development to be recognizable. There are two characters that appear to be entirely distinctive, one of which relates to the atlas and the other to the presence or absence of an anterior basal cusp on the superior sectorial; associated with these is the lack of curvature of the femur and no disposition whatever towards the formation of a postero-internal cingular cusp on the superior molars,—features which have characterized all lines of Canide in some stage of their development. From the Viverride, they can be distinguished not by any essential characters of fundamental importance, but only by the posses- sion of such primitive and archaic features as we should reasonably expect to find in the ancestors of the modern viver- rines. The more important of these relate to the separate condition of the scaphoid, lunar, and centrale of the carpus ; the presence of a third trochanter on the femur ; the large size and internal position of the lesser trochanter ; the slight groov- ing of the astragalus, and the large size of the deltoid crest of the humerus. To this may be added the greatly inflated and modernized condition of the otic bulla in the living Viverride. From the Mustelide, as tar as at present known, the Viver- 144. Wortman—Studies of Hocene Mammalia in the ravide are distinguished not only by the possession of the primitive characters above enumerated, but by two very trenchant dental peculiarities which the mustelines exhibit, viz.: the great antero-posterior enlargement of the internal portions of the crowns of the superior molars, and the absence of the deep slit-like notch separating the two blades of the superior sectorial. These, together with the absence of the anterior basal cusp on the superior sectorial, the great breadth of the base of the skull, the small rugged bulla, and long bony spout-like meatus, constitute the real distinguishing features by which the Mustelidee may be separated from the Viverride. The skeletal characters are otherwise much alike in the two families, the manner of perforation of the transverse process of the atlas being variable in the mustelines. : With reference to the Felide, it may be said that our knowledge is not so perfect regarding their origin and Eocene representatives as it is with respect to other living families, so that it is well nigh impossible to make a satisfactory compari- son. If the Paleonictide are the forerunners of the felines, which seems so exceedingly probable, then the two groups belong to different phyla and may be distinguished by the structure of the superior molars. In the Paleonictide, it is the posterior part of the crown of the first superior molar which is elongate; whereas in the Viverravide, it is always the anterior part of the crown which is the longer. Viverravus Marsh. Viverra, a civet cat; and avus, a grandfather. This Journal, August, 1872; Didymictis Cope, Tertiary Vertebrata, 1884; Viverravus Wortman and Matthew, Bull. Amer. Mus. Nat. Hist., 1899, p. 136. A group of small or medium-sized civet-like Carnivores, with a number of species distributed throughout the Eocene from the Torrejon to the Bridger inclusive. ‘They are charac- terized by having the dental formula I. 8, C.4, Pm. 4, M. 2, of which the two superior molars of the dentition have broad tubercular crowns, with great transverse extension of the anterior border, the two external cusps being unequal and placed well inwards from the external margin of the crown; the second inferior molar, tubercular or becoming so, and much smaller than the first; the inferior sectorial having a high trigon, with oblique principal shear, a well-developed posterior shear, and a relatively large, more or less basin-shaped heel; and _ pre- molars having posterior accessory cusps. In his original description of this genus, Professor Marsh says: “ A much smaller Carnivore, about the size of the mink, is represented in our collections by two lower jaws with teeth, and a sectorial upper molar of one individual and portions Marsh Collection, Peabody Museum. 145 apparently of several others. The lower jaws in this genus _ are long, slender and compressed; the last two lower molars are tubercular. Both have the posterior part of the crown quite low and the anterior half elevated and composed of three angular cusps. The four teeth anterior to these are much compressed. The upper flesh tooth closely resembJes that in some of the Viverride and the genus should probably be referred to that group.”* The species thus far referable to this genus are numerous, of which one, V. haydenianus, from the Torrejon; four, V. leptomylus, protenus, massetericus, and curtidens, from the Wasatch; and two, V. gracilis (dawkinsianus) and altidens, from the Wind River beds, have been described by Cope. It will thus be seen that the genus has a very great vertical range, greater, in fact, than any known contemporary group of mammals throughout the whole Eocene. With the close of the Bridger epoch, according to our present knowledge, the genus disappeared from this country, since no remains of it have as yet been found either in the Uinta, White River, or John Day deposits.t It is pos- sible, however, that the group may have continued to exist to a much later date on this continent, and that they retreated to the southward along with the tropical fauna which disap- peared from Wyoming at the close of the Eocene. It is pos- sible, therefore, that their remains will yet be found in the Miocene of the South, but this, of course, is merely conjectural. Viverravus gracilis Marsh. Didymictis dawkinsianus Cope, Tertiary Vertebrata, 1884, p. 310. The type, figures 18, 19, consists, as Professor Marsh has stated, of parts of both mandibular rami and a superior secto- rial, but there are at least twenty individuals of the species represented in the collection by various fragments. Of the type the right ramus is the more perfect of the two, and carries the third and fourth premolars and the first and second molars. The alveoli for the first and second premolars, together with that for the root of the canine, are represented. All the pre- molars are two-rooted, even the first, which among the Carni- vora is very generally a single-rooted tooth. The third and fourth premolars have high pointed crowns, with anterior and posterior basal cusps, together with a distinct and trenchant accessory cusp. The sectorial has the trigon much elevated * Loc. cit, p. 7, of separata. + A possible exception to this statement may be found in the imperfectly known genus Bunelurus of Cope, from the White River Oligocene of Colorado. When more fully known, it will not be at all surprising to find that this genus is a direct descendant of Viverravus. 146 = =Wortman—Studies of Hocene Mammalia in the and of greater antero-posterior length than the heel; the ante- rior and internal cusps are equal in height, but the external was evidently much higher; it is somewhat damaged and does not display its full length. The principal shear is oblique, and there is a posterior shear which bites against the anterior edge of the first molar. The heel is relatively short, wide, and basin- shaped ; the external part of the basin is the thicker and more elevated. The second molar is much smaller than the first, but displays practically the same structure; the trigon, however, is much less elevated, and the shears are imper fect. The external and internal cusps are of equal size and height, but the anterior is much smaller and lower. The heel is relatively longer than in the first molar, beg obtusely pointed behind, where it is terminated by a low but. distinct cusp; from this cusp a low ridge is continued forwards and inwards to enclose the imner FiguRE 18.—Right lower jaw of Viverravus gracilis Marsh; outside view; natural size. (Type.) FIGURE 19.—Superior sectorial of same; outside and inside views; natural size. (Type.) FIGURE 20.—Superior sectorial of Viverricula sp.; natural size. part of the basin. Externally there is a stronger cusp situated in advance and to the outer side of the last mentioned cusp, which furnishes the external boundary of the basin. The superior sectorial displays the typical laniary structure of the more highly-developed Carnassidents ; it is composed of a pair of external cusps, laterally flattened and elongated in such a manner as to form a pair of very effective ‘shearing blades. Of these, the anterior is the larger and more elevated, being separated from the posterior by a deep vertical noteh, which appears as a narrow slit upon the lingual surface. At the antero-external angle of the larger cusp is seen a lower, but very distinct, anterior basal tubercle. Internal and oppo- site to the anterior edge of the main external blade is placed a relatively large, pointed internal cusp of about the same pro- portions of that of the Genet. In fact, the whole dentition, as far as known, is strikingly like that of ‘this living species. In his description ot V. dawkinsianus, Professor Cope states that the first premolar of the inferior series is a single-rooted tooth. If this is true (which I am inclined to doubt), the two Marsh Collection, Peabody Museum. 147 species are distinct and we have the anomalous condition of the latest known species of the genus being less specialized in this particular than its immediate predecessor. The species is distinguished from all the others by its small size and the short broad heel of the lower sectorial. The principal meas- urements of the type are as follows: Length of inferior molars and premolars. ----- CUS a Menreth-orinieror molars... 22222 522252... 9°5 Weneuh Gl sectoral’ 22. 02. she eo 5°2 Transverse diameter of sectorial__.-----..--- 3°4 Antero-posterior diameter of superior sectorial, 7: Transverse diameter of superior sectorial._--. 3°5 The type specimen was found at Grizzly Buttes, by G. G. Lobdell; other specimens were obtained at various levels from the upper to the lower part of the horizon. Viverravus minutus sp. nov. This species is represented in the collection by remains of at least six individuals, of which the lower jaws alone, in varying degrees of completeness, are preserved. That which is here 21 FIGURE 21.—Left lower jaw of Viverravus minutus Wortman; outside view; three times natural size. (Type.) selected as the principal type, figure 21, is a left ramus carry- ing the molars and the three posterior premolars in good pre- servation. The specimen in question is broken just posterior to the base of the coronoid and also in front in the region of the symphysis, but it includes the base of the canine alveolus as well as that for the first premolar. The crowns of the molars are somewhat worn, so as not to display very clearly the char- acter of the cusps of the heels. On this account, I associate with it another specimen as a cotype, also a left mandibular ramus of the left side, in which the heels of these teeth are more perfectly preserved. A third fragment of a jaw carries the canine in perfect condition. 148 Wortman—Studies of Hocene Mammalia in the The character of the species, as exhibited by these three specimens, may be stated as follows: It is considerably smaller than V. gracilis; the first premolar is two-rooted ; the third and fourth premolars have anterior and posterior basal cusps and posterior accessory cusps which are absent on the second ; the heel of the sectorial is proportionally narrower, and not quite so distinctly basin-shaped, as in V. gracilis. The last molar has a distinctly narrower, more trenchant heel and does not exhibit the two cusps seen in V. gracilis. The canine is long, slender, and considerably recurved. The remains indi- cate an animal of the size of the common weasel, the slender jaws and sharp cutting teeth giving evidence of a very highly carnivorous habit. Length of inferior molars and premolars . ---- 2 Length of inferior molars: 22s 4) (feos as de Length of first molar: 228.0553.) as eas Wradth of first: molars. eh iia eae 2°5 Depth of jaw at interior sectorial._-.__--._-- 4°5 The principal type specimen was found on Dry Creek, Bridger Basin, by Messrs. Lamothe and Chew, although other specimens are from Grizzly Buttes. Oddectes herpestoides gen. et sp. nov. O6n, an egg; and dectes, a biter, in allusion to the habits of the Ichneumon. A genus of small viverrine-like animals, having the dental formula J. 2,C.4, Pm. +, M. 2, with trigon of infertomumelars® high, the cusps sub-equal, and the principal inferior shear high and very oblique; inferior molars with trenchant or slightly basin-shaped heels; transversely extended superior molars, with tubercular crowns ; anterior and posterior external angles of first molar about equally extended, and antero-external angle of second more extended than posterior; a superior sectorial, with sharp blade-like principal cusps, and premolars without posterior accessory cusps. The remains upon which this genus and species are founded consist of the larger part of the skeleton of one individual, considerably broken, but at the same time with nearly all the characteristic portions represented. I associate with this type two other individuals, one of which is represented by a left lower jaw, broken away at the base of the coronoid, but con- taining all the teeth, with the exception of the incisors, in a state of good preservation; the third specimen is the posterior por- tion of a left mandibular ramus of an immature individual, in which the fourth premolar was just comin into position. Marsh Collection, Peabody Museum. 149 Dentition.— W ith the exception of a single isolated tooth which apparently belongs to the upper series, the incisors are not preserved in any of the three specimens, but the bases of the alveoli for those of the lower jaw can be indistinctly made out. They were three in number and arranged apparently as in the Paradoxures, without having the second one pushed back out of line, as is frequently found in the Canidee and some Viverride. The canine, figure 22, is relatively large, high, pointed, and recurved; there is a faint vertical external groove, and a broader shallow one upon the internal face of the crown; on the posterior and external surface is an exten- sive worn area, where the. tooth impinged upon its fellow of the upper series. 22 FIGURE 22 —Left lower jaw of Oddectes herpestoides Wortman; outside view; three times natural size. (Cotype.) The premolars are four in number, with rather short, stout, thick crowns, having more or less of a tendency towards the development of strong internal cingula and cusps as seen both in the Paradoxures and Herpestine. ‘The first is small, single- rooted, with an obtusely-pointed crown directed well forwards like the corresponding tooth in Herpestes griseus, and sepa- rated from the canine by a short interval. The second is abruptly larger, two-rooted, and has a small, indistinct poste- rior basal cusp. The third and fourth are still larger; their crowns are relatively thick, with obtuse points and inconspicu- ous anterior and posterior basal cusps; there are no posterior accessory cusps present, except a slight indication of one in the last premolar, as in Herpestes, to which the series of teeth, including the canine, bears a very striking resemblance. The molars show many peculiarities for a Carnivore, and at first glance might indeed be readily mistaken for an Insecti- 150 Wortman—Studies of Hocene Mammalia in the vore; but a careful examination clearly reveals their carnivo- rous character. As is usual among the Carnassidents, the first molar is the largest of the series, which decreases in size pos- teriorly. Where the full number is present, it ig a very general rule that the first greatly exceeds the second in size; but in the present case the decrease is much more gradual. The crowns have the anterior part much elevated and divided into the three usual cusps. In- stead, however, of the external one being much larger and higher than the other two, they are more nearly equal (figure 23). _ They are placed at the points of a triangle whose sides are sub-~ eee ee ete of equal, the shortest side, or base, Obes Herpes Workman it being’ direeted inwards and a lit (Cotype.) tle forwards. Between the outer and anterior cusps a short more or less imperfect shear is developed, and a posterior shear is also present between the outer and inner cusps. The heel is composed of a central fore and aft secant ridge, on the inner side of which the border is slightly raised, thus giving the first step in the formation of the basin-shaped talon. The succeed- ing teeth are alike in structure, with the exception that the principal shear is proportionately less developed on account of the reduction in size of the anterior cusp of the trigon. Of the superior dentition, figure 24, the single incisor, if it is correctly referred to this category, has a rather narrow, pointed crown, somewhat flattened upon its posterior surface, and corresponds most nearly with the second. The canine — resembles that of the lower jaw, showing about an equal degree of stoutness and curvature. The first premolar is not preserved in any of the specimens. The second has an obtusely-pointed crown, with a small posterior basal cusp and an indistinct. cmgulum upon the inner margin of its base. The third is thicker, somewhat triangular in cross-section, with a more pronouneed posterior basal cusp and a stronger internal cingulum. In many of the viverrines, more particu- larly the Herpestine and Paradoxurine sections, the third pre- molar has a tendency to develop an internal cusp, and the formation of the internal cingulum may be regarded as the initial stage in this process. At first sight it would appear that the fourth superior pre- molar is of a too highly developed sectorial character to corre- spond with that of the lower jaw; but, curious as this may seem, there can be no possible doubt of the association. This tooth, however, betrays its- primitive character in the large size Marsh Collection, Peabody Museum. foal of its antero-external cusp, which is relatively high and coni- eal, with its posterior portion little drawn out into a cutting surface. The posterior cusp is sharp and _ blade-like, but pro- portionally small. The two are separated by the usual vertical fissure. There is a distinct anterior basal cusp, which has an unusually external position. The presence of this cusp is a very constant feature of the viverrines, and its external position recalls the Herpestine section of the family. The internal cusp is of only moderate proportions and is placed well for- wards, as in the civets in general. The first molar is symmetrical in respect to the extension of the external angles; they are both equally extended and the two sub-equal external cusps are placed well inwards from the 24 FIGURE 24.—Right upper jaw of Oddectes herpestoides Wortman; crown view; three times natural size. (Type.) crown margin. The internal cusp is large and lunate, and there is a well-developed anterior and a very faint posterior intermediate cusp. The second molar has the antero-external angle well extended, but the posterior is short; with the exception of this difference its structure is very like that of the first. The third molar is not preserved, but the alveoli indicate not only its existence but its goodly size as well. Vertebre.—While the vertebral col- umn is by no means complete, yet a number of the vertebre are preserved and serve to give some idea of this part of the skeleton. A portion of the atlas, figure 25, shows that the perfo- rations for the vertebral artery are the same as those in the civets. The body weure 25.—Portion of of the axis is rather long, sharply atlas of Oddectes herpestoides keeled below, and there is a well-de- Wortman; top view; three : é halves natural size. (Type.) veloped peg-like odontoid process. The bodies of the remaining cervicals are short, depressed, and keeled ; they are smaller in proportion to the size of the lumbars than in any of the living civets. The lumbars increase in size 25 152 Wortman—WStudies of Hocene Mammalia in the progressively backwards, showing elongated inferiorly keeled centra. The third lumbar, figure 26, has a spine of moderate height, and simple, cylindrical poste- rior, and hollow, half cylinder-like an- terior zygapophyses, and distinct ana- pophyses and metapophyses. There is apparently no trace of the double tongue and groove pattern of certain of the contemporary Creodonts. The ae sacrum 1s composed of three anchy- NG losed vertebrae, and there was a long Figure 26.—Third Jumbar 20d powertul tail. vertebra of Oodectes her pestoides Fore limb.—The scapula is repre- icrinans sice view; Umeesented Dy the proximal! “endyamaa mm wie ree eee damaged condition. The glenoid cavity is elliptical and cup-shaped; the neck is very short, the unusually heavy spine rises close to the glenoid border, as in the Binturong, and there was a prominent metacromion present. The humerus, figure 27, is complete, but considerably crushed laterally ; it exhibits the following characters: The head is well rounded, pyriform, and overhangs the axis of the shaft pos- teriorly; the greater tuberosity is inconspicuous, and does not reach the level of the head; the deltoid crest is large and extends well down the shaft; the distal extremity is broad, with large supinator ridge and internal condyle, and there is an entepicondylar foramen. : In its proximal portion, the ulna, figure 29, exhibits some peculiarities of structure to which that of the Binturong makes a very decided approach. The chief peculiarity is seen in the upward curvature of the under surface of the olecranon and its great lateral breadth. Upon its inner or radial side, it is produced in such a way as to form a broad shelf-like projec- tion, as in the Binturong; it is relatively short and thick as in this latter species, in marked contrast with its elongated form in many of the contemporary Creodonts. The posterior wall of the greater sigmoid cavity has comparatively little elevation, giving to the cavity a shallow appearance, but the anterior boundary or coronoid process is prominent and well extended upon its radial side. Just in front of this latter process is seen the deep muscular impression for the attachment of the tenden of the anterior brachial muscle. The shaft is considerably flattened from side to side and traversed by broad shallow longitudinal grooves, which continue to the distal end. In the lower fourth of its extent, the shaft becomes sharply triangu- lar in cross-section by reason of the development of a sharp ridge from the more or less rounded internal surface. This Marsh Collection, Peabody Museum. 153 ridge is very highly developed in the Binturong and less so in Herpestes. The distal end is not preserved. The radins, figure 28, as compared with the ulna im size, holds about the same relationship as that seen in the civets in general. The head is cup-shaped and has an imperfectly cir- cular outline, indicating thereby complete power of rotation. FIGURE 27.—Humerus of Oodectes herpestoides Wortman; front view. FIGURE 28.—Radius of same species; front view. FIGURE 29.—Ulna of same species; front view. All figures are three halves natural size. (Type.) The shaft is slightly curved, somewhat compressed, and dis- tinctly trihedral in its lower fourth. The distal end is pro- vided with a well-developed styloid, deep tendinal grooves, and a well-excavated articular surface for the scaphoid and lunar. 154 Wortman—Studies of Hocene Mammalia, ete. The manus, figure 30, is sufficiently preserved to furnish a fair idea of its or canization. All the carpal bones are present with the exception of the lunar, magnum, and pisiform. The scaphoid, centrale, trapezium, and trapezoid, with the proximal portions of the first three metacarpals attached, were found in the position shown in the accompanying figure, and there is reason to believe that the positions of the bones are substan- tially correct. The scaphoid is relatively large, and is articu- lated distally with the trapezium, trapezoid, and centrale. The trape- zium is of moderate size and sup- ports the pollex in the usual way, which is not to any very great degree opposable, if at all. ~The trapezoid is rather large, of an im- perfectly triangular form, and sup- ports the second metacarpal. The centrale may be said to have an out- line intermediate between a quad- rate and triangular pattern ; it rests unequally upon trapezoid and mag- num, and lies under the junction of the scaphoid and lunar. The unci- form resembles that of the modern | civets, and presents a lateral facet Figure 30.—Left manus of for articulation with the lunar. The pee pe ems cuneiform is flattened from above Si, (Cae) downwards, and articulated with the unciform in quite the usual way. Of the metacarpals, that of the pollex is the heaviest, the second and third being more slender, with rather narrow ‘proximal extremities. On the upper portion of the shaft of the second is seen a prominent tuberosity, and a less conspicu- ous one upon the third near the proximal extremity. These bony protuberances, if normal, probably served for the attach- ment of the long and short carpal extensors, since, In position, they correspond nearly to the insertion of these muscles. The length of the metapodials cannot be ascertained, but they were presumably short. Their distal ends are hemispherical and keeled like those of the modern civets. Some fragments of the proximal row of phalanges are preserved, and these are again like those of ferpestes and the Binturong. One perfect phalanx, figure 31, belonging probably to the second row of _ the fore foot, is long and slender, notably more so than in the GBinturong, wherein they are longer than usual in the Viver- ride. No ungual phalanges are known. [To be continued. | Adams— Electromagnetic Effects, ete. 155 Art. X VI.—The Electromagnetic Effects of Moving Charged Spheres ; by EpwIn P. ADAMS. THE magnetic effects due to moving charges were first shown experimentally by Professor Rowland* in 1876. Dr. E. Lecher,t in 1884, thinking that the importance of the experiment rendered its repetition desirable, did so, but with negative results. Insufficient data regarding his experiment make it difficult to point ont the cause of his failure to obtain the effect. Since then, Professor Rowland’s results have been fully confirmed by Professor W. C. Rontgent in 1885; by Rowland and Hutchinson§ in 1889; and by Professor F. Him- stedt| later in the same year. . The next experimental attack upon this problem was by M. V. Cremieu§ at Paris in 1900-1. His experiments were originally undertaken to determine whether a changing mag- netic field exerts a mechanical force upon an electrically charged body. The negative results obtained led him to undertake a series of experiments on the magnetic effect of moving electric charges. The results of these experiments are apparently all opposed to the results obtained by the above observers. The data which have thus far appeared do not give sufficient details to render it certain that positive results should have been expected. Cremieu himself states that he is convinced that the effect does not exist. The great importance of the experiment would seem to render further investigation desirable, and the present paper contains a description of an experiment with this end in view. All previous experiments have been made with rotating disks. With one exception the direct effect of a charged rotating disk upon a magnetic needle has been examined. The exception referred to is the method employed by Cremieu in one of his experiments where he sought to observe the inductive effect of charging and discharging a rotating disk upon a neighboring coil of wire in circuit with a sensitive galvanometer. The use of charged spheres was suggested by Professor J. J. Thomson** in 1881, and he gives a calculation of the mag- * This Journal (3), xv, p. 30, 1878. + Rep. d. Phys., xx, p. 151, 1884 ¢ Sitz. d. Berlin Akad., p. 198, 1885. § Phil. Mag. (3), xxvii, p. 445, 1889. || Wied. Ann, xxxviii, p. 560, 1889. ml Comptes Rendus, cxxx, p. 1544; cxxxi, p. 578; exxxi, p. 797; cxxxii, p. 327. +a Eni. Mae xi.) py, 236, 1881, Am. Jour. Sci.—Fourts Srrizes, Vou. XII, No. 68.—Aueust, 1901. si! 156 a responds to 500 revolutions of the axle. Adams—Electromagnetic Lifects of Aro -----+-> 2O6Smne22 25. netic force which would be produced by the motion ~ of a sphere charged to the highest possible potential. In many respects this seems the most natural method of procedure, and was adopted in this experiment. The description of the appara- tus employed follows. A hollow brass shaft, AA (figs. 1 and 2), is sep- arated into two portions by the hard wood bar B. ~ The shatt turns in fiber bearings, and the pulley and belt at D, fig. 2, communi- cate power from the coun- tershaft FF. , The spheres which carry the electric charges are spun out of sheet copper into hemi- spheres, soldered together. There are two sets of '‘ spheres, 16 in each set. Brass rods pass through the hollow spheres, and are soldered to them. These brass rods are screwed into collars carried on the axle. The two sets of spheres are thus insulated from each other by the hard wood bar, 8@ in length: The electricity is commu- nicated to the two por- tions of the axle by the cop- per brushes CO, fig. 2, and the spheres, being in con- tact with the axle, thus be- come charged themselves. The speed-counter E, fig. 2, is directly attached to one end of the axle. The gear- ing is so proportioned that one revolution of the crown-gear and dial I, cor- Moving Charged Spheres. 157 The magnetic system upon which the direct effect of the moving charged spheres is observed, is enclosed in the brass tube H, closed on the bottom by a glass plate coated with tin- foil. This tin-foil is cut into strips parallel to the axle for the purpose of preventing conduction currents flowing in it ina direction in which they could cause a deflection of the needle. The needle-system is carried on a piece of mica, 75™ long and 1:25™ wide. Pieces of well-hardened magnetized watch- spring are cemented at the lower edge and near the upper edge of the mica so as to form an astatic system. The needles are placed perpendicularly to the axle. The mirror is also cemented to the mica, a little above its center, and is observed through an opening in the brass tube at S, covered with a thin glass plate. The whole needle-system is suspended by a quartz fiber, 32 in length. The magnetometer tube H is carried in a brass collar P, which is screwed to the brass plate M. This plate is provided with levelling screws, and rests on the board shelf N, sup- ported at its ends on two brick piers. Underneath the level- ling screws are placed pieces of felt, to take up any mechanical 158 Adams—Klectromagnetic Lffects of vibrations. By means of control magnets placed on the brass plate M any required degree of sensitiveness can be given to the needle-system. All metallic parts of the magnetometer tube and supports are earthed. A single turn of wire at K serves to determine the needle-constant. The deflections are read by means of the telescope and scale T, placed at a dis- tance of 3 meters from the mirror. Power for driving the spheres is furnished by a 4 horse- power motor, at a distance of 7 meters from the magnetome- ter. A heavy iron casting, L, in front of the motor, gives additional screening of magnetic disturbances due to the motor. - The motor is belted to the steel countershaft, FF, which turns in hangers placed along the cement floor. A rigid wooden framework is built up from the floor to carry the revolving spheres. The axle AA is at a distance of 1 meter from the floor. The brick piers which support the magnetometer are entirely separate from the floor and from the framework which carries the spheres. The electricity for charging the spheres is furnished by the battery of 10,000 storage cells used by Professor Trowbridge in investigations in spectrum analysis, ete., and which has been described by him in this Journal. The battery is on the third floor of the laboratory, while the apparatus for this experiment was set up in the basement. The wires leading down from the battery are well separated from each other and from sur- rounding walls, except for a short distance, where they are carried in thick-walled glass tubes. A commutator is inserted to reverse the sign of the charges of the spheres. | A good deal of difficulty was met with due to the wind pro- duced by the revolving spheres. These have a velocity of about three miles a minute, and the wind was sufficient to shake the shelf supporting the magnetometer so that it was impossi- ble to take readings owing to the continual vibration of the needle-system. This made it necessary to build a shield around the spheres to keep the wind from blowing directly on any part of the magnetometer support. In order to bring the needles as close as possible to the spheres, the top board of the shield has a hole cut in it into which the lower end of the magnetometer tube is placed. The hole is closed on the bot- tom by a very thin glass plate, which the spheres just clear when revolving. Not the slightest movement of the needle could be observed when the motor alone was run, or when the motor and steel- shaft along the floor were run. But the cutting of the earth’s magnetism by the brass axle was sufficient to produce a deflec- tion of several centimeters. As long as the speed remained perfectly constant this gave no trouble. But if the speed Moving Charged Spheres. ue 159 varied by even a small amount this was very troublesome, and was one of the principal sources of error in the experiment. The two sets of spheres are charged oppositely, one being connected to the positive pole of the battery and the other to the negative pole. On charging them while at rest a very small deflection was observed. This was not due to direct electrostatic effect, but to the rush of current flowing in to charge the spheres. It.is only an instantaneous effect, the needle coming back to its original position of equilibrium almost immediately. This deflection was entirely gotten rid of by inserting a large water-resistance in series with the battery. With this resistance in, and the spheres at rest, no effect could be observed on the needle when the spheres were charged, or when the charge was reversed in sign. When the spheres were set revolving, and the electrification was reversed, a distinct deflection of the needle was produced. It was difficult always to get satisfactory readings of this detlection due to slight changes in the speed and the conse- quent change of the zero-point. But the qualitative effect was unmistakable. The deflection was in the direction to be expected; that is, a positively charged sphere gives rise to a magnetic force in the same sense as a current flowing in the direction of motion. At first sight there seems to be a possible alternative explana- tion of this effect. It may be that the charges are continually swept off from the rapidly moving spheres by their motion through the air, and that thus a continual flow of electricity is produced in the wires which connect the battery to the spheres. But the strength of such a current, if it existed, would be far too small to produce the observed effects owing to the very large resistance of the water in series with the battery. With this consideration, and the absolute regularity with which the effect was observed, there appears to be no reasonable doubt that this is an actual magnetic effect due to moving charges. All observations were made between the hours of one and five in the morning. It was impossible to get any satisfactory readings in the day time, owing mainly to the magnetic dis- turbances produced by the electric cars. Furthermore, the speed of the motor was more constant at that time than during the day, since the load on the mains of the power-plant varied less. It was found most satisfactory to have the sensitiveness of the needle such that a deflection of 5-15 millimeters was produced on reversing the electrification. Much greater sen- sitiveness could easily have been obtained, but the zero-point varied so much that the readings were less reliable. The speed used was about 50 revolutions per second. At this speed no trouble was experienced from wind or from vibrations commu- 160 Adams—FElectromagnetic Hifects of nicated directly to the magnetic system. But when the speed exceeded 60 revolutions per second the needle began to vibrate | sufficiently to make observations difficult. The following is the method used in taking a series of read- — ings. The motor was started, and the time of 2500 revolutions of the axle determined. The spheres were then charged ; two elongations of the needle on one side of the zero-position were read, and one on the other side; the electrification was then reversed, and similar readings taken. Sometimes the readings were not taken until the needle had come nearly to rest, and its new equilibrium position estimated. This was repeated until about ten reversals had been made. The speed was then again determined. The average of the deflections, and the average of the two values of the speed were taken as repre- senting the series. The needle constant was determined before and after a number of series. An attempt will be made in the following to compare the results obtained with the results expected from theory. From reasons which will appear later this comparison can be regarded only as approximate, and is given merely to show that the observed results are of the right order of magnitude. Row- land’s method of using the experiment to determine V, the ratio of the units, will be used. The magnetic force produced by a moving charge, g, travel- ing at velocity v, is :* H— qv Sin e p- where p is the radins vector drawn from the charge to the point at which H is measured, and ¢ is the angle between p and the : direction of motion. This expres- sion holds only in case the veloc- ity v is small compared with the velocity of light. This may also be taken as the magnetic force produced by a moving charged sphere, the charge being supposed concentrated at its center. The force acts in a direction perpen- dicular to p and to. the direction of motion. The magnetic force at either the upper or lower needle, due to one of the spheres at any point in its path, is found as follows: The two sets of spheres revolve * J. J. Thomson, Phil. Mag., xi, p. 236, 1881; Heaviside, Electrical Papers, vol. ii, p. 505. Moving Charged Spheres. 161 in two parallel circles, distant 6 from each other. The plane of revolution is taken perpendicular to the plane of the paper (fig. 3). The needles lie in one of the planes of revolution. ‘The force at P due to the sphere at A is required. pak A, 6 = OB. GS eles = OC = OA = radius of revolution. = angle between p and tangent at A. = angle between vertical radius and radius to A. pad? +0? +c'—2cr/d' +0" cos cos d=cos 6 Age Cah esas p =a’ +b’ +c’—2de cos 6 d sin 6 COS == cin e=4/ —| (d cos 6—c)* +0? d’ +b’ +c’—2de cos 6 The force acts in a direction perpendicular to p and the tangent at A. The component of this force in the direction of the normal to the plane of revolution is required. Let Ww be the angle between the direction ot the force and the normal to the plane of revolution. d cos 6—ec cos GV= ——_____ ‘/(d cos 0—c)? +0" VY =27rcN where N is the number of revolutions per second. idence - 2a Neg(d.cos ons V[d?+0'+c°—2de cos 6]? X is the component of the force at P in the direction of the axle due to the sphere at A. V is the ratio of the units. The capacity of the spheres and their potential are measured in electrostatic units. Figure 4 is plotted from this expression, and shows how the force varies with the position of the spheres. The uppercurve gives the resultant force at the lower needle due to both sets of spheres, and the lower curve, which is nearly a straight line, gives the force at the upper needle. Let the mean value of 162 Adams—Llectromagnetic Lffects of the force at the lower needle obtained by time-integration of the curve be: 2aNq v A and the mean value of the force at the upper needle : eel oa See esse ae ee oe. Seeee Zeveeeeeeeeee Cees See enees PTL TT ee Hes SGGeSGReeeGRSGSORee Then the effect on the needle will be the same as if constant forces of these magnitudes acted upon it. The same result could be obtained by imagining two coils of wire passing through the centers of the two sets of spheres through which a current was sent in opposite directions of such a magnitude that the same amount of electricity passed any point per second as in the case of the charged spheres. Moving Charged Spheres. 163 Force at lower needle due to calibrating coil : Qr Th? = es TC, (A? + 2”)? h being the radius of the coil, and « its distance from the plane of the needles. Force at upper needle due to calibrating coil: Qa ri[2 (> ) P.(cos4)— 2 ah ne =) Px(cos4) + Re ass |=2 ID r being the distance of the center of the coil to the upper needle, and @ the angle between the axis of the coil and 7. Let M be the moment of the lower needle, and H the earth’s horizontal magnetic force at its center; M’ and H’ the corre- sponding values for the upper needle. Let @ be the angular deflection of the needle-system produced by the current in the calibrating coil, and ¢ the angular deflection produced by the moving charged spheres. Equating the couple acting on the needle. system due to the earth’s field to the couple acting on the needle-system due to 7 current in the calibrating coil, and putting M/M’=1, we ave: HM—H™M’ 271(C—D) Pr CERN = tanhd Similarly, equating the couple acting on the needle-system due to the earth’s field to the couple acting on the needle-system due to the revolving charged spheres, we have : HM—H'M’ _ 2Ng(A—B) M ee Va tanec Hence vy —A-B Ng tan 0 me DW cane Let 6 be the scale deflection on reversing the current I in the calibrating coil, and A the scale deflection on reversing the charges of the spheres, Then : d and d’, the distances of the centers of the lower and upper needles respectively from the axle, were determined by means of a cathetometer, the distance of the mirror from the center of the axle being ‘directly measured, from which d and d’ were obtained. The current sent through the calibrating coil for determining the needle-constant was measured bya Weston milliammeter. 164 Adams— Electromagnetic Liffects of The value found was accurate to at least one-half of one per cent. which is sufficient for this purpose. The charge of the spheres is the most uncertain element in the quantitative determination, and it is this uncertainty especially which makes the method of revolving spheres far less suitable for quantitative work than the method of rotating disks, particularly as employed by Rowland in his second experiment. If a single set of spheres had been used, charged to the same potential, an equal opposite charge would have been induced on neighboring conductors which would have travelled with the charges on the spheres. It would have been difficult to determine just what the resultant effect should be. For this reason two distinct sets of spheres were used, charged oppositely, the spheres always keeping the same relative posi- tions. It was then assumed that the only moving charges were those carried on the moving spheres. The capacity of the spheres was calculated on the assumption that they were the only conductors present. The charge on any one sphere was calculated by the method of images, the charges on all the other spheres being regarded as concentrated at their centers. We then have (Maxwell, volume i, section 159): + gq = charge of each sphere. + P = its potential. a = its radius. J.J ete. = the distances of the centers of the spheres from one another. Then the charge on any sphere is given by: q = Pa—qa(= ope : ) + 9a(+ 7 a where the odd subscripts refer to spheres of the same set and even subscripts to spheres of the other set. Numerous determinations have shown that the potential of the individual cells making up the battery averages almost exactly two volts when freshly charged. On this basis, assum- ing perfect insulation, the potential used was 20,000 volts. It was thought best, however, to get a closer estimate of the potential, since the insulation was not perfect. This was measured by means of a guard-ring electrometer. If the radius of the movable disk is R, the inside radius of the guard- ring Rk’, D the distance between the moveable disk and the fixed disk, W the weight required to balance he attraction of the two disks, and g the acceleration of gravity,* * Maxwell, volume i, section 217. Moving Charged Spheres. 165 gW P= 4p V ip + R” Re = 9 -55933 Ris ~6:033 Co — 980° The following determinations were made: D G P 3°0 2°078 » 64:0 ee 2°780 62:0 20 4°505 63°0 These values of the potential are in electrostatic units. To convert into volts multiply by 300, and the potential as meas-: ured by the electrometer is 18,900. As a further check the maximum sparking distance between two metallic spheres was measured. With polished brass spheres, 2°6™ in diameter, this distance was found to be 0°58. According to the observations of Baille,* this corresponds to a potential : P= 63 Baille’s observations were with spheres of different sizes. In the region of a spark-length of this magnitude, he found that spheres of 3° diameter gave nearly the same results as spheres 1™ in diameter, so that his results for spheres of 3° diameter can be used with very small error for spheres 2°6™ in diameter. From these determinations, the value = 63 in electrostatic units is taken as the potential of the spheres. The great advantage in using a storage battery as the source of electricity is that one measurement is sufficient to determine the potential. The battery when used was always freshly charged, and the variations in its potential from one exper!- ment to another were very small. Substituting the following numerical values: Radius of spheres = 1°35 4 ae 20°38 d pees A | d' = 2998 b SOG ae = 19°30 A =r we find A-B == 41 C-D == 0;000232 q = iio ue * J. J. Thomson, Recent Researches, p. 77. 166 Adams—EHlectromagnetic Effects of Below are given the successive equilibrium positions of the needle in six series of reversals, as an illustration of the results that have been obtained : No. 1. No. 2. No. 3. 63 165 130 60 66 a2 178 109 1292 64 76 188 223 122 130 71 19 200 220 112 130 69 76 213 Ot 140 69 146 156 164 171 162 The needles were then moved a little closer to the apHEle> giving: C—O G=98 34 A-B = 1°834 No. 4. No. 5. No. 6. 92 114 164 108 84 106 135 148 156 13 118 120 135 148 149 108 155 142 149 eo 143 12 53 135 164 192 139 155 147 In all these cases, the first column gives the readings when the spheres directly under the magnetometer are positively charged, and the second column when they are nega charged. In No.1, the successive deflections are: 38, 6, 2, 12, DSO ad sais giving an average of 6:7. In No. 4 they are : 21, 13, 0, 8, 18, 18, 9, 19, -6, 10, -8, 7,9; giving an average of 9. It is evident that with variations such as are present here the measurements can be regarded only as a rough approximation. But the most important fact to be observed is that when the spheres directly under the magnetometer are changed from negative to positive a deflection toward the small figures of the scale takes place; and when changed from positive to negative a reverse deflection takes place. - When on account of changes in the zero-point some of the deflections do not apparently follow this rule, they are entered with a nega- tive sign in finding the average deflection. There has invari- ably resulted a positive deflection on taking the average. Earlier experiments were made with four spheres in each set. With this number it was possible to get much higher speeds—75-85 revolutions per second. Similar qualitative results were observed, but when comparison with theory was attempted it appeared that fair agreement could be obtained if Moving Charged Spheres. 167 the maximum value of the magnetic force were used instead of the average. This was largely an accidental result, a sufficient number of reversals not having been made. It would have been desirable to have made a larger number of reversals in each series, but after the apparatus had been running for some time the bearings heated so much that it-was impracticable. The following table gives the values of the ratio of the units obtained from the above readings : No N A if 6 V 1 42 6°7 "00364 26 2°6 Or 2 55 10°6 "00355 31 2°6 3 55 9°0 "00355 31 3°] = 49 11°3 00298 29 2°9 5 BE Miaets D "00280 15 at | 6 48 7°0 "00280 15 2°6 JACEE (0 2) RE ge ee 2°8 LOS These results may be taken as fairly representing all that have been obtained. The agreement between theory and experi- ment is fully as good as could be expected when all the uncer- tain elements in the determination aretaken into consideration. These uncertain elements are: (1) The actual charges carried by the spheres and the effect of surrounding bodies, especially the plate coated with tin-foil covering the lower end of the magnetometer tube. This tin-foil is cut into strips abcut 1 millimeter in width and its effect must be small. (2) The non-uniformity of distribution of electricity upon the spheres so that the charges cannot be regarded accurately as concen- trated at their centers. (8) Errors in reading the deflection of the needle due to outside disturbances. Experiments have also been made with the direction of motion of the spheres reversed. The results obtained are similar in every respect to those given above, except that they are reversed. Experiments using only a portion of the 10,000 cells of the storage battery gave results which agree fairly well with the preceding. The deflections were too small, however, to expect very close agreement. Jefferson Physical Laboratory, Harvard University, Cambridge, Mass. 168 Dewar—The Nadir of Temperature Art. XVIL—TZhe Nadir of Tenperature and Allied Prob- lems, by James Dewar, LL.D., F.R.S. [Bakerian Lecture (abstract) read before the Royal Society,* June 13, 1901.] 1. Physical Properties of Liquid and Solid Hydrogen. 2. Separation of Free Hydrogen and other Gases from Air. 3. Electric Resistance Ther- mometry at the Boiling Point of Hydrogen. 4. Experiments on the Lique- faction of Helium at the Melting Point of Hydrogen. 5. Pyroeleectricity, Phosphorescence, etc. Details are given in this paper which have led to the follow- ing results :— The helium thermometer which records 20°:5 absolute as the boiling point of hydrogen, gives as the melting point 16° abso- Inte. This value does not differ greatly from the value pre- viously deduced from the use of hydrogen gas thermometers, viz., 16°°7. ‘The lowest temperature recorded by gas thermom- eter is 14°°5, but with more complete isolation and a lower pressure of exhaustion, it will be possible to reach about 13° absolute, which is the lowest temperature that can be com- manded by the use of solid hydrogen. Until the expermments are repeated with a helium gas thermometer filled at different pressures, with the gas previously purified by cooling to the lowest temperature that can be reached by the use of solid hydrogen, no more accurate values can be deduced. The latent heat of hquid hydrogen about the boiling point as deduced from the vapor pressures and helium-thermometer temperatures, is about 200 units, and the latent heat of solid hydrogen is about 16 units. ; The order of the specific heat of liquid hydrogen has been determined by observing the percentage of liquid that has to be quickly evaporated under exhaustion in order to reduce the temperature to the melting point of hydrogen, the vacuum ves- sel in which the experiment is made being immersed in liquid air. It was found that in the case of hydrogen the amount that had to be evaporated was 15 per cent. This value, along with the latent heat of evaporation, gives an average specific heat of the liquid between freezing and boiling point of about 6. When liquid nitrogen was similarly treated for comparison, the resulting specific heat of the liquid came out 0°48 or about 6 per atom. Hydrogen therefore follows the law of Dulong and. Petit, and has the greatest specific heat of any known substance. The same fine tube used in water, liquid air, and liquid hydrogen gave respectively the capillary ascents of 15°5, 2 and 5°5 divisions. The relative surface tension of water, liquid air, and liquid hydrogen are therefore in the proportion of 15:5, 2, 0-4. In other words, the surface tension of hydrogen at its * From an advance proof received from the author. and Allied Problems. 169 boiling point is about one-fifth that of liquid air under similar conditions. It does not exceed one thirty-fifth part the surface tension of water at the ordinary temperature. The refractive index of liquid hydrogen, determined by measuring the relative difference of focus for a parallel beam of light sent through a spherical vacuum vessel filled in succes- sion with water, liquid oxygen, and liquid hydrogen, gave the value 1:12. The theoretical value of the liquid refractive index is 1-11 at the boiling point of the liquid. This result is sufficient to show that hydrogen, like liquid oxygen and nitro- gen, has a refractivity in accordance with theory. Free hydrogen, helium, and neon have been separated from air by two methods. The one depends on the use of liquid hydrogen to boil the dissolved gases out of air kept at a tem- perature near the melting point of nitrogen; the other on a simple arrangement for keeping the more volatile gases from getting into solution after separation by partial exhaustion. By the latter mode of working something like 1/34000th of the volume of the air liquefied appears as uncondensed gas. The latter method is only a qualitative one for the recognition and separation of a part of the hydrogen in air. In a former paper on the “‘ Liquefaction of Air and the Detection of Impur- ities,’* it was shown that 100 ce. of liquid air could dissolve 20 ec. of hydrogen at the same temperature. The crude gas separated from air by the second method gave on analysis— hydrogen 32°5 per cent., nitrogen 8 per cent., helium, neon, &c., 60 per cent. After removing the hydrogen and nitrogen the neon can be solidified by cooling in lquid hydrogen and the more volatile portions separated. There exists in air a gaseous material that may be separated without the liquefaction of the air. [or this purpose air has to be sucked through a spiral tube filled with glass wool im- mersed in liquid air. After a considerable quantity of air has been passed, the spiral is exhausted at the low temperature of the liquid air bath. The spiral tube is now removed and allowed to heat up to the ordinary temperature, and the con- densed gas taken out by the pump. After purification by spectroscopic fractionation the gas filled into vacuum tubes gives the chief lines of xenon. The spectroscopic examination of the material will be dealt with in a separate paper by Professor Liveing and myself. A similar experiment made with liquid air kept under exhaustion, the air current allowed to circulate being under a pressure less than the saturation pressure of the liquid to prevent liquefaction, resulted in cryp- ton being deposited along with the xenon. A study of fifteen electric-resistance thermometers as far as the boiling point of hydrogen, has been made, and the results * Chem. Soc. Proc., 1897. 170 Dewar—The Nadir of Temperature reduced by the Oallendar and Dickson methods. The table [here omitted] gives the results for seven thermometers, viz., two of platinum, one of gold, silver, copper, and iron, and one of platinum-rhodium alloy. Of these the lowest boiling point for hydrogen was given by the gold thermometer. Next to it came one of the platinum thermometers, and then silver, while copper and the iron differ from the gold value by 26 and 32 degrees respectively. The gold thermometer would make the boiling point 23°°5 mstead of the 20°-5 given by the gas ther- mometer. Then the reduction of temperature under exhaus- tion amounts to only 1° instead of 4° as given by the gas thermometer. The extraordinary reduction in resistance of some of the metals at the boiling point of hydrogen is very remarkable. Thus copper has only 1/105th, gold 1/30th, plati- num 1/35th to 1/17th, silver 1/24th the resistance at melting ice, whereas iron is only reduced to 1/8th part of the same initial resistance. The real law correlating electric resistance and temperature within the limits we are considering is unknown, and no thermometer of this kind can be relied on for giving accurate temperatures up to and below the boiling point of hydrogen. The curves are discussed in the paper, and I am indebted to Mr. J. H. D. Dickinson and Mr. J. E. Petavel for help in this part of the work. Helium separated from the gas of the King’s Well, Bath, and purified by passing through a U-tube immersed in liquid hydrogen, was filled directly into the ordinary form of Cailletet gas receiver used with his apparatus, and subjected to a pres- sure of 80 atmospheres, while a portion of the narrow part of the glass tube was immersed in hquid hydrogen. On sudden expansion from this pressure to atmospheric pressure a mist from the production of some solid body was clearly visible. After several compressions and expansions, the end of the tube contained a small amount of a solid body that passed directly into gas when the liquid hydrogen was removed and the tube kept in the vapor of hydrogen above the liquid. On lowering the temperature of the liquid hydrogen by exhaustion to its melting point, which is about 16° absolute, and repeating the expansions on the gas from which the solid had separated by the previous expansions at the boiling point or 20°°5, no mist was seen. From this it appears the mist was caused by some other material than helium, in all probability neon, and when the latter is removed no mist is seen, when the gas is expanded from 80 to 100 atmospheres, even although the tube is surrounded with solid hydrogen. From experi- ments made on hydrogen that had been similarly purified like the helium and used in the same apparatus, it appears a mist can be seen in hydrogen (under the same conditions of expan- sion as applied to the helium sample of gas) when the initial and Allied Problems. al temperature of the expanding gas was twice the critical tem- perature, but it was not visible when the initial temperature was about two and a-half times the critical temperature. This experience applied to interpret the helium experiments, would make the critical temperature of the gas under 9° absolute. Olszewski in his experiments expanded helium frem about seven times the critical temperature under a pressure of 125 atmospheres. If the temperature is calculated from the adia- batie Expansion starting at 21° absolute, an effective expansion of only 20 tol would reach 6° "3, and 10 to 1 of 8°°3. It is now safe to say, helium has been really cooled to 9° or 10° absolute without any appearance of liquefaction. There is one point, however, that must. be considered, and that is the small refractivity of helium as compared to hydrogen, which, as Lord Rayleigh has shown, is not more than one fourth the latter gas. Now as the liquid refractivities are substantially in the same ratio as the gaseous refractivities in the case of hydrogen and oxygen, and the refractive index of liquid hydrogen is about 1:12, then the value for liquid helium should be about 1:03, both taken at their respective boiling points. In other words, quid helinm at its boiling. point would have a refractive index of about the same value as liquid hydrogen at its critical point, and as a consequence, small drops of liquid helium forming in the gas near its criti- eal point would be far more difficult to see than in the case of hydrogen similarly situated. The hope of being able to liquefy helium, which would appear to have a boiling point of about 5° absolute, or one- fourth that of quid hydrogen, is dependent on subjecting helium to the same process that succeeds with hydrogen; only instead of using liquid air under exhaustion as the primary cooling agent, ‘liquid hydrogen under exhaustion must be employ ed, “and the resulting liquid collected in vacuum vessels surrounded with liquid hydrogen. The following table em- bodies the results of experience and theory: Initial temperature. t inne) ELISE Boiling points. emperature. temperature. Mane melvin tf lh lo ae 20% 1? peatenbyGroeen 85-2255) 5522 ke 15 6 4 LULL 9 (eee ee 20 8 5 (He?) Baliausicd tquid airs) __ 2.2 +. 75 30 20 (H) 2 Ae et le ke es a 325 130 86 (Air) eturinet THoat 29805 2. sot et. 750 304 195 (COz) The first column gives the initial temperature before con- tinuous expansion thre ough a generator, the second the critical point of the gas that can be ‘liquefied under such conditions, and the third the boiling point of the Rare liquid. It will AM. Jour. Sci1.—FourtTH ee Vou. XII, No. 68.—Avgeust, 1901. 12 172 Dewar—The Nadir of Temperature, ete. be seen that by the use of liquid or solid hydrogen as a cooling agent we ought to be able to liquefy a body having a critical point of about 6° to 8° absolute and boiling point of about 4° or 5° absolute. Then, if liquid helium could be produced with the probable boiling point cf 5° absolute, this substance would not enable us to reach the zero of temperutnre; another gas must be found that is as much more volatile than helium as itis than hydrogen in order to reach within 1° of the zero ot temperature . If the ees get comprises a substance having the atomic weight 2, or half that of helium, such a gas would bring us nearer the Het goal. In the mda tine the production of liquid helium is a difficult and expensive enough problem to long occupy the scientific world. A number of miscellaneous observations have been made in- the course of this inquiry, among which the following may be mentioned. Thus the great increase of phosphorescence in the case of organic bodies cooled to the bouing point of hydrogen under light stimulation is very marked, when compared with the same effects brought about by the use of liquid air. A body like sulphide of zine cooled to 21° absolute and exposed to light shows brilliant phosphorescence on the temperature being allowed to rise. Bodies lke radium that exhibit self- luminosity in the dark, cooled in liquid hydrogen maintain their luminosity unimpaired. Photographic action is. still active although it is reduced to about half the intensity it bears at the temperature of liquid air. Some erystals when placed in liquid hydrogen become for a time self-luminous, on account of the high electric stimulation brought about by ‘the cooling causing actual electric discharges between the crystal mole: cules. This is very marked with some. platino-cyanides and nitrate of uranium. Even cooling such erystals to the tem- perature of liquid air is sufficient to develop marked electrical and luminous effects. Considering that both liquid hydrogen and air are highly insulating liquids, the fact of electric dis- charges taking place under such conditions proves that the electrical potential generated by the cooling must be very high. When the cooled crystal is taken out of either liquid and allowed to increase in temperature, the luminosity and electric discharges take place again during the return to the normal temperature. A erystal of nitrate of uranium gets so highly charged electrically that, although its density is 2-8 and that of liquid air about 1, it refuses to sink, sticking to the side of the vacuum vessel and requiring a marked pull on a silk thread, to which it is attached, to displace it. Such a erystal rapidly removes cloudiness from liquid air by attracting suspended particles to its surface. The study of pyro-electricity at low temperatures will solve some very important problems. Miscellaneous Intelligence. 173 SCIENTIFIC INTELLIGENCE. 1. Magnetic effect of Hlectrical Convection. —The magnetic effect produced by the motion of an electrified body, first proved by Rowland in 1876, and, though questioned by Cremieu, since then repeatedly confirmed (see the article by Edwin P. Adams in the present number, pp. 155-167, of this Journal), has been also established by Harotp PENDER at the Johns Hopkins Physical Laboratory. An account of his experiments is given in a recent issue of the Johns Hopkins University Circular (No. 152, May- June, 1901). The method of Cremieu, which, however, gave him negative results, was employed. To produce the convection, two micanite discs (diam. 30 cm.) gilded on both sides and charged from a Voss machine and battery of six gallon Leyden jars, were driven at a speed of 75 to 100 revolutions per second. An inter- rupter made it possible to reverse the charge 12 to 25 times per second. EHarth-connected condensing plates were fixed opposite each face of each disc and one centimeter distant. Between the two inside condensing plates was suspended a coil of 1295 turns of No. 21 copper wire connected through a commutating device with an extremely delicate astatic galvanometer. The coil, cir- cuit, and galvanometer were enclosed in earth-connected metallic shields. The arrangement of the commutator was such that the alternating current induced in the coil by the reversal of charge in the rotating discs gave a steady deflection of the galvanometer. Tbe apparatus was used in two ways: (a) the two discs were rotated in the same direction and at any instant charged alike; and (b) the discs were rotated in opposite directions and charged oppositely at any instant; of these, the second method gave steadier deflections and was used more frequently. ‘The direction of deflection, repeatedly tested, was found, as expected, to be always in accordance with Ampére’s rule, i. e., the’ motion of a positive charge always produced the same effect as that of a con- duction current flowing in the direction of motion of the charge. The quantitative results are discussed, as follows: “The strength of a convection-current is defined as the quantity of electricity carried convectively past any point in unit time. On the assumption that a convection-current is magnetically equivalent to a conduction-current of the same strength, the current induced in the suspended coil on reversing the sign of electrification of the discs can be readily calculated from the dimensions of the apparatus and the difference of potential between the discs and plates. This calculated value of the current can then be compared with the observed value as deduced from the deflection of the galvanometer. The formula for the caleulated value of the current involves the ratio of the two systems of electric units, so that instead of comparing directly the observed and calculated values of the current the two can be equated and the vaiue of the ratio » thus determined. The value of this ratio thus found is a test for the accuracy of the assump- tion that a convection-current is equivalent magnetically to a 174 Scientific Intelligence. conduction-current. From 17 sets of observations, each set con- sisting of 18 separate determinations of the deflection and the other quantities involved, the mean value of v thus found was 3°05 <10*°, the determinations which differed from this the most being 2°75X10" and 3°24 X10", respectively. The value of this constant is known to be 3°00 x 10".” Similar results were obtained when discs and condensing plates were divided each into six sectors by radial scratches, thus showing that the effect observed was not due to conduction-currents in their surfaces. To prove, further, that the deflection observed was actually due to the magnetic action of the rotating charged discs, a further experiment was tried. The discs were rotated in the same direc- tion and at any instant charged oppositely ; and, again, rotated in opposite directions and at any instant charged alike. Their magnetic effects on the coil should then annul one another pro- vided the two discs rotated with the same speed, and this was the result obtained. For example, when the two discs were rotating in same direction with speeds of —86°9 and —88°8 revolutions per second, respectively, and charged alike, the observed deflection was —66:2; when they were rotating oppo- sitely under the same conditions, with speeds +880 and —89°6, the observed deflection was —1:0. It is concluded, therefore, that the results obtained “show beyond any doubt thai electrical convection does produce magnetic action, or, more exactly, that when the sign of electrification of a moving charged body is changed, an electric conduction-current is induced in a neighboring circuit, of a strength equal to that which would be induced in this circuit by reversing the direction of a conduction-current in a circuit coinciding with the path of the convection-current.” | 2. American Association for the Advancement of Science.— The fiftieth annual meeting of the American Association will be held in Denver, August 24 to 31. Professor C. G. Minot of Cambridge will preside as President. The address of the Perma- nent Secretary, Mr. L. O. Howard, is Washington, D. C. (Cosmos Club) until Aug. 15, after that, Brown Palace Hotel, Denver. Information in regard to transportation, hotel accommodation, etc., may be obtained from the Local Secretary, Mr. Arthur Williams, Chamber of Commerce, Denver. OBITUARY. Dr. Josepn LuConte, Professor of Geology and Natural His- tory at the University of California, died on July 6, at the age of seventy-eight years. A notice is deferred until another number. . Proressor Prrer Gururie Tarr, the distinguished Scotch mathematician and physicist, died on July 4, at the age of sev- enty years. He had held the chair of Natural Philosophy in the University of Edinburgh since 1860. His original contributions, chiefly in mathematical physics, were numerous and important; he was also a lucid and suggestive writer in his treatment of scientific subjects from the popular standpoint. THE AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES. | a Art. XVIII.— The Discharge-current from a Surface of large Curvature ; by JoHN E. Aumy, Ph.D., Instructor in Physies, University of Nebraska, Lincoln. 1. THE discharge of electricity from a surface of large curvature has been studied by Warburg,* and others.f Inas- much as the nature and form of the discharging surface, when a point, or pointed wire, is used, is necessarily more or less prob- lematic, it seemed of interest to study the discharge from a fine wire of uniform dimension. The discharge from a platinum wire of very small size to a con- centric, circular cylinder has been the subject of study. 2. The discharge apparatus is shown infig. 1. The brass cylin- der, C, cemented to the glass plate, is further insulated by the supports of sealing-wax, s, s.- The discharge wire, passing through the capillary tube, T, embedded in parafiin, hangs ver- tically, stretched by the weight of a short piece of tube, T’, in which its lower end is embedded. Paraffin discs, D, D’, serve to keep the wire at the center of the cylinder. The gas is removed and supplied by tubes P and * Warburg, Wied. Ann., Ixvii, p. 82; Ann. d. Phys. (4), ii, p. 295. + Sieveking, Ann. d. Physik. (4), i, p. 299, ete. Am. Jour. Sct.—FourtH Series, Vou. XII, No. 69.—SEPTemMBER, 1901. 176 Almy—Discharge-current from a : p. The cork, 8, which closed the upper end of the cylinder, was coated with sealing-wax, surface-leakage, even when the glass tube became coated with moisture, being thus prevented. The electric connections are sketched in fig. 2. W,a Wims- hurst machine, was driven bya small motor. A galvanometer (D’Arsonval), G, measured the discharge current; potential dif- ferences between wire and cylinder were read on a ‘ Braun,’ — later a Kelvin, ‘ vertical,’ voltmeter, V. Potential differences were regulated by a fine discharge-point shunting the poles of the Wimshurst. As air, which was the gas used throughout the first portion of the observations, shows some deterioration with use, when the discharge current attains to a measurable value (by “ deterioration ” is meant a decrease of its capacity to carry on the discharge), fresh, dried, gas was continuously drawn through the discharge apparatus. As only relativity was wished for, the voltmeter and galva- nometer were not standardized; nor were the variations of atmospheric pressure, from day to day, noted. Observations. 3. First, the proportionality of total discharge to length of the discharge wire was tested. It was found, that so long as 2 E iS the length of wire is greater than the diameter of the cylinder, and with the paraffin discs restricting the volume of the gas which takes part in the discharge to the solid cylinder of the length of the wire, the current is directly proportional to the length of the wire, with a given applied potential, and the minimum potential difference that will produce measurable discharge current is independent of the length of the wire. In case short lengths (as compared to cylinder diameter) Sunface of large Ourvature. LTT were used, or the discs were dispensed with, exact propor- tionality no longer holds. A series of observations illustrating this proportionality are given : | TABLE I. Values of the quotient. Current. Wire length. Potentials. To —=93 lem 2, = 25m, iO es i —— OC 14 16°7 Ti ges 16°0 16 36°1 30°4 36°6 34° 18 60°6 59°3 60°6 60° 20 90:0 90°2 92°6 88° 22 120° 118°2 WBE V205 24 mee fe 160° 162° 1633. 26 cae Safes 214: 216° Diameter of cylinder, 9°4°™. Diameter of wire, 0:0047°. 4. The relation holding between the applied potential and the resulting discharge current is found to be expressed by the quadratic : , I=aLV(V—d) I, V, and L, having the usual significance, and @ and 6 are constants, depending upon the cylinder, wire, and gas. The constant } is found to approximate, more or less closely, to the “minimum-potential” of Rontgen, i. e., the least potential which will give the measureable discharge. This law was found applicable for cylinders of diameters, from 2 to 10™, giving agreement within errors of observation, except for potentials which differed only very slightly from the “ minimum potential.” Even after the gas had been used for a considerable time without renewal and also at widely varying pressures (from 20 to 80°" of mercury) the relation between current and potential remains similar, with, naturally, different constants. As illustration I cite a series, taken at random from the considerable number of observations made : TABLE II, : V al V il (arbitrary units.) (cale.) (observ.) (of cylinder.) (calc.) (obs.) 13 13°77 17:0 —14 15°38 7 14 26°5 Dies —16 36°2 36° 16 56° 5a° —18 62° 61° 18 pA? 92° —-20 92° 92° 20 as 5)9 138° —22 126: 122° 22 184° 186° —24 166° GZ 24 240° 240° —26 210° ZO: 26 313° 320° —28 260° 251: a= ‘0438; b= 11%. —30 313° Bilis Coe Op = 12915. 178 Almy—Discharge-current from a Length of wire, 15°"; diameter, 0:0047™,; diameter of cylinder, 9:45 pressure of the gas (air), 74:7 of Hg. The variation of the constants, a, and 6, when the sign of the dis- charge is reversed, is, in the case of air, not large; but in case hydrogen is used a very large difference exists. The equation given above seems applicable to the discharge under all the various conditions of the gas. Even after the gas had suffered any amount of “ deterioration,” so that the discharge that passed was decidedly less, the relation between potential and current still exists,—with different values for the constants, @ and 0. 5. The applicability of this relation at widely different gas pressures (varying from 20°" to 80° of Hg.) was also estab- lished. ‘The variation of the discharge current, with gas pres- sure, for a given apparatus, does not seem capable of any simple formulation. It seems probable that this was due to a difference in the rate and extent of the deterioration of the gas at the different pressures, so that the constants obtained repre- sent different conditions of the gas, beside differences of pres- sure. The constants obtained are : Pressure. a. b. 20°™ of Hg. 3°55 3°20 AO ess “ 1:05 8°53 TOS ee "09—'65 11°7—12:2 The dimensions of the apparatus were those given in the pre- ceding section. 6. For determination of the law of the variation of current with the radius of the cylinder, three cylinders of suitably dif- ferent radii, with discharge wires of the same size and length, were mounted as the one used heretofore. Hydrogen, gen- erated from hydrochloric acid with zinc, purified and dried by the usual process, was passed into the cylinders in place of air, as this gas shows less rapid deterioration with continued use. Still, under the most favorable conditions that could be obtained, the current given by a certain potential was subject to considerable variation. It was found that to a certain degree of approximation, the total discharge current is inversely proportional to the cube of the radius of the cylinder, other dimensions being the same. It is to be noticed that this relation holds most accurately when the current in the cylinders compared is of the same order of magnitude. It seems probable that the deviation which occurs when the current in one tube is large compared to that in the other may be due to more rapid deterioration of the gas with increased discharge; in fact, the deviation is always most pronounced when the discharge is largest in both tubes. i Surface of large Curvature. 179 A typical series of observations is given below: the lengths of the discharge wires were each 10™; the radii of the cylin- ders were 5°0™, 3:2 and 1°5™ respectively ; that of the dis- charge wire was 0°0034™. TABLE III. Applied Current. Ratios. potentials. I, I, ic dey ie in for for for = Te volts on: f= 372 rab : E 3500 4°0 16°0 152 4:0 9°5 3800 wears 24° 200 charg es 8°3 4000 9°0 36° 290 4°0 8°0 4500 1h 74: 480 3°9 6°5 5000 eS 118° baie iis 3°8 gage 5500 55° 189 eee 3°4 TY Mean values, 3°82 8:1 The ratios 7,°/r,° and 7,'/r¢ have the values 3°8 and 9°%, respectively. The variations obtained were usually within the limit of those shown in this series. 7. In résumé, then, it has been found that the current dis- charging from’ a fine wire to a concentric surrounding cylinder is given by the equation,— I= = aV (V—2) where, I = the discharge current = the potential difference between wire and cylinder L= length of the discharge wire ry = radius of the cylinder 6=the minimum potential necessary to produce a meas- urable discharge, and, @= a constant depending upon the size of the wire, ¢ the discharging gas, and the sign of the discharge. Physical Laboratory, University of Nebraska, May, 1901. 180 Robinson— Octahedrite and Brookite. Art. XIX.— On Octahedrite and Brookite, from Brindle- town, North Carolina; by H. H. Ropinson. THE two minerals described in this article were kindly sent to Prof. S$. L. Penfield by Mr. W. E. Hidden, and are from the original locality on the north slope of Pilot Mountain near Brindletown, Burke County, North Carolina, discovered by him in 1879.* They occur, according to Mr. Hidden, in detached erystals scattered through the gold-bearing gravels of the dis- trict, having been derived from the disintegration of the local schists. As accompanying minerals, are mentioned zircon, monazite, xenotime, samarskite, fergusonite, and many others. It was thought that brief descriptions, with figures, of these two minerals would be of interest both because of their rare occurrence in the United States and because no careful investi- gation of the crystals from this locality seems to have been made. 1. Octahedrite. The crystals studied vary in size from 2 to 5™™ in diameter on the horizontal axes and from 0°50 to 1:°25™™" in thickness. The larger ones are of a deep bluish-black color and then they possess a metallic-adamantine luster, while a smaller one is quite colorless. Under the microscope, the color is seen to be due to a dark pigment which entirely pervades some crystals, while in others transparent spots are left, around the borders of which the pigment gradually fades out. A small transpar- ent crystal, representing a section perpendicular to the vertical axis, was dark in all positions between crossed nicols. In con- verging polarized light, however, it was seen that the mineral was not perfectly uniaxial, as the axes of the interference figure opened slightly on revolving the section. In habit the crystals are tabular owing to the prominent development of the basal pinacoid, ¢ (001). This Mr. Hidden mentions as being the common habit. The simplest crystal examined consists of the forms ¢ (001), ¢ (101), « (103), and p (111), and is shown by figures la and 16, of these, la is a horizontal projection (on ¢) and 16 shows the crystal in the usual perspective.t In figures 2a and 2b aslightly more com- * This Journal (3), xxi, 160, 1881. + In this article the method has been adopted of giving with each clinographic drawing of the crystals a plan, which represents a true horizontal projection of the lower figure. In most cases the drawing of such a plan greatly facilitates the construction of the customary crystal figure, giving a complete control over it, and the two figures taken together give a better idea of the crystal shape than either one alone. Since there exists the closest relationship between the two figures, it has been considered advisable to represent them in the position in which they are drawn. Accordingly each plan is shown as revolved a certain angle (18° 26’) from a horizontal position. Robinson—Octahedrite and Brookite. 181 plex combination is represented with two additional faces, o (107) and vw (117). These two figures show all the prominent forms as well as their relative development. The complete list of faces observed is as follows: ¢ (001) o (107) y (902) y (1:1°28) e (101) d (301) p (111) p (1:1:40) x (103) u, (5°0°19) ili) The errors in the measured values of the above faces as com- pared with the values given by Dana* range from 1’ to 10’, the average being 33’. Though most of the faces give rather poor reflections, the measurements are all close enough to clearly establish their identity. { (a. The form v (11-28) was first observed by Seligmann,t who identified it with but moderate certainty, owing to the poor development of the faces on the crystals examined by him. The angles measured on two of these crystals leave no doubt as to the correctness of the indices established by him, as is shown by the following: Calculated. Measured (1). Average. Measured (II). Average. CAV, 5° 72! Sra -3a preks gp ag Fey 2 br TEE. inc dy WL 2 bigs The above measurements and the fact that the face lies in the zone cap are sufficient to establish the identity of this * System of Mineralogy, 6th ed., 1892, 240. + Zs. Kr., ii, 337, 1886. 182 Robinson—Octahedrite and Brookite. form. The faces are all of small size, yet the measurements from five give an average error of but 14’. On one crystal three out of the possible four forms were observed at one extremity, on a second only two. 2. Brookite. Five crystals of this mineral were examined and they are of special interest because they possess a habit quite different from that ordinarily observed. The habit of all the crystals is prismatic. Their develop- ment on the a and 0 axes is well shown in the plans of figures 3-6. The length in the direction of the vertical axis is ‘from four to six times their greatest width. Being relatively so long and slender, it did not seem best to represent them as doubly | terminated. 5 (a) The unusual feature of these crystals is the prominent devel- opment of the pyramid s (822). This form was first observed by Kokscharow* as a very small face, and the only notice of its having a prominence at all like that on the crystals from this locality is in an article by Streuver, “Sulla Brookite di Buera nell’Ossola.”+ The habit of the crystals described by him is, however, quite different in that it is short and tabular parallel to a (100), the proportion between the width and thickness being as 1: $. | *Vh. Min. Ges., St. Petersburg, 1848-49. - + Proc. R. Acc. d. Line. (4), vi, 77, 1890. Robinson—Octahedrite and Brookite. 183 Figures 3a and 30 show the simplest type, the crystal being terminated by the single pyramid s (822) in combination with a _ small basal pinacoid. In figures 4@ and 40, s (822) is in combi- nation with a dome y (104) and a much larger basal plane. Figures 5a@ and 50 show the pyramids e (122) and 2 (112), which are common on brookite, about equal in their development to s(822). The pyramid @ (54:10) is new, occurring, as shown, asa narrow face in a zone with e, z, and w Of the domes, y (104) and #(102) are common, while 7 (101) is new. In figures 6a and 66 the general habit is more like that ordinarily observed in brookite, the domes y, 2, and ¢, and the pyramid e all being common forms. The pyramid s (322), o (824), and v (326) are in azone and o (324) isnew. The pyramid pv (146) is also new. Inthe prismatic zone a (100), m (110), and 7 (210) are largely developed, while a (320) and & (410) are present as narrow faces. All the crystals are of small size, ranging from 1°25 to 2°50™™ in width, 0°75 to 1:75™™ in thickness, and 3 to 7™™ in length. Their color is a dark reddish-brown and renders them too opaque for optical examination in natural sections. Vertical striations on the prismatic faces, with the exception of m, are always very marked, giving them a dull luster. The terminal faces are all somewhat striated, though not sufficiently to greatly injure the quality of the larger ones. The faces observed, with their measurements as made on a two-circle goniometer and their calculated values, are given in 184 Robinson—Octahedrite and Brookite.’ the following table. The scheme is that of Goldschmidt as given in his “ Krystallographische Winkeltabellen.”’ Faces. Measured. Calculated. Error. 9 p vy) p Dorn e (001) Lows ee aie Bll Soe a (100) 89° 58’ 89° 58’ 90° 0 OOF 0" 2 a m (110) 49 55 Be ia 49 55 reuaye Ona a (320) 60 50 crierage 60 42 AP Nise i ¢ (210) Gy 3 oe OM KO RG wbnite a iae k (410) vamos Sai Ane 78 O07 Ca aes Hea uke: d (043) i By rs ice 51 32 at t (021) nee GH O Dae 62 06 sic) gh y (104) oo Di We ay 90° On 15240 S.A x (102) Sry Se 29 15 rea tae 29 18 ola 7 (101) (new) macs 48 18 aes 487174 2 3eanus v (146) os 16 304 33 234 16 324 33 18 lS é (122) 30 47 47 38 30 438 47 Al 4 38 2 (U2) 49 53 36 12 49 55 36 15 Pgh 2) o (111) 49 54 55 47 He eer MESS Dy ee B (54:10) (new) 55 55 34 05% 56 03. 34 045° ene v (826) 60 43 382 44 60 42 32 45 Ls) bee g (324) (new) 60 464 43 49 i ete ye) 4410 8 60 42 62 35 Bed 62 36 OS Average error, Se In calculating the values of the four new forms given in the foregoing table the axial ratio of a:b: c= 0°84158 : 1:0°944389, as determined by Kokscharow,* was used. The form 6 (5°4-10) was observed on one crystal (figure 5), three faces being pres- ent on one end and two on the opposite, out of a possible four in each case. The dome 7 (101) was observed but once on the same crystal. The pyramids o (824) and v (146) were observed on but one erystal (figure 6), the first form being represented by two and the second by three faces. Sheffield Laboratory of Mineralogy, Yale University, New Haven, Conn., May, 1901. * Mineralogie Russlands, I, 61, 1853. ~B. Davis—Small closed Cylinders in Organ-pipes. 185 Art. XX. — On the Behavior of Small closed Cylinders in Organ-pipes ; by BERGEN Davis, Ph.D. WHILE experimenting with stationary sound waves in organ- pipes, the following striking effect was obtained. A consider- able number of small gelatine capsules, such as are used for medical purposes, were thrown in a promiscuous pile in the center of the pipe, and when the pipe was blown so as to give its first overtone quite strongly, the small cylinders immediately moved to the middle of the loop of the stationary wave and _ there arranged themselves in rows across the pipe as shown in the figure. The spacing between the rows was quite regular and the capsules acted as though there was a strong attraction at their ends, in a direction perpendicular to the vibration, and a repulsion at their sides in a direction par- allel to the vibration. An investigation was undertaken to determine the effect of the size of the cyl- inders, of the number of rows, and of the amplitude of vibration, upon the dis- tances between the rows. The stationary wave was that produced in a stopped organ-pipe speaking its first overtone. Ths side of the pipe was removed and a glass plate substituted for it. At the node nearest the mouth a thin rubber diaphragm was placed across the pipe, which protected the portion back of the diaphragm from disturbances arising from blowing. This region of the pipe, from the diaphragm to the stopped end, enclosed one-half of the sta- tionary wave. The amplitude of vibration was measured by means of the force acting on a small hollow cylinder closed at 186 BL. Davis—Small closed Cylinders in Organ-pipes. one end, as was described by the writer in this Journal.* The torsion balance carrying this cylinder was placed at the dale of the loop. ? The small cylinders whose spacing in the sound wave was to be measured were of two kinds, consisting of small gelatine capsules and of paper tubes. These two kinds of cylinders. were used in independent series of experiments. Three sizes of gelatine capsules were used, commercial num- bers of which are Nos. 00, 2 and 5. The dimensions in centimeters of these cylinders are given below. Length. Diameter. IN OO Oe sha ae on ot 2°42 °83 ING. Gi) Eee Ios) 63 IN O20 1s Ge03 fae 1°03 "48 A. sufficient number of one of the above sizes of capsules were placed in the organ-pipe to form a considerable number of rows. In order to obtain a desirable amplitude of vibra- tion the torsion head was turned from the zero position by an amount corresponding to a desired amplitude, then the pres- sure of blowing was increased until the force acting on the measuring device just balanced the torsion previously given to the wire; at the same time the capsule cylinders arranged themselves in rows across the pipe. This previous setting of the torsion head enabled me to reproduce the same amplitude of vibration at will. In the tables below are given the average distances in centi- meters between the rows for the various sizes, with increasing amplitudes of vibration. The amplitude 2A is here used to denote the total excursion of the vibrating air particles. No. 00. Capsule cylinders. Amplitudes. Number of rows. —————— SeaaN (CRRA SS mimEe Sa 2A 2 3 6 10 39 4° 3°O 226 De 434 3°O 3°3 2°6 aL 476 3°3 3°9 26 2°1 "548 3°95 seth yyaT| 2°2 No. 2. Capsule cylinders. Amplitudes. Number of rows. oo a a sa hae aa TA 2A 2 3 6 10 39 . 2°8 2 2°8 2° "434 2°4 2°5 2°65 1°95 “476 2°5 2°5 2°55 2° 548 2°95 2°5 2°45 2° * This Journal, September, 1900. B. Davis—Small closed Cylinders in Organ-pipes. 187 No. 5. Capsule cylinders. Amplitudes. Number of rows. an OOOO eS OO 2A 2 3 6 10 °39 2 25 2°4 eg) "434 1°8 2°5 2°1 1°55 “476 2 2°1 2-1 1°6 548 2 2°5 Dee 1°66 The most striking result obtained from the above tables is that the distances between rows slightly decrease with increas- ing amplitudes of vibration. The average distance between the rows decreases in general as the number of rows increases, excepting in the case of the smallest, No. 5 cylinders, in which case the spacing increased when three and six rows were used. This tendency to increase with these particular rows is also exhibited by the No. 2 cylinders. The general rule nay also be deduced, that the larger the cylinders, the more the spacing decreases with the number of rows, as will be evident by com- paring the two-row and the ten-row columns in the three tables. At the higher amplitudes the capsule cylinders were quite violently agitated, as though the position of the loop were somewhat unsteady, which increased the difficulties of accurate measurement. The corresponding experiments were performed with paper cylinders of various diameters. These cylinders were each 63°" in length and were open at both ends, on account of the circumstance that the length of one of them was nearly equal to the diameter of the pipe. Since the effect to be observed was a result of the forces acting at the sides and not at the ends, the open ends did not affect the experiment. The open ends were of advantage in that they lessened the force with which the cylinders in all cases adhered to the walls of the ipe. | 4 The diameters of these cylinders in centimeters are given below: Gs Pac neater epgenn UNeal Weee 5 SN 6 eae aa ta LO er al NANI REE Rr sg 9 1:00 NO a Aah giao ee ten 1°4 These paper cylinders were introduced into the closed cham- ber of the pipe and the same experiments performed as with the capsule cylinders. Each cylinder now corresponds toa row as described in the previous experiments. The same amplitudes of vibration were retained for the purpose of com- parison. The results are given in the following tables : 188 BL. Davis—Small closed Cylinders in Organ-pipes. No. 1. Faper cylinders. Amplitudes. Number of rows. 2A 2 3 6 10 oy) 3°2 3°2 2°6 2a. 434 2°6 3°2 2°5 1°9 “476 2°3 3°2 2°5 1°9 548 2°3 BOW 2°4 leg) No. 2. Paper cylinders. Amplitudes. Number of rows. FSS Cae ae SS — 2A 2 3 6 10 39 6°5 3" Di 2°3 "434 6° 3° 258 2°3 "476 6° 3°2 2°9 22 "548 6° 3°2 2°8 2°2 No. 38. Paper cylinders. Amplitudes. Number of rows. Ooo ooo eee eee oo 2A 2 3 6 10 °39 6° 4°3 3°9 2 454 o°9 4° 3° eee "476 5° 4° eer tS arene "548 5° 4 singe pina No. 4. Paper cylinders. Amplitudes. Number of rows, a RTT TTT. Grae ge on) —_——_—_ SasN 2A 2 3 6 10 39 ue o°5 3°3 2°8 434 6°5 5° Beat. a 476 6°5 Pa tere eRe Mts? 548 pho le cee eis Here again it will be noticed that the space between the cylinders slightly decreases with increasing amplitudes of vibration. The space between the rows decreases as the num- ber of rows increase. This result is somewhat different from that obtained with the No. 2 and No. 5 capsule-cylinders, while the same result was obtained as with the large No. 00 capsule-cylinders. ‘The blanks in the columns with the larger cylinders when several rows were used, are due to the fact that the tone passed over into the next overtone before the required amplitude was reached. This was probably due to the effect of friction. The presence of so much obstruction tended to form a node at this point, which is near the natural B. Davis—Small closed Sons in Organ-pipes. 189 position of the node of the next overtone. When this over- tone occurred, the rows would divide into two portions, which moved toward the middles of the two loops of the new stationary wave. A striking effect was obtained by placing a number of the smallest capsule-cylinders a short distance from the node. When the pipe spoke, they immediately ran rapidly to the middle of the loop, and there assumed a regular arrangement. The effects described in this paper are of course of the same nature as the Kundt dust figures. The individual capsule- cylinders may be considered as dust particles in which the size has been much increased, the particles still remaining light enough to respond readily to the delicate forces to which ~ they are subjected. Prof. Rood suggested as an explanation of the spacing that a sound-shadow is formed on the two sides of the cylinders at each half vibration, alternately. These shadows, being regions of less motion, press so to speak against the cylinders. When there are two rows the shadows between the cylinders press them apart until the force just equals that pressing against the outside or nodal sides of the two cylinders. This also explains their rapid movement from the nodes toward the loop. The average velocity on the side nearest a node is less, and hence the pressure greater than on the side nearest a loop, where the velocity is greater. The behavior of the capsule-cylinders illustrates the distribution of the forces acting upon rigid bodies in moving fluids, the mathematical analysis of which has been so fully developed by W. Koenig.* The effect here described can be easily reproduced as a lec- ture experiment. The ordinary stopped organ-pipe found in lecture cabinets will suffice for the purpose. The smaller cap- sules will perhaps be found io give the effect more strongly in case the pipe is not a powerful one. Physical Laboratory of Columbia University, June 1, 1901. * Wied. Ann., xlii, pp. 353, 549, 1891. 190 Wells and Willits—Coasium-Tellurium Fluoride. ArT. XXI.—On a@ Casiwm-Tellurium Fluoride; by H. L. WELLS and J. M. WILLISs. SEVERAL tellurium double fluorides have been described : NaF:TeF, by Berzelius, KF-TeF,, NH,F:TeF, and -BaF,: 2TeF,H,O by Hégbom.* It is noticeable that all these fluo- rides belong to a type which is different from that of the double chlorides, bromides and iodides of tellurium, e. g., 2KCl'TeCl, , 2RbBr: TeBr, and 2CsI°Tel,, etc., which have been thor oughly studied in this laboratory by “Wheeler. t We have undertaken, therefore, an investigation of the combination of cesium fluoride with telluriam fluoride, with the expectation that possibly several types of double fluorides might be obtained. After a systematic examination of the matter, however, we were able to prepare only one double fluoride, CsFTeF’,, which corresponds in type to the previously known fluorides. A concentrated solution of Tel’, was prepared by dissolving about 10% of pure TeQO, in an excess of strong hot hydro- fluoric acid, and to this cesium fluoride was added in small portions, the liquid being concentrated by evaporation and cooled after each addition. At the same time small portions of tellurinm fluoride were added to a concentrated solution of about 50% of cesium fluoride in hydrofluoric acid and this solution was evaporated and cooled in the same manner. Under the widest range of conditions, however, only a single double salt was obtained. 1: 1 Cesium-tellurium fluoride, CsF:Tel’,.—This salt erys- tallizes beautifully in large, transparent, colorless needles. The presence of free hydrofluoric acid is necessary for its forma- tion, for it is decomposed by water. Several crops, made under widely varying conditions, were analyzed with the fol- lowing results: Calculated for Found. CsTeF,. I, 10 inate 166 Cesium. eae ae 37°36 36°59 37°50 38°56 Bs Mellunrimmss 3.2 oes 35°96 30°51 36°45 35°82 35°60 Bilmorine 24520022). 2 26°68 26°76 24°51 26°18 eon Fluorine was determined volumetrically by converting it . into Sif’, collecting this in water, and titrating with a stand- ard solution of potassium hydroxide. In another portion, after evaporating with concentrated sulphuric acid and dissolv- ing the residue in hydrochloric acid, tellurium was precipitated with sulphur dioxide, collected on a Gooch crucible and weighed as metal. From the filtrate from the tellurium ceesium was obtained, and weighed as normal sulphate. Sheffield Scientific School, May, 1901. * Bulletin, xxv, 60. + This Journal (3), xlv, 267. Wells and Willis—Chlorides of Cesium and Thorium. 191 Art. XXII.—On the Double Chlorides of Cesium.and Tho- : rium ; by H. L. Weis and J. M. WIntIs. NEARLY all of the known double halogen salts of quadriva- lent metals belong to a single type, of which 2K CI-PtCl, and 2K F-SiF, are examples. It has been shown, however, by Marignac* and by Wells and Footet that the double fluorides of zirconium exist in a variety of types. Therefore, since thorium is somewhat closely related to zirconium, we have undertaken an investigation of some thorium double halides, and have selected the cesium salts as being the most prom- ising. ison attempting to prepare cesium thorium fluorides we found that thorium fluoride is practically insoluble even in concentrated solutions of cesium fluoride containing hydro- fluoric acid. There is no doubt that the two fluorides com- bine under these circumstances, but since we obtained only finely divided precipitates as products and there was no cer- © tainty as to their purity, further work on the fluorides was abandoned. Chydenius{ has previously described two potas- sium thorium fluorides, 2K F:ThF,4H,O and KF:ThF,4H,0O, but on account of the insolubility of thorium fluoride and of these double salts it is probable that there may be some doubt in regard to the correctness of these formule. We have prepared two czsium-thorium chlorides, to which we assign the formule 38CsCl:ThCl,12H,O and 2CsCl-ThCl,: 41H,O. The amount of water of crystallization in these com- pounds is somewhat uncertain, since they form very small hygroscopic crystals, and it is difficult to dry them by pressing on paper. The search for double chlorides was made syste- matically by starting with a solution of about 65% of thorium chloride in hydrochloric acid, adding 2 to 4® of cesium chloride at a time, and evaporating and cooling after each addi- tion until finally, after dividing the solution and using a part of it, a very large excess of ceesium chloride was present. In analyzing the salts, chlorine was determined as silver chloride, sometimes in separate portions, in other cases in the filtrates from which thorium hydroxide had been precipitated ; thorium was weighed as oxide after precipitation with ammonia, and the cesium in the filtrates was converted into normal sulphate and weighed as such; water was determined by dif- ference. * Ann. Chim. Phys. (3), lx, 257, + This Journal (4), i, 18; iii, 466. ¢ Pogg. Ann., exix, 43. Am. Jour. Scil.—FourtH Series, Vou XII, No. 69.—SepTEMBER, 1901. 14 192 Wells and Willis—Chlorides of Cesium and Thorium. 3; 1 Cesium-thorium chloride, 3CsCl:ThCl,12H,O.—This salt was produced from solutions containing about 128 of - thorium chloride and from 80 to 110% of cesium chloride. It forms colorless crystals of feathery structure upon cooling very concentrated solutions. Three different crops made under somewhat varied conditions gave the following results upon analysis: . Calculated for ' Found. CssThCl,"12H.0. it. IL Ty Cresiuni ane 36°45 36°21 36°14 Oe Thorium 5. ee 21°20 PX OP 7((0) 21°68 21°05 Chlonimesee as ee 22°61 23:09 me OB aT Water ela 19°74 [20-00] : “sas 2:1 Cesium-thorium chloride, 2CsCl:ThCl,11H,O.—This salt was obtained in colorless crystals, somewhat resembling the previous salt, but not nearly as feathery in appearance. It was formed in concentrated solutions containing about 65% of thorium chloride and from 30 to 100% of cesium chloride. The following analyses were made of different crops : Calculated for Found CSE NA ONEROe IE ick IV. est) 22 se 2 Re ok 29°26 29°84 29°19 28°92 ee CRinorauin | oe eee eee Cee DSO) 25°41. 25°70 25'22 Chlorine .__.- apa ae: VOLAS 23°40 DATs DSP 3 V5) 93°55 WN hater) ac. peat see 21°78 [21:34] [20°74 | [22°03] The salt loses water slowly in the desiccator over sulphuric acid. A sample dried in this way lost 6 per cent in two days, Al per cent after one week, and 20 per cent, corresponding to practically all the water, after one month. | The two chlorides that we have obtained are different in type from the potassium salt KCl-2ThCl,18H,O described by Cleve* and from the ammonium salt 8NH,Cl-ThCl,8H,O described by Chydenius.t It seems certain that the ammo- nium salt just mentioned represents a mixture, for it is described as a sintered mass made in the dry way. Sheffield Scientific School, May, 1901. * Bulletin, xxi, 118. + Pogg, Ann., cxix, 43. Wortman—Studies of Hocene Mammalia, ete. 193 Art, XXI1I.—Studies of Hocene Mammalia in the Marsh . Collection, Peabody Museum ; by J. L. WorrTMan. [Continued from p. 154.] Hind limb.—The greater part of the pelvic bones of both sides are present, from which an accurate idea of this portion of the osteology can be formed. In comparison with the ilium, the ischium, figure 32, is elongate, being 73 per cent of the length of the former bone; in Herpestes it is 79, in Crossarchus 69, in the Binturong 66, and in the domestic cat 65 per cent of the length of the ilum. The peduncu- lar portion of the ilium exhibits about the same degree of constriction as that seen in the Binturong, and is somewhat greater than in Herpestes. The gluteal surface is little expanded, and is occupied by two longitudinal grooves separated by a promi- nent rounded ridge; of these the superior is the wider and deeper, the inferior being | Ae ee long and narrow. This division of the o¢ gigectes herpestoides gluteal surface is a characteristic feature of Wortman; dorsal view: the pelvis of all Eocene Carnivora, and in two and one-fourth some species it persists into the Oligocene. ce erat In the living forms, it has very generally dis- ~~ appeared, only a trace of it remaining in some of the less special- ized types. The meaning of its universal presence in the early types is to be accounted for on the supposition that the primitive 31 FigURE 32.—Portion of left half of pelvic girdle of Oddectes herpestoides Wortman; side view; three halves natural size. (Type.) Marsupial-like ancestors hada type of ilium similar to that of the living Opossum, in which it consists of a simple elongated tri- hedral bar attached to a single vertebra. Slight expansion of the dorsal and ventral borders of this bar would give the con- ditions seen in the Eocene types, wherein the longitudinal ridge corresponds to and is the remains of the primitive bar. 194. Wortman—Studies of Eocene Mammalia in the A little in front of, and below the acetabulum is a relatively large rugose area, which served, to give origin to the rectus Jemoris ; it is rather large and this part of the bone somewhat Figure 33.—Left femur of Oddectes herpestoides Wortman; front view. FIGURE 34.—Teft tibia of same species; front view: FIGURE 25.—Left patella of same species; front view. All figures are three halves natural size. (Type.) Marsh Collection, Peabody Museum. 195 deep, resembling in this respect the corresponding part of the pelvis of the Binturong. The ischial tuberosities were little _ developed, as in the Ichneumon, and there is but slight mdica- tion of the ischial spine. The femur, figure 33, is relatively stont and has a straight shaft. The hemispherical head is set upon the shaft by a short, thick neck, and the fovea capitalis for the terete liga- ment is distinct. The trochanter major rises to the level of the head, and the digital fossa is deep, narrow, and slit-like, as if it had been compressed from before backwards, as in certain of the carnivorous Marsupials, notably the Dasyures and Opos- sums. The trochanter minor is broad and placed upon the inner side of the shaft just below the neck. The distal end of the bone is characteristic, being relatively broad and flat. The condyles are well separated and have comparatively little poste- rior extension; they are sub-equal in size, and the inner is not produced downwards to any perceptibly greater extent than the outer. The rotular groove is broad and shallow and has but shght upward extension upon the anterior surface of the shaft. In all its characters, with the exception of the third trochanter and the large size of the trochanter minor, the femur bears a very strong resemblance to that of the Bintu- rong, and differs from the viverrine and herpestine representa- tives in the character of the distal end, the shape of the greater trochanter, and the position of the lesser, which is more on the inner border of the shaft. The patella, figure 35, moreover, is short, flat, thin, and scale-like, having more the shape of that of the Binturong, in marked contrast to its narrow, elongate, thickened form in the other sections of the family. The tibia, figure 34, does not closely resemble that of any ot the living Viverridz on account of the preponderance of primitive characters which it possesses. These are seen in the relatively small size, the slender and rounded shaft, its great lateral curvature, and the small development of the. cnemial crest. The head has the appearance of being flattened from before backwards; the two tuberosities have the usual form, but the spine is very much less divided than in the modern species. The entire form is quite as much like that of the Marsupials as that of the living types. Near the middle of the shaft or what may be taken as the extreme lower end of the cnemial crest is seen a prominent roughened area for tendinous attachment. This area is also pronounced in the tibia of many Marsupials; and is likewise present, but less strongly marked, in the Binturong. If it served for the attachment of the tendon of one of the inner hamstring muscles, which is in all probability the case, it indicates an unusually low position for this insertion. A short distance above, and io the inner side, is seen the point of insertion of 196 Wortman—Studies of Eocene Mammalia in the the remaining tendons of the inner hamstring. The distal end is unknown. With the exception of some fragments of the middle portion of the shaft, the fibula is not preserved. These fragments indicate that it was comparatively little reduced, a character which again takes it back towards a Marsupial stage of devel- opment. The pes is represented by but a few inconsiderable ieee which do not, unfortunately, serve to give a very clear idea of its organization. A distal extremity of a caleaneum corresponds much more nearly with that of the Binturong than either LTerpestes, Crossarchus, Viwerra, or Hupleres. The articular surface for the cuboid is rather flat, the astragalar facet much arched, and the sustentaculum. widely projecting, with a slightly excavated facet. There was apparently no contact between the fibula and caleaneum. The cuboid has about the same relative height as is found in the viverrines in general; but in corre- spondence with the flat distal end of the caleaneum its proximal surface is likewise considerably flattened. The navicular again exhibits a very decided resemblance to that of the Binturong, and differs sharply from the viverrine, herpestine, and eupler- ine sections of the family. In the Binturong, the posterior margin of the bone is divided by a deep notch or sulcus into two sub-equal projections. In Herpestes, Crossarchus, and Lupleres, no such sulcus exists, and the postero-internal angle is enlarged and elevated. In the fossil this sulcus is present, but not so deep as in the Binturong, yet the arrangement is essentially the same. No other part of the pes is known. The principal measurements are as follows: Length of three posterior premolars and first two superior molars 2 2.0. es 2 eee Length of supenor sectoral (222,05. - 9s toes ip Transverse diameter of superior sectorial --_. - 3°8 Length of first and second molars (outside)... 6°7 Transverse diameter of first molar (anterior).. 5°8 Transverse diameter of first molar (posterior). 5°9 Transverse diameter of second molar (anterior) 5°5 Transverse diameter of second molar (posterior) 4°5 Length of inferior dental series from anterior borderjof canine, 4. set ok Sle 30°5 Leneth of first loweramolar _.. 2432. 32 -4e - 4°5 Length of second lower molars 2..--.--22) 22: 377 Depth of jaw at anterior border of first molar. 8:2 Length of centrum of third (?) lumbar vertebra 15: Length of centrum of last lumbar vertebra ... 15° Length of centrum of penultimate lumbar ver- ‘tebrate oro . COS RE Oma eer 2k 17% Marsh Collection, Peahody Museum. 197 Length of centrum of axis including odontoid. 16° ™™ Length of centrum of third cervical -.---._-- 8° Hemet Ol MEMMeNUS: es eee os, 3s 2: BB I0 Transverse diameter of humerus (distal end).. 16°5 Beneth ot ulna, (estimated) 277 2 OG Monee cnn O RAMUS” ee eee on he ope 58° Length of carpus and metacarpus (estimated). 24° ene th-or phalanx, second row = »2e-22) 20.2210 Henoth-of pelvis’). 2. 2T8 SETS EO) 74: Transverse diameter of illum ___..---------- Bs encthrol Temurliiien vad (Ra SIs Oi). Te Beat oO: Transverse diameter of proximal extremity_.. 16°5 Transverse diameter of caput femoris ......-- 8:3 Transverse diameter at lesser trochanter ....- 15: Transverse diameter of distal extremity ---- -- 16° Antero-posterior diameter of condyles ...---- 12. ome uhrornatelasys 48 hs ee at ae 6° Transverse diameter of patella -_.-..-....- REP Menein of tibia (estimated je oe a J a 81S Transverse diameter of head of tibia -...-_-- 14°5 Antero-posterior diameter of head of tibia .-.. 11° Discussion.—In the foregoing description of this family, frequent reference has been made to their relationship with the Viverride, the characters in which the two families resemble as well as differ from each other have been pointed out as far as they are known, and the position has been generally assumed that the one stands in direct ancestral relation to the other. I will now proceed to summarize the arguments upon which this assumption is based, but it is first necessary to observe that the idea is by no means a new one. Professor Marsh, in giving the name Viverravus to the species first described, expressed this relationship, although he did not follow it up with the necessary evidence to establish it upon a firm foundation. In my work on the “ Comparative Anatomy of the Teeth” (1886), I gave utterance to the same view, but did not attempt a demonstration of the same. Flower and Lydek- ker, in “ Mammals Living and Extinct,’ 1891, p. 539, state: “The North American genera JLiacis and Didymictis (Uintacyon and Viverravus) are generally regarded as representing a separate family—Miacidee—with affinities to both the Viverridz and the Canide.” In our paper on “The Ancestry of Certain Members of the Canide, the Viverride, and Procyonide” (Bull. Amer. Mus., 1899), Matthew and myself reiterated this opinion, and from a study of the skeleton of a species of Viverravus brought forward some evidence in its support. From a consideration of the materials at present before us, it may be regarded as clearly established that there were at 198 Wortman—Studies of Hocene Mammalia in the least two generic types living in this country during the depo- sition of the Bridger sediments, the sum total of whose char- acters, as far as they can be at present ascertained, come nearer to the living Viverridee than to any other known group of the Carnivora. These characters, moreover, not only do not interpose any difficulties in the way of deriving the one group from the other by direct descent, but they furnish just such a combination of resemblances and differences as we should be reasonably led to look for on @ priort grounds, in the ~ Eocene ancestors of the existing Viverride. The living repre- sentatives of this family include three and perhaps four widely divergent types, so different from each other that at various times they have been considered by some very good authorities © to represent distinct family modifications. One of these, the Viverrine, for which there are excellent reasons to believe that they come nearest to the original stem form, has fossil rep- resentatives in the Upper Eocene of Europe, whose remains are indistinguishable generically from those living to-day. This fact indicates a most remarkable persistence or stability of structure which we are not at liberty to suppose began abruptly in the Upper Eocene, but must have, in some degree at least, belonged to its earlier Eocene progenitors. Now it is either a striking coincidence or a highly significant fact that this very character is one of the preéminently distinguishing features of Viverravus. We have already seen that it passes with but very little change from the Torrejon to the Bridger inclusive, or, in other words, it did not change more than specifically during the deposition of between six and seven thousand feet of sediment. ‘There is another fact which must not be lost sight of just here, and that is, that in the Eastern Hemisphere these forms appear abruptly in the upper stages of the Eocené,. while in the Western Hemisphere they disappear quite as sud- denly in the age preceding. So much, then, for the general con- siderations touching the possible relationship of the two groups. Let us next inquire into the special or particular resem- blances and differences which they present. In the Viverravus- Vwerra series we note, (1) that the dental formula is identically the same in the two groups; (2) that the structure of the various teeth, including all the details of arrangement with respect to the component cusps, ridges, ete., is surpris- ingly similar; (8) that the skull of Viverravus exhibits the same more or less compressed, post-orbitally elongated form as that of the Viverrine; (4) that the atlas has the same arrange- ment of the perforations for the passage of the vertebral artery—a highly characteristic viverrine feature; (5) that the succeeding vertebre are very similar in the two groups, including the long and powerful tail, which is still retained ; Marsh Collection, Peabody Museum. 199 and (6) that there is such a general similarity in the structure of the limbs that one could easily have been derived from the other. The main points of difference are not of a fundamental character, but relate to the assumption of modernized features, such as (1) increase in size of the brain; (2) development of tympanic into a large, two-chambered, otic bulla, with consequent modification of the contiguous cranial foramina and paroccipital process; (3) decrease in relative size of the lumbar vertebree; (4) reduction in size of the deltoid crest of the humerus; (5) union of the scaphoid, lunar, and centrale of the carpus; (6) loss of third trochanter of the femur, and reduction in size of the lesser trochanter; (7) development of enemial crest of tibia, and (8) grooving of astragalus and loss of fibular contact with the caleaneum. These characters, as we know from so many other lines, pertain to all the primi- tive Carnivora, and involve just such parts of the osteological structure as have been profoundly and progressively modified by development. Instead, therefore, of offering any difficul- ties in the way of the derivation of the Viverrine from Vwerravus, they are of such a nature as one would reasonably assume to belong to the primitive type. The evidence, then, in favor of this derivation, as it at present stands, is strongly presumptive if not absolutely demonstrative. If well founded, it establishes for this phylum an antiquity greater than that of any group of living Carnivora thus far known. The position of the genus Oddectes with reference to the living forms is not so clear. The possession of three true molars in the superior and inferior series would hardly have been expected in a Bridger representative, seeing that in Viver- ravus they had already been reduced to the modern formula as early as the Torrejon. The genus, therefore, can clearly have nothing whatever to do with the ancestry of the Viverra series. That it is viverrine in its affinities, however, is so strongly suggested by almost every feature of its structure that it cannot be consistently placed in any other group. Among living forms, it seems to bear a more decided resemblance to certain of the Paradoxures than to any other section of the family, although the relatively short, thick premolars, with the disposition to the development of internal cusps, recall very forcibly certam members of the Herpestine. This character, however, is also found in the Paradoxures. In the inferior premolars it is important to note that in their lack of develop- ment of a distinctively trenchant and sectorial character, they are well fitted to give rise to the various types of teeth of this group, even including members the most aberrant and highly modified in this respect; as Avctogale and Arctictis. Indeed, the skeletel parts, as far as they are preserved, show many significant resemblances to the Binturong. These are 200 Wortman—Studies of Hocene Mammalia in the especially noticeable in the ulna, femur, patella, navicular, caleaneum, and cuboid. These resemblances, however, may be accidental and deceptive, and. until we can connect the two groups somewhat more closely in point of time, their relations cannot be regarded as finally settled. In this connection I wish to draw attention to three genera described by Professor Marsh from various fragmentary materials (mostly teeth and portions of jaws), the validity of which is either questionable or the remains are insufficient to indicate their true position. They are as follows: Triacodon fallax Marsh.* The type of this genus and species, figure 36, consists of the anterior portion of a tooth crown of the first lower molar or sectorial, which agrees in every particular with the corresponding tooth of Viverravus gracilis. I therefore do not hesitate to refer it to this species. Professor Marsh errone- ously supposed that it was an entire tooth Ficure 3¢6—a Of the premolar series, and this at first sight right lower sectorial would seem possible, since the broken sur- of Triacodon fallax face where the heel joins the trigon is worn eon aes in such a way as more or less to conceal the fracture. I give herewith a cut of the tooth seen from the inside. A second species, 7. grandis,t was also described by Pro- fessor Marsh, and this again is founded upon the trigon of a sectorial molar of some larger species of Carnivore, presuma- bly Limnocyon verus, although the specimen agrees equally well with some other species, notably Sinopa agilis. A third. species, 7. nanus, is of the same character, and is referable to’ some small carnivorous or insectivorous species. 36 Ziphacodon rugatus Marsh.{ This genus and species was established upon the anterior portion of a lower jaw, figure 37, carrying the second and third premolars with the roots of the first and fourth, together with the alveolus for the canine ; and a part of that for the first Figure 37.—Left lower jaw of Ziphac- jneisor. As far as one is able odon rugatus Marsh; outside view; three halves natural size. (Type.) to judge, this jaw Tragment agrees perfectly with Veverra- * This Journal, vol. ii, August, 1871, p. 15, separata. + This Journal, vol. iv, August, 1872, p. 32, separata. + This Journal, vol.iv, August, 1872, p. 25, separata. Marsh Collection, Peabody Museum. 201 vous gracilis. The principal points of similarity are seen in the two-rooted first premolar, which is characteristic of ail _the Bridger species of Viverravus, the form of the premolars, and the agreement in size. The name may, therefore, I think, be properly placed as a synonym of V. gracilis. Harpalodon sylvestris Marsh.* This small species of Carnivore was founded upon a frag- ment of a left lower jaw, figure 38, supporting the third and fourth premolars and a part of the crown of the first molar or sectorial, including the complete heel. The structure of the premolars agrees most nearly with those of Uintacyon edaz. The heel of the sectorial is somewhat less trenchant, having the inner portion slightly developed so as to form an incip- ient basin. The jaw is broken below so as not to display its character very well, yet it has the appearance of being much more slender, though lackmg the thick heavy structure of this latter species. 5. 39 — Port . . — Portion When more fully known, this may prove 6 left lower jaw of Har- to be a distinct form. With the present palodon sylvestris Marsh ; material, however, I am persuaded that the oer Wiens WETee ’ . es 1alves natural size. best course to pursue is to place it in the (qn) doubtful list, smee there are no characters by which it can be properly placed or distinguished. | A second species of this genus, 1. vulpinus, founded upon a lower jaw fragment, figure 39, containing the last premolar, I think clearly belongs to Viverravus gracilis, and I have no hesitancy in placing it in this category. 38 Family Paleonictide Osborn and Wortman. Paleonictide Osborn and Wortman, Bull. Amer. Mus. Nat. Hist., vol. iv, no. 1, 1892. A family of little known, primitive Carnassidents placed ancestral to the Felines, ranging in time from the Wasatch to the Bridger inclusive, and charac- terized by the early and rapid reduction of the molar dentition, the short face, and cat-like lower jaw. Of this family, three genera are known, of which Palc«onictis comes from the Wasatch of Europe Figure 39.—Fragment of left and America, Amblyctonus from lower jaw of Harpalodon vulpinus the Wasatch of New Mexico. and Marsh; outside view; three halves ; Ze natural size. (Type.) Atlurotherium from the Bridger. 39 *This Journal, vol. iv, August, 1872, p. 25, separata. 202 Wortman—Studies of Hocene Mammalia in the A well-preserved anterior portion of a skull of Palwonictis americanus was secured by the writer in the Wasatch horizon of the Big Horn Basin, in 1891, which formed the basis of the description of this species and is the best specimen of the group thus far recovered. Amblyctonus is known from a few fragmentary remains of teeth and jaws only. No other skele- tal parts have been found. Af lurotherium latidens Marsh. Limnofelis latidens Marsh, this Journal, vol. iv, August, 1872; Patriofelis leidy- anus Osborn and Wortman, Bull. Amer. Mus., vol. iv, no. 1, 1892; Patriofelis lecdyanus Wortman, Bull. Amer. Mus., 1894, p. 164; dlurotherium leidyanum Adams, this Journal, June, 1896, p. 442. Professor Marsh, in describing Zamnofelis latedens, states : ‘“‘ A second very large Carnivore, but inferior to the preceding 40 FiguRE 40.—Left lower jaw, with milk teeth and first permanent molar, of Ajlurotherium latidens Marsh; inside view; three halves natural size. (Type.) in size, is indicated by a last upper premolar and probably by some other fragmentary remains. . . .. Another specimen apparently of this species is a left lower jaw of a young indi- vidual. It contains the canine and three (tour) molars, the last of which is still nearly enclosed in the jaw.” It is now evi- dent that the fourth premolar above mentioned belongs to Patriofelis ferox, which was previously described by Pro- fessor Marsh; and that the second specimen, or that of the immature individual, figures 40, 41, is a member, not only of \ Marsh Collection, Peabody Museum. 2038 another genus, but another family. It is therefore the type of the genus and species under consideration. The specimen con- sists of an incomplete mandibular ramus containing the decid- uous canine, second and third deciduous molars, the alveolus for the first, and the crown of the first permanent molar, which was just being erupted. Imme- diately posterior to this tooth is seen a part of the bony crypt which lodged the imperfectly calcified germ of a second molar, so that the inferior molar and premolar formula can be determined. The three deciduous molars imply a like number of permanent premolars, and as Figure 41. — Crown there is evidence of no more than two View of first lower molar molars, we may fairly assume that the 0.“ (audens; three number was two. The formula would (type. peen bel (?) ©... Pm 3, M.-.. The deciduous canine was not fully erupted and the point of the crown is missing. Like the corresponding milk tooth of the lion, the root is considerably compressed. from side to side and of great antero-posterior diameter. The crown is small in comparison with the root, and the enamel extends mueh further down on the buccal than upon the lngual side of the crown. There is a deep vertical sulcus, flanked by a dis- tinct ridge in front, on the inner face, the base of the crown, which is surrounded by a distinct cingulum. The first deciduous molar, as indicated by its alveolus, is placed behind the canine without diastema, somewhat internal to the tooth line and slightly overlapping the second tooth. The tooth was small and single-rooted. The second tooth is abruptly larger; it is implanted by two strong roots and has a crown composed of a large, laterally compressed, central cusp, with strong anterior and posterior cusps. There is no basal cin- gulum. In structure, this tooth agrees very closely with that of the lion. The crown of the third deciduous molar is much damaged and does not display its structure satisfactorily. Enough remains, however, to enable me to state that its crown was thoroughly sectorial in organization. The anterior blade is preserved and is rather short antero-posteriorly. The great proportional length which remains indicates without much doubt that there was a rather large heel and probably an inter- nal cusp as well. In the milk sectorial of the lion both these elements are present; and as far as this tooth is preserved in the fossil, its resemblance to that of the cats is much more marked that it is to any other family of the Carnassidentia. 41 204 Wortman—Studes of Hocene Mammalia in the As a whole, the milk dentition of the fossil differs from that of both the Canidee and the Viverridee in having three instead of four milk molars ; from the Hyzenidee in the larger canine, the small, single-rooted first molar, and the development of three strong cusps on the second. In the hyena, the first deciduous molar is large and two-rooted, and the anterior basal cusp on the second is rudimental or wanting. As far, how- ever, as we are able to judge, in its imperfectly preserved con- dition, it agrees perfectly with that of the cats. The crown of the sectorial is of much interest, inasmuch as it exhibits a stage of development intermediate between that of Palwonictis of the Wasatch and Dinictis of the Oligocene. The three usual cusps of the trigon are present, but they are flattened and their edges drawn out to a much greater extent. than in any other Carnassident of its time. The internal cusp is relatively small, has a very posterior position, and there is no evidence that the posterior shear between it and the ante- rior edge of the first superior molar was functional, which leads to the conclusion that this latter tooth was considerably reduced. The heel is composed of a single, rather large trihe- dral cusp, with the anterior edge produced forwards toward the base of the trigon, thus very strongly foreshadowing the trenchant heel of the corresponding tooth of Déinictis, Wim- ravus, and so many other Felines of the Oligocene and Miocene. As already noted, a part of the bony crypt which served to lodge the germ of a second molar is to be seen just posterior to the sectorial. Aside from the important fact of demonstra- ting the existence of a second molar, it furnished the further information that this tooth was much reduced in size, which is of still greater moment. The jaw has greater vertical depth in front than behind, the symphysis is large, but there is no trace of a flange as in Dinictis. Altogether it may be remarked that it is just such a type as preceded that of the Sabre-tooth Tigers. | The measurements are as follows: Length from anterior base of canine to posterior part of first permanent molar ____-__-..-- SOE Length of second and third deciduous molars. 33° Length of first permanent molar.._-_------- = 1835 Depth of jaw at first deciduous molar-.-_-_----- 32° Depth of jaw behind first permanent molar... 23°5 This important type specimen was found by Mr. G. M. Keasby of the Yale party of 1871, in the Henry’s Fork Bad Lands of the Bridger Basin, Wyoming. Marsh Collection, Peabody Museum. 205 Aflurotherium bicuspis sp. nov. A second species of this genus is indicated by a first lower molar or sectorial tooth, figures 42, 48, of the left side. It is somewhat smaller than this tooth in &. latidens and differs further from it in the possession of two cusps on the heel instead of one, which gives it a more decidedly basin-shaped 42 Figure 42.—Left lower sectorial molar of Mlurotherium bicuspis Wortman ; outside view. (Type.) FIGURE 43.—Crown view of same. Both figures are three halves natural size. character. Of the two cusps comprising the heel the larger is median and the smaller is external. This species is more closely allied to Palwonictis, in which the heel of the first lower molar is tubercular. The principal measurements are as follows: mAptero-posterior diameter - 2.22.22. -.-2 2.2: leceam Transverse diameter! 5.2) 22) ee 8°5 Herht-ot crown at’ middle-cusp_ 222° =... 22-12 The specimen was found’ by L. LaMothe on Henry’s Fork, Bridger Basin, Wyoming. Discussion.—Doubtless some objection will be raised to the placing of this family in the Carnassidentia, and did we have the genus Paleonictis alone to consider, it could be possibly regarded as amore or less questionable procedure, since the development of the fourth superior premolar and the first inferior molar into highly specialized carnassial teeth, to the exclusion of all the others, had not reached quite that degree of perfection displayed by either the contemporaneous viver- rine or canine phylum; but at the same time, the tendency in this direction was so clearly evident as to leave no room for doubt that the genus should be arranged with those forms which finally developed the typical sectorial dentition of the modern Carnassidentia. There seems, moreover, to be entire unanimity of opinion that Z#lurotherium is not only a mem- ber of the Paleonictide, but is the direct descendant of 206 Wortman—Studies of Eocene Mammalia, ete. Paleonictis. If this is trae—and it would seem to be a very probable assumption from the evidence thus far obtained*—it then follows that the whole series are true Carnassidents, since the first lower molar of the Bridger species is practically as well developed and as exclusively a sectorial as it is in any of the Oligocene Felids. The genetic relationship of Zlurotheriwm to the Oligocene Felide is a matter of the greatest moment, for the reason that, if once established, it gives us the long sought key to the solu. tion of the pr oblem of feline ancestry. Although the material upon which our knowledge at present rests is scarce and frag- mentary, and the evidence correspondingly meager and incom- plete, yet such as this affords appears to be of neither an indifferent character nor uncertain significance. We have already seen (1) that. the milk dentition of the lower jaw is surprisingly feline in the make-up; (2) that the molar and premolar formula for the inferior series is the same as in cer- tain of the Oligocene types, notably Denictis; (8) that the structure of the molars and premolars coincides almost exactly with those forms, and (4) that the form of the lower jaw exhibits unmistakable evidences of relationship. The speci- mens do not afford any evidence whatever of a contrary nature, and until such is forthcoming we must look upon this source of feline origin as not only possible but extremely probable. If the conclusions herein set forth are well founded, it follows that in the Bridger epoch there were clearly established three of the main lines of modern Carnassident descent, a fact which in itself implies an origin at a much earlier date. Indeed, as the evidence accumulates we are forced to conclude that the beginnings of these phyletic lines are incomparably more ancient than we had formerly thought possible. *See Ancestry of the Felide, Bull. Amer. Mus. Nat. Hist., 1892, p. 94, in which I fully discussed this subject. [To be continued. | Liveing and Dewar—Separation of the Gases, etc. 207 Art. XXIV.—On the Separation of the Least Volatile Gases of Atmospheric Arr, and their Spectra ; by G. D. LIvEING, M.A., Se.D., F.R.S., Professor of Chemistry in the Uni- versity of Cambridge, and JAMES Dewar, M.A., LL.D., F.R.S., Fullerian Professor of Chemistry, Royal Institution, London. [Read before the Royal Society of London, June 20, 1901. ] Our last communication to the Society* related to the most volatile of the atmospheric gases; that which we now beg leave to offer relates to the least volatile of those gases. The former were obtained from their solution in liquid air by frac- a JU G a, = el Goo bene Fah [sn : iS \ 2 > r= >= tional distillation at low pressure, and separation of the con- densible part of the distillate by cooling it in liquid hydrogen. The latter were, in the first instance, obtained from the residue of liquid air, after the distillation of the first fraction, by allowing it to evaporate gradually at a temperature rising only very slowly. The diagram, fig. 1, will make the former process intelligible. A represents a vacuum-jacketed vessel, partly filled with liquid air, in which a second vessel, 6, was immersed. From the bottom of B a tube, a, passed up * Proc. Roy. Soc., vol. Ixvii, p. 467: this Journal, xi, 154. Am. Jour. Sci1.—FourtTH SERIES, Vou. XI, No. 69.—SEPTEMBER, 1901. 1 208 Liveing and Dewar—Separation of the Least Volatile through the rubber cork which closed A, and from the top of £ a second tube, b, passed through the cork and on to the rest of the apparatus. Each of these tubes had a stopcock, m and nm, and the end of tube a was open to the air. A wider tube also passed through the cork of A and led to an air-pump, whereby the pressure above the liquid air in A was reduced, and the temperature of the liquid reduced by the consequent evaporation. To keep the inner vessel, 6, covered with liquid, a fourth tube, 7, passed through the cork, and its lower end, furnished with a valve, y, which could be opened and closed by the handle g, dipped into liquid air contained in the vessel (. As the pressure above the liquid in A was less than that of the atmosphere, on opening the valve » some of the liquid air was forced through 7 into A by the pressure of the atmosphere, and in this way the level of liquid in A main- tained at the required height. Since 6 was maintained at the temperature of liquid air boiling at reduced pressure, the air it contained condensed on its sides, and when the stopcock n was closed and m opened more air passed in through the open end of a, and was in turn condensed. In this way 6 could be filled completely with liquid air, the whole of the most volatile gases being retained in solution in the liquid. The tube 6, passing from the top ‘of B, was connected with a three-way stopcock d, by which it could be put in communi- cation with the closed vessel, J, or with the tube e, and by which also / and e could be connected. The tube e passed down nearly to the bottom of the vacuum-jacketed vessel £ and out again through the cork; and so on to a gauge f, and through a sparking tube g to a mercury pump /.* The stop- cock m being still closed, the whole of the apparatus between m and the pump, including the vessel 1, was exhausted, and liquid hydrogen introduced into #. The three-way cock d was then turned so as to connect 6 with JD, and close e, and then, n opened. £ was thereby put in communication with D, which was at a still lower temperature than £, and the gas dissolved in the liquid in B, along with some of the most vola- tile part of that liquid, distilled over, and the latter condensed in a solid form in Y. When a small fraction of the liquid in B had thus distilled, the stopcock d was turned so as to close the communication between / and 6 and open that between Dande. Gas from YD passed into the vacuous tubes, but in so doing it had to pass through the portion of e which was immersed in liquid hydrogen, so that condensible matter car- ried forward by the stream of gas was frozen out. , * The Sprengel in figure is simply diagrammatic. Gases of Atmospheric Air, and their Spectra. 209 _ For separating the least volatile part of the gases, the vessel FE, with its contents, was dispensed with, and the tube 6 made to communicate directly with that connected with the gauge, sparking tube, and pump; and generally several sparking tubes were interposed between the gauge and pump, so that they could be sealed off successively. The bulk of the liquid in B consisted of nitrogen and oxygen. ‘These were allowed grad- ually to evaporate, the temperature of 46 being still kept low so as to check the evaporation of the gases less volatile than oxygen. When a great part of the nitrogen and oxygen had thus been removed, the stopcock m was closed, and the tubes partially exhausted by the pump, electric sparks passed through g, and the gases examined spectroscopically. More gas was then evaporated from 4, and the spectroscopic examination repeated from time to time. The general sequence of spectra, omitting those of nitrogen, hydrogen, and compounds of carbon, which were never entirely removed by the process of distillation alone, was as follows: The spectrum of argon was first noticed, and then as the dis- tillation proceeded the brightest rays, green and yellow, of krypton appeared, and then the intensity of the argon spec- trum waned, and it gave way to that of krypton until, as pre- dicted by Runge, when a Leyden jar was in the circuit, the capillary part of the sparking tube had a magnificent blue color, while the wide ends were bright pale yellow. Without a jar the tube was nearly white in the capillary part, and yel- low about the poles. As the distillation proceeded, the tem- perature of the vessel containing the residue of liquid air being allowed to rise slowly, the brightest of the xenon rays began to appear, namely, the green rays about A 5420, 5292, and 4922, and then the krypton rays soon died out and were superseded by the xenon rays. At this stage the capillary part of the sparking tube is, with a Jar in circuit, a brilliant green ; and is still green, though less brilliant, without the jar. The xenon formed the final fraction distilled. Subsequently an improved form of apparatus was used for the fractionation. It is represented in fig. 2. A gasholder containing the gases to be separated, that is to say, the least vclatile part of atmospheric air, was connected with the appa- ratus by the tube a, furnished with a stopcock ¢. This tube | passed on to the bulb 4, which in turn communicated through the tube 6 and stopcock d with a sparking tube, and so on through the tube ¢, with a mercurial pump. Stopcock d being closed and ¢ opened, gas from the holder was allowed to pass into 6, maintained at low temperature, and there condensed in the solid form. Stopcock ¢ was then closed and d opened, and gas from £ allowed to pass into the exhausted tubes between 210 Liveing and Dewar—Separation of the Least Vclatile £& and the pump. The tube e was partly immersed in liquid - air in order to condense vapor of mercury, which would other- wise pass from the pump into the sparking tube. The gas passing into the sparking tube would, of course, have a pres- sure corresponding to the temperature of 46, and this was further ensured by making the connecting tube pass through the liquid in which & was immersed. The success of the operation of separating all the gases which occur in air and which boil at different temperatures depends on keeping the temperature of 6 as low as possible, as will be seen from the following consideration :— The pressure p, of a gas G, above the same material in the liquid state, at temperature 7, is given approximately by the formula B p22 where A and B are constants for the same material. For some other gas G’ the formula will be log p= A— B log p, = A,— Pp? | | el eee Ba Bs and DS euon ae A,+ 7a Now for argon, krypton, and xenon respectively the values of © A are 6°782, 6°972, and 6-963, and those of B are 339, 496-3, Gases of Atmospheric Avr, and their Spectra. 211 and 669°2; so that for these substances and many others = is considerable 1 _A—A, is always a small quantity, while and increases as 7’ diminishes. Hence the ratio of p to p, increases rapidly as 7’ diminishes, and by evaporating all the gases from the solid state and keeping the solid at as low a temperature as possible, the gas first coming off consists in by far the greatest part of that which has the lowest boiling point, which in this case is nitrogen, and is succeeded, with comparative abrupt- ness, by the gas which has the next higher boiling point. The change from one gas to another is easily detected by examining the spectrum in the sparking tube, and the reservoirs into which the gases are pumped can be changed when the spectrum changes, and the fractions separately stored. Or, if several sparking tubes are interposed in such a way as to form parallel communications between the tubes } and e, any one of them can be sealed off at any desired stage of the fractionation. The variation of the spectra of both xenon and krypton with variation in the character of the electric discharge is very striking, and has already been the subject of remark, in the case of krypton, by Runge, who has compared krypton with argon in its sensitiveness to changes in the electric discharge. Runge distinguishes krypton rays which are visible without a jar and those which are only visible with a jar discharge. The difference in the intensity of certain rays, according as the dis- charge is continuous or oscillatory, is no doubt very marked, but, with rare exceptions, we have found that the rays which are intensified by the oscillatory discharge can be seen with a continuous discharge when the slit of the spectroscope is wide. Runge used a grating, whereas we have, for the sake of more light, used a prism spectroscope throughout, and were there- fore able to observe many more rays. There is one very remarkable change in the xenon spectrum produced by the introduction of a jar into the circuit. With- out the jar xenon gives two bright green rays at about 4917 and 24924, but on putting a jar into the circuit they are replaced by a single still stronger ray at about X 4922. In no other case have we noticed a change so striking as this on merely changing the character of the discharge. Changes of the spectrum by the introduction of a jar into the circuit are, however, the rule rather than the exception, and there are changes in the spectrum of krypton which seem to depend on other circumstances. In the course of our examination of many tubes filled with krypton in the manner above indicated, we have found some of them to give with no jar the green ray A 5571, the yellow ray X 5871, and the red ray A 7600 very 212 Liveing and Dewar—Separation of the Least Volatile bright, while other rays are very few, and those few barely visible. Putting a jar into the circuit makes very little differ- ence; the three rays above mentioned remain much the brightest, nearly, though not quite, so bright as before, and the — blue rays, so conspicuous in other tubes, though strengthened by the use of the jar, are still very weak. In other tubes the extreme red ray is invisible; the rays at \ 5571 and 5871 abso- lutely, as well as relatively, much feebler, while the strong blue rays are bright, even brighter than the green and yellow rays above named. In one tube the blue rays could be seen, though not the others. This looks very much as if two different gases were involved, but we have not been able to assure ourselves of that. The caseseems nearly parallel with that of hydrogen. There are some hydrogen tubes which show the second spec- trum of hydrogen very bright, and others which show only the first spectrum ; the second spectrum is enfeebled or extin- guished by introducing a jar into the circuit, while the first spectrum is strengthened ; and the conditions which determine the appearance of the ultra-violet series of hydrogen rays have not yet been satisfactorily made out. It is to be noted that putting the jar out of circuit does not in general immediately reduce the brightness of the rays which are strengthened by the jar discharge. Their intensity fades gradually, and is generally revived, more or less, by reversing the direction of the current, but this revival gets less marked at each reversal until the intensity reaches its minimum. The rays strengthened by the jar discharge also sometimes appear bright, without a jar, on first passing the spark when the elec- trodes are cold, and fade when the electrodes get hot, reappear- ing when the tube has cooled again. Moreover, if the dis- charge be continued without a jar, the resistance in the krypton tubes increases rather rapidly, the tube becomes much less luminous and finally refuses to pass the spark. With an oscillatory discharge the passage of the spark and the bright- ness of the rays are much more persistent. This seems to point to some action at the electrodes, which is more marked in the case of krypton than in that of xenon. The wave-lengths of the xenon and krypton rays in the tables below were determined, in the visible part of the spectrum, with a spectroscope having three white flint-glass prisms of 60° — each, by reference to the spark spectrum of iron, except in the cases of the extreme red ray of krypton, which was referred to the flame spectrum of potassium, and its fainter neighbor, which we saw but did not measure. The indigo, violet, and ultra-violet rays were measured in photographs, taken with quartz lenses and two calcite prisms of 60° each. The spec- trum of the iron spark was photographed at the same time as Gases of Atmospheric Air, and their Spectra. 213 that of the tube, the former being admitted through one-half of the slit, and the latter through the other half. The xenon spectrum is characterized by a group of four conspicuous orange rays of about equal intensities, a group of very bright green rays of which two are especially conspica- ous, and several very bright blue rays. The only list of xenon rays we have seen is that published by Erdmann, with which our list does not present any close agreement except as to the strongest green lines. The number of xenon rays we have observed is very considerable, and some of them lhe very near to rays of the second spectrum of hydrogen, but inasmuch as these rays are more conspicuous with a jar in circuit than with- out, which is not the character of the second spectrum of hydrogen, and, moreover, many of the brightest of the hydro- gen rays are absent from the spectrum of the tubes, we con- clude that these rays are not due to hydrogen. Certain rays, which we have tabulated separately, have been as yet observed in only one tube: they include a very strong ultra-violet ray of unknown origin, and due either to some substance other than xenon, or to some condition of the tube which has not been repeated in the other tubes. Our krypton rays agree much more closely with Runge’s list, but outnumber his very considerably, as might be expected when prisms were used instead of agrating. Prisms, of course, cannot compete with gratings in the accuracy of wave-length determinations. We think that the krypton used by Runge must have contained some xenon, and that the rays for which he gives the wave lengths 5419-38, 5292-37, and 4844°58 were really due to xenon, as they are three of the strongest rays emitted by our xenon tubes, and are weak in, and in some cases absent from, the spectra of our krypton tubes. Tables of the approximate Wave-lengths of Xenon and Krypton Rays. Rays observed only with a Leyden jar in circuit have a * prefixed, those observed only when no Leyden jar was in cir- cuit have a t+ prefixed. The intensities indicated are approximately those of the rays when a jar is in circuit, except in the case of the two rays to which a + is prefixed, which are not seen when a jar is in cir- cuit. Rays which are equally intense whether a jar is in cireuit or not have a | prefixed to the number indicating their intensities ; those which are less intense with a jar than with- out have a < prefixed to the number expressing their intensi- ties. The rest are, in general, decidedly more intense with a jar than without. 914 Liveing and Dewar—Separation of the Least Volatile Xenon Rays. ae Intensity. ee Intensity. ee Intensity. Rees, Intensity. *6596 4 5025 5 00 00 90 00 2 F Heh O°5 Sle lS 90: 00 of OZ We 1] G hoa" LO Aven os oe 47 304 es 1 N 201 TAAL 22, es 144 28 &¢ 2 R 11°0°1 173 10 s UID. Br es 1 i Although the discrepancy between the measured and calcu- lated values is considerable in some cases, the agreement is sufficiently close to indicate that the forms are identical with those of the first occurrence. A 18 19 similar irregularity of contour is p-£ also noticeable, as shown by the De figure. AX Occurrence No. 3. Crystals Jrom the Moon Anchor Mine.— Our study is confined to two small crystals detached from a gangue specimen sent us by Mr. Rickard. The crystals are lath- shaped, measuring 1°5™™ in length and 2 by 0°3™™ in cross ‘section. The development of the crystals is represented by 4 “4 figures 18 and 19. The gangue upon which the crystals occur seems to be a decomposed ande- site, the calaverite having been deposited along with quartz and fluorite in a crevice. The forms together with their measured and calculated angles are recorded in the following table: | &S 234 Penfield and HFord—Calaverite. Measured. Calculated. lam) —_—-“~——- | ae —“_—. “= Times — Symbols. Vertical. Horizontal.|} Vertical. Horizontal.| observed. b 010 00° 00’ 00° 00’ 00° 00 00° 00 2 m 110 00 OO 31 28 00 00 31 3804, 4 li 40%? 9 30 14 482 ie 33°80°'4 He ones ee | q MO 32 25 15 33 32 o2 15 303) 3 p TWA Gy Maes ts} 46 57 54 48 46 544) 1 fs 10°15°22 HAL NG 52 49 TMU BS O22 -38 il WwW palais Way ANG 46 47 124 55 46 484) 3 U Tecra Oee ea yy 480) 3l 6 66 WOR es Be 31 28 2 7 _ 7156 128 50 53 30 129 Vs tod, oul 1 S 11°62°6 WA al oe i) oa lod L144 ee ee ys ee B Old 9 30 90 00 i B 801 So) I 90 00 i a a D 304 62 8 Ge 62754: Be D, c 001 89 51 ee 89 474 ee Z ft, POEL ie 2 oe IS) ak of 1 f, 15°0°1 liom gs Ws Bs he 1 On these two crystals, as indicated by the figures, attention may be called to the occurrence of the clino-pinacoid } (010), or polar face, and to the prominent developments of m, w and q. The forms g, and 7 have been observed by us on only the one crystal represented by figure 18. There is some uncer- tainty concerning o (2° 10: 11), figure 19. It is a small face, apparently in the zone m, g, but it is very near to s (11°62 6) of occurrence No. 1, which is in the zone D,p,oandg. The dome B is probably (801), which is in the same zone as m and g. In cross-section these crystals have a far more regular con- tour than those of the occurrences previously cited. Occurrence No. 4.—The occurrence thus designated is that of a suite of specimens sent by Mr. Bixby with no special designation of the mine from which they came. The speci- mens resemble very closely those from the Moon Anchor mine. The crystals are lath-shaped, some of them very thin, attached to what appears to be an andesitic gangue. They have been deposited in crevices, and are associated with quartz and fluor- ite. Many of the crystals are much bent and cracked. Although there was an abundance of calaverite on the speci- mens only one crystal was found which seemed to offer any possibility of measurement. The end of the erystal when detached from the gauge measured about 1™™ in length and only 0°6 by 0-15" in cross-section. Although so very minute, the reflections from its faces were excellent: in facet this Penfield and Ford—Calaverite. 235 erystal may be regarded as the best we have examined. Without a two-circle goniometer it would have been almost impossible to have measured more than a few of the angles of such a small crystal. With a two-circle goniometer, on the other hand, the measurements were made without difficulty. Altogether seventeen terminal faces were observed on this crystal, the area of whose cross-section is less than that of a hyphen on this printed page. The crystal, which is the most symmetrical in its development of any we have seen, is represented with ideal symmetry in figure 20. The clino- pinacoid } (010) is present, and monoclinic sym- metry is proved by the occurrence of the forms Mm, 0, P, f, and s as pairs of faces, making in each case almost identical angles on either side of (010), as indicated by the horizontal circle measurements. A complete list of the angles measured on this crystal is given in the following table: Measured. Calculated. ease Cie aay ee at ae 55) Symbols. Vertical. Horizontal. Vertical. Horizontal. b 010 G0= 00 007 06. 00% 00;= 005 00! m ee @) ‘S 31 32 oe 31 303 m 110 os 31 32 5 31 304 0 a 22 0 AT AO 2812 47° 304 28 18 0 pls 22-10 oh 28 22 ee 28 18 p it SA Oa AT 8 54 48 46 544 p il et 46 41 “ 46 544 af 1 Es RO ole oGdl Sr (O29. 361 40 5 i 112 “ 61 36 es 61 40 g 4‘7°10 Waat53) o2°AG HAN T, Spe 22 Vv ZA EO S204T 938-25 81 48 38 44 k 1632-21 LOS? 2a, Blk 8 10815 18n 131 6 UU 11:18:10 12% Qi 33052 127 32, 23h 428 Uu 11:18°10 é 31 00 3 re 31 28 a9 i20°6 143 52 °22 36 142. £AS A283) wD Ss 11°62°6 ae Oe S i 50 s 11°62°6 ee & 00 se 7 50 B, 17°0°2 9° 26 90, 00 2-30. 90: (00 B, 17°0°2 9 45 6¢ 66 66 D a ae 63 38 os 625-4 - H 5°0'11 Gee lel 4 71 36 ‘ I 2°O°11 82 25 82 25 a Ee 11:0°2 165 58 e 165 28 a i PPG? 165 26 o S s 236 Penfield and Ford—Calaverite. Most of the terminal faces, b, m, u, p, v, u, @ and s, are identical with those observed in previous occurrences. Of the new forms, 7, g and #, the two former fall in a zone with D (304) and v (2°11°10), to which zone also ¢ (13°20°4) belongs ; compare the stereographic projection, page 242. The forms s and £B, are somewhat doubtful. They may be o (2-10-1) and #& (801); compare page 234. 11:62°6, 110 and 17-02 are in a zone, as are also 210-1, 110 and 801. Occurrence No. 5.—This material includes the crystals from the Prince Albert mine sent to us by Dr. Hillebrand, and pre- viously examined erystallographically by one of us, as already referred to. The crystals average from 4 to 1™™ in diameter and 2™" in length. Most of them are twinned according to laws which will be described later, but three which show no indication of twinning are represented by figures 21 to 23. 21 22 23 Figure 21 represents a left-hand termination, the others right- hand ends. One, figure 23, is a doubly terminated crystal, showing similar forms at both extremities of the axis of sym- metry. The prominent forms are those observed on other occurrences. The new forms 7 and p are each represented by one small face. The symbols of the forms and the measured and calculated angles are as follows : Measured. Calculated. pics at Gs RE ices —~— — | Times Symbols. Vertical. Horizontal. | Vertical. Horizontal. | Observed. m 110 00° 00’ 00° 00’ 31° 304’ 3 Ge| e295 SPR Eo) 15) SiS) 82 Oa elion y hOn, 1 0 | 13°22°10 47 af es 1 AT 304 28 18 3 p ital 54 50 46 50 54 48 46 542 6 cig | Sa pA SIA a 59) 56) 2 b6R 10 GOP 4s «56-53 a y | 296 76 35 30 35 76 38 30 54 1 p | 44°21°2 UG) 432 eee Oe ShS e527 9 1 Penfield and Ford—Calaverite. 237 Twinning—Two laws of twinning have been determined which may be stated as follows: 1. The twinning plane is in the ortho-dome zone, at 90° to 101. 2. The twinning plane is the ortho-dome 101. According to both laws the symmetry axis and the positions of the p faces are the same for both the normal and twin positions. Penetration twins have also been observed with the symmetry axes, or the striated zones, crossed 24 25 26 at angles of about 90°, but no erystals have been found from which sufficiently accurate measurements could be obtained for determining the exact law of twinning. The first law of twinning has been observed only on crystals from the Prince Albert mine (Occurrence No. 5). The twin- ning is partly of the nature of two individuals meeting along the twinning plane and partly of the nature of twin lamellae, revealed by a series of striations crossing the terminal faces at right angles to the intersection of two p faces. Figures 24 to 26 represent three individuals which illustrate this law of twin- ning, and although the figures do not show the twinning at all distinctly, it is doubtful whether any other kind of illustration 238 Penfield and Ford—Calaverite. would answer the purpose better. The stereographic projection, figure 27, is better adapted for the purpose. The diameter of the circle, passing through 101, 101 and the poles lettered 7, represents the twinning plane, and symmetrically on either side of it are the poles of the two m and two o faces, making small angles with one another. The poles of the four m faces, two in normal and two in twin position, serve to make the orientation clear. In figure 24 the faces in the upper right-hand portion are in normal position, except for the occasional crossing of twin lamellee, as indicated by the striations. The faces m and o below and to the left are in twin position, and make reéntrant angles with the corresponding planes of the normal erystal. The two p faces are common to both individuals. The angles man and oo, measured over the twinning plane, are as follows: Measured. Calculated. NAN Bet OY 3° 14! OA O Tih vo 6 54 The form @ was observed only on this erystal and appears as a very narrow, somewhat tapering face, intersecting p and disappearing against a twin lamella. The measured and eal- culated angles of this crystal are as follows: Measured Calculated ~ ee cee, = Se = Symbols. Vertical. Horizontal. Vertical. Horizontal. Dp 111 BAST AS! iG 50) 54° 26) Gm 543’ ” ili oe le mo ce 4G ae m0) 11:14:10 52 it ae 34 a2 12: 38 Iles n of twin Dil ON ae Orme ne 57 94 Beene O 18°22°10 47 26 28 6 47 304 28 18 O of twin 62 QO 28 2 62 5° 28 18 w AQ)? SL 40 16 47 12 4.0 QiGuae aan 8 Digel LOS il Di) Oe 4 57 18 7o8 29 On the crystal represented by figure 25 only the small o and m faces to the left are in twin position, the o faces making a: reéntrant angle at the twinning plane. A few of the angles of this crystal are as follows: Measured Calculated. (iam SSS = (3 ——s =a Symbols. Vertical. Horizontal. Vertical. Horizontal. p iol BAT 5015 AG ee ie 54°. 48’ 46° 542) p it ; oo AGG ee Ce AG 13°22°10 AT Gras D 47 304 28 18 0 of twin 62-4 T28 Ras 5 62 5. (289 sie m 110 00' 7 00731: piged 00 - 00 , 815 %38@; m of twin 109 - 45 31 35 L097) Sia earl 304 Penfield and Ford— Calaverite. 239 The erystal represented by figure 26 appears for the most _ part like a normal individual, crossed, however, by twin lamel- lae as indicated by the striations. The form lettered s seems to be in twin rather than in normal position. The second law of twinning, where 101 is the twinning plane, has been observed on two crystals, one from the Monu- ment the other from the Prince Albert Mine. These are illus- trated by figures 28 and 29, respectively. In the crystal rep- resented by figure 28 the faces 0, m and part of s above are in twin position. The twinning plane separates the two o faces, runs through s, and disappears in that portion, stippled in the figure, where the crystallization has been interrupted. The m faces make a reéntrant angle. The s faces in normal and twin position fall almost together, making, according to calculation, an angle of only 0° 44’ with one another. In the erystal under consideration s was not a very good face. The meas- ured angles are as follows: | Measured. Calculated. Wes ents ae ere ele aay Cie ae Symbols. _ Vertical. Horizontal.| Vertical. Horizontal. m 110 pe 0050005 30252) 00; 00! ~ 3h 303) m of twin Mondo oO oy 70 26 81 304 0 13-22-10 ees. 28 015 47 304 28 18 O of twin Pavel 23utos Lip vobey 2aer 1s :" 10°44°15 Pies slo te | lta 18. Foe Spropaply of twin | 1466532 S88) *147 420-5" 750 In the erystal from the Prince Albert Mine, figure 29, both laws of twinning are combined. The first law is indicated by twin lamellz crossing the lower portion of the crystal. The faces in the lower left-hand portion of the figure may be regarded as belonging to a single individual in normal position. To the right the two m faces meet along the twinning plane, which seems to run for a ways through s and then disappears. The salient angle made by the two m faces, measured over the Am. Jour. Sct.—FourtH Series, Vou. XII, No. 69—SEPTEMBER, 1901. 17 240 Penfield and ies laverite. twinning plane, is 35° 2’, caleulated 35° 4’. The measured angles are as follows: ' Measured. Calculated. — sae {i - ee Symbols. Vertical. Horizontal. Vertical. Horizontal. m 110 00° 00’ SSS) 00 00 Bh be 303’ m | of twin | 109 41 31 40 109 38 31 30% t 13°20°4 94 14. 93 19 93. 384 23 32k t of twin 86 1 23 31 Se et 23 32% O loro 2 Ap ial Titer. ad bl 47 30 28 18 p EVI 54 49 46 49 54 48 46 544 qd da 22955 32 9 Lag 2k oo. | oe 15 304 S of twin ase ye REED 147 20 1-50 As far as the angles of the crystals are concerned, the sec- ond law (twinning plane 101) might be considered as alone sufficient for explaining the two kinds of twinning. Instead of the first law, as stated above, it might then be said that the crystals were twinned about 101, but were united by a compo- sition face at right angles to the twinning plane. Since we have in figure 29 a combination of the two kinds of twinning it has seemed to us simplest to explain it as according to two laws. Summary.—There have thus far been presented in the tables of, angles sufficient data to indicate that there are certain planes which ocenr on calaverite crystals repeatedly, and with such constancy in their angles that they can not be regarded as in any way accidental. There are no forms more frequent in their occurrence and more prominent in their development than those designated as m (110) and p (111), unless it is, per- haps, 0 (13°22°10). If it be accepted that m (110) and p (111) have been well chosen as fundamental forms, then the erystal- lographic relations between calaverite and sylvanite may be expressed as follows:—Both are closely related not only in chemical composition, but also in their axial ratios and in their axial inclinations 8.—Both are alike in erystallizing in the monoclinic system, and in having the dome (101) as twinning plane.—Calaverite differs from sylvanite in having no distinct clino-pinacoid cleavage, and in having, with few exceptions, different forms, most of which must be designated by unusually complex indices. The symbols assigned by us to the calaverite forms are those which seem to correspond most nearly with the results of our measurements. No one can appreciate more fully than the writers that many of the symbols are so complicated that it is almost impossible to believe that they are true. On the other hand, we fail to find any way of simplifying them which does Penfield and Ford— Calaverite. 241 not cause discrepancies in the angles greater than the character of the reflections would seem to warrant. It is probable that some of the symbols will need revision and change. The crystals which we have been able to examine are such as might be designated as good, though they are not of the very best quality. It should be explained also that all of the crystals in our possession have been studied and all of the results given, not merely a selection of best values. If one could have a larger variety of occurrences to select from, it is likely that some exceptionally good erystals might be found from which more exact and reliable measurements could be obtained. A careful study of such crystals would be of great value, for the correctness of the complex symbols indicated by the measure- ments made by us needs verification. One especially note- worthy feature of the symbols as given by us is that eleven, or some multiple of it, appears in many of them; though the significance of this, if there is any, is not apparent. It is possible that by adopting another orientation some sim- plification of the symbols may result. The chances for bring- ing about much simplification, however, do not seem to be very promising, and any change in orientation must necessa- rily do away with the apparent similarity in crystallization between calaverite and sylvanite, as indicated by us on page 240. ; It was noted in a recent number of Nature* that Mr. G. F. Herbert Smith had discussed crystals of calaverite before the Mineralogical Society of London, and described them as “ tri- clinic, but pseudo-monoclinic owing to twinning about an axis parallel to the prismatic zone.” Upon observing this, we immediately communicated with Mr. Smith, sending him the results of our investigation, and he in reply has kindly sent us a brief statement of his work. He expects soon to publish his results, and we do not feel at liberty to discuss them at this time. It may be stated, however, that he has in the three- circle goniometer recently described by him* an instrument, regarded as especially adapted to the study of such complex crystals as those of calaverite, by means of which he is able to discover zonal relations which ordinarily would escape detec- tion. From the study of these he is led to assign simple indices to the majority of the forms, which necessitates, how- ever, the assumption of triclinic symmetry and a peculiar twinning. As far as the angles and the distribution of the faces are concerned, his crystals, like ours, evidently satisfy the conditions of monoclinic symmetry. We quote the following * Vol. Ixlii, p. 555, 1901. + Min. Mag., vol. xii, p. 175, 1899. 242 Penfield and HFord—Calaverite. from his letter: —“ Unless, however, the crystals are regarded as triclinic twins it is impossible to obtain simple indices. This, I allow, is the only argument in favor of this view; there are no reéntrant.angles. I do not know whether I shall persuade any one to share this view. If the crystals are twinned there must be extraordinarily intimate penetration.” The problem may, perhaps, resolve itself as follows :—Kither in the assumption of monoclinic symmetry (which the crystals apparently possess) with complex symbols, or triclinic sym- metry (wholly obscured by twinning and resulting in pseudo- monoclinic symmetry) with simple indices. | When the present article was nearly completed we learned from Dr. Palache of Cambridge that he and Professor Moses of New York were also engaged upon the study of calaverite crystals. They both have found the same forms as observed by us, with few exceptions, and also have crystals which are far more complex. It may be expected that the results of their investigation will appear in due time. Penfield and Ford—Calaverite. 243 It is not claimed that the results as set forth in the present communication furnish an explanation of the crystallization of ealaverite. They are presented rather as an accumulation of data which has enabled us to figure and, to a certain extent at least, to describe the crystals, as also to point out close similar- ities in axial ratios and twinning between calaverite and syl- vanite, which, if wholly accidental, are certainly remarkable coincidences. Although our results may not be conclusive nor wholly satisfactory, they are presented in such shape that others may make use of them in formulating any theories they may entertain concerning the crystallization of this remarkable mineral. For convenient reference all of the forms which have been recorded in the foregoing pages are shown in stereographic projection, figure 30, and are given in the following tables, together with their calculated two-circle angles: Terminal Forms. Symbols. Vertical. Horizontal. Symbols. Vertical. Horizontal. Cr O10 COPOOT OUR OOC Gr MONS 22" 71 36 2 52" 38) Tee FTO 00 00 31 304 |g; 4:7:110 (A BO De ee 1930, 14°98) 4, 296 FO BS BO Bt My 1 38°80°4 OPA it Od Oo -MNOr, Fs Bl. 48), 688 484 ; ue A Mere 9-35-83 5. Vk. -10'32:91 108 184, 31-6 pee 20726 2 82 1 ir, (10:44:15. 114 54. 18 10d Gee lis: 2 074 DE BSE DE SOA Won, lil 124755) 460485 Grek ts29"5 oO le SON Peel Oe elo Tso) sol 228 Pep etsT ti 36 40° 23°30} 7, 756 129 a sron eels LI 40 26°47 38 +a, 1206: 7 14224-3235 Pages 2210 474305) 2818s * 11-62°6) 1424 7 50 eiate OL S212) 38 17ers Bl? 144.98 747 foes aba 54.48 46 54¢/6, 2:10°1? 144 28 8 33 Sy ld Ser i BU LE TS OO Wiese SIs = 18) ates eae OG) 12 mueeeale 27 160 “> - 56 B81 \2 15-2971 - 174 34... 99°46 mii LL? DODO Te Ole Orson A422 ee 17 6) Beeb Ae 59 Forms in the Zone 100-001. Symbols. Vertical. | Symbols, Vertical. Symbols. Vertical. oe 00 00 | Go 13-0710 47 S0e4 N11 06 149° 4’ Pee Ol & 692") PD, 304 C2 ae Nee Onl nit 4's Pee 2 9308 F501 1?) 71. 36 INP Ol 146) 47, ees 10” See) I 20-11) 82 425 Pict) 2a oor 28 Peas Oe 1148) ce, 0 01 889° 475) P5.25-0-4. 167 | 9 eee eo, o35 1 0490) Aan (Ry 901 1717-0 So 35) | M9010 22 eRe DLO 17234 By dO 197-91 M101, eas Fe 57051 Li4 8a Ee tl Os 32. 2 el M. 11-0710 1277.32 244 Penfield and Ford—Calaverite. Considerable time has been devoted to the study of the zonal relations of calaverite, with the results shown in the stereo- graphic projection, figure 30, which may be summarized as follows: 1. D (304), ¢(10-7-11), p (111), (11:14:10), 0 (18:22"19), g (11°29°5) and s (11°62°6). scab of the zone (4 1 3). This is the principal zone, and almost the only one which is apparent on simple inspection of the crystals. A possible simpler symbol for s is (2°11°1), also in this zone, but most of the measurements indi- cate the more complex symbol. 2. D (304), f(112), g(4°7:10), v (211710) and ¢(13-20°4). Symbol of the zone (4 2 3). 3. ¢ (001), eee p (111), 7 (110) and w(111). Symbol of the zone (1 1 0). 4. m (110), i (10-32'21), w (111), j (756), and N (201). Sym- bol of the zone (112). 5. M (101), g (4°7°10), p (111) and p (4421-2). Symbol of the zone (121). 6. m(110), 2 (11:14:10) and g (4°7:10). Symbol of the zone (10°16:3). 7. m(110), 0 (13°22°10), v (2°11°10) and M, (90°10). Symbol of the zone (10°10°9). 8. B(801), g (11:29°5), o (2°10°1)? and m (110). Symbol of the zone (118). 9. B, (17:0°2), s(11'62°6) and m(110). Symbol of the zone (2-217. ay B, (901), p(111) and v (2:11:10). Symbol of the zone ye (010), s (11°62°6), « (11-20°6) and N, (11-0°6)) SWiicmas the only case where two pyramidal forms have been observed in a zone between 6 and a form in the striated zone. Furthermore it is exceptional to find forms in the striated zone having the same vertical circle angles as the pyramids. In addition to the foregoing there are a number of very close approximations to zones, some of which are indicated by dashed great circles in figure 80. These are drawn from (901) and (1:0°4), and although probable zonal relations seem to be indicated by the projection, the symbols assigned to the forms do not quite satisfy the zonal equations, and possible changes, indicating greater simplicity of the symbols, do not suggest themselves. Chemical relations of Sylvanite and Calaverite——Both minerals conform to the poictal formula [Au,Ag]Te,, silver being regarded as isomorphous with gold. In sylvanite the proportion of gold to silver is near 1:1, so that the formula may be written AuAgTe, The percentage of silver Penfield and Ford—Calaverite. 245 demanded by the formula is 13'4, whereas the published analy- ses show from 11°50 to 13°86 per cent. An analysis of sylvanite from Cripple Creek by Palache* yielded 12°49 per cent of silver and aratio of Au: Ag=1:0°87. As most analy- ses of sylvanite indicate less than 12 per cent of silver, their agreement with the formula AuAgTe, is only approximate. Calaverite, on the other hand, is a nearly pure gold telluride, conforming closely to the formula AuTe,, though always con- taining some silver, and at times, as indicated by the analyses, as high as 3°5 per cent. Although chemically there is not a very great difference between calaverite containing 3°5 per cent of silver and sylvanite containing 11°5 per cent, still there is evidently a tendency for sylvanite to conform closely to the formula AuAgTe, and calaverite to the formula AuTe,. Thus from a chemical standpoint alone calaverite may well be regarded as a distinct mineral species. The relation between the two minerals is analogous to that existing between calcite and dolomite, and as the latter are closely related crystallo- graphically, so calaverite and sylvanite are closely related as indicated by their similarity in axial ratios and twinning. In order to distinguish between sylvanite and calaverite, the following method of testing, which has been applied to all of the occurrences examined by us, may be recommended: The powdered mineral when boiled in a test-tube with concentrated nitric acid is quickly oxidized, the silver and tellurium going into solution, gold being left. The clear solution, decanted into another test tube, diluted, and tested with hydrochloric acid, gives a precipitate of silver chloride, which is consider- able if the mineral is sylvanite, but is slight, or amounts per- haps only to a turbidity, if the mineral is calaverite. After washing the gold by decantation it may be dissolved in a few drops of aqua-regia, indicating a complete separation of gold and silver. Specific gravity determinations and quantitative estimations of gold and silver have been made on specimens from the Monument Mine and from the unknown Cripple Creek Mine designated as Occurrence No. 2. Of material from the first source a single fragment weighing 0°7729 grams was employed, while from the second source a number of fragments weighing 1:2552 grams were available. The specific gravity determina- tions were taken with much care on a chemical balance. The results by Penfield are as follows: * This Journal (4), x, p. 422, 1900. 246 Penfield and Lord— Calaver ite. Monument Mine. Occurrence No. 2. Theory. Sp. Give: 9°328 Ratio 9-388 Ratio AuTe, Ae Bee 40°99 2.0.9) ie ect ieee 2 rei paler, 44°03 aaa 174-016 ramet ents 30 + 22 ae Meher ee aie Pas SE sh ae [26:75] Vea ae “Deed Gangue- 0°02 0°08 -_—— 100°00 100-00 100:°00 The ratio of Au-+-Ag: Te in both analyses is 0°98 : 2:00, or very nearly 1:2. The second occurrence approaches very near to a simple gold telluride, the per cent of silver, 0°40, being the smallest thus far recorded in any published analysis of calaverite. The specific gravity determinations are a few tenths higher than generally given; from the nature of the material, however, and the care exercised in taking them they must be very exact. Of the other occurrences examined crys- tallographically by us, No. 5, from the Prince Albert Mine, contains 3°23 per cent of silver, as determined by Hillebrand, and the others, occurrences 3 and 4, are probably abort like the material from the Monument Mine, judging from qualita- tive tests. In all of his publications on ealaverite Genth describes the mineral. as having a bronze-yellow color. This seems to us misleading, for, although the crystals have at times a yellowish cast, the brightest and best of them are silver-white. Some of the dull crystals examined by us have a gray color, very like that of tarnished silver. ; It is interesting to note the production of gold from the Cripple Creek region, as communicated to us by Mr. T. A. Rickard of Denver. During 1900 the production amounted to 877,972 ounces, valued at $18,147,681, and it is probable that the most of this vast amount was derived from calaverite. It was practically all derived from telluride ores, as very little free gold is found in the district, and that only in the upper- most portions of the veins, near the surface. Sheffield Laboratory of Mineralogy and Petrography, Yale University, New Haven, July, 1901. Miscellaneous Intelligence. i 247 SC LEN TIN LOG ONL LLGEN CE. 1. On Temporary Set; by C. Barus.—Following the sugges- tions of my last paper* I have since been able to map out the occurrence of what I shall call temporary set, distinguished from permanent set inasmuch as the former may be shaken out of a metal by molecular agitation (without heat, as for instance by magnetization of iron). ‘The latter cannot, of course, be shaken out of the metal. Temporary set (d@) is instantaneously imparted ; it begins with the zero of twist (@), increasing at an initial rate d6/0=-:004, the rate gradually diminishing to zero when the (elastic) obliquity of the external fibre of the twisted wire exceeds ‘002 radians, after this temporary set becomes unstable, passing into permanent set. The maximum amount of temporary set may be estimated as ‘0015 of the maximum twist: within the elastic limits. Its variation depends not on the impressed strain but on the antecedent strain, i. e., the strain between two successive molecular agitations and which may lhe on both sides of zero. 2, Some new rock-types—KENYTE. This name has been applied by J. W. Gregory to “ liparitic representatives of an oli- vine-bearing nepheline syenite, consisting of anorthoclase pheno- erysts with or without some augite and olivine phenocrysts and a glassy or hyalopilitic groundmass which varies in color from grayish green to a deep sepia-brown. Aegyrine, if present, occurs in small granules; zenigmatite and quartz are absent.” The rocks occur as surface lava flows on Mount Kenya in Kast Africa. They are supposed to be closely related to pantellarites but no analyses are given of them, and as the giassy base is stated to gelatinize with acid and to contain abundant soda, it is difficult to see why they are not glassy phonolites especially as they are held to be surface representatives of the nephelite syenite of the central core.— Y. J. Geol. Soc., lvi, p. 211, 1900. KEDABEKITE. EK. von Federov has given this name to a rock occurring in dike form in the vicinity of the Kedabek mines in the Government Elisabethpol, Transcaucasia. It is very fine granular, of a dark gray with atone of green. It con- sists of basic plagioclase, a lime-iron garnet (“aplome”) and a strongly pleochroic pyroxene called violaite (conf. p. 86 of this volume). The type is held to be remarkable in that it unites such different kinds as augite-garnet rock and diabase. The occurrence is in close relation with an augite-garnet rock and is held to be a local facies of a neighboring diabase. The analysis by A. Kupffer gave: SiO. 8% 11:58 | 0:79. 10-126] 0:000152 | 0°00119)| + 800,,)0;00 aman mem & | 1300 | 0°0107 (2 705 se | 8400 | 070107 | 705 B: 1:09 | 0-74 |,0°125) 0000101 | 000079, 4) 0-005 25 aiapad Ai| Sater | 3°85 | 0°52 '|'0"115| 0°000200 |-0°00157 3400 0 04eR = Fane A, | <2) 2:29 | 0°57 /'0-115| 02000185 | 000105) 1557) O:O1Gemnmee A, | se) } 1:67) | 07527 10:115| 0-000100' 000078415) “2 tO-Wiier aes A,| §“ ©|'0:80.| 0°50 |:0:115)0:000046: ||0 000861) 25 7” "0; 0Gsinanimnm A,| “ | 1:53 | 0°254) 0-114) 0000044 | 0-:000845| “ |0:0049 | 111 C, | Boker's | 3°13 | D= | 0-22 | 0:000119 | 0-00094 |. “ | 0-0211 | 177 _ C,| wire | 1:14 | D= | 0-22 | 0-000043 | 0-:00034 | “ | 00024 | 55 J,| Extra | 2°20 | D= | 0-078] 00000105; 0-000082) “ | 0-00284 | 270 J, | Stet | 1°21 | D= | 0-069| 0:0000045| 0:000035! “ | 0:00076 1168 Galvanometers of High Sensibility. 257 Saturation of Magnets.—The experiments with magnet B, _ (see preceding table), magnetized in fields of 2200 and 5200 lines, and with magnet B,, magnetized in fields varying from 800 to 3400 lines per sq. em., show that a limit to the intensity of magnetization attainable is soon reached, practically in a field of 1000 lines of induction per cm’. We undoubtedly have the isthmus effect coming in when these small magnets are magnetized by placing them between the faces of an elec- tromagnet; thus in B, which was magnetized in a field (undis- turbed) of 2200, an average intensity of magnetization of 284 is reached, which must correspond to an average remanent induction of at least 3500. | Liffect of Length and Cross Section of Magnets.—The experi- ments made on the group of fine hair-spring magnets H, to H.,, all taken from one piece tempered glass hard in water, shows the variation of the magnetic moment with the length. These results are plotted in a curve, fig. 6 (p. 260), which shows that, for short magnets, the magnetic moment decreases much more rapidly than in proportion to the length. The effect of cross section on the intensity of magnetization is shown by comparing, e. g., H, and B,, the average intensity of magnetization in H, (although shorter) being more than 2°7 times that in b,. Similarily A, (width reduced by grinding), although shorter, has a considerably greater intensity of mag- netization than A,. This strongly emphasizes the fact that probably the greatest gain in sensibility at present at our dis- posal is in the use of a number of magnets of small cross sec- tion instead of one (or so) of large section, i. e., if you have a given mass of magnetic material (to be made up of magnets of a given length), the system will be considerably more sen- sitive if you use a number of magnets of small cross section. This brings up at once the question whether, when you divide up the magnetic material into a number of thin magnets, you do not in the end lose what you gained in magnetic moment by subdivision, by the fact that in galvanometer systems the magnets must be placed very near together and so tend by their mutual action to weaken one another. Mutual Action of Magnets.—This question was investigated for many sizes of magnets and under various conditions ; one or two experiments will, however, be sufficient to show that the effect is much smaller than is usually supposed. To study this the magnets B, were fastened to one of the brass inertia discs, magnetized, and period determined, in an exhausted receiver, when the distance between the centers of the magnets was 3°37"; the magnets were then pushed together until the dis- tance between the centers of the magnets was 1:°2™", when the 258 Mendenhall and Watidner— period was again determined; magnets then pushed apart, ete. The results are shown in the following table: ae M Distance between Time of Vibration Magnetic Moment Centers of Magnets. of Disc. C. G. 8. Units. By og oovan 2°6 18secs 0°1951 12 oo TS °1160 3°2 2°659 71212 1:2 2°725 °1155 3°2 2°663 "1209 ihe” PNT . S1los 3°2 2°666 "1206 These experiments show that the first near approach of the magnets produced a permanent decrease in the magnetic moment amounting to about 4 per cent (this was to be expected inasmuch as these magnets had not been aged). After that it will be seen that the magnets practically regain their former magnetic moment on separation. The temporary decrease in magnetic moment due to mutual action is less than 5 per cent. A similar set of experiments with the shorter magnets, B,, gave a decrease of 5°5 per cent due to mutual action when the center lines of the magnets were 1™™ apart. When the mag- nets were pushed together so that they touched their neighbors throughout their length, the magnetic moment continued to decrease for some minutes until the decrease amounted to 19 per cent, and on separating them they did not recover but were permanently weakened about 10 per cent. Quality of Steel.—Under similar conditions magnets of Jonas & Colver’s special magnet steel* (probably tungsten steel) have moments about 25 per cent greater than those made from watch hair-spring, which is the next best material. Lliffect of Jars, etc.—Magnets B,, 4™™ long, dropped twenty times though 30" on toa olass plate diminished only | per cent in magnetic moment; with magnets B*, 1™™ long, the change was less than 1 per cent. Boiling immediately after magneti- . zation reduces the moment by 10 per cent or 15 per cent, and most of the loss occurs during the first few minutes of boiling. Paschen*+ has called attention to the fact that to increase the sensibility, for a given period, we must increase the total mag- netic moment M and decrease the moment of inertia, K. Assuming that the magnetic moment is proportional to the. volume, i. e., for a given cross section proportional to the length of the magnet, and that the inertia of the non-magnetic * From Boker & Co., N. Y. yy et Wied. Anny, xlviii, 282) Galvanometers of High Sensibility. 259 material could be neglected, Paschen showed that if two systems were made with magnets of the same cross section but Y of lengths 7 and i respectively, then the second should give a deflection n’ times as large as the first system, for if the magnetic moment of the first system = M ; and the moment of inertia “ s¢ KY then ! - the magnetic moment of the second system = a and the mom. of inertia “ es c = n The deflection of first system is proportional to K anne) =< €6 mecond) .!. proportional to = ae i. e., the deflection of the second system is 7° times that of the first system. The assumption (that M « / for given cross section) on which the deduction is based is not even approxi- mately true for short magnets, as the preceding experiments show, but it does at least indicate an advantage in favor of short magnets. This deduction is supported by experiment. Thus Paschen found by testing three similar systems having magnets 4™™, 2™™, and 1:°3™" long respectively, made from the same fine hair-spring, that the second gave three times the deflection and the third six times the deflection of the first system. From curve 1 it will be seen that, if the moment of inertia of the non-magnetic parts of the system can be neglected in comparison with that of the magnetic (a condition which can perhaps be realized even with systems as light as 2 mg. total weight) the sensibilities of two systems having magnets 2™™ and 1™™ long respectively, of the same mater‘al and cross section, would be proportional to 2 K = ey for system 1 K = a for system 2 z.é. the second system would be about twice as sensitive as the first, instead of four times as deduced by Paschen on the assumption that M « / for given cross section. If it were always possible to use enough magnets, so that the non-magnetic moment of inertia could be neglected, then the 260 Mendenhall and Waidner— sensibility would continue to increase as the magnets decreased in length. This not being the case in general, it is possible to use too short magnets. To show this we have calculated the sensibilities to be expected with four systems made of the same material and cross section as magnets I, (see table), which showed a greater intensity of magnetization than any others of the same length. Furthermore we assumed that only four such magnets were to be used in each group, to be -65™™ long in the first system, and 1™™, 1°5™™, 27” in the others; that the mirror weighed °23 mg., and was 1" x 1-1"" in area, the staff and shellac weighed (09 mg., and the hook -06 mg.—these values being found by trial. Except for the omission of the hook, which is readily dispensed with, it would be difficult to reduce the non-magnetic mass much below this. The results are shown in curve 2—where ordinates are proportional to sensibility, and abscissee to length of magnets used. A very decided maximum is shown, for a length of 11". It must be said, however, that the magnets could be put closer together without undue demagnetization, and that two sets could be Galvanometers of High Sensibility. 261 mounted, one on each side of the staff, thus increasing the number from eight to twenty; this would require shorter magnets for maximum sensibility. We have not yet exactly realized this system in practice, but give the result obtained with a system having only six magnets, of the same material and cross section as I,, and an average length of about 1:16"; the non-magnetic moment of inertia was probably a little less than that assumed in the above calculation. The sensibility actually obtained, and also the sensibility reduced to the Ayrton—Mather scale, is given in the following table, together with the sensibilities of a few other recent galvanometers, taken from the table of Ayrton and Mather (Phil. Mag., Oct., 1898). It should be noted that with these very light systems the “period” when it is at all long—say 5* to 10° complete—is very difficult to determine, on account of the great damping. Moreover, it is unfair to compare systems which were tested at considerably different periods, and reduced to the same period by assuming the sensibility « T*’—for this is on the basis of a Sree period.* In computing moments of inertia for the above systems, we have assumed that the axis of rotation passed through the center of gravity of the system and was parallel to the staff ; if the system is badly constructed so that the axis of rotation departs ‘2™™ from the staff at its upper extremity, the moment of inertia will of course be increased, and the sensi- bility (for a given period) decreased by about 25 per cent. The moments of inertia of the four systems computed for use in plotting curve 2 varied between 44x10° and 529x10° C.G.8. units. TABLE. Deflections per micro- Bee ee = ampere when scale Galvanometer and Con- S ‘an Sst Sensibility ane cones, | dist.=t1o000 scale divs., stants when tested. uspenacion. py stem. | complete period = tested. F Io secs. Resistance =r ohm. Snow (Wied. Ann.| 6 small watch-|C=1°5 x 10amp.| xlvii, p. 218; Phys.| spring magnets in per mm. Rev., i, p. 37). each group. 3 on|Scale dist.=300cm. | 4 coils, 30mm. ext.| each side of staff.;T (complete)= and 6mm. int.diam.| Magnets 3-4 mm. 10 secs. | Each coil wound) long. R=140 ohms. | 470 in 2 sections, with} Mirror 5mm. 1800 turns. dia., 0°14mm. thick. | R = 140 ohms Quartz fibre 40cm. | (series) | long. | Wt. of system= 80mg. * See Ayrton and Mather, Phil. Mag., pp. 366, 1898. 262 Mendenhall and Waidner—Galvanometers, ete. Galvanometer and Con- Sensibility and condi- tions of when Deflections per micro- ampere when scale dist.=1o000 scale divs. ‘68me., length of magnets ==1:15mm. Plane mirror, ‘9mm. x Imm. ' sizes of wire, 500 turns on each coil. =3 ohms (in parallel) only one turning point). Zero stable. R=3 ohms. stants when tested. Suspension System. ented Use complete period = : Io secs. Resistance =I ohm. Wadsworth(Phil.! 10 small magnets | C= 4 x 10~!4 amp. Mag, xxxviii, p. in each group (5 on per mm. 553). ‘each side of staff, | Scale dist.=100cm. 4 coils, 50mm.| varying from 2~—|T (complete) = ext. diam.,2mm.int., 3mm. in _ length) 20 secs. and 40mm. deep. | made from smallest R=86 ohms. Each coil wound in) size sewing needles 615 5 sections, with (untreated). 2396 turns. | Glass 150mm. R=86 ohms (series)| long, weighs 5mg. | Mirror (concave) 2°5mm. diam. and weighs 12mg. Nichols EK. L.and| Wet. of system= | C=1:3 x 10-!°amp. HUES (Phys. Rev., | - 48mg. | per mm. 1,00: 2): | Seale dist = 150em. 9900 R=9°3 ohms | _T (complete) = ¥ (parallel) | | 10 secs. R=9°3 ohms. 4 Paschen Wied.: Hach group has; C=3:3 x 10-Pamp. Ann., xlviii, p. 284).|13 magnets 1 to per mm. 4 coils, 40mm.'/1°5mm. long, on! Scale dist.=300cm. external and 5mm.| both sides of glass; T= i5secs. (ape- : internal diam. staff 03mm. apart;| riodic, i.e, has Wound with graded) Mirror 2mm. dia.| only one turning 5800 wire, about 1200) 003mm. thick. point). turns in each coil. 5em. quartz fibre.| R=60 ohms. R=60 ohms. Wt. of system = | omg. | Mendenhall and! 3 magnetsineach | C=5°6x 10-Uamp. Waidner. | group. /per mm. deflection. 4 coils, ext.diam.| Total wt.1mg. | Scale dist. =200em. =15mm., internal; Wt. of magnetic) T=9 sec. complete diam. = 2mm. 6/part of needle = | aperiodic (i. e. has 6090 Thompson Physical Laboratory, Williams College. Davis—Locating Nodes and Loops of Sound, etc. 268 Art. XX VIII—On a Method of Locating Nodes and Loops of Sound in the Open Arr with Applications ; by BERGEN Davis, P#.D. | THE experiments and determinations given in this article constitute an application of a new phenomenon recently described by the writer in this Journal.* It is now my wish to point out how the device previously described can be used to locate nodes and loops of sound waves outside of organ- pipes, and in general how the mechanical effects produced by sound waves, at a distance from this source, can be studied. It having been found that a small hollow cylinder which is closed at one end, will move across a stationary sound-wave in a direction perpendicular to the stream-lines, a small mill-like arrangement was constructed by placing four hollow cylinders on the ends of card-board arms, in such a manner that the closed ends pointed'in the same angular direction. This mill was provided with a glass pivot at its center, and was supported on the point of a fine needle. Thecylinders were made from No. 00 gelatine capsules, and were each 1:7 long and -79™ in diameter. This system was mounted in the mouth of ares- onator, with the plane of the system parallel to the mouth, and hence perpendicular to the direction of vibration. The aperture of the resonator, with the mill in place, is shown in figure 1. The resonator was in unison with an organ-pipe of considerable power, and when the pipe was blown the mill was found to rotate with a high velocity, and the’rate of rotation was different for differ- ent positions in the room. By carrying the resonator around a large room, the positions of the nodes and loops could be located with considerable accuracy. The resonator containing the mill was next carried to another large room, on the floor above that where the pipe was located : there was no opening in the floor or ceiling between the rooms, and the doors of both rooms were tightly closed, but in spite of this, the mill was observed to rotate as before, but not so rapidly. Nodes and loops could be located here also. * Bergen Davis, this Journal, Sept., 1901, p. 185. 264 Davis—Locating Nodes and Loops of Sound, ete. In order to avoid reflection from walls, and the conseqnent formation of stationary waves, the pipe was then carried out of doors, and compressed air led to it by a long rubber pipe. . Here, in the open air, the mill was found to rotate very rapidly when near the pipe, and the rate decreased with the distance from the pipe, ceasing to rotate at about 60 feet from the pipe. This distance could probably be very much increased by the use of more delicate apparatus, especial care being taken with the pivot and needle point. This furnishes a means of study- ing the decrease of intensity with distance, and, with the aid of the formula developed by Lord Rayleigh,* of measuring the actual amplitude of the vibration at various points in the open air. | A very sensitive sound detector of this character might be made by suspending a system of very small cylinders in the mouth of the resonator by means of a quartz fiber, and then observing the deflection by a mirror and telescope. Such an instrument ought to be as sensitive as the one constructed by Boys, who used a suspended dise for the same purpose.t An instrument of this kind, perhaps, might be useful in investigat- ing the acoustic properties of buildings, also in the study of the reflection, refraction and absorption of sound. Physical Laboratory, Columbia University, June 15, 1901. * Theory of Sound, II, pp. 195-200. + Boys, Nature, vol. xlii, p. 604. | Pee . Winton-—Anatomy of the Fruit of Cocos nucifera. 265 Art. XXIX.—The Anatomy of the Fruit of Cocos nucifera ;* by A. L. WiInTon. [Contribution from The Connecticut Agricultural Experiment Station, New Haven, Conn. | J. MorpPpHoLoGy AND MACROSCOPIC STRUCTURE. SINCE the general structure of the cocoanut fruit has been treated by numerous writers on systematic and economic bot- any, only such facts are here given as are essential for a clear understanding of the relation of the parts and the microscopic structure. The flowérs are arranged in spikes branching from a central axis and inclosed with a tough spathe usually a meter or more in length (fig. 1). A single female flower is borne near the base of each lateral axis, and numerous male flowers are dis- tributed on all sides of the axis between the female flower and the apex. After the male flowers drop, each naked lateral axis persists and is a prominent appendage of the fruit (figs. 2 and 3, 8). Only one ovule of the three-celled ovary comes to maturity, but the tricarpelary nature of the fruit is indicated by its triangular shape as well as by the longitudinal ridges and the three eyes or germinating hole of the nut. The epicarp of the fruit (fig. 3, Apz) is a smooth tough coat, of a brownish or grayish color. The mesocarp (fig. 3, J/es), consists of a hard outer coat but a few mm. thick and a soft portion usually 3-4™ thick on the sides and much thicker on the base. Imbedded in the meso- carp are numerous longitudinally arranged fibers, varying in size from slender hairs to large, sparingly branching and anastomosing, flattened forms, 2-5" broad. The large fibers are situated chiefly in the inner layers, with their flat surfaces parallel with the surface of the nut. Oftentimes the inner layers of the mesocarp become impreg- nated with a brown fluid, which on drying, gives the thin tis- sue a mottled brown appearance. * European microscopists have studied the foods and aduiterants which have come under their observation but have overlooked a number of distinctly Ameri- can products. The writer has undertaken to fill in some of these gaps by a series of papers, of which this is the second. The first paper, on the anatomy of maize cob, was published in the Cesterreichische Chemiker-Zeitung, 1900, p. 345, and also in the Conn. Experiment Station Report, 1900, p. 186. Each paper will describe from the purely scientific standpoint the macroscopic and histological structure of the material investigated, and also in a final chapter point out the application of this knowledge to the detection of adulteration. The last chapter is not strictly suited to the pages of this Journal, but is so dependent on the scientific descriptions which precede it that it would be almost valueless if published separately. Am. Jour. So1.—Fourts SERIES, Vou. XII, No. 70.—Octoser, 1901. Le 266 ©=Winton— Anatomy of the Fruit of Cocos nucifera. Fic. 2. Half grown cocoa- nut fruit with calyx, and axis from which the male flowers have fallen. x 4, Fic. 4. Inner surface of a cocoanut shell with adher- Fig. 1. Inflorescence of the cocoanut show- ing outer testa. At the left ing spathe inclosing the spikes, each with the raphe, from which pro- numerous male flowers above and a single ceed veins forming a net- female flower near the base. x #4. work over the surface. x4. Winton— Anatomy of the Frurt of Cocos nucifera. 267 The endocarp, or shell (fig. 3, Hnd), consists of a hard, dark brown coat, 2-6" thick, with numerous fibers adhering to the “surface. Three nearly equidistant ridges (often indistinct) pass from base to apex, where they unite to form a blunt point. At the basal end, between the ridges, are the three depressions or eyes, the tissues of which are much softer and thinner than of the rest of the shell (fig. 3, A). Through the softest of these eyes the embryo, embedded in the endosperm directly behind it, escapes in sprouting. Fie. 3. Ripe cocoanut fruit. S, lower part of axis forming the stem; A, upper end of axis with scars of male flowers; Hpi, epicarp; Mes, mesocarp with fibers; nd, endocarp or hard shell; 7, portion of testa adhering to endosperm; Alb, endosperm surrounding cavity of the nut; K, germinating eye. x 2 = The testa of the anatropous seed (fig. 3, Z, and fig. 4) is a thin coat of a light brown color, closely united with the endo- carp without and the endosperm within. Embedded in the outer portion and extending from the principal eye nearly to the apex is the raphe, consisting of a thin band of vascular tis- sues about 1™ broad, which sends off branches in all directions, forming a network about the seed. The endosperm with the inner portion of the testa may be separated from the outer testa and endocarp by introducing a knife blade between the layers. By this operation the veins are split, part of the vas- cular tissue adhering to the convex surface of the inner testa, and the remainder to the concave surface of the outer testa, so that both surfaces are covered with reticulations. 268 Winton—Anatomy of the Fruit of Cocos nucifera. The endosperm or meat of the cocoanut (fig. 38, Add.) is a white, fleshy layer, 1-2™ thick, in which, near the base, is embedded the small embryo. While immature, the nut is fled with a milky liquid and has no solid endosperm, but as the ripening proceeds the endosperm is gradually formed and at the same time the milky liquid diminishes in quantity or entirely disappears. Cocoanuts yield food for man and eattle, oil, fiber, and other useful products. The epicarp and mesocarp are cut away from nuts designed for export, although invariably a small amount of the mesocarp with its fibers remains attached to the shell. In removing the meat, the outer testa, as has been stated, also adheres to the hard shell, so that cocoanut shells consist not merely of endocarp, but also of a certain amount of mesocarp and testa. II. Histonoey. The microscopie structure of the cocoanut seed is described by Hanausek,* Harz,t Moeller,t Koenig§$ and other authori- ties on foods and applied microscopy. Cocoanut fiber (coir), which has long been extensively employed in making mats and cordage, and also cocoanut shell, which has been used for making knobs and other turned articles, were studied by, Wiesner| nearly thirty years ago, but his work was designed chiefly to distinguish the fiber from other commercial fibers and the shell from the similar shell of Attalea funifera. Von Hoehnel{ describes briefly the histology of coir, but, like Wiesner, does not appear to have understood the true nature of the steomata. Weiss,** Engler and Prantl,tt+ and some other authors refer briefly to the microscopic structure of parts of the cocoanut, but their descriptions are of little value in diagnosis. 1. EHpicarp. The epicarp or epidermal layer is about :015™™" thick and is made up of tabular cells with dark brown contents. In surface view the cells are usually square, rectangular or tri- * Die Nahrungs- und Genussmittel aus dem Pflanzenreiche, Kassel, 1884, p. 155. + Landwirthschaftliche Samenkunde, Berlin, 1885, p. 1120. ¢ Mikroscopie der Nahrungs- und Genussmittel aus dem Pflanzenreiche, Ber- lin, 1886, p. 241. g Die Untersuchung landwirtschaftlich u. gewerblich wichtiger Stoffe, Berlin, 1898, p. 291. || Die Rohstoffe des Pflanzen-Reiches, Leipzig, 1873, pp. 436 and 789. (A new edition is being published in parts, but the chapters on the cocoanut have not yet. appeared.) . ] Die Microscopie der technisch verwendeten Faserstoffe, Leipzig, 1887, p. 52. ** Anatomie der Pflanzen; Wien, 1878, 1 Band. ++ Die natiirlichen Pflanzenfamilien, II Theil, 3 Abteilung, p. 22. Winton—Anatomy of the Fruit of Cocos nucifera. 269 angular, with double walls about -005™™" thick and are arranged. _ with some regularity in rows. | 2. Mesocarp. (a) Hard ground tissue.—This tissue consists of thick-walled cells which are often tangentially-transversely elongated. In the first few layers the walis are about the same thickness as in the epidermis, without evident pores, but further inward they are more strongly thickened (double walls often -015™™ thick) and conspicuously porous. Still further inward they pass into the parenchyma of the soft ground tissue. (b) Bast-fiber bundles.—In the hard ground tissue the bun- dles have no phloem or xylem but are composed entirely of bast-fibers with cell walls often thicker than the lumen. The number of fibers seen in cross section varies from two or three 53 CASsOrie. AA aa \ SOLON CWoannug, MIAN CZ COLIN OOO ph Ph OL, aeRO OPS \ CGO Sa oe as abst. SS WiSG DOKI ig . ee « eee 7. \ Se c IAD. AZORES ROS © ee ay SOGNOce ASRS = SSSS Citane : Oeeste® eS 0 Bea sleet: wintece, CCM aS e S is) fr oo) Oy Sox i ; . CANO a\s: cele a Ser PORE poy RAS es Uae tO oNe 3 RAT ee 6 OP o 1©, See ae OW EOWA ING Ui RRS . Y oS se a INS \f Fic. 5. Transverse section of a large flattened (mesocarp) fiber of the cocoa- nut. ste, stegmata; 7, sheath of bast fibers; ph, two phloem groups; x, xylem; p, parenchyma of ground tissue; a, rudimentary bundle belonging to small branch. x 90. up to a hundred or more. Transitional forms between fibrous and fibro-vascular bundles occur further inward. (c) Soft ground tissue.—The thin-walled parenchyma cells of the soft ground tissue are in some parts isodiametric, in other parts longitudinally elongated, and in still other parts trans- versely-tangentially elongated (fig. 8, w). Wherever the brown liquid previously referred to has penetrated the inner layers of the mesocarp, groups of the parenchyma cells here and there, being impregnated with this material, are of a rich, brown color and appear thicker-walled than the others (fig. 8, 270 Winton—Anatomy of the Fruit of Cocos nucifera. br.). This brown substance is quickly changed to a reddish color by caustic potash, but is not affected by alcohol, ether or the specific reagents for proteids, fats and resins. No imme- diate effect is produced by ferric chloride solution, but on long standing the color is changed to olive green. (d) Fibers (Coir).— These are fibro-vascular bundles with a strongly developed sheath of bast-fibers.- Toward the xylem side of the bundle, particularly in the large fibers, the sheath usually diminishes in thickness and the vascular portion, as seen in cross section, is more or less eccentric, surrounded by a crescent-shaped sheath with the horns connected by a narrower stri ie the smaller fibers there is but one group of phloem ele- ments, but in the larger flattened: fibers there are usually two, or occasionally more, groups ay from each other by a continuation of the sheath (fig. 5). Normally the xylem is near the inner flat side and the two phloem groups are approxi- mately symmetrical with reference to the shorter axis of the elliptical cross section; but often the xylem is near one of the narrow sides and the phloem groups are symmetrical with ref- erence to the longer axis, and still more often the arrangement is diagonal or otherwise invegnlar, Mohl* in 18381 noted that the phloem in the stem of Cala- mus was normally divided into two distinct groups, and Knyt as well as other authors have since found the same arrangement ina number of palms. By the study of many sections, the writer has demonstrated that a cocoanut fiber with two phloem groups has also a double xylem, aithough in most sections no separation is evident, and the whole fiber consists of two sim- ple bundles united side by side, which may completely sepa- rate further on in their course by the forking of the fiber. Serial sections cut through such compound fibers show that at the place of forking the phloem groups are still further sep- arated and the xylem also is divided by bast-fibers, thus form- ing two distinct bundles which pass into the two branches. The phloem in each branch is at first entire, but further on, if the branch is large it usually divides, and still further on the whole bundle may split up, with the formation again of two fibers. Occasionally a fiber which has no evident division of the xylem has four groups indicating that the fiber is com- posed of four united bundles, which, on branching, form two tibers each with a double bundle. Large fibers not only fork but also send off small lateral branches. The rudimentary bundles belonging to such * De Palmarum Structura, Translation in Ray. Soc. Reports and Papers, 1849; ae + Verhandl. d. Bot. Ver. Proy. Brandenburg, Bd. xxiii, 1881, pp. 94-109. Winton—Anatomy of the Fruit of Cocos nucifera. 271 branches may often be seen in cross sections of the trunk fiber below the place of branching (fig. 5, a). a. Stegmata (figs. 5 and 6, ste)—As seen in surface view these are circular or elliptical cells from -008 to 020" in diam- eter, which extend in longitudinal rows over the surface of the fibers. Longitudinal sections show that the cells are biconvex, fitting into depressions in the bast-fibers, and that the outer walls are exceedingly thin, while the inner and side walls are strongly thickened, thus bringing the cell cavity near the outer surface. Inclosed in each cell and filling it almost completely, is a silicious body, from -006 to -012™" in diameter, with wart- like protuberanges on the surface which fit into corresponding depressions in the cell walls (fig. 7). That they are composed 2 Dagan eO0%nt Dog h) vet es helen Con. yo Fig. 6. Longitudinal section of a large (mesocarp) fiber of the cocoanut. sfe, stegmata; S7,silicious body; f, bast fibers; ¢, tracheids with small pits; ¢’, trache- ids with large pits; sp, spiral trachea; 7, reticulated trachea; sc, scalariform trachea; s, sieve tube; c and c’, cambiform cells. x 300. of silica is demonstrated by their incombustibility, their insolu- bility in hydrochloric and nitric acids and their complete solu- bility in hydrofluoric acid. Their appearance is particularly striking in tangential sections which have been heated on a cover glass until thoroughly carbonized and finally treated with hydrochloric acid on the slide. The heating should be y performed at dull redness, since ata _‘ Fig. 7. Silicious bodies from higher temperature the bodies lose the stegmata of cocoanut fiber. their characteristic appearance. ete: Wiesner* refers to these stegmata as “ bast parenchyma,” and from his description it would appear that he considered them * Loc. cit , pp. 436-438. 272. Weinton—Anatomy of the Fruit of Cocos nucifera. silicefied cells and did not understand that they are sclerenchy- matized cells with selicious contents. Von Hoehnel,* who uses, however, the term “stegmata,’ also appears to have fallen into the same error. . Rosanofi+ found stegmata in twelve species of palms, and Kohl,{ who has made an exhaustive study of the subject, in twenty-three additional species. Neither author mentions Cocos nucifera, but Kohl found in C. flexuosa stegamata with silicious contents which answer the description of those in coir fiber. &. Bast-fibers (figs. 5 and 6, 7) completely surround the bundle. They vary in length up to 2™™ and in diameter up to 03""_" The double cell walls are from one-half to one-sixth the breadth of the lumina, with conspicnous pores and diag- onal markings. In longitudinal section the walls adjoining the stegmata are sinuous in outline, due to the depressions into which the stegmata are fitted. On the edge of the xylem the bast-fibers pass into tracheids (fig. 6, 2). y. Xylem (fig. 5,2; fig. 6). The elements are trachez, tracheids and various forms intermediate between tracheids and bast-fibers, and tracheids and parenchyma. The trachese range in diameter up to ‘05"™, the larger (found in large fibers) being reticulated (fig. 6, 7) or sealari- form-reticulated (se), the smaller (found both in large and small fibers) being spiral or reticulated spiral. Among the spiral tracheze one finds considerable variation both as to their size and the steepness of their spirals. As might be expected, those in the protoxylem often have delicate spirals with turns wide apart. An intermediate form is shown in fig. 6 (sp). The tracheids, distinguished from the tracheee by the trans- verse or diagonal partitions and by their smaller size and thin- ner walls, likewise display an interesting diversity of size and form. Among these are forms with large pits and curious reticulations (fig. 6, z’), also transitional forms between tracheids and bast-fibers (¢) on the one hand, and tracheids and paren- chyma on the other. 6 Phloem. Sieve tubes and cambiform cells make up the phloem (fig. 5, pA}. Measured in cross sections, the diameters of the sieve tubes vary up to :(03™". In longitudinal sections it may be seen that the sieve plates areeither at right angles to the walls or oblique and that oftentimes they are covered with callus through which run a few indistinct pores (fig. 6, s). ITO, Olin, (Ds DAs + Bot. Ztg., 1871, p. 749. { Kalksalze und Kieselsaure in der Pflanze, Marburg, 1889, p. 289. Winton— Anatomy of the Fruit of Cocos nucifera. 273 Cambiform cells occur singly, in rows and in groups among the sieve tubes and also at the edges of the phloem. Those among the sieve tubes are for the most part small (about -003™ in diameter), prismatic and with abundant protoplasmic con- tents (fig. 6, c’). They correspond to the “ gelettzellen” of Wilhelm, Tschirch* and other authors except that the walls adjoining the sieve tubes, so far as the writer has observed, are not pitted. At the edges of the phloem, particularly adjoining the xylem, the cambiform cells are larger (often -01™™ in diameter) and are often empty. The differences between these forms are, however, so slight and perplexing that the writer, follow- ing the example of De Bary and Strassburger, prefers to group them all under the head of Cambiform cells. (ce) Intercellular spaces, such as occur in the protoxylem of many monocotyledinous plants, are seldom, if ever, seen in coir fibers, but oftentimes, although less commonly than in the hard shell, the phloem and part of the xylem are destroyed during growth, leaving a channel in the bundle. 3. Endocarp. This coat, known commonly as the shell (fig. 8, end), is a dense aggregation of stone cells, among which run longitudi- nally partially destroyed bundles. (a) The stone cells with their thick, deep yellow walls, branching pores, and dark brown contents, present a striking and characteristic appearance. They are either isodiametric or strongly elongated, the latter (often 0-2™™ long) being usually spindle or wedge-shaped, although hammer-shaped, hooked and various other curious forms abound. A study of sections shows that the elongated cells are arranged in groups, commonly with the longer diameters in tangential-transverse directions and are best seen in cross sec- tions of the shell (fig. 8, gst), but in some groups, particularly those adjoining the bundles, they pass longitudinally about the shell (fig. 9, Zs¢). It is evident from fig. 8 that more than half - of all the stone cells are tangentially-transversely elongated. Those which appear isodiametric (/st) are partly cells which are isodiametric in three dimensions and partly longitudinally elongated cells in section. Groups of thinner-walled cells with dark brown contents are occasionally met with. The brown contents of all the endocarp cells react the same as the brown impregnating material of the mesocarp. (b) Vascular bundles are studied with difficulty in the mature shell. By the rupture of the phloem and part of the * See Tschirsch, Angewandte Pflanzenanatomie, Wien, 1889, p. 349. 274 Winton—Anatomy of the Fruit of Cocos nucifera. xylem during growth, passages are formed, which, in shells transversely cut or broken, are evident to the naked eye as » Mes 4 1 = = = TF. ire —> LSS ae hi ——— = 227; CZ a eres yh Z a) = CIN Se SE a, SSeS => SS E Zo ; IG: See ee —— Se Sas Oe a CURED Cpus ee SSS SOs = SS LO oe Le SS: 3 ee eI Se SS Eee Se ONO Meee PSS ae SSC oh em CASS AES = D: eS SS eS UE ee BEE en Sg POSS SF I CSS 6S Sys > End RS ee Ee SI ee EES CLS Ist---«B& eee SS SSeS > — SSS SS) a pS ti &,X,MATOLONY: WIEN. eg) k SSS SSG o_O S Fic. 8. Transverse section of a cocoanut shell. nd, endocarp or hard shell ; Mes, adhering mesocarp; 7, adhering cuter testa; w, colorless parenchyma of mesocarp ground tissue; br, same as w but impregnated with a brown sub- stance; g, vascular bundles, in the endocarp with phloem and xylem partially obliterated ; Ist, longitudinally elongated and isodiametric stone cells; gst, trans- versely elongated stone cells. x 60 Winton-—_Anatomy of the Fruit of Cocos nucifera. 275 minute holes. The structure of the bundles is still further obseured by the presence of fungus threads and spores. In structure the bundles differ from those of the mesocarp fiber, the bast-fibers being replaced by forms intermediate between fibers and tracheids (fig. 9, 7). The vascular elements are chiefly spiral tracheze (sp.), and pitted tracheze (g), the lat- ter being especially noticeable. Fie. 9. Longitudinal-radial section of cocoanut endocarp through the stone cells and edge of bundle. gst, transversely elongated and isodiametric stone cells; (st, longitudinally elongated stone cells; jf, thick-walled porous cells ; g, pitted trachea; sp, spiral trachea. «x 300. 4, Testa. Several microscopists have studied the testa, but, owing doubtless to differences in the material, hardly two of them agree as to the number of coats or the character of the elements. The description which follows is based on the examination of numerous specimens. . (a) Outer testa. This coat consists of a ground tissue of large, variously shaped cells, crossing one another in all direc- tions, (fig. 8, 7, fig. 10), between which ramify the veins. 276 Winton—Anatomy of the Fruit of Cocos nucifera. Most of the ground tissue cells have colorless double walls, — from ‘004 to ‘010™™" thick, with conspicuous (sometimes large): pores, but in the inner layers they often have thmner walls without evident pores and except for their shape bear no resemblance to the other cells. As a rule, the cells are empty, but some here and there con- tain a brown substance apparently the same as is contained in the mesocarp and endocarp, which often takes the form of spheres (fig. 10, £), disks, or films with circular openings (wv). Colorless stone cells (fig. 10, st) are present in the outer layers and contrast strikingly with the deep yellow stone cells of the endoearp. Fig. 10. Tangential section of the outer testa of the cocoanut showing the ground tissue of thick-walled porous cells. Most of these are empty, but a few contain brown contents in the form of globules, (x) or films with circular open- ings (v). st, colorless stone cell: sp, spiral trachea. x 300. The conspicuous elements of the veins are spiral trachee, pitted tracheze and elongated cells intermediate between pitted trachese and the porous cells of the ground tissue, and are not distinguishable from the same elements of the endocarp bun- dles. (See fig. 9, sp, g and 7.) 3 In breaking away the meat, the separation is through the middle of the veins and the inner iayers of the outer testa, Winton—Anatomy of the rut of Cocos nucifera. 277 nearly all the ground tissue and about half of the vascular ele- _ ments remaining on the inner surface of the shell. (b) cnner testa. F irmly attached to the endosperm are from ten to twenty layers of small isodiametric or slightly elongated eells. The double walls are about -003™™" thick and free from pores. These cells contain a material varying in color from light yellow to dark brown, which either fills them completely or occurs in globules, films, etc., as in some of the cells of the outer tests. In the layer’ adjoining the endosperm the cells are smaller and have darker brown contents than the cells in the other layers. 5. Endosperm. Although the microscopic character of the endosperm has been fully explained by Harz, Hanausek and Moeller, a brief description is here given to accompany the descriptions of the other parts of the fruit. In the outer layers the prismatic cells are nearly isodiametric¢ (about °05™" in diameter), but further inward they are radially elongated, often reaching a length of °3"™. Cell partitions are about :0038™™ thick, without pores. The cells contain bundles of needle- shaped fat erystals and lumps of proteid matter, each Inmp- containing, as a rule, a erystalloid. Ether and alcohol readily dissolve the fat crystals and strong potassium hydrate solution saponifies them. The proteid bodies give the usual color reactions with iodine, Millon’s reagent and dyes. IIL THE DETECTION OF POWDERED COCOANUT Es IN GROUND SPICES. The adulteration of ground spices with powdered cocoanut shells was brought to notice in 1885 by W. H. Ellis,* public analyst, Toronto, Canada, and has since been frequently detected by A. McGill 4 of Ottawa and food analysts in differ- ent parts of the United States. The extent to which this fraud is practiced is indicated by the following summary of results obtained by the writer during the years 1896-7 in the examination of samples collected in the State of Connecticut. Black pepper. Cloves. Allspice. Samples examined --.__--.--- 147 37 24 Samples adulterated (total).. 47 V7 val Samples adulterated with ground cocoanut shells... 21 7 6 *Dept. Inland Revenue, Rep. on Adult. of Food for 1885, Ottawa, 1886, pp. 67, 79. + Laboratory of the Inland Rev. Dept., Bull. No. 20, 1890, pp. 7-11. 278 Winton—Anatomy of the Fruit of Cocos nucifera. It is stated on credible authority that in Philadelphia at the present time about six hundred tons of shells, obtained as a by-product in the preparation of dessicated cocoanut—an arti- cle much used in pastries and confectionery—are annually reduced to a powder in mills of peculiar construction and sold to spice grinders. This powder, without further treatment, is mixed with ground allspice, which it closely resembles in appearance. By cautious roasting the color of ground cloves and nutmegs is matched, and by roasting at a higher tempera- ture a charcoal is obtained which, mixed with starchy matter, is a clever imitation of black pepper. Fig. 11. Cocoanut shell powder. sé, dark yellow stone cells with brown con- tents; 7, reticulated trachea; sp, spiral trachea; g, pitted trachea; w, colorless and br, brown parenchyma of mesocarp; ff, bast-fibers with stegmata (sie). x 160. ; Powdered cocoanut shells appears to be a distinctively American adulterant. ‘The leading treatises on the microscopy of foods in the German, French and English languages, even those of recent publication, make no mention of it, and a num- ber of prominent European food chemists and microscopists have declared to the writer that they had never heard of its use. On the other hand, cocoanut cake (the residue from the oil presses), which in Europe is commonly employed, both as a cattle food and as an adulterant of human foods, is almost un- known in America. All the tissue elements of the mesocarp, the endocarp and the outer testa are present in cocoanut shell powder, but the stone cells of the endocarp make up the bulk of the material. (fig. 11, st). These cells are characterized by their porous, brown-yellow cell walls, their dark brown contents which Winton—Anatomy of the Fruit of Cocos nucifera. 279 become a reddish brown on treatment with potassium hydrate solution, and the predominance of peculiar elongated forms. They differ in one or more of these characteristics from the stone cells of pepper, allspice, clove stems, walnut shells, almond shells, Brazil-nut shells, hazel-nut shells, peach stones and olive stones. | The outer testa or lining of the shell also forms a consider- ‘ able part of the powder, the most striking elements being the thick-walled, porous cells (#) and the vascular elements. Colorless cells of the mesocarp ground tissue (w) are not dis- tinguishable from the parenchyma of many other plants, but when impregnated with the brown substance which has been described they are striking objects (67). Potassium hydrate changes the color of these brown cells to a reddish brown, but ferric chloride does not produce any immediate effect, thus distinguishing them from the brown cells of allspice seed, the color of which potassium hydrate removes and ferric chloride changes at once to a green. Spiral, reticulated, and pitted tracheze (sp, ¢ and g), from the mesocarp, endocarp and testa bundles, are also frequently met with in the powder, the pitted trachea being quite unlike any vascular elements of the spices. The stegmata (ste) of the mesocarp fibers with their silicious contents are characteristic, but they are difficult to find owing to the great preponderance of other tissues. Bast-fibers (/) are more liable to be encountered than the stegmata, but they furnish less conclusive evidence. Spices adulterated with charred cocoanut shells show under the microscope black, opaque fragments which are not bleached by aqua regia or nitric acid and potassium chlorate. Except in cases where some of the stone cells or other elements have escaped charring, this material cannot be distinguished from other forms of charcoal. Chemical analysis is a valuable adjunct to the microscopic examination and often determines approximately the extent of the adulteration, but since other nut shells have a similar com- position, the microscope is essential for the identification of the particular adulterant present. As was pointed out by the writer* five years ago, the crude fiber obtained in the process of analysis is particularly suited for the microscopic detection of stone cells and other tissues. The radical difference in composition between cocoanut shells and the spices to which they are added is shown by the following results by Winton, Ogden and Mitchell.t * Conn. Agr. Expt. Sta., Rep. 1896, p. 34. + Ibid., Rep. 1898,. pp. 198-211. 280 Winton—Anatomy of the Fruit of Cocos nucifera. Mobalaslyisote ees. a0 eae Ash soluble in water .._- ....---- -Ash insoluble in hydrochloric acid Volatile ether-extract ._.-...-___- Non-volatile ether extract ._-_-.- -- PNICOMOIEebTACt ae eee Reducing matters by direct inver- Slomyealc. nas starcn. see Starch by diastase method -_-_-_-- Crade*iber 22020 Han ire Te Motal nitrogen: 2s eee ee Oxygen absorbed by aqueous extract Quercitannic acid equivalent to O. absorbed Spey een ate Pe ep G ae 3 Black analyses.) Allspice. | Nutmeg. |Cocoanut en of 3 |(Av. of 3/shells. (1 analyses. )|analysis.) pepper. | Cloves. (Av. of 14|(Av. of 8 analyses, )|@nalyses.) fe % 1196 7°81 4°76 5°92 2°54 3°08 0°47 0:06 1°14 IASI) 8°42 6°49 9°62 14°87 38°63 ore) 34°15 2°74 13°06 8°10 2°26 OS) Abe 2°33 Hae SP % 9°78 4°47 2°47 0°03 4°05 5°84 — 121-79 18-03 | 3-04 22°39 0:92 1:24 oil o, 3°63 2°28 0°86 0°00 3°02 36°70 HO ihe 25°56 23°72 Pe 5). | —= % 7°36 0-54 In conclusion, the author takes this opportunity to thank his highly esteemed instructor, Prof. Dr. Josef Moeller of Graz University, Austria, for kindly assistance in the early part of this investigation: The work was begun in Prof. Moeller’s laboratory during the autumn of 1899, but after a_year’s inter- ruption was finished at this Station. Acknowledgment is also due Prof. E. Gale of Mangonia, Florida, who generously furnished material for study, and also Herr F. X. Matalony of Vienna, who skillfully reproduced on wood the author’s drawings. Wortman—Studies of Hocene Mammalia, etc. 281 Arr. XXX.—Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum; by J. L. Worrman. With Plates I-IV. [Continued from page 206. | SUBORDER OREopoNTA Cope. In the first part of this paper, I have given the more impor- tant characters by means of which this group of the Carnivora is distinguished from the Carnassidentia. It now seems desir- able to enter somewhat more fully into a discussion of these particulars, and I shall therefore consider some of the charac- ters of less importance, from the point of view of classification. The first of these which it is necessary to emphasize as a broadly distinguishing feature is that the Carnassidents have been suc- cessful, and have left numerous modified descendants which constitute a large and important part of the living mammalian fauna in almost all regions of the globe; whereas the Creo- donts failed, and have completely died out, without possibly the aquatic Pinnipedia represent them in the living Carnivore fauna. In the organization of the skull, there are a number of characters that are more or less distinctive, among which may be mentioned the large lachrymal spreading out upon the face, as is invariably the case among the carnivorous Marsupials and a number of the Insectivora. This is true of JdLesonyz, Dromocyon, Harpagolestes, Sinopa, Proviverra, Limnocyon, Thereutherium (?), and very probably also of Oxyena and Patriofelis. In some species, notably the Mesonychide, the nasals are broad posteriorly, as in the Marsupials, and almost exclude the frontals from contact with the maxillaries. The posterior border of the palate is terminated by a thickened ridge, recalling to a certain extent the structure of the marsu- pial palate. Tn the posterior part of the palate, moreover, there are very generally a large number ot foramina of variable size, but usually minute, situated behind the posterior palatine canals, which undoubtedly represent the remains of the vacu- ities so common to the marsupial skull. In certain forms, especially Sznopa, and in some species of Hywnodon, as Scott has shown,* there is an additional foramen in the base of the skull, just in advance of the condyloid, which is also a con- spicuous feature of the marsupial skull. The zygomatic arch, with its component bones, is heavy, and the malar usually extends well back toward the glenoid cavity. In the superior molars, if either border is elongated it is always the posterior, * Jour. Acad. Nat. Sci., 1888, p. 178. Am. Jour. Sc1.—FourtH Series, Vou. XI, No. 70.—OcToBer, 1901. 20 282 Wortman—Studies of Hocene Mammalia in the and if any of the molars and premolars are especially enlarged, it is never exclusively the fourth premolar above and the first molar below. The scaphoid, lunar, and centrale are very gen- erally free; but according to Cope,* they are united in some French species of H/7ywnodon from the Miocene. In some forms, such as certain of the Arctocyonide,t there may also be union of the scaphoid and centrale, as in many carnivorous Marsupials, and the pollex and hallux may be more or less opposable. In some species, the ungual phalanges are rela- tively broad, flattened, and fissured .at their ends, while in others they are laterally compressed, curved, and pointed. If we now contrast these characters with those of the Car- nassidentia, we observe that in them the lachrymal is never spread out on the side of the face, but is confined within the orbit. The nasals are not especially broad posteriorly ; the pesterior edge of the palate is not thickened; the numerous posterior palatal foramina are absent; there is never a double condyloid foramen, and it is always the anterior border of the superior molars that is elongate,if either. If there is enlarge- ment and specialization of any of the teeth it is always the fourth superior premolar above and the first true molar below. From the Upper Eocene stage onward, the scaphoid, lunar, and centrale are always united, and there is never union of scaph- oid and centrale alone. The bony claws are always curved, compressed, and pointed. Present evidence points to the fact that the two groups probably arose side by side from the Mesozoic Marsupials,| although we do not know any true Carnassidents earlier than the Torrejon, while certain of the Creodonts come from the * Puerco. | Quite recently, Matthew4 has classified the Creodonta as follows : * Tertiary Vertebrata, p, 256. + See Matthew, Bull. Amer. Mus., Nat. Hist., Jan., 1901, p. 14. t+ The only exception to this statement with which I am acquainted is found in the viverrine Hupleres gondati, from Madagascar, which otherwise resembles the Insectivores. S An apparent exception to this is found in Palwonicis, in which the posterior border of the first molar is slightly longer than the anterior. || I here use the term Marsupials in its larger sense, or as equivalent to the Metatheria of Huxley. It would apvear to be more or less doubtful whether the existing Marsupials furnish a stage of embryonic development immediately ante- cedent to that of the appearance of a distinct allantoic placenta. Just how far they may have departed from their Mesozoic ancestors in this respect we shall never, perhaps, be able to ascertain, but there cannot be the slightest doubt of the fact that a preplacental stage existed, from which the placental had its origin. While it may not have been exactly like that seen in the living Marsupials, it must have, according to the very nature of the case, resembled it in a great many important particulars. #| Loements, sp aie Marsh Collection, Peabody Museum. 283 “J. CrEopontTa Primitiva. No specialized carnassial ; trituber- cular upper and lower molars; shear rudimentary or absent. Claws unknown. Oxycleenide. Includes some genera with Lemuroid affinities in the dentition. IJ. CrEroponta Apaptiva. Carnassial, when present, on p.* and m.;. Claws, where known, of modern type and carried probably more or less free of the ground. Scapholunar-centrale early uniting (podials tending towards true Carnivore type). 1. Post-carnassial teeth disappearing ._-.------- Paleonictide. 2. Post-carnassial teeth becoming tubercular_ ._-.. Viverravide. 3. No carnassials; molars becoming flat-crowned ; premolars disappearing Vio.) i022) i... sole Arctocyonide. Ill. Creroponta Inapapriva. Carnassial, when present, not on _ p.£andm.;. Claws, where known, blunt, hoof-like, resting on the ground. No tendency of union of the carpals (podials tending toward Ungulate type). 1. Carnassials m.4 shearing teeth..-..--.-------- Oxyenide. 2. Carnassials m.2 shearing teeth _.__-------- Hycenodontide. _ 8. No carnassials; teeth with high, round, blunted cusps; upper molars tritubercular; lower molans, premolarvtorim. oo. 5. Mesonychide.” This classification is of course based upon the older concep- tions of the limitations of the group, and differs materially from that herein proposed. For example, the Paleonictidee and Viverravide have been removed to the Carnassidentia, the reasons for which have been fully set forth on a preceding page. This leaves only the single family Arctocyonide in Matthew’s group “Creodonta Adaptiva.” This family, as has already been suggested, stands much nearer to the carnivorous Marsu- pials in the organization of the feet than to any Carnassident. The union of the scaphoid and centrale, the marked tendency to opposability of both pollex and hallux, the character of the fibulo-astragalar articulation, as well as the compressed, curved, and pointed claws, are almost exactly paralleled in the case of the living Dasyures. Moreover, itis highly probable that this family will be found to be further characterized by the extra- orbital extension of the lachrymal and the double condylar foramen. In the light of these facts neither the genus nor the family can be regarded as having any very close affinities with the Carnassidentia. In this connection, it is important to state that Cope, } in his description* of the fragmentar y skeletal material of Clenodon Jerox at his disposal, compared it carefully with the carnivor- ous Marsupials Didelphys, Sarcophilus, and Thylacynus ; he - pointed out a number of similarities of structure to these forms * Tertiary Vertebrata, 1884, p. 330. 284 Wortman—Studies of Kocene Mammalia in the and concluded that “its near est living ally is the Zhylacynus cynocephalus of Tasmania.” In opposition to this view, Matthew observes*—‘“ Cleenodon has, however, no marsupial characters except such as may be considered an inheritance from a common stock, which gave rise to both Marsupials and Placentals. Its progressive characters are placental carnivore.” Just how this latter author explains the remarkable resem- blances of foot structure in the two groups, including the marked tendency towards opposability of pollex and hallux, or what constitutes the so-called progressive characters of which he speaks, other than these same marsupial characters, [ am at a loss to imagine. Jn my judgment Cope’s position rests upon very sound anatomical reasoning, the force of which is rendered all the more patent by these later discoveries. As regards the third group, | am compelled to add that I find myself unable to accept the definition as accurate, or the grouping as natural. If by the term carnassial is meant the most specialized cutting teeth of the series, then the definition is fairly satisfactory, but the character would have been much better and more clearly expressed, had it been stated what teeth are sectorial. The character of the ungual phalanges, and their supposed manner of articulation, I do not regard as of any very great value in the classification of the major divis- ions of this group, for the reason that within the limits of one of the families, at least, their structure ranges from the com- pressed, curved, and sharp- pointed variety, to the flattened, depressed, and fissured type. The statement that the podials are tending toward an “ Un- gulate type” is such an extraordinary one, and is so at variance with the facts, especially as regards the Oxyeenidee and Hyzeno- dontide, that Tcan har dly believe that the author intended it to apply to these families. It is true that the feet of the Mesonychidee, in the later members, assumed a more or less “Ungulate type,” because of a highly developed running habit, just as, among the living Carnassidents, the limbs of the mod- ern Canid have taken on a similar structure from a like cause. But to suppose that there is even the faintest suggestion of anything “‘ Ungulate” in the foot structure of such forms as Sinopa, Limnocyon, Oxyena, Patriofelis, or Hycenedon, other than the fissured claws in the last three genera, is certainly quite beyond the facts. The association, moreover, of the Mesonychidee with the Oxyeenidee and Hy senodontidee is not very apt, since in point of structure it is widely separated from these two families, and constitutes one of the most distinct and separate groups of * Bull. Amer. Mus. Nat. Hist., 1901, p. 12. Marsh Collection, Peabody Museum. 285 the suborder thus far known. I define the five families as follows: Oxyclenide. No specialized carnassials ; tritubercular upper and lower molars; shear rudimentary or absent (Matthew). Genera: Oxyclenus, Chriacus, Protochriacus, Deltatherium, and Zricentes. All from Lower and Middle Eocene. Arctocyonide. No carnassials; molars flat-crowned;_pre- molars becoming progressively reduced in size; scaphoid and centrale early united; hallux and pollex more or less opposable ; claws compressed, curved, and pointed. Genera: Arctocyon, Clenodon, and Anacodon. From Lower and Middle Eocene. Mesonychide. No carnassials; molars with characteristic, high, bluntly conical-cusps, superior tritubercular, inferior becom- ing premolariform ; claws depressed, little curved, and fissured ; limbs in later forms becoming much modified in accordance with running habit. Genera: Zriisodon, Goniacodon, Sarcothraustes, Dissacus, Pachyena, Mesonyx, Dromocyon, and LHarpagolestes. .From Lower, Middle, and Upper Eocene. | Oxyenide. Pm.4 and M./, carnassial, of which M4 are largest and most specialized; claws, as far as known, depressed, little curved, and fissured. Genera: Oxycna, Patriofelis, Limnocyon, Oxycenodon, and Therenitherium. From Middle and Upper Eocene. Hycenodontide. Pm.4 and M.;3; carnassial, of which M.2 are »2> the most specialized, especially in the later types; claws depressed, little curved, and fissured; or compressed, curved, and pointed; in early types a double condyloid foramen as in Marsupials. Genera: Sinopa, Proviverra, Hycwnodon, Pterodon, Quercy- therium, Cynohycnodon, Pulewosinopa, and Didelphodus. From Middle and Upper Eocene and Oligocene. Hamily Mesonychide Cope. This family is represented at the very beginning of the Tertiary by the genus Z7zzsodon, with three well-marked species from the Puerco of New Mexico. In the suczeeding Torrejon beds of the same region, three generic modifications make their appearance; namely, Sarcothraustes, Goniacodon, and Dissacus. The first of these has been properly regarded as the direct descendant of 7Z7rizsodon, while the last repre- sents the beginning ot the J/esonyx line of succession. ‘The first three of these genera—TZ7iisodon, Sarcothraustes, and Goniacodon—have been considered by Cope and Scott to rep- resent a distinct family, Trisodontidee, but I have shown* that their classification is probably best accomplished by placing * Bull. Amer. Mus. Nat. Hist, June, 1899, p. 146. 286 Wortman—Studies of Hocene Mammalia in the them as a subfamily of the Mesonychide. The differences which distinguish them from the Mesonychine are not great, and consist in the deep, heavy lower jaws, with powerful sym- physis, together with the much wider, more typically trituber- cular upper molars, as well as the more subequal size and normal arrangement of the cusps of the trigon of the lower molars. While the Trisodont division became extinct at the close of the Torrejon,in Dissacus we have the beginning of the AZesonyx line, which continued through the whole of the Eocene, being represented in the several stages by a number of species and genera, and towards the close of its career, as we shall pres- ently see, developed a limb structure almost equaling, if not actually surpassing, that of the modern dogs in point of special- ization for a running habit. The connections between these two extremes are close, the interval being completely bridged by the Wasatch Pachyena. In Dissacus, the feet are rela- tively short and stout ; they are pentadactyle and the toes were spreading, a condition which was gradually modified into the four-toed, compressed, elongated feet of the cursorial J/esonyzx. I define the known genera of the Mesonychine as follows: I. Digits, 5-5; humerus with an entepicondylar foramen. a. Internal cusps of inferior molars distinct ; posterior external cusps of superior molars smaller than anterior; M.3, Pm.4__-.Dissacus. b. Internal cusps of inferior molars vestigial ; external cusps of superior molars equal ; Ms Pm 3 es gare te eee Pachycena. ce. Internal cusps of inferior molars vestigial ; external cusps of superior molars equal ; M.2, Pm.2 ---- -..-..-_--2+..--.-.. Hapa If. Digits 4-4; humerus without entepicondylar foramen. a Pige s Eae e es eee a Ma eal Aha De acl 8 les Maes Me Dromocyon. Be Me ee ee INS ea eee Mesonyx. Subfamily Mesonichine. Hurpagolestes macrocephalus gen. et sp. nov. The remains upon which this genus and species are founded consist of the greater portion of a skull, together with a com- plete humerus of the right side, a distal end of a femur, and a centrum of an axis, all belonging to one individual. With the exception of considerable vertical crushing, the anterior portion of the skull is well preserved, including most of the teeth. The posterior part of both mandibular rami are present, with Marsh Collection, Peabody Museum. 287 most of the teeth 7m sttu. These remains indicate an animal somewhat larger than Mesonyx uintensis of the Uppermost Kocene stage, which has hitherto been considered the largest Carnivore of the Eocene. The skull exceeds that of a full- grown Grizzly Bear, but the body was not as large in propor- tion. The principal generic characters have already been stated in the foregoing analytical table. Those which distinguish it most sharply from the contemporary Dromocyon and Mesonyx are seen in the reduced number of superior premolars and the presence of an entepicondylar foramen in the humerus. es of FA eclitn i at x i id : " AR ‘ (CV. x, ry ie i Me Sj . u ia Fe Week we j 2 i ? y i 'y E ¢ ay ee " ’ ent ran 5 ‘ ' i jak Am. Jour. Sci., Vol. XII, 1901. Plate Ill. EXPLANATION OF PLATE III. Skull of Dromocyon vorax Marsh; posterior view; three-fourths natural size. (Type.) ns, nuchal spine; po, postorbital process ; pop, paroccipital, process ; fm, foramen magnum ; oc, occipital condyle: smf, stylomastoid foramen. Am. Jour. Sci., Vol.. XII, 1901. Plate IV. EXPLANATION OF PLATE IV. Under view of the skull of Dromocyon vorax Marsh; one-half natural size. (Type.) apf, anterior palatine foramen ; ppf, posterior palatine foramen; mt, mas- seteric tubercle ; as, alisphenoid canal; fo, foramen ovale; flm, foramen lacerum medius; gf, glenoid fossa; eam, external auditory meatus ; m, mastoid ; pp, paroccipital process; flp, foramen lacerum posterius ; oc, occipital condyle; cf, condyloid foramen; ty, tympanic; frm, foramen rotundum ; sf, sphenoidal fissure ; op, optic foramen. Wood—Crinoid from the Hamilton of Charlestown, Ind. 297 Art. XXXI—A new Crinoid from the Hamilton of Charles- town, Indiana; by EtvirA Woop. With Plate V. THE following paper was prepared in the laboratory of the Geological Department of the Massachusetts Institute of Technology: the specimen having been furnished for study through the courtesy of that institution. I take pleasure also in expressing my indebtedness to Prof. William H. Niles, head of the Geological Department, for the opportunity thus offered. The specimen which serves as the type of the species described below, although deprived of the free arms and the eolumn, is exceptionally well preserved, showing the more delicate structural features of the test. It is now in the collec- tion of the Massachusetts Institute of Technology. Genneocrinus carinatus, sp. nov. Calyx sub-globose. Arm regions prominent with deep fur- rows between them. Surface ornamented by delicate, sharply elevated carine and acute spines. Basals three, pentagonal. The five hexagonal radials fol- lowed by 2X5 costals. First and second costals slightly, if at all, smaller than the radials. Distichals 1x10. Of the pri- mary palmars immediately following the distichals, that nearest the median line of the ray is an axillary plate bearing 3 x 20 secondary palmars; the outer is followed by two additional primary palmars, making 1x10 on the inner, and 3x10 on the outer side of the ray, or forty primary palmars in all. Interpalmars three, following one another in vertical succes- sion. The third, or most distal, primary and secondary palraar in each row gives rise to an arm making six arm bases in each ray. The free arms are not preserved. First interbrachial plate followed by two in the second and three in the third row. Plates of the fourth row are variable in number. The suc- ceeding plates merge into those of the tegmen. The heptagonal primary anal of the posterior area is suc- ceeded by three plates and these by four with a minute tri- angular plate intercalated between the two rows. Beyond this the plates are irregular in shape and difficult to distinguish, but there are probably five plates in the fourth and three in the fifth row, as represented in the diagram (fig. 1). Tegmen moderately elevated, made up of small irregularly arranged plates. Ambulacral regions convex and separated by deep sulci extending half way to the summit of the tegmen. Just in front of the arms in line with the center of the ray is a strong spine 3 to4™"™ long. Immediately surrounding this Am. Jour. Sci.—Fourty Series, Vou. XIJ, No. 70.—Ocrtoper, 1901. 298 Wood—Crinoid from the Hamnvilton of Charlestown, Ind. spine is a variable number of plates which bear low spines or nodes. Other plates of the tegmen appear to have been smooth. Anus located half way between the summit and the periphery, its plates not preserved. Base of the calyx flat, a feature due to a thickening of the basals. This thickening extends over about two-thirds of the surface of each basal and is produced in a thin rim whose margin is parallel with that of the plate. Diagram to show arrangement of calyx plates of Genneocrinus carinatus, sp. nov. The lower half of each radial and the primary anal plate are ornamented by a projecting crescentic ridge thickened at the center and dying out or terminated by a spine at the sides. From the outer curve of this ridge carine pass to the upper sides and angles of the radial, where they meet similar caringe of the higher brachials. The brachials and iaterbrachials bear strong carine which pass from the center to each side of the plate, and others, less prominent, directed toward but not reaching the angles. | Column circular, occupying one-half the area of the basals. Axial canal pentalobate and having a diameter one-third that. of the column. | Observations.—This species is remarkable for its elaborate rand delicate surface ornamentation, the thin carine rising at right angles to the surface and sometimes a millimeter or more in height. Wood— Orinoid from the Hamilton of Charlestown, Ind. 299 The whorl of erescentic ridges on the radials when viewed from below bears some resemblance to a six-petaled flower. Each of three alternate petals has a small erect spine at the center. The carina following the median line of the ray and its branches is stronger than the others and increases in promi- nence from the radials to the arm bases. This carina is straight to a point above the middle of the second costal, where it divides, each branch crossing a distichal and branching again below the upper angle of the plate. A third branching takes place on the inner primary palmar. Spines are an important feature of the surface ornamenta- tion. In addition to those already mentioned there are often minute spines in the angle formed by the finer and coarser earine of the brachials and interbrachials. These are some- times elongated, showing a tendency to form a new carina in this position. The second costal bears a very small spine in the angle formed by the branching of the strong carina, and two similar spines are present on the inner side of the dis- -tichal below the branching and in line with the straight por- tion of the carina. Three much larger spines, decreasing in size upward, are present on the interpalmars, one at the center of each plate. The upper plates of the calyx are somewhat variable in shape. The second costals of the postero-lateral rays are hep- tagonal, those of the antero-lateral rays pentagonal and the anterior second costal is hexagonal, having three edges on one side, above the base, and two on the other; that is, one repre- sents the heptagonal and the other the pentagonal type of plate. The distichals also vary in shape in the different rays. Those resting upon heptagonal costals have six and those on pentagonal costals seven sides. The distichal on one side of the anterior ray is thus heptagonal and that on the other side hexagonal, corresponding with the irregularity of the adjacent costal. The primary palmars resting upon the distichals are heptagonal on the inner and pentagonal on the outer side of the ray. The first interpalmar is an inverted pentagonal plate, the second hexagonal and the third obscurely pentagonal. The first interbrachial is in each case regularly hexagonal, but the interbrachials of the second row are six or seven-sided according to the heptagonal or pentagonal shape of the adjoin- ing second costal. The middle plate of the third row is hexa- gonal, wider at the lower than at the upper end. The other two plates of this row are hexagonal when the nearest dis- tichal has the same number of sides and five-sided when the latter is a heptagonal plate. 300 Wood —Crinoid from the Hamiiton of Charlestown, Ind. The plates of the posterior area are irregular in shape and have five, six, or seven sides. Of the three plates resting upon the primary anal two are hexagonal and one pentagonal. The plates of the tegmen do not vary greatly in size, the orals being only slightly larger than the others. Formation and tocality. This species is found in lime- stones of the Hamilton Group at Charlestown, Indiana. The species here described is closely related to Actinocrinus (Genneocrinus) cornigerus Lyon and Casseday, but it differs from the latter in the presence of three instead of two inter- palmars, the absence of a central spine on the tegmen and of crescentic ridges on the radials. According to the original description of G. cornigerus the plates of the posterior + area resemble those of the Actinocrinide, that is one hexagonal primary anal followed by two plates ; but this species is regarded by Wachsmuth and Springer as synonymous with G. kentuckiensis, in which the primary anal is heptagonal and bears three plates as in G. carinatus. G. kentuckiensis differs from the latter species in the number of distichals which is 210 instead of 1x 10, and in the arrangement of the palmars and interpalmars the axillary distichals supporting 24 palmars in each ray, and the three interpalmars being in two rows instead of one. From both G. kentuckiensis and G. eucharis the present species differs in the nearly equal size of second costals and radials, in the number and arrangement of the higher plates of the posterior area, and in several other important respects. Actinocrinus asper Lyon bears an angular ridge on the lower portion of the radials which under favorable conditions of preservation might resemble the petal-like ornamentation of the radials in Gennewocrinus carinatus, but the basal plates are not thickened as in the latter species, the costals are very small in comparison with the radials, and each plate of the tegmen bears a strong node. Genneocrinus carimatus appears, therefore, to possess an association of well marked and peculiar features which entitle it to be considered a distinet species. Geological Department, Massachusetts Institute of Technology. THE HELIOTYPE PRINTING CO., BOSTON. ELVIRA WOOD DEL. GENN AZOCRINUS CARINATUS. b VENTRAL DISK. C DORSAL VIEW. @ POSTERIOR SIDE. Browning— Estimation of Cesium and Rubidium, ete. 301 Art. XX XII.—-On the Estimation of Cesium and Rubidium as the Acid Sulphates, and of Potassium and Sodium as the Pyrosulphates ; by Puttie E. BROWNING. [Contributions from the Kent Chemical Laboratory of Yale University—CIL | BUNSEN* is authority for the statement that the acid sul- phate of rubidium does not lose sulphuric acid at a heat approaching redness. It is statedt in the literature that the acid sulphates of cesium and rubidium when subjected to a low red heat pass into the form of the pyrosulphates. R. Webert found that by treating the dry sulphates of potassium, cesium, rubidium and thallium with sulphuric anhydride in a closed tube and heating on a water bath two layers separated. In the lower layer he obtained crystalline bodies which proved to have the constitution R,O-8SO,. On strong heating he obtained from these substances bodies of the form R,O-28O, and finally R,O-SO,. He also notes that in the case of the cesium salt the removal of the excess of sulphuric anhydride was attended with greater difficulty. Baum§ states that the pyrosulphates of the alkalies may be obtained by heating the acid sulphates under atmospheric pres- sure at low redness, or under diminished pressure at a tem- perature between 260-320° C. In a recent paper| from this laboratory I have shown that thallium may be estimated as the acid sulphate by evaporating a thallous salt in solution with an excess of sulphuric acid and bringing the residue to a constant weight at a temperature of about 250° ©. The similarity which thallium bears in some of its combinations to the alkaline metals suggested the study of the sulphates of these elements under the same general condi- tions of procedure. My first experiments were made with a pure cesium salt as follows: A weighed amount of the nitrate was placed in a previously weighed platinum crucible and treated with an excess of sulphuric acid. The crucible was then placed upon a steam bath until the water and nitric acid were largely expelled and then removed to a radiator, consisting of a porce- lain crucible fitted with a pipe-stem triangle sO arranged that the bottom of the platinum crucible would be about ‘midway between the top and bottom of the porcelain crucible. This improvised radiator was set in an iron ring and a thermometer placed so that the mercury bulb would be on a level with the * Ann. Chem. (Liebig), exix, 110. + Graham. Otto Lehrbuch d. Chem., iii, 278, 269. t Ber. Dtsch. Chem. Ges., xvii, 2497. S$ Ber. Dtsch. Chem Ges., xx, 752. | This Journal, ix, 137, 1900. — TO Ore CW bo o/2) 10 302 Browning—Estimation of Cesium and Rubidium bottom and close to the side of the platinum crucible. An ordinary Bunsen burner served as the source of heat and the temperature was kept so far as possible between 250° ©. and 270°C. After the fuming attending the removal of the large excess of sulphuric acid ceased, the crucible and contents were removed to a desiccator, and after being allowed to cool, weighed. This process of heating was continued for half- hour periods until the weights were constant. The results shown in Table I were obtained by this method of treatment. In experiments (1), (4) and (9) it will be notieed that the weights were constant somewhat above the conditions of the acid sulphate, a fact which would go to show a tendency on the part of the czesium salt to hold an excess of sulphuric acid over the amount necessary to form the ordinary acid sulphate. The results show that by regulating the heat at a temperature TABLE I. CsHSO, Ist 2d Error Cs2SO4 Error CsNO3 ealeu- coustant constant on caleu- Cs2SO4 on taken. lated. weight. weight. CsHSOx,. lated. found. Cs2SQ,. erm. erm. erm, orm, orm. erm. orm. erm. 0°1706 0°20138 0°2054 0°20620 0°0007 + 071706 0°2013 0°2010 0°0003 — O10382) OA Dee 1201 0°0016— 071032 071217 1252 0°1222 .0:0005-+ —0:0961 00948) 0-012 — 0°1218 0°14387 0°1458 0°0021+ 0°1180 O°1118 0:00 — 0°1214 06:1435 0°1430 0°0005— 071214 -0°1485 0°1422 0°0013— 071150: 071356 -0°1350 0°0026 — On056. 0:1245 071272, 01243-00003 O:1056— 071245, 01252 0:0007 + between 250° C. and 270° C. cesium may be brought with a fair degree of certainty to the condition of the acid sulphate. As a check upon the results the acid sulphate was, in a few cases, treated with a little ammonium hydroxide, the excess of this was removed upon a steam bath and the neutral sulphate was obtained by ignition at a red heat to a constant weight. These determinations agreed fairly well with the theory. The same ‘procedure was followed with rubidium, a pure rubidium chloride having been chosen as the starting point. The results are given in Table Ii. No tendency was observed on the part of this element to hold sulphuric acid in excess of the amount necessary for the formation of the acid sulphate. When the same method of treatment was applied to the sodium and potassium salts, pure chlorides being used as the starting point, a somewhat different result was obtained, in that the weight of the final product appeared to indicate the formation of the pyrosulphate. | as the Acid Sulphates, ete. 303 The results given in Tables III and IV, in which the sodium and potassium: salts are calculated as pyrosulphates, are sufii- ciently satisfactory for purposes of quantitative estimation. As in the ease of the cxsium and rubidium salts, a number of determinations as the neutral sulphates were made by igni- tion of the sodium and potassium pyrosulphates with results which are recorded. In Table V two determinations are recorded, in one of which the cesium and rubidium salts were treated together, and in the other the sodium and potassium salts. An application of this same general method TABLE II. RbHSO, Rb2SO, RbCl ealeu- RbHSO, ealcu- Rb.SO, taken. . lated. found. Error. lated. found. Error. erm. orm. erm. orm. CAIN 2) CHRIS = orm. doe O21959 -Or1889> O-1878 0:0011— 2 071212 O:1829 0°1840 O:0011+ 0°1460 0°1460 0:°0000+ 38 071230 0°1856 0°1850 0:'0006— Aa 12 Ss0 Oooo, 0 l8os 070002 7 0:1357 0°1350°. 0:°0007— OG Ome OL2 +30 O:2416 O:00L4— O1777 O1772 0:0005— 6 0°1860 0°2052 0°2032 0:0020— 0°1501 0:°1490 0:0011— TABLE IEI. K.S,0, K.SO,4 KCI. - ealeu- =§-_KeSe0, calcue Kos0u taken. lated. found Error. lated. found. Krror. erm. grm. erm. erm. orm, erm. erm. 1 00-2172 0°-3704 0-3698 0:0006— ZOD ome O- 290900 0:2886..7.0°0023——" 7 O:1993' - O21972. 0:0021 — a0 11920-20382" 0°2022 o:00l0— 0°1393 O'°r381 0°0012— Poort 0-830 021823 " 0:0007— 5 01096 0°1868 0°1860 0:0008— to a lithium salt gave no evidence of the existence of a stable acid sulphate or pyrosulphate. The results may be semmed up as follows: Cesium and rubidium salts of volatile acids when treated with sulphuric acid in excess and brought to a constant weight at a temperature between 250° C. and 270° C. form acid salts of the type RHSO, and the neutral salts of the type R,SO, on ignition. Some tendeney of the cesium salt to hold more ‘sulphuric acid than corresponds to the formation of the acid sulphate RHSO, was apparent at temperatures between 250° C. and 270° C., ‘but upon raising the temperatures above 300° C. the loss sis excessive and showed a tendency on the~ part of the acid sulphate, at this temperature, to pass towafd the condition of the pyrosulphate. Sodium and potassium salts when heated under the conditions described give pyro- sulphates of the type R,S,O, which on ignition go into the neutral 304 Browning—LEstimation of Cesium and Rubidium, ete. sulphates of the form R,SO,. Lithium gives neither salts of the type RHSO, nor of the type R,8,O, under the conditions of these exper iments. TABLE IV. Na.S.0,; Na.S0O, NaCl cealcu- Na2S.0, ealeu- NasSO, taken. lated. found. Hrror. lated. found. Error. orm. erm. erm, erm. orm. erm. orm. tT '0°1042 0°19%8 O°1972 0:0006— 012 56. *O-12540 00 — 2 RO TO26s O11 952 Ooo O20000 3° 0°1093 0°9075° 0°2065 0-0010— 0°13828 0:1320° Oo -00es— 4. 0°1402: -0:2662 0°2651 0°001T1— 01703 “021696 40-00G7—— TABuE We RbHSO, + Rb.SQ,+ CsHiSOn Rb Ss On Cs250. Rb:SO7-e RbC1+CsNO; calcu- CsHSO, ealcu- CseSO, taken. lated. found. Error. lated. found. Error. orm. erm. erm. erm. erm. erm. erm. RbCl 0:1428 (1) CsNO, 0°1264 0°3646 0°3666 0:0020+ 0°2749 '0:27525 0 :000s— N2S20, + Na.S0O,z + Rosy Na.S.O07 + ES ON Na.sSO, + NaCl+KCl ealeu- GSH OF, caleu- KeSOu taken. lated, found. Error. lated. found. Error. erm. erm. erm. erm. grm. erm. erm. NaC@ly0:1233 (2) KCl 0°1340 0°4627 0°4630 0°0003-+ 0 38062 -0°3040 0°0023 — March, 1901. C. f. Keyes Provincial Carboniferous Terranes. 305 Art. XXXIII.—TZime Values of Provincial Carboniferous Terranes ; by CHARLES R. KEYEs. THERE have recently* been discussed at length certain remarkable features relating to the development of the Car- boniferous rocks in the province of the Mississippi valley. The enormous thickness of the Carboniferous of the region— probably upwards of 25,000 feet—and the nicety with which it is separable into serial and minor subdivisions tend to make the general geological section of the Continental Interior basin the standard one for America. Peviods ,NorTHERN Section SoutuHern Section TA ieiss Oklahoman) 00 0 DesMoines Fi. - Figé, 1.—Relative Formational Development of the Carboniferous. In describing the various peculiarities of the formations only the stratigraphical relations and features were considered. The importance attached to the marked inequalities in the development of the five series recoguized which was made such a prominent factor and the absence of all reference to the relations of the several series to the time standard, appears to have led to the inference that the time ratios were to be regarded as approximately proportional. It was not the intent to unduly emphasize the stratigraphical disparities at the expense of the time representatives. The latter, for reasons not necessary to specify in detail here, were simply thought to be phases of the subject not to be considered in the article inentioned. * Bull. Geol. Soc. America, vol. xii, 1901, pp. 173-196. 306 C. R. Keyes—Time Values of The stratigraphical features considered are summed up in the diagram above. It represents a meridional ideal section — across the Carboniferous basin of the Western Interior region. The north and south ends are drawn to scale. The white rep- resents the relative formationa! development; and the black the portions now missing. The very unequal development of the several series has given rise to a tendency to misconstrue the taxonomic ranks of the formations. For example, the Des Moines series. has been sometimes given a rank subordinate to that of series; while in the case of the Oklahoman the suggestion has been made to have its rank that of a system. If the stratigraphical values of the series be represented according to their relative developments, they appear about as follows: Giimaienomen ore ss Ls ei Okalvomain® = oe os ee oe eee 2 Missourian tee ee a ee cree 4 Des. Moines.2) 2 a25 1 Arkansan._ oS a es Mississippian... <_.. =, el ees In the general geological scheme, which is dual in character, the time ratios are as important as those represented by the stratigraphy. TFaunal and floral comparisons of the different sections are not possible at this time because exact information is scant. We have, therefore, to rely upon data derived from all the available sources. Assuming that the time occupied in the formation of the Des Moines series, according to the cor- rected valuation hereafter considered, as unity, the factors tending to modify the accepted values ‘indicated by the thick- ness of the several series may be briefly enumerated. The Mississippian series is predominently a limestone forma- tion. Taken alone, its lithological composition would indicate a long time element. As a matter of fact, its formation was probably much more rapid than is usual among limestones. The major member is made up of a crinoidal and shell breccia. The accumulation of its material would thus take place many times faster than the formation of fine-grained, non-fossilifer- ous beds through ordinary precipitation. Another feature indicating that the period during which the Mississippian beds were laid down was not so long as might appear at first glance, is the undoubted oscillation of the sea-floor, which grew quite marked towards the close of the period. In these changes of elevation the uprisings at length predominated over the down- sinkings, finally resulting in the complete emergence of a large part of the province ¢ above sea- level, and closing the Mississip- Provincial Carboniferous Terranes. 307 pian period. Over the entire area now known as the Upper Mississippi valley a prolonged period of subaerial erosion took the place of sedimentation. West of the line now occu- pied by Mississippi river the land border extended southward beyond the latitude of the present Missouri-Arkansas boundary, as lately clearly shown by Marbut.* The full significance of this position of the old shore-line at the close of the Lower Carboniferous is considered elsewhere.t Taken in all its vari- ous aspects, the data at hand indicate that the length of time occupied in the formation of the Mississippian was but. little more than that during which the Des Moines series was laid down. Compared with the Des Moines, the Mississippian should probably be placed at about one and one-half. The enormous development of Coal-measures in central Arkansas has been an anomaly in the stratigraphy of the Car- boniferous. The thickness, estimated by Brannert{ to be over 20,000 feet, is much greater than that of the entire Carbonit- erous elsewhere on the American continent. The conditions permitting this vast accumulation were certainly very local and unusual. ‘hey are briefly stated§ as follows: South of the latitude of the Boston mountains in northern Arkansas, Car- boniferous sedimentation continued on without interruption from the Mississippian; while in the north erosion took place. The sediments from the northern land area, where erosion was going on vigorously, were carried sonthward and dumped off the shore, rapidly building the latter outward. There may have been a great land area in northern Lou- isiana, and probably was. If so, what is now the Arkansas River valley was a broad deep estuary opening out to the west, and the sediments came in from both sides, as well as from the head towards the east. The conditions were then similar to those presented now by the lower Mississippi plain. Only the great embayment opened to the west instead of to the south. The present Arkansas valley, however, has probably been formed through erosion largely, if not entirely, since Tertiary times and by a system of drainage in no way dependent upon the Carbcniferons drainage. When the great uplift of Mis- sourl and Arkansas rose, the northern part, embracing the so-called Ozark isle, and the southern portion, comprising the Ouachita mountains, were made up of resistant limestones, and yielded less quickly to erosion than the central soft shales; and the Arkansas river, which happened, in the old peneplain, to traverse the central part of the uplifted area, was able to main- * Missouri Geol. Sur.. vol. x, 1896, p. 83. + bull. Geol. Soc. America, vol. xii, 1901, p. 173. t This Journal (4), vol. ii, 1896, p. 235. S Bull. Geol. Soc. America, vol. xii, 1901, p. 195. 308 | C. I. Keyes—Time Values of tain its old course. The present uplift, which is due to one general movement, is now apparently divided into two elevated regions separated by a broad valley. While in the south there is this prodigious record of the strata, in the north there is no record at all in sediments. The period of erosion is represented by only a thin irregular plane of unconformity from which alone no time value can be deter- mined. In spite of the enormous thickness of the Arkansan, the time element must have been very much less than would ordinarily be inferred from the figures relating to the vertical interval occupied. The period might have been as extended as the Mississippian; but, all things considered, it would seem much more likely that it was actually considerably more lim- ited than that of the series immediately below. Probably the time ratio would be very nearly the same as for the Des Moines. Although the Des Moines series is so thin over such a large part of the area oceupied by it, it is actually very much more important as a terrane than its relative thickness would indi- cate. North of the Boston mountains the formation is per- haps nowhere over 500 feet in thickness. The beds composing the series were laid down along a shore which was on the whole gradually sinking.* Many oscillations permitted sub- aerial erosion to take place time and again, so that the newly formed sediments were frequently removed or worked over almost as fast as they were formed. It was a period when erosion contended against deposition for supremacy, with a result of one making advances at one time and the other at another. The net result was finally a slight gain for sedimen- tation.t The period was manifestly a long one, for it is believed that in Arkansas no less than 3,500 feet of strata are referable to it. Its time ratio is certainly not very much less than that of the Mississippian. The Missourian series is essentially a marine formation inter- calated between two terranes which were laid down in shallow waters. Its limestones are nowhere the coarse-grained breccias such as occur so frequently in the Mississippian. They are fine-grained and often earthy, and are separated from each other by important beds of shale, which are moreover usually much thicker, reaching measurements of often several hun- dreds of feet.t It is not improbable that part of these beds 1 laid down while some of the nearer shore sediments of the Des Moines series were being carried into still waters.§ * Towa Geol. Sur., vol. ii, 1894, p. 113. + This Journal (3), vol. xh, 1891, p. 273. t+ American Geologist, vol. xxiii, 1899, p. 298. § lowa Geol. Sur., vol. 1i, 1594, p. 162. Provincial Carboniferous Terranes. 309 In point of the time taken for its formation, the Missourian series may be regarding as ranking with the Mississippian. The so-called Permian of the Western Interior basin (Okla- homan and Cimarron, the latter generally known as the Red -Beds) is composed largely of shales and shaly sandstones. Unusual conditions prevailed during the formation of these beds. In the main, the deposits indicate the presence of shal- low waters, in strong contrast to the marine conditions which prevailed previously in the same region. The sediments were laid down largely in more or less closed basins, which may have often had access to the sea, but which finally became altogether dry. The conditions existing were identical with those under which the original Permian beds of Russia were formed.* Nothwithstanding the comparatively great development of the beds of these series, the formation of them was evidently rapid. The time occupied was probably not more than twice as long as that of the Missourian. Careful comparison of all available data indicates that the actual time ratios for all-the series is very nearly equal. All things considered, this appears quite remarkable. The values seem best to accord with the following figures: 7 NOTA Oey yee pea ee 8 il OO MIa eS Wee Be ec 1 WiiSsOUIAMes cep Oe ae 2 ce, ile Wess Vioimies. 235 hn oe oe) 1 PRRRCAIN SAM Been ee kee el tte INSSISSIp Prange 200228 Pee 13 * Journal Geology, vol. vii, 1899, p. 329. 310 J. Trowbridge—Spectra of Hydrogen Art. XXXIV.—The Spectra of Hydrogen and some of its Compounds ; by JOHN TrowsripGr. (With Plate VI.) © In a late paper I expressed the conviction that the so-called line spectrum of hydrogen cannot be considered apart from the spectrum of water vapor; and that one can never be sure that one is observing with a condenser discharge a pure spectrum of hydrogen. [am convinced from further experimentation that this conclusion is correct; and I am also led to the con- clusion that a certain amount of water vapor is essential in all electrical discharges through gases. Just as aqueous vapor seems to play an important role in most chemical reactions, so, it seems to me, its presence in rarified gases contained in ordi- nary glass tubes, enables a dissociation to take place which — determines the strength and character of the electrical dis- charges. , I am led, moreover, to the conclusion that pure hydrogen is a perfect insulator, and that the passage of electricity through a gas depends upon the dissociation of the hydrogen and oxygen, by means of which change in the distribution of energy the gases are made luminous. Before proceeding to an account of my experiments, I will state some ‘of the grounds upon which I base my belief that pure hydrogen is an insula- tor of electricity. V. Schumann, in an important paper, has shown that a column of pure hydrogen at atmospheric pressure transmits ~ the ultra-violet rays as well as the most perfect vacuum he has been able to obtain. Now Maxwell’s electromagnetic theory of ight demands that the space between us and the sun, or in other words the vacuum of space, should bea perfect insulator, otherwise the electromagnetic waves would be completely absorbed and the earth would remain in darkness. This observation of Schumann seems to me one of the most impor- tant in physical science, for it proves, I believe incontestably, that hydrogen cannot be a conductor. Professor Dewar has also shown that liquid hydrogen is an insulator. The experiment sometimes shown, in which a wire, rendered incandescent by a current of electricity and sur- rounded by an atmosphere of carbonic dioxide, is suddenly di- minished in brilliancy by supplanting this atmosphere by one of hydrogen, can be explained, in my opinion, not by the better con- ductibility of hydrogen for heat, but by the increased resistance of platinum due to the occlusion of this gas by platinum. A * Ann. d. Physik, No. 3, 1901, p. 642; this Journal, May, 1901, p. 394. and some of its Compounds. 311 palladium wire increases often as much as fifty per cent. by the occlusion of hydrogen. The increased length of the electric spark in an atmosphere of hydrogen is not due to an increased conductibility, but to a dissociation of water vapor which is analogous to that which takes place in a voltaic cell. These are some of the facts which lead me to believe that hydrogen is an insulator and that the water vapor, therefore, plays a controlling part in the passage of electricity through gases. lam conscious that the conclusions in this paper are somewhat radical; and J have, therefore, worked assiduously, during the past three years, to test them in every way which my mind suggested, for it is not probable that many investi- gators have at present twenty thousand storage cells, which would enable them to repeat my experiments. The strength of currents and the voltage I have employed have certainly reached the limits of glass tubes to withstand such powerful discharges. The form of tube figured in my previous article* is the only one which | have found capable of withstanding steady currents of one-tenth to one-fifth of an ampere, and instantaneous condenser discharges of many hundred amperes. The great advantage of the use of a storage battery over the employment of a Ruhmkorf coil; in the study of the ionization and molinization of gases, is now generally recognized. This advantage is forcibly seen in the first experiment which [I shall bring forward in support of my view of the importance of the role played by water vapor in the passage of electricity through gases. A wide tube of the type I have referred to, the narrow portion being approximately one centineter, was provided with massive copper-ring electrodes, oue inch in outside diam- eter and one-eighth of an inch thick, which were heavily electroplated with copper in order to avoid the impuri- ties of commercial copper. The glass tubes were then exhausted and filled with hydrogen made by the electrolysis of distilled water and phosphoric pentoxide. The gas was sent through tubes filled with caustic potash and many drying tubes filled with phosphoric pentoxide. The gas was kept in the drying tubes many hours; and its flow was delayed by partitions of glass wool. More than a litre of the gas was used in the process of flushing out the spectrum tubes; so that the entire pump and connecting tubes were for several hours presumably filled with hydrogen gas. When the tubes, having been exhausted to the most lumi- nous stage, were excited by a condenser discharge and were examined by a straight-vision spectroscope, the ordinary four- * This Journal, x, 222, 1900. 312 J. Trowbridge—Spectra of Hydrogen : line. spectrum of hydrogen alone seemed to be present. When, however, the invisible portion in the violet was photo- — sraphed, the bands at wave lengths 3900 and 4315 were invariably present, unless the tube: had been maintained, during the process of filling, to a temperature of more than 350° ©. After such a process of heating. the spectrum became that represented in Plate VI, fig. 2, while before heating it was that shown in Plate VI, fig. 1. In both plates the normal spectrum is above the gaseous spectra. Further toward the ultra-violet, under all conditions, there were also nitrogen bands. Long heat- ing diminished the strength of these bands. This ie of experimentation shows that mere eye inspection of glass tubes filled with rarified gases is generally fallacious; we might con- clude from this eye study that the presence alone of the four lines of hydrogen denoted that we had this gas in a pure state; whereas the photography of the invisible portion would show: that this was far from the truth. When the glass tubes filled with rarified hydrogen were submitted to the influence of a steady current of electricity, it was found that perfectly pure copper was deposited in a lus- trous state on the glass walls of the tube which surrounded the negative terminal, while an olive-green oxide of copper covered the walls around the positive terminal. The same tube was excited by a Ruhmkorf coil, and no difference could be detected in the deposits around both terminals; they were both rusty green with here and there, it may be, streaks of pure copper. The mirrors produced by astrong steady current at the negative terminal were very lustrous and showed no trace of an oxide of copper. It was evident that the current had dissociated water vapor in the presence of an excess of hydrogen, and had reduced the copper at the negative pole and had set free oxygen at the positive pole, which had, in turn, combined with copper. The rarified gases thus acted like a voltaic cell. When we examine the photograph of the discharge repre- sented on Plate VI we see an interesting exhibition of ionization and molinization. The hydrocarbon bands at wave length 43815 show a series gradually decreasing in length of waves, while another band beginning at wave length 3900, due probably to water vapor, shows a series increasing in length of waves. It would seem that the carbon in one case endeavored to throw off the hydrogen from the hydrocarbon molecule; and in the other case the hydrogen became loaded with oxygen molecules. This to and fro ionization and molinization continues until the oxide of. copper at the positive terminal has taken up a large share of the oxygen of the water vapor present. There is thus a critical point in the tube at which a sudden increase of resist- He EST Oe x Am Jour. Sci., Vol and some of its Compounds. 313 ance takes place. It is possible to exhaust glass tubes to such a degree by the mere passage of a strong steady current that X-rays begin to manifest themselves. When a similar tube, filled with hydrogen with great care and prepared by long heating at a temperature a little below 500° C., is submitted to electrical discharges, the water vapor bands become far less pronounced; and the hydrocarbon band at wave length 4315 entirely disappears, while the light of the tube greatly diminishes in brilliancy. The hydrocarbon or eyanogen band at wave length 3884 is present in all the tubes I have employed; and with whatever gas is submitted to these strong discharges. Strong heating does not cause it to disap- pear, and it seems to be due to carbonaceous matter introduced into the tubes in the process of blowing for I cannot trace it to impurities coming from the pump. Professor Hartley, in a late communication in Nature, has called attention to the con- stant presence of hydrocarbon spectra in Geissler tubes. Ata -Jater point in this paper I shall return to a further study of these spectra, due to the combination of hydrogen and nitrogen with carbon. At present I desire to dwell upon the point I wish to make: that all discharges in rarified gases contained in glass vessels are conditioned by the amount of water vapor present; and that a steady current passes through a gas at comparatively low pressure much in the same manner that it does through an electrolyte. In an article on the production of the X-rays by a steady battery current, | dwelt upon the phenomena presented in highly rarified tubes which represent, to my mind, the disso- ciation of water vapor; and [| will refer again at this point to the phenomena already described. According to this hypothe- sis the rarified water vapor is dissociated at the surface of the antivathode, which is thus greatly heated; the occluded hydro- gen plays a part in this phenomenon. The behavior of large aluminum electrodes in glass vessels filled with ammonia gas is also an interesting example of the ' dissociation of water vapor. ‘The gas was obtained by heating ammonium chloride, passing it over freshly slaked lime and through drying tubes filled with phosphoric pentoxide. A sufficient amount of ammonia gas was thus obtained for the purposes of spectrum analysis. When a large condenser, charged to a difference of potential of twenty thousand volts, was discharged through the rarified ammonia gas—there being practically no self-induction in the circuit, and the main effect, therefore, was due to the pilot discharge—the light of the tube changed from a brilliant white to a rosy red ; and eye inspection with a straight-vision spectro- Am. Jour. Sci1.—FourtH Series, Vou. XII, No. 70.—OoctoprEr, 1901. 22 314 J. Trowbridge—Spectra of Hydrogen scope showed only the line spectrum of hydrogen. One would conclude from this inspection alone that there was pure hydro- gen in the tube. One might also surmise that the oxygen of the water vapor always present on the walls of the glass vessel had combined with the aluminum terminals, setting free the hydrogen which then carried the current. The pressure, how- ever, in the tube increased, and therefore gas must have come from the aluminum. In the exhaustion of X-ray tubes pro- vided with aluminum cathodes much time and long treatment with condenser discharges is necessary to drive ont the gases from this cathode. The principal gas seems, from the experi- ment with ammonia gas, to be oxygen. Thesame phenomenon is seen in tubes supplied with magnesium terminals, but to a much less extent. It is not seen when the terminals are of copper, iron, silver, platinum or carbon. This behavior of aluminum toward oxygen is very suggestive in regard to the ready passage of the X-rays through this metal. I have been unable, with the conditions under which I have worked, namely, the use of very powerful discharges, to obtain the spectra of hydrogen apart from water vapor and hydro- carbons. The study, therefore, of the spectra of hydrogen compels one to carefully study the spectra of the hydrocarbons and that of cyanogen ; for I am forced to the conclusion that the combination of hydrogen with oxygen is a controlling fae- tor in all discharges through rarified gases. ‘The following isa _ preliminary study of some of these compounds, which is added at this stage of my inquiry to illustrate this theory. When various gases are put in tubes provided with carbon electrodes and these tubes are exhausted to a pressure of from one to two millimeters the resultant spectra are very similar. The follow- ing gases have been studied in the neighborhood of the great HH lines of the solar spectrum : Hydrogen Oxygen Nitrogen In the hydrogen tube the only lines that appeared in the region from 4320 to 3200 were: 4268 very intense 3922 faint. The tube was very thoroughly heated while it was being ex- hausted. The above mentioned lines do not generally appear with hydrogen in tubes with metallic electrodes ; but witha tube with platinum electrodes, filled with hydrogen, and heated for two hours during exhaustion at a temperature of and some of its Compounds. 315 300° C., the same two lines appeared, and in addition the fol- lowing very faint lines: 3871 3886 These same four lines also appear in carbon tubes when filled with oxygen. In addition, the following lines are present : 3936 3971 4077 All these lines appear in the nitrogen tube, and in addition: 3883 3876 3868 3856 3849 3841 None of these lines appear in tubes with metallic electrodes filled with nitrogen; and they are therefore not nitrogen lines. - All the lines in the nitrogen tube are more intense than those in the oxygen tube, and it is possible that with a longer exposure these additional lines would come out in the oxygen. To study the effect of the carbon terminals, the following gaseous compounds were put into wide tubes provided with copper terminals and rendered luminons by condenser dis- charges : Cyanogen Carbon monoxide Carbon dioxide Acetylene. Cyanogen was prepared by heating mercuric cyanide and pass- ing the gas over sulphur to remove any traces of mercury vapor. Carbon monoxide was prepared by heating potassium oxalate with concentrated sulphuric acid and passing the gas over potassium hydroxide, and collecting over water. Carbon dioxide was prepared by treating potassium carbonate with dilute sulphuric acid and:collecting over water. All of these gases were allowed to remain in contact with pentoxide before introduction into the tubes. With acetylene, carbon monoxide, and carbon dioxide, con- denser discharges being employed, the results appear to be identical with those obtained with hydrogen in the tube with carbon electrodes. With cyanogen the same lines appear; and in addition, the bands which are characteristic of these gases 316 J. Trowbridge—Spectra of Hydrogen with continuous currents, which will be described later. In general, with condenser discharges, all these spectra are the same ; the differences which occasionally appear may be due to changes i in pressure, time of exposure, etc. The line 4268 in all these cases is by far the most prominent line present in the region studied, and may be taken as charac- teristic of hydrogen, oxygen, and nitrogen in tubes with ear- bon terminals, and of gaseous carbon compounds in tubes with metallic terminals. This line does not usually appear in hydrogen in tubes with metallic terminals; occasionally it appears very faintly. It appears, however, very strongly ina tube provided with platinum terminals which is filled with hydrogen and heated for two hours during exhaustion to a temperature of 350° C. The spectrum, in this case, appears to be identical with the spectrum of hydrogen in a tube provided with carbon terminals. Were it not for this fact, it would seem as if this line were due to carbon in some form; but even with this fact, it is possible that there was enough foreign e¢ar- bonaceous matter present in the platinum tube to produce the result noted. Eder and Valenta* find, among others, the following lines in the spectrum of an induction coil between carbon terminals: 4268 3921 Apparently these are the same lines found in the tubes pro- vided with carbon electrodes; and also in tubes with metallic electrodes which are filled with carbon compounds. Observed visually, with a straight-vision spectroscope, all the above cases appear identical. When, for instance, hydrogen was put into the tube with carbon terminals and submitted to discharges from an induction coil, at first the line spectrum of hydrogen appeared. After the discharge had passed for some titne, this gradually changed into the characteristic band spectrum of ear- bon. To the eye alone the change was equally noticeable; the light being, at first, reddish, and then changing to a white. Similar changes were noticed when nitrogen and oxygen were used in the carbon tube. It seems to me that the following conclusions can be drawn provisionally from the above: When various elementary gases are introduced into wide tubes with carbon electrodes and exhausted to a pressure of 1-2™, and submitted to condenser discharges, compounds of carbon with the various gases are formed. With nitrogen this compound is probably cyanogen ; with hydrogen acetylene; but when a photograph of the spec- trum in each case is taken, we get, not the spectrum of the * Beiblatter, xviii, 1894, p. 753. and some of its Compounds. 317 compound, nor that of the elementary gas, but a carbon spec- trum. This, however, does not mean that we get the line spectrum of elementary carbon; for it is certain that there is water vapor present in the tubes, notwithstanding the temper- ature to which it has been subjected. The carbon may then unite with the oxygen of the water vapor, forming either carbon monoxide or dioxide; the hydrogen being oceluded by the terminals or the glass walls. Just as the spectra of gaseous carbon compounds in wide tubes with metallic terminals appear identical with the spectra of elementary gases in tubes with carbon electrodes submitted to condenser discharges, so we should expect that the spectra produced would be the same in the two cases. This is found to be true. The general appearance of the photographs obtained with continuous currents is very different from those obtained with condenser discharges. In the former case there | isa marked band appearance in addition to a line spectrum. The most prominent of these bands in the region studied is the one beginning at 3884. With the dispersion used this band consists of five prominent lines crowding together toward the ultra-violet. A somewhat similar band, apparently of six lines, begins at 4216. Another band, rather faint, consisting of a large number of fine lines shading off toward the ultra-violet, begins at 4126. The bands beginning at 4216 and 4884 appear to be the same as the bands which Kayser and Runge* designate respectively as the second and third cyanogen bands in the are spectrum of carbon in air. Besides these bands a number of single lines appear which are common to all the gases. Among these, the most promi- nent are: 3652 4048 4080 4360 In the tube filled with cyanogen many of the same lines that appear in tubes with metallic terminals filled with nitrogen ‘are present besides the above. ‘This is true to a certain extent of the other gases; and it is not surprising, since it is to be expected that some atmospheric air is aL present as an impurity. The conclusions to be drawn from these experiments with steady currents are similar to those we have deduced from condenser discharges. When elementary gases are introduced into tubes with carbon terminals and exhausted to a pressure * Abhandlungen der Akademie der Wissensch. zu Berlin, 1889. 318” J. Trowbridge—Spectra of Hydrogen, ete. of 1-2™™, and are snbmitted to continuous currents, we obtain the spectrum of carbon or some compound of carbon. From the above results very little information can be obtained as to what this compound is. The same spectrum is obtained what- ever gas 1s introduced into the tube; and moreover, this is the same spectrum which is given by gaseous carbon compounds in tubes with metallic terminals. What, then, are the conclusions to be drawn from the pres- ent stage of this investigation with gases submitted to powerful electric discharges? It seems to me they are as follows: 1. Hydrogen is an insulator. 2. The passage of electricity through hydrogen, nitrogen, oxygen and their gaseous compounds, is conditioned by the water vapor present. 3. The dissociation of this water vapor in the case of tubes’ filled apparently with pure hydrogen, under the effect of a strong steady current of electricity, shows an electrolytic action closely analogous to that of the voltaic cell. In the case of electrolytic copper: terminals in an atmosphere of hydrogen, pure copper is deposited from the negative terminal; and a suboxide of copper at the positive terminal. 4. Under the effect of powerful condenser discharges oxygen is an free from commercial aluminum and magnesium. Certain carbon bands are always present in glass tubes filled with hydrogen, nitrogen, oxygen and ammonia gas not- — withstanding the greatest care which may be taken in submit- ting them, during “the process of exhausting, to a high tempera- ture, when powerful discharges are employed. 6. The brilliancy of the light of tubes filled with hydrogen diminishes as the process of the dissociation of water vapor . goes on and the resistance of the tube increases. It is possible to raise such a tube to the X-ray stage from a pressure of 1-2™™" merely by the application of a strong steady current. 7. The X-rays excited by the application of a steady ecur- rent are due to the radiations set up by the dissociation of highly rarified water vapor. Jefferson Physical Laboratory, Harvard University. Chemistry and Physics. 319 SOHN PLE LCT INP Ei blhG YN Ch: I. CHEMISTRY AND PHysiIcs. 1. Induced Rudio-activity produced by Salts of Radium.— An interesting series of experiments has been made by P. CurtE and A. DEBIERNE, which shows that radio-activity is trans- mitted in the air through short distances and induces activity in other substances. This induced activity 1s much more intense when it is produced in completely closed vessels than when the action takes place in the open air. A radiferous substance was placed in a small open bulb in a closed vessel, and plates of various substances (lead, copper, aluminium, glass, ebonite, card, parafiin, all of which appear to be acted on in about the same manner) were placed in various positions in the vessel, and in all positions, even behind a lead screen, they became active to about the same extent after a day’s exposure, provided that the air of the vessel had free access to their surfaces. A plate resting on the bottom of the vessel became active only on its exposed sur- face. With very active barium chloride (radium), plates exposed for several days in this way attained an activity 8000 times greater than a plate of metallic uranium of the same dimensions. When exposed to the free air, the plates lose the greater part of their activity in a day, but the activity disappears much more slowly when the plates are left in the closed vessel after the active substance has been removed. If the experiment is made with the bulb containing the active substance completely closed, no induced activity is obtained. Other experiments showed that this induction is rapidly carried through capillary tubes connect- ing small closed spaces. It was found that the action is pro- gressive, finally reaching a limit which depends upon the activity of the active body. ‘The action is more rapid the smaller the vessel containing the bodies. These phenomena have been observed with various salts of active barium, such as the chloride, sulphate, and carbonate. Salts of actinium also produce induced activity, but, on the other hand, salts of polonium, even when very active, produce no such effect. The authors believe that this circumstance may have a connection with the fact that polonium does not emit rays that are deviated by the magnetic field. In a subsequent communication the same authors state that water can be rendered radio-active either by distilling it from a solution of radium chloride in an air-tight vessel, by placing a dish of water in a closed space containing a solution of a radium salt, or also by enclosing a solution of a radium salt in a her- metically closed celluloid capsule and plunging the latter into distilled water in a closed vessel. In the last case, no trace of the salt passes into the water, but the activity of the solution is to a great extent communicated to it. This induced activity can- 320 Scientific Intelligence. not be transmitted by air through a wall of dry celluloid, but it passes easily if the wall is wet with a drop of water. Water can be rendered very strongly active, but when kept in a sealed tube it loses the greater part of its activity in a few days. When kept in an open vessel the water loses its activity much more rapidly, the loss being the more rapid the greater the surface exposed to the air. Solutions of radium salts in open vessels behave in the same way as the water just mentioned, but in this case the loss of activity is not absolute, for if such a solution is put into a sealed tube it gradually acquires its original activity in the course of ten days or so. The authors look upon radio- activity as being analogous to heat in being dissipated by radia- tion and by conduction. In the latter case it passes through gases and liquids, but not through solids. If a solid radio-active body is left free to the air, its activity does not diminish sensi- bly, and the authors have shown that a solution of a salt pro- duces much more intense phenomena of induced radio-activity (twenty times greater) than the solid salt itself. When the salt has been in solution several days, the radio-active energy is divided between the salt and the water, and if the latter be then distilled off, it contains a great part of the activity, while the solid salt is much less active (ten or fifteen times, for example) than before solution. When left to itself the solid salt regains, little by little, its original activity. The communication of the activity of a radium salt to its water of solution is very slow, and equilibrium is obtained only after about ten days.— Comptes Kendus, cxxxii, No. 9; cxxxii, No. 5. H. L. W. . 2, A New Method of Quantitative Analysis.—R. W.THatTcHER has devised a method for determining the weights of precipitates without separating them from the liquid from which they are precipitated. The proposed method consists in determining the weight of a measured volume of the precipitate and mother- liquor, and then determining the specific gravity of the mother- liquor alone. From these data, if the specitic gravity of the precipitate is known, the weight of the latter may be calculated by the formula, d(a—bd') d—d ” where a is the total weight, 6 the total volume, @ the specific gravity of the precipitate, and d’ the specific gravity of the liquid. In order to apply the method, the author determined the specific gravities of a number of precipitates by using known quantities of soluble compounds of the substances to be deter- mined, precipitating as usual in this process, and determining the specific gravities of the mixture and of the mother- -liquor. A modification of the formula that has been given then served to give the required specific gravities. The analytical operations were carried out by the use of two Geissler specific gravity bot- tles of 100 and 50° capacity. A precipitation was first made in Chemistry and Physics. | 321 a volume less than that of the larger bottle, the whole was trans- ferred to the latter at a temperature below 20°, which was used as the temperature for weighing, and the bottle was filled with water, the contents thoroughly mixed, and the whole finally weighed. ‘To obtain the mother-liquor free from the precipitate, the contents of the bottle were transferred to a test-tube and whirled in a centrifugal machine. For settling most precipitates less than one minute was required, while in no case was it neces- sary to spin the apparatus more than three minutes. The clear liquid was then drawn off by means of a dry pipette, and its specific gravity was taken in the smaller Geissler bottle. The author has applied the method to the determination of chlorine as silver chloride, sulphur as barium sulphate, calcium as oxalate, phos- phoric acid as ammonium phosphomolybdate and invert sugar by means of the cuprous oxide precipitated, with very satisfactory results. The method requires that the composition of the pre- cipitate should be constant, although it need not be exactly known. It was found that the specific gravity of aluminium hydroxide varies with the time that elapses after precipitation, so that the method is not yet applicable to the determination of this substance. It is expected that the method will find con- siderable application, particularly for rapid work in technical analysis, but it requires very accurate specific gravity determina- tions where quantities of about one gram are taken, a circum- stance which will probably limit its adoption by analytical chemists.—/our. Amer. Chem. Soc., xxiii, 644. H. L. W. 3. Huropium, a New Hlement.— By a long series of fractiona- tions with magnesium nitrate Demargay has separated an earth, intermediate between gadolinium and samarium oxides, of suffi- cient purity to show no samarium lines, and only mere traces of gadolinium lines in the spectrum. The solutions give absorp- tion-bands, and when traces of calcium sulphate are added to the oxide a brilliant phosphorescent spectrum is produced. The approximate atomic weight 151 is mentioned, but no details in regard to the determination, nor any account of the chemical properties of the substance are given. Numerous spectrum lines _are described, however, and the name given above is proposed for the element.— Chem. News, |xxxiv, 1. H. L, W. 4. Research Papers from the Kent Chemical Laboratory of Yale University ; edited by Frank Austin Goocn, Professor of Chemistry in Yale University. In two volumes: Vol. J, pp- xiv, 411; Vol. II, pp. x, 415. New York, 1901 (Charles Scribner’s Sons, price $7.50 net).—This work forms part of the “Yale Bicentennial Publications,” a series of volumes prepared by the Professors and Instructors of Yale University and issued in connection with the Bicentennial Anniversary to be held Oct. 20-23 of the present year ; the series is designed as “a par- tial indication of the studies in which the University teachers are engaged.” The Kent Chemical Laboratory established at New Haven 322 Screntific [ntelligence. through the liberality of Mr. Albert Emmett Kent, was opened in 1888 under the direction of Professor F. A. Gooch. The or iginal research work accomplished in the laboratory since that time’ by Professor Gooch and his students is now presented together in these two volumes. One hundred and eight papers are included, the larger part of them upon analytical processes in mor ganic chemistry. A critical review of these papers is not called for in this place since almost all of them have been published in the pages of this Journal from vol. xxxix, 1890, to the current volume. ‘This collection of papers speaks most strongly for the large amount of excellent original work which has been accom- plished in this laboratory. 5. Magnetic Hffect of Electrical Convection.—A discussion of the experiments of Rowland and of Cremieu upon this subject is given by H. A. Wixson in the July number (pp. 144-150) of the Philosophical Magazine ; the object of the paper being to point out that Cremieu’s ‘failure to confirm Rowland’s results is to be attributed rather to the methods employed than to the non- existence of these effects. The author’s conclusions are stated as follows: ‘*(1) That in Cremieu’s attempt to detect the electrostatic effect of a varying magnetic field, the effect of the steady magnetic field on the charging currents was left out of account, and that his latter effect is equal and opposite to the former. (2) That, consequently, Cremieu’s negative result constitutes an indirect proof of the existence of an electrostatic effect of a varying magnetic field, of the amount usually predicted theoreti- eall (3) That in Cremieu’s attempt to repeat Rowland’s experiment, his addition of a metallic screen placed close up to the fixed sec- tors should cause a current to be produced which should almost entirely compensate the desired effect. (4) That Cremieu’s partial failure without the additional screen, is guite possibly due to defective insulation of his sectors.” ‘The experiments of H. Pender upon the same subject, briefly stated in the Johns Hopkins University Bulletin No. 152 (this Journal, xii, p.173), are discussed at length in the August num- ber of the Philosophical Magazine. 6. The Kffect of Amalgamated Gases on Resistance ; by Wit- LiAM Ro.urs. (Communicated.)—As the result of experiments with X-light tubes, I concluded that we depend on gases amalga- mated with the terminals of a vacuum tube for an efficient cathode stream for producing X-light. It, therefore, seemed desirable to see if the partial removal of these gases would inany way change the resistance of the metal forming the terminals. Spiral ter- minals were made, the ends of each spiral being brought to the outside of the tube so that a current could be sent through the spirals before the X-light tube was pumped and after as much gas as possible had been removed from the metal by heat, elec- tricity and pumping. Mr. Oelling and Mr. Heinze constructed Miscellaneous Intelligence. 323 the tubes and made the tests during the summer of 1900. They reported that the resistance of a wire was increased by removing the amalgamated gases. In one experiment the resistance of the wire forming a spiral terminal in an X-light tube was ‘068 of an ohm before and ‘080 after. he instruments for making the experiments were not of a high grade. On this account the results were not reported, as I hoped to repeat the experiments with better apparatus and with terminals made of all the common metals. As this opportunity has not come, I now report the experiments that some one better equipped may extend them. Il. MiscELLANEOUS SCIENTIFIC INTELLIGENCE. 1. American Association.—The fiftieth annual meeting of the American Association for the Advancement of Science was held at Denver, Colorado, during the week beginning August 24th. The President of the meeting was Professor Charles 8. Minot of Cambridge. The attendance was larger than might have been anticipated considering the distance of Denver from Eastern centers, the registration reaching 311; nearly one-third of those in attendance came from the Atlantic coast. The interest taken by the citizens in the proceedings and their efforts for the enter- tainment of the guests contributed essentially to the success of the meeting. About 220 papers were presented for reading at the sessions of the Association and those of the affiliated societies, which were also well attended. The address of the retiring President, Professor R. 8S. Woodward, was delivered on the evening of August 27 on the subject ‘‘ The Progress of Science.” The address is printed in full in the issue of Science for August 30. The following numbers of the same journal, which is now the official organ of the Association, contain a detailed account of the Denver meeting with the addresses of the Vice Presidents before the several sections. | The place selected for the meeting of 1902 is Pittsburg, the meeting to embrace the week from June 28 to July 3. The officers elected are as follows: President, Asaph Hall. Vice Presidents: Section A, G. W. Hough, Northwestern University ; Section B, W.S. Franklin, Lehigh University ; Sec- tion C, H. A. Weber, Ohio State University ; Section D, J. J. Flather, University of Minnesota; Section HE, O. A. Derby, Sao Paulo, Brazil ; Section F, C. C. Nutting, lowa State University ; Section G, D. H. Campbell, Leland Stanford Uni- versity ; Section H, Stewart Culin, University of Pennsylvania ; Section I, C. D. Wright, Commissioner of Labor, Washington ; Section K, W. H. Welch, Johns Hopkins University. Permanent Secretary, L. O. Howard of Washington ; General Secretary, D. T. MacDougal, New York Botanical Gardens; Treasurer, R. S. Woodward, Columbia University. A meeting of the Association will also be held in Washington 324 Scientific Intelligence. during ‘‘convocation week,” that is the week of Jan. Ist, 1903. This mid-winter meeting is a decided change of policy, but it is left for the future to decide whether the time of meeting shall be permanently changed or whether there shall be two meetings annually. 2. British Association.—The meeting of the British Associa- tion for the Advancement of Science was held in Glasgow, Sept. 11 to 18. The presidential address was delivered by Prof. A. W. Riicker (see Nature, Sept. 12). 3. Catalogue of the African Plants collected by Dr. Friedrich Welwitsch in 1853-61. Vol. II, Part LI, Cryptogamia, British Museum (Natural History), 1901. Pp. 261-565.—The final por- tion of this work is especially valuable since it contains accounts by no less than ten specialists of all the divisions of cryptogams from a region where the lower plants had been seldom collected. The groups most fully treated are the fresh water alge by W. & G. 8. West and the lichens by E. A. Wainio. It is stated that the earliest collections of algze made in Africa are more extensive and representative than any hitherto described. Of the orders of fresh water alge the Desmidiacee and Myxophyce are best represented. It is of interest to note the occurrence of JVosto- chopsis lobatus Wood, first discovered in the United States, and later found in South America. In looking over the list of fresh water species from Africa, one finds a confirmation of the fact that fresh water algee are more cosmopolitan in their range than any other plants except perhaps the alge of brackish water. The marine algze which were worked up by Miss E. 8S. Barton are, as might have been expected, much less numerously represented, only forty species being enumerated, and some of these were collected at Madeira and the Cape Verde Islands. The lichens” include a large number of new species which are described in detail. The fungi were determined by Miss A. L. Smith, and include the species of Welwitsch previously described by Curry and Lagerheim. Of the four Mycetozoa enumerated by Arthur Lister, two were first found in the United States. The mosses by Antony Gepp, and the hepatics by F. Stephani include a number of new species, and the vascular cryptogams are given by W. Carruthers with the description of but one new species. There is also a general index of both volumes of the catalogue. Ww. G. F. 4. Leitfaden der Wetterkunde, gemeinverstdndlich bearbeitet ; von Dr. R. Bornstern. Pp. 181 with 17 plates and numerous text figures. Braunschweig, 1901 (Fr. Vieweg u. Sohn).—The number of those interested in the science of weather predictions and desirous of having an intelligent knowledge of the meteoro- — logical principles upon which they are based has increased very largely in recent years in consequence of the admirable work done by the various Government Bureaus. To them, as also indeed to those who have already had a training in this branch of science, the author’s concise and systematic treatment of the subject will be of great value. He discusses all the topics Miscellaneous Intelligence. 325 involved with uniform clearness throughout and—so far as space allows—with much thoroughness. At some points the work con- tains so much that it will cail for close study from those not already well grounded in physical science. An interesting chap- ter is that giving a brief account of the work of the Weather Bureaus of the various governments. The book is well illustrated, the charts at the end including, for example, the representations of the different types of clouds, being particularly good. 5. Studien tiber die Narkose ; von Dr. E. Overton. Pp. 195. Jena, 1901. G. Fischer.—This monograph contains a review and critique of the more prominent theories of the mode of action of narcotics and anaesthetics. The greater part of the book is taken up with a discussion of the theory of narcosis which has recently been advanced by Professor Hans Meyer of Marburg, and advo- ‘cated independently by the author. Jriefly, this theory assumes that all chemically indifferent compounds which are soluble in fats or similar substances act as narcotics on living protoplasm to the extent to which they are distributed therein. This action will obviously be most pronounced on those tissues which are richest in fat-like compounds, viz., the nervous tissues. ‘lhe com- parative efliciency of various substances as narcotics (anaesthetics) must be dependent on their relative affinity for these fat-contain- ing tissues and the tissue fluids respectively. Dr. Overton has attempted to ascertain this relationship—the coéflicient of distri- bution—for a number of organic compounds which act as narcotics. Many experiments on plants and animals—mostly lower forms— are reported in detail by the author, who is an assistant in botany at the University of Ztirich. L. B. M. 6. Ueber Harmonie und Complication ; von Dr. Victor Gotp- SCHMIDT. Pp. 136, large 8vo. Berlin, 1901 (Julius Springer).— Starting from some of the fundamental principles of crystallo- graphy, the author has attempted to show that analogous laws exist both in the harmonic relations and in their combination in other departments of science—as in sound and light. Thus the so-called harmonic series of numbers of which a simple example is given by 0°3:1:2°0, which in crystallography defines the position of the planes in a crystal in terms of the molecular forces, is believed to fix the positions of most of the notes of the musical scale and also those of the chief Fraunhofer lines in the solar spectrum. It would be impossible in a brief space to make clear the methods of the author in arriving at this conclusion, still less to explain his manifold deductions. The reader cannot fail to appreciate the industrious ingenuity exercised even if not convinced as to the correctness of all the views advanced. 7. Natures Miracles: Kamiliar Talks on Science ; by E1isua Gray, Ph.D., LL.D. Vol. II. Electricity and Magnetism. Pp. 248, 12mo. New York, 1901 (Fords, Howard and Hulbert).— In this little volume the writer presents some of the important topics in electricity and magnetism in simple language and with familiar illustrations, designed to interest the general reader to whom scientific treatises are closed books. 326 Scientific Intelligence. OBITUARY. CuHaRLES ANTHONY ScuoTr, for many years a prominent member of the staff of the United States Coast Survey, died on July 31 at the age of seventy-five years. Born at Mann- heim, Germany, in 1826, he came to this country at the age of twenty-two and at once entered the service of the United States Coast Survey, in connection with which he was to do such impor- tant work. A man of keen, well-trained and original mind, endowed with unlimited energy and industry, his labors, carried on for half a century, have contributed no small part to the grand total accomplished by this Bureau. He was eariy attached to the Computing Division and in 1855 he assumed charge of this department, a position which he retained until January, 1900; in 1856 he was advanced to the rank of assistant. He is best known for his numerous and important contributions to Terrestrial Magnetism, frequent references to which will be found in these pages. It was in recognition of his services in this direction that in 1898 he received the Wilde Prize from the French Academy. After being relieved of his duties as Chief of the computing division, early in 1900, he was assigned to the dis- cussion of the arc measurements in the United States; the volume entitled ‘‘ The Transcontinental Triangulation and Ameri- can Arc of the Parallel” was prepared by him (see this Journal, vol. xi, p. 172, 1901). This important work and another, “The Eastern Oblique Arc of the United States,” now about to be issued, form a fitting close to the labors of a long and most useful career. The following just tribute, adopted Aug. 1 by the members of the Survey, well expresses the estimate of the man formed by those most closely associated with him: “He was enthusiastic, faithful, and diligent in all duties he was called upon to perform, and through his learning and probity earned a reputation extend- ing over two continents which 1s most worthy of emulation. Conscientious and expert in his specialties, geodesy and terres- trial magnetism, his labors added immeasurably to the reputation of the Bureau and of his comrades who gathered the material he so ably discussed. The methods of computation now in use in the Bureau are an indelible record of his ability. His high ideals of duty and his tireless and persistent striving for them made him stand forth as a noble example of the best type of public official, and his uniform kindliness endeared him to those who knew him as a friend.” Baron Avorr Erik NorpEnski0Lp died at Stockholm on August 12 in his sixty-ninth year. He was alike renowned for his numerous and important contributions to the Mineralogy and Geology of Scandinavia and even more for his intrepid explora- _ tions in the far north. Dr. WitHELtm Scuur, Professor of Astronomy at Gottingen, well known both for his researches and his work as a teacher, died the past summer at the age of fifty-five years. Baron vE Lacaze-Durutsrs, the celebrated French zoologist, Professor in the Sorbonne, died on July 21 in his eighty-first year. - Aba AD AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES.] Art. XXX V.—On the Effect of Temperature and of Moisture on the Emanation of Phosphorus, and on a Distinction in the Behavior of Nuclei and of lons; by C. Barus. Introduction. © 1. Object, etc.—Endeavoring to differentiate the properties of the nucleus and the ion, it occurred to me that the effects of temperature, when worked out simultaneously by the volu- metric and by the electrical methods, would probably present a contrast. If the two functions relating to condensation and to electrical conduction are different, then their thermal varia- tions are not likely to be the same. The temperature which insures the maximum production does not also necessarily insure maximum instability. The results of the following paper bear out this surmise. Again, if phosphorus is to be used as an ionizer, some defi- nite knowledge as to the cause of its variable intensity is essen- tial from a practical point of view. The substance is so remarkably adapted for the purpose in many ways, that the endeavor to put it in control quantitatively is well worth while. This too, I think, has been accomplished. Finally I have shown that the low number of ions (n=2 x 10° per cubic centimeter) in the saturated phosphorus emanation, found from the experiments with the tubular condenser, is due to non-saturation. [ have been able to nearly double this number, putting these results in accord with the data of plate and spherical condensers. Incidentally certain curious condi- tions under which the emanation produces permanent conduc- tion in the condenser is identified with the occurrence of traces of moisture. This behavior so closely resembles the effects of radio-activity, that the extreme caution needed before such a property can be predicated, becomes apparent. Am. Jour. Sci.—FourtH Series, Vou. XII, No. 71—Novempsr, 1901. 23 328 C. Barus—Lffect of Temperature and of Moisture Volumetric Comparisons. 2. Apparatus.—The apparatus to investigate the relation of the emanating activity of phosphorus to temperature is shown in figure 1, the thermal part consisting of a coil of thin lead pipe ($ inch bore), Z, submerged in a large water bath of cop- per, AL, 13™ high, 15°" broad, and 20° long. There were 21 turns of lead pipe, each turn 6™ in diameter. The air com- ing from the gasometer train by way of a desiccator at D, and a stop cock, /’ (fine screw valve), traversed this considerable length of slender tubing, fully taking the temperature of the water bath, thereafter, to be discharged into the central straight pipe of brass, ab, 1:2 in diameter, containing the ionizers (not shown). The charged air is finally conveyed into the influx pipe of the color tube, C, by the removable short neck, G. A thermometer, 7, is placed in the water bath; another may be inserted into the end, 6, through a perforated cork, so as to be in contact with the ionizers. Care was taken that all changes of temperature should be slow. Thus it took 3 hours for the temperature to rise from 5° to 18° in the following experiments, for instance. The ionizers, as usual, were strips of wire gauze, holding thin pellets of phosphorus between them. They were inserted into, or removed from the tube, a through 6. If saturation is aimed at, an excess of freshly cut phosphorus surface should TABLE I.—EKffect of the temperature of phosphorus on its emanation. Steam jet method: p=4°5°™; influx air at 27°-28°. 10? x 10? 10? x Color. dV /dt. 7) Color. adV/dt 6 Color. dV/di 6 liter/min °C None chars 5°} Bl+ 70 128° Ble 85 27° |None ie 6) Bl 65 12°8 Bl+ 110 None ie 7 | opaque 80 12°8 a Puffs faint 8:2 Bl 75 1320). Bl 100 30°8 Pufis strong, 86 90 30°7 ereenish t SPO Ae e3 85 176 Bl 95 28°6 Puffs bluish 93 | opaque 105 17°6 100 28°6 Do. 9°8 Bl 95 25°2 Do. No perma- } 10:2 Bl 90 33 90 25°2 nent color [| BIsES = ak 30°8 Bl 80 22°3 Greenish permanent 10°7 Bl 95 306 Bl+ 90 22°3 Faint yl. 380 10:9 Bl 30 32°6 Bl— 61 18°8 green. Bl+ 105 32°4 Bl 90 18°8 Blue gray 280 128) Sh 100 32 2 Bl 75 18°8 Bl-— 350 12:2 | Bl+ 100 35°3 Bl + 115 18°8 All colors came out Bl 110 35:2 Bl+ 60 13°2 intensely includ- P28: eel 90 34 9 Bl 60 13°2 ing opaque Bl 70 13:3 - opaque 100 L3°3 * Bl, blue: Bl+, dark blue; BI—, light blue; yl, yellow. on the Emanation of Phosphorus. 329 be used. This was only done when specially called for in the present work, where the form of the temperature function is the chief consideration. 3. Method and data.—The usual method of experiment was adopted, the liters per minute (¢@ V/d?) of saturated phosphorus air necessary to produce the fiducial blue of the cclor tube, being observed at different temperatures. The data are given in Table I, in the first part of which observations for falling temperature, and in the second for rising temperature, are recorded. The pressure of the steam jet was about p=4 to 6. The inflowing air showed a temperature of 27° to 28°. The table contains some other colors (including opaque) for orientation. The chief data of the table are reproduced in the chart figure 3 (p. 334) and show the sudden cessation of reaction at 12°-138°. 4. Discussion.—For the sake of preliminary comparisons with the corresponding electrical charts given below, it is well to lay off 1/(d V/dé) in its variation with temperature: for this reciprocal runs parallel to the concentration of the emanation producing the color. The construction is given in figure 4, in which the sudden rise of activity in producing nuclei is apparent at 13°, and the subsequent gradual decline thereafter as far as examined, is again manifest. Anticipating data of subsequent paragraphs, I may add that the maximum ionizing activity is at 20°, showing the two thermal relations to be non-coincident. The charts show in the first place, that as temperature falls from the highest admissible values, say 35°, the emanation of phosphorus actually increases* at a rate of about 2 per cent per fall of 1° C. The maximum activity occurs at about 13°, and is upwards of 25 per cent greater than at 30°. Between 12° and 13°, however, the emanation is quenched at an enormously rapid rate, falling just short of suddenness. Practically, there- fore, the reaction begins at about 13°, with full if not greatest intensity. Below 12°, the emanation is insignificant and the maximum permanent colors obtainable are faint blue grays, even when the gasometer flow is forced to, say 400 liters/minute. There are no opaques. Below 10° there were no permanent colors discernable. 5. Here, however, and slightly above and below this tem- perature, definite puffs of color or of darkness are obtained immediately after opening the faucet suddenly. The pbhe- nomenon may be repeated indefinitely by closing the faucet *This increase maybe due to the gradual thorough desiccation of the phos- phorus by the dry current of air. These grids were not dried preliminarily over calcium chloride. 380 C. Barus—Lffect of Temperature and of Moisture for a period of 2-5 seconds (longer at the lower temperatures), and then suddenly opening it again. The puffs are at first vaguely recognized at about 8°, or even below. They become more marked as temperature rises. They are still marked at even 12°, when the fainter colors are beginning to be perma- nent. They show a maximum degree of darkness depending on temperature, beyond which they do not increase even if the cock is closed indefinitely. From a theoretical point of view this result is noteworthy. Below the reaction temperature (say 12°5°), what may be called the vapor pressure of the reaction is a definite quantity, but decreases with temperature at an enormously rapid rate; above the reaction temperature, the vapor pressure is relatively con- stant as temperature increases. All this recalls the well known analogy appropriated by the physical chemist. Now if we suppose the nuclear velocity to be a relatively constant quantity, within a short range of temperature, while the emanating activity decreases, the density of the emanation formed within the ionizer will clearly diminish as temperature decreases below the reaction temperature, however long the air is in contact with the ionizer. Hence the color of the putts should gradually become fainter with, decreasing temperature, as actually observed. If WV nuclei are produced per super- ficial square centimeter of phosphorus, if & be the corresponding average nuclear velocity and 2 the number present per cubic centimeter, V = kn. Thus depends on the ratio of JV and k. The vapor pressure analogy suggested is not wholly tenable inasmuch as nuclei are actually absorbed at the walls of the vessel (tube ad, figure 1), so that VV vanishes with » in the lapse of time. Since # is of the order of one unit, VV and n may be regarded as about of the same order, roughly speaking. The number of particles generated per square centimeter of the phosphorus will not greatly differ from the number present per cubic centimeter of the emanation. 6. Above the reaction temperature, if the rate of production, JV, were regarded as relatively constant, the means of comput- ing the increase of speed of the nucleus with rising temperature would be at hand. Ifm be the mass of the nucleus and mk*/2 varies as absolute temperature, *,/h, = V(t,+273)/(,+ 273). Turning now to the chart, figure 3, let ¢,= 30° and ¢, = 20°. Then #,/k,= 1:02, whereas the chart gives #,/k, = 1:25. These two results being out of keeping with each other, the thermal variation of % is insignificant compared with the cor- responding decrease of JV. 7. Data for low temperatures.—After finishing the electri- cal investigation presently to be discussed, it seemed desirable on the Emanation of Phosphorus. 331 to corroborate the above results with experiments made near the temperature at which phosphorus becomes active. Asa whole the new data agreed with the above inferences. It was discovered, however, that permanent though faint colors could be obtained even below the limits stated above (18°), by very gradually increasing the speed of the charged air current from zero until the field showed. the limiting coloration for the low temperature selected. When the air current is further increased, however slightly, the field of the color tube at once clears, almost with a flash. It is thus possible to “blow out ” the emanating activity of the phosphorus with a current only a trifle faster than the one which produces the corresponding a a Lig: fh. 6 color maximum. The puffs of color obtained above are the same phenomenon. Below 138°, opaque did not oceur. At 14° the full activity was accentuated. Electrical Comparisons. 8. Apparatus.—It is now desirable to compare these data with the results obtainable in measuring the radial currents when the tubular condenser is made the channel of communi- cation between the pipe, ab, of the water bath, figure 1, and the color tube, C. In other words, the tubulure, G, is now replaced by the condenser, KX, figure 2, for discharging ionized air into C, by fitting the tubulure 0 to the end a of the ionizer, figure 1. Details of adjustment are given in my \ 332 C. Barus—Lffect of Temperature and of Moisture earlier papers.* The slender condenser, AX, was effectively 50° long, °32™ in internal, and °60™ in external diameter. The inner face (surface of the rod ed) is charged to about 40 volts. The tube AA, insulated at the ends from the rod, is put to earth at # The electrical discharge takes place radially from rod to tube, and should oceur only when the emanation passes in the cylindrical shell between the faces (14™ thick and 50™ long) entering at 6 and leaving at a. It is difficult in so slender an apparatus and in view of the use made of it to avoid conduction through the insulators, altogether, particu- larly in a damp atmosphere. Hence in the following tables the insulation when the medium is ordinary air, is given; but even if ignored it will not probably affect the relation to tem- perature. In a warm dry room the insulation is perfect, and advantage was frequently taken of this observation. A steam or water jacket, /, surrounds the condenser for special experiments. § 22. Steam enters and. leaves by the tubulures, s. 9. Method.—The method consisted in reading the efflux volumes, V, at the gasometer or aspirator bottle, before and after the series of electrical measurements. As the latter were always duplicated, three volume measurements were made at stated times. From these d V/dt was obtained graph- ically. The fall of potential at the electrometer (capacity of the latter 60%, in parallel with that of the condenser, 59°") was observed at intervals of 15 seconds apart. Hight readings in two series were made between the volume readings. The initial potential being about 41 volts, and equivalent to 87 or 80 scale parts, respectively, each scale part is equivalent to about half a volt; but the absolute values are without interest. As usual, care was taken to await constancy of temperature in the water bath. 10. Data. —In the following Table II, the time of observa- tion in minutes, ¢, the reading of the gasometer, V, in liters, the reading of the electrometer in scale parts, s (zero at s=250), and the temperature, 0, of the water bath are given in succes- sive columns. In the second and third columns, moreover, the rate of efflux of the air, d V/dt, in liters/min., and the initial radial electrical currents ds/dé are tabulated. , is the initial potential difference in volts. As a rule, two values of d V/dé are entered for each tem- perature, one for a moderate current of about °50 liter/min. through the condenser and the other for the stronger current of about 1:0 liter/min. From brevity, an example of these data corresponding to a single temperature, 9 = 153°, will only be given. *This Journal (4), xi, p. 310, 1901; Phil. Mag. (6), vol. i, p. 672, 1901; ibid. (6), ii, p. 40, 1901; and elsewhere. on the Emanation of Phosphorus. 333 TABLE IJ.—Radial currents in the condenser. Phosphorus ionizer at different temperatures. Example of method. Time, ¢ Vx 10? ad V/dt (mean) s ds/at 6 lit {lame 70 27 — lass min. 9 30 125 12 30 205 S 0§ . 163 12—— min, 15 166 30 169 45 172 ht 13™ 308 280 *83—— i5-ou min. 15 30 450 | Re 610 § 0s 163 27—— min. 15 170 30 176 45 183 Il. Diseussion.—The data of these tables might be con- structed without further reduction in a. graph where the abscissas are temperatures and the ordinates, ds/dt, propor- tional to the radial currents. Two curves are suggested, one for the high, and another for the low velocity, dV /dt ; but in view of the slightly different values of d@V/dt implied in each, it is better to reduce to two volume standands; d V/dé = 45 liter/min. and 1:00 liter/min. were selected as most nearly coincident with the observations as a whole. The reduc- tion was made compatibly with the results of my earlier paper (I. c.), linearly from two values of dV/dt and ds/dt at each temperature. This linear relation is again incidentally shown in figure 5 at about 19°. Theslopes of these lines vary with temperature. If Q be the charge, s the deflection, C the effective capacity, f-the potential difference, A the factor of the electrometer, Y= CAs; 1=dQ/dt = CA-ds/dt, where 7 is the radial cur- rent. Thus 2/Q = (ds/dt)/s. ; | For the initial currents, as alone measured in this paper, one may always assume the simple exponential relation Q=Q,e—™—), or 5 = s,e—*'—), where the subseripts zero refer to the initial charges, deflections, temperatures, ete. Hence, 7,/Q,= (ds/dt), /s, = — x, is an appropriate variable for comparing the data. This may also be computed as 3384 OC. Barus—lffect of Temperature and of Moisture — «x= d(logs)/dt, but the approximate method of computing ds/dt from observations 15 seconds apart is more convenient. — The corrected values of — a =(ds/dt),/s,, when ad V/dt is ‘45 and 1:00 liters per minute, respectively, are given in the graph, figure 6. Different dots correspond to different series. The curves are smoother than the uncorrected results would have been, and the values for low efflux are naturally more certain. Jor apart from instrumental difficulties, there is at high velocities a danger of interfering with the temperature of the ionizing phosphorus. Swift currents are not so easily cooled in the water bath and intense action of the ionizer con- tributes its own temperature error. In both curves the con- duction of the insulators prevents the curves from actually reaching the abscissa. ; x 4 20 6 q 5 a Pas : ee fee fe} 5 i) Ls tv || | Heg4 | 24g Ile ge O10 26° 380 AP DERREDA TOUT re 40° q5° 20° 25° if 35° 12. Contrast with the color data.—The character of these curves may now be examined in comparison with the color data of figures 3 and 4, the latter being specially available. In both there is a rise of activity from about 9° through a maxi- mum, and an eventual less pronounced decline of activity toward 35°; but in their details, the two sets of curves are very different. The nuclei of fioure 3 are suddenly produced in maximum concentration at about 13° C., as suggested by the arrows ¢ in figure 6 et seq.; they then decline in number regularly and very gradually as ‘far as observed. In figure 6, however, the ions show a gradual increase of number, even as far as 20°, after which their number also falls off to the limits of observation. on the Hmanation of Phosphorus. 335 One may argue, therefore, that the nuclei as first produced are but weakly ionized in spite of their maximum condensa- tional activity. As temperature rises, the latter property of the nucleus declines, but the ionization increases as far as about 20°. Thereafter both properties decline. As the number of nuclei decreases from the reaction at 13° onward with increas- ing temperature, one may infer that the ionization increases with temperature ; from another point of view, that the ioni- zation increases as the property of the nucleus to induce con- densation diminishes. It is then with the nearly non-ionized nucleus that the maximum of condensational activity resides, just as if ionization were the result of a dissociation or a dis- aggregation of the nucleus. ! 13. Electrical experiments repeated.—It is doubtful whether the color experiments can be much improved. These results are bound to lack sharpness; but the electrical experiments are open to further development in the first place by retaining a constant velocity d V/dt throughout. This may be done by inserting a second stopcock, #” (not shown), to check the air current to a fixed value, even when /, figure 1, is quite open. In the second place the weakly ionized emanation at low tem- peratures should be tested directly as to its condensational power. One may inquire whether the reduced condensing power of the positive and the negative ionizations differ ; whether at a given temperature definite ionization is obtainable quantitatively, ete. In Table III results obtained by the same method as above are summarized for brevity. The volumes, dV /d¢ (liters/min.), . of air charged with phosphorus emanation traversing the con- denser, are a nearly constant quantity in view of the second stopcock already mentioned. @ is the temperature at which the fall of potential, ds/dt, was observed, s being the deflection in scale parts of the electrometer used above. The condenser was given a positive and a negative charge alternately, with the outer face put to earth. Four readings for the negative charge were included between similar sets for the positive charge. The conduction of the insulators is given. The last column contains the datum # =— (ds/d),/s,. 14. Discussion.—The initial currents (w) are shown in the graph, figure 7. As a whole the results are much more defi- nite than in figure 6, seeing that no reduction for volume dif- ferences is now needed. Though there is a small difference between the currents corresponding to the positive and the negative charges, this difference lies within a scale part and may be taken as an error of observation. The position of the maximum of nuclei is again shown atc. The range of the new data after deducting the error in insulation is smaller than 336 C. Barus—Lffect of Temperature and of Moisture TABLE III.—Radial currents in the condenser. Phosphorus ionizer at different temperatures. My, = 41 volts. sp = 73. 6, ete. adV/dt ds/adt So/So Insulation 00 eons 05 14:0° "48 — 5:5 +:07 14:0 © "49 — 6:0 —°08 16°4 | "48 + 13°3 +°18 16°4 48 —14°0 — 19 ~ 19:8 Phas ores +1870 +25 19°8 48 —18°0 —'25 2258 48 + 16°5 +°23 22°8 “48 —17°0 —°23 26°6 48 +15°5 +°21 26°5 "A7 —15°0 —'21 30-7 48 +13°8 +119 30°7 "47 —145 — ‘20 34°9 - "48 +12°8 +°18 34°8 48 —13°0 —'18 above, a circumstance presently to be considered and attribut- able to moisture and leakage errors, or in general to the neces- sarily unsaturated condition of the ionization within the condenser. Moreover the form of figure 7 differs from figure 6 and has approached more nearly to the color results of figure 4. The maximum, however, is still near 20°, so that the inferences above on the earler appearance of the nuclei, is sustained. 15. Permanent conduction produced by the emanation.— At this stage of my work, I encountered a peculiar and puz- zling series of phenomena which were not noticed in my earlier work, probably because the room temperature was purposely kept high and the atmosphere dry. After the air was passed over phosphorus freshly put into the tube ad, figure 1, the condenser receiving the emanation was thereafter found to remain permanently conducting even with the air current shut off, precisely as though it had itself become radio-active. This conduction was relatively so enormous that the electric cur- rents could not be measured by the same electrometer and the occurrence of an internal metallic contact or break in the insu- lation was immediately suggested. I, therefore, overhauled the condenser carefully, inserted an internal bushing, replaced the internal rod by a new one, etc., all without effect. The condenser showed good insulation after putting it together, but became a conductor after the passage of the first phos- on the Emanation of Phosphorus. son phorus emanation. Permanent leakage due to dislocation of the solid parts was thus ont of the question. This conduction vanished over night. It was reproduced as soon as fresh phosphorus air passed through the condenser. It then remained permanent though gradually diminishing for hours, and was nearly gone again next day. Hence two causes are suggested : either a film of residual moisture aspirated off from the phosphorus grid (which however was as usual care- fully dried by squeezing in a press between folds of blotting paper, and then exposed to the air, so that only traces of moisture can be in question) was precipitated in the condenser to the detriment of the hard rubber insulation; or else some form of emanation given off from the phosphorus made the condenser radio-active. Incidentally I may advert to the extreme caution needed before such radio-activity can be assumed, the behavior in both cases being essentially alike. Warming the condenser seemed to be useless. Moderate amounts of dry air (say 7 liters flowing out in about 10 minutes) passing over the phosphorus were nearly ineffective. It was no remedy to remove the phosphorus and pass dry air alone in the forward direction. Separating the condenser from the water bath did not change its conduction. Thus I found, for instanee, for the condenser alone and free from air current, ds/dit = 33; an hour later, ds/dt = 25; next day ds/di =‘; good insulation, ds/dt = 2. eAt 9 = 30° (water bath temperature), the tendency of the condenser to conduct permanently was at first accentuated but soon completely wiped out. The electric current reached a normal value. This appeared so much like a moisture error that I further teated it by passing the air current backwards, through the condenser first and then over the phosphorus into . the atmosphere, in this way drying both parts. Insulation of the condenser was thus at once restored. Again on passing a considerable volume of dry air (say 15 liters, slowly) over the phosphorus, this too lost its power to make the condenser permanently conducting. Hence in the experiments of the following table the phosphorus was first dried in this way in a current of dry air. The work then progressed smoothly, show- ing the relation of the emanating activity of phosphorus to temperature in a new light. I may add again that in none of my earlier experiments were like discrepancies encountered. Possibly a corroded ae grid may be hygroscopic, something like platinum black. 16. Specially dried phosphorus grids.—In Table IV, @ is the temperature of the water bath, d V/dt the volume of dry air in liters passed per minute over the phosphorus: ds/dt is 388 OC. Burus—Lfect of Temperature and of Moisture the corresponding initial radial electric current in the con- denser in arbitrary units, when the potential difference, #,, is about 40 volts and the initial deflection, s,, as stated. In both cases the insulation of the condenser in the absence of the air current, is measured for each temperature. This is then deducted and the corrected electrical currents tabulated. In Table IV, on the left, a channeled hard rubber bushing guards (unnecessarily as it afterwards proved) against metallic contact of the core and envelope. In the data on the right this has been removed. The phosphorus in both cases is dried pre- liminarily, as stated, by acurrent of dry air from the desiccator, entering the condenser first and passing thence over the phos- phorus grid into the atmosphere. The last columns give the currents (ds/dt),/s, =%,/s, = 7,/@Q,, supposing that initially Q = Q,e~“—) as stated in $11. | TABLE IV.—Ionizing activity of dried phosphorus. #&,=4lvolts. s»=73°5. . dV/dt = 52 liters/min. 6 ds/dt $0/S0 0 ds/dt $0/80 19-2" 6 “O09 25°8° 7 "10 18°0 oy ‘06 21°2 if "10 16°5 - "05 20°0 9 "12 15°0 0 "00 18°8 9 12 dfs 8 12 16°8 6 ‘08 16°6 8 “LI 18°2 11 pile 15°5 6 08 iL Oe2 13 "18 14°0 2 “03 20°6 13 18 13°3 1 ‘91 22°4 2 "16 Color experiments made at 2772 10 14 13°5° (here) showed blue 33°8 9 12. equivalent to ‘8 lit./min. After the color experi- ments, for dV / dt = ‘93 lit / min. results were as = ~ = tollows: Channelled hard rubber bush- 14-0 4 06 ing in condenser. 14:0 + "06 = —— —-4-- Bushing removed. 17. Discussion. The results of Table IV are shown graph- ically in figures 8 and 9, where the currents (ds/dt),/s, are as usual the ordinates. In a general way the character of figures 6 and 7 has been preserved, inasmuch as there 1s maximum ionization at about 20°: but the details of behavior are again different. In the first place the scale of the phenomenon is gradually reduced as the emanating body is repeatedly subject to desiccation. This merely means deficient phosphorus sur- face, as I take it. on the Emanation of Phosphorus. 339 There is no certain tendency of the maximum to move into smaller temperatures in the later experiments. Thus when d.V/dt = °50, nearly, Figure 6, maximum (ds /dt),/s, =°20 at 20° Figure 7, Ei) 19° Figure 8, =e LS p40 Figure 9, cB 19s Whether the phosphorus is being actually consumed, or whether merely a superficial change is in question will be investigated below; but the charged air current is gradually further removed from saturation and will contiuue on the decline in the following experiments. Moreover here is an explanation of the differences of slope shown in the volu- 340 C0. Barus—lifect of Temperature and of Moisture metric and electrical curves of an earlier paper (1. c.). For the degree of “ dryness” reached wholly determines the electrical curve without in the same degree influencing the volume curves, as will presently be further manifest. In all cases the dependence of the electrical results on tem- perature remains quite different from the corresponding dependence of the color data on temperature. Special experi- ments made at the end of Table IV with identical apparatus showed strong color activity at 13°6°, viz: blue corresponding to 80 lét/min. while the subsequent electrical measurements at 14° (see Table VI) reproduced the original exceptionally low conductions. The position of the phosphorus grid in the tube, ad, of the water bath, figure 1, showed an effect insufficient to be of moment in relation to the phenomena under discussion. Thus Rear position, furthest from end @ and condenser, (ds /d¢),/s,="138 (see 7, figure 9). Front position, nearest to condenser, (ds/dt),/s, = °17 (see /, figure 9). Nevertheless there is nearly one-third more saturation when the phosphorus grid is nearest the condenser than when remote, a circumstance which, as already intimated, makes it difficult to investigate saturation in this way. Any connecting tube between ionizer and condenser is an absorber, particularly if bent. 18. Corroborative experiments with the color tube—A series of experiments were now begun with the steam jet, to ascer- tain the difference between the character of the emanation immediately after the phosphorus grid has been prepared (without preliminary desiccation) and after a large volume of dry air (20 to 30 liters) has been passed over it. If the rela- tively enormous currents obtained in the condenser in the first ‘instance are due to nuclei, there must be a corresponding result in the volume per minute of the saturated emanation neces- sary to produce a fixed color (blue) in the color tube. ‘The reverse is the case, as will be seen in Table V, where nuélei and ion production are in a measure reciprocal occurrences. ain other words, the initial enormous conductions are accompanied by an abstraction of nuclei. tg Three phosphorus strips were as usual dried in press between folds of bibulous paper and then exposed to air for some time. They were then inserted into the tube, ad, of the water bath, figure 1. The initial (apparent) ionization as tested by the condenser was invariably too intense to be measurable. The condenser was then removed and a short tube, G, added to on the Emanation of Phosphorus. 341 obviate excessive absorption before discharging into the color tube. In Table V, the liters of dry air which have passed over the phosphorus strips are given under Z. The successive liters of emanation per minute to produce the standard blue are given in the third column (dV/dt); the fourth gives the current (often estimated) when the emanation passes through the con- denser at the fixed rate of -5 liter/min., selected for con- venience. TaBLE V.—Condensation producing activity of initial ‘: wet” and final ‘ary: ” phosphorus emanation. Temperature @=18°. dV/dt =°5 lit/min., s.=73, for condenser. Color, ete. L adV/dt ds/dt Remarks. Insulation 0 0 ...-| Condenser discharged within 5%¢¢ Bl Oem L000 ee. & Bl 6 160 ~..-.- | Condenser discharged within 10% 13 Sets 100 | Current estimated. 2 iy seal 5) 15 | Phosphorus “dry.” Insulation 6 25 140 .-.--!Tonization constant. %,/s,='21 Another experiment. Grid scoured. Temperature 21°. BIl— 0 115 __..|Condenser discharged within 75° 4 ....---|Nuclei insufficient to produce 3 opaque. Bl Oe: 100 200) Opaque now attained. Con- Bl === 90 ----| denser discharged within 20°, Bl 8 103 -.--| Current estimated. Bl 15 90 9 | Insulation ds/d¢ = 1, nearly per- Bl 20 90 fect. Bl se gal Ionization constant. §,/s,=°'13 The first part of Table V shows definitely that when the currents are too large to be even estimated, the emanation needed to produce the standard blue is larger than at the end of the experiment, where the radial currents have fallen off to their small fixed value. Thus the high conduction is without nuclear condensing effect. } In the second part of the table the volumes are nearly con- stant except at the beginning, where it was found impossible to obtain opaque or even full blue, whatever volume is passed through the ionizer. The necessary number of nuclei was not forthcoming. As in the preceding table, however, less than ten liters of air are snfficient to dry the phosphorus into full activity so far as the color is concerned, whereas the conduc- tion still retains abnormally large values. Another reciprocal relation is shown on the table. dV/dt here happens to be unusually large, so that the phosphorus is 342 OC. Barus—EHffect of Temperature and of Moisture for some reason weak as a nuclei producer. I, therefore, washed and scoured the surface of the grid, obtaining the usual order of values in the second part of the table. On the other hand, the currents in the first part are larger than those of the second part. Here again, therefore, the tendency to produce nuclei reciprocates in intensity with the tendency to produce ions, or better to produce conduction in the condenser. The latter is facilitated by the presence of traces of moisture, but nuclei are not so produced. 19. Corroborative experiments with damp paper. — Another method of throwing light on the inquiry will be a comparison of the conduction produced in the condenser by air passing over damp filter paper with the corresponding case of air pase: ing over phosphorus. The wet paper (omitting the data) behaves in a less intense way something like the phosphorus. The rise of conduction, however, is ovadual, the conduction at best moderate and the return of the condenser to the original degree of insulation relatively rapid. With phosphorus, the conduction after the first minute or so has risen to the immeasurably large values and when the air current ceases the condenser shows similar conduction. Much more dry air is needed to dry the con- denser and the phosphorus to normal values (fully twice as much as in the preceding case). Eventually the currents also return to the normal, relatively small limit and the insulation of the condenser is nearly perfect again. Qualitatively the two phenomena run in parallel; quantita- tively, they are enormously different. Inasmuch as the paper is obviously wet, whereas the phosphorus grid has been dried short of desiccation, inasmuch as any emanation must behave like a water evaporation, I think that the volatile body is probably of the nature of a hydrophosphide. Some electri- cally active substance is distilled in the presence of moisture and precipitated in the condenser. 20. Corroborative experiments with desiccators.—The final test made to detect the character of the emanation was one of direct desiccation over chloride of calcium, before insertion. The day happened to be damp and the insulation poor. The experiments, however, are none the less definite. (1) Phosphorus dried in air and inserted into the dried fee of the water bath, a, figure 1; the condenser was at once dis- charged on passing the air current through it. On removing the phosphorus the condenser showed too large a leakage to admit of the measurement of current. All appurtenances were now dried in a current of dry air and the final insulation deter- mined. (2) The phosphorus grid having been placed for about 15 on the Hmanation of Phosphorus. — . 343 minutes in the desiccator was again inserted into the tube, ad. ‘The current now obtained was ds/dt = 37; the insulation proved to be ds/di = 30. Hence the current due to ionized air was but ds/d¢ = 7, an abnormally small value, indicating the absence of moisture. The phosphorus grid was once more put in the desiccator for 15 minutes. After réplacing it in the water bath the cur- rent observed was 40; the insulation 27. Hence the leakage due to ionized air is here ds/dt = 18, agreeing with the usual order of ionized values above. Owing to the unfavorable condition of these experiments, not much definiteness was to be anticipated from them ; but they show clearly that the enormous initial emanation from fresh phosphorus is all but wiped out, relatively speaking, after the phosphorus has been dried preliminarily over calcic chloride. Whether in the rigorous absence of all moisture phosphorus would cease to ionize air, remains to be seen. It is also a question whether the desiccation over calcic chloride may not be accompanied by detrimental chemical action, refer- able to the chloride. At least the following work continues to show that the phosphorus grid repeatedly treated in this way continually loses strength as an ionizer, in spite of the inter- mediate submersion in water to keep it over night. 21. Hffect of prolonged drying.—A final attempt was made to see if after continued drying over calcic chloride the ioniza- tion would be wiped out altogether. The followiug table shows this to have been unsuccessful. The room was favor- ably dry and warm and the leakage errors in the condenser not appreciable. After nearly 5 hours the potency of the ionizer is not diminished (see figure 11). It has rather increased, due possibly to the attraction of traces of moisture even within the permanently dried tube of the apparatus, figure 1. An extra tube of calcic chloride was attached. In a second experiment the temperature effect is tested for this specially dried phos- phorus. ‘The corresponding graph is shown in figure 10, and the ionization is weaker than in any earlier experiment. Nevertheless the results show maximum activity in the neigh- borhood of 20°, though even at 12° the ionization is not quite extinguished. The exceptionally low ionization is not accounted for except as due to deficient phosphorus, the natu- ral result of long continued consumption. 22. Promiscuous experiments. — Having investigated the effect of the temperature of the body of phosphorus on its emanation, I next purposed to ascertain the dependence of the emanation itself on temperature. This could be done by sur- rounding the condenser by a steam jacket (shown in figure 2, Jj; 8, 8, being the influx and efflux pipes) and noting the effect AM, Jour. Scl.—Fourts SERIES, Vou. XII, No. 71.—Novemper, 1901. 24 344 OC. Barus—Lfect of Temperature and of Moisture of the rise of temperature of nearly 100°. After repeated trials, however, I found that the high temperature so far diminished the insulation of the hard rubber bushings of the condenser, that no measurements would be trustworthy. On cooling the condenser, the insulation again became perfect. Quartz insulators suggest themselves as probably alone avail- able. | 23. Comparison of old and new grids.—Trial was made with freslily cut phosphorus, the grids used in the above work being as much as a year old. The results show a like order of values for both, in spite of the intense fuming of the new grid. Nevertheless the latter is apt to be from two to three times stronger than the old grid. Thus after thoroughly drying the new grid over calcic chloride and testing it at 21°5°, the data were AV /dt = ‘42, (dsi/ dt). [s= 20. 43, = 20: Exposure to the air of the room for about half an hour after desiccation was without effect. This experiment proves that the air in all the earlier experi- ments was undersaturated, agreeing with an earlier paper (1. ¢) from which values are quoted in the next paragraph. 24. Older data compared with the present data.—It is finally worth while to adduce the corresponding data, Table VI, of an earlier paper. The equivalent colors of the steam tube are the yellows and crimsons of the second order. The temperatures are 20°-30°. TABLE VI.—Relevant data from earlier papers, o= 40 volts. Color. Temperature, 0. 10? x dV/dt 10? x (ds/dt)o/so Vellowmy eee ee 30. DOK 16 Mellow, ges is 30 65 . 17 Crimsone a eae 30 43 ta Yellow-green, ----- 21 iO 13 Viellows tet se anaess 24 58 2% Crimson, 38 ae es 42 14 Clearly then, for d V/dt = :48, nearly, the order of values for the currents is the same here (the experiments made over a year ago but with the same grids) as in the above experi- ments. The room being very hot insured dryness in that work without preliminary desiccation. Large volumes of dry air were passed over the grids, moreover, for colors of the first order were principally observed, requiring even as much as 3 liters/min. per observation. The arrows (, new and a, old) in figure 9 show the relation of the results in the last table to the values for freshly cut phosphorus in excess given at the beginning of paragraph 23. on the Emanation of Phosphorus. 345 Since from the old values the number of nuclei was computed as 2 =2xX10*, the new values would make them approach m=4X10*, the datum found with plate and with spherical condensers by entirely different methods, remembering that from the occurrence of an absorbing influx pipe, the air within the tubular condenser can nowhere be quite saturated. Experiments on the maximum number of nuclei per cubic centimeter producible by phosphorus under favorable condi- tions, were made at some length; but as they yielded no new results beyond those of § 28, they ‘need not be instanced here. 25. Conclusion and summary.—The experiments with phos- phorus show that nuclei adapted to condense atmospheric moisture are produced most abundantly at about 13°C. Below this, the rate of production decreases with enormous Tapidity (just short of suddenness), probably ceasing at about 8°. This holds true for very different velocities of the dry air passing over the phosphorus. Above the reaction temperature, the activity decreases slowly as temperature rises, having not decreased more than 25 per cent at 35°. Since the phosphorus is superficially heated by the reaction, wen) statements are out of the question. Below 13°, the opaques of the color series are absent and the maximum tints are fainter and of increasingly higher orders. ‘The strongest permanent color may be approached by very gradually increasing the charged air current. A limiting velocity decreasing with temperature may thus -be reached, beyond which all colors vanish and the reaction is quite “blown out,” as it were. Suddenly opening the stopcock after a period of quiescence shows puffs of color under these conditions, which also vanish as temperature decreases below 13°. Putting NV =kn, where WV is the number of particles generated per second per superficial square centimeter of phos- phorus, m the number per cubic centimeter of the ionized medium, #& their (average) velocity in any cardinal direction, the variation of JV is to be ascribed to n, & being relatively constant. In contrast with the color data (nuclei), the ionization of air passing over phosphorus increases with temperature to a maxi- mum at about 20°, after which there is a less pronounced decline. This ionization is not an arrival comparable in sud- denness with the appearance of nuclei, nor are the maxima identical as to temperature. One may infer that the nuclei as first produced are weak in ionization but of normal strength in condensational activity; that thereafter the latter property declines because (probably) the ionization increases as far as 20°. Finally both properties decline. This reciprocity is accounted for if the ionization is a result of the dissociation of 346 C. Barus— Effect of Temperature, ete. nuclei. The degree of ionization is here independent of the sign of the charge. If the phosphorus grid is not preliminarily quite dry, traces of moisture are apt to escape with the emanation and produce permanent conduction in the condenser. Considerable varia- tion of the electrical coefficients may thus ensue, though the color results are, relatively speaking, but slightly affected. As the electrical discrepancy seems to be out of proportion with the quantity of moisture present, it is probable that the ema- nation escapes in some combination with it. The whole phenomenon vanishes on thorough desiccation both of the phosphorus and the apparatus. Grids frequently treated in this way show a gradually decreasing ionizing intensity, prob- ably due to the continued consumption of phosphorus or to a removal of effective surface. Throughout the experiments the relation of the color curves to the electrical curves remains practically unchanged, in spite of the different degrees of satu- ration (ionization). The relatively enormous conductions associated with non- desiccated phosphorus are without a color effect when tried in the steam tube. They rather give evidence of an abstraction of nuclei. Moreover such reciprocal properties are manifest in other instances. $18. Air passed over damp paper behaves similarly to the emanation from non-desiccated phosphorus, with a difference of intensity in favor of the latter. Desicca- tion over calcic chloride removes the incidental conduction entirely. The emanation may thus be dried to a limiting degree of ionization which is not then further reduced on dry- ing. In practice all operations should be made with desiccated phosphorus ; otherwise the bafiling discrepancies encountered in the case of plate and spherical condensers may be antici- ated. : By using freshly ent fuming phosphorus in excess in the ionizer, it was possible to increase the radial currents in the condenser to nearly twice their usual value, remembering that the emanation within the condenser is in all parts essentially unsaturated. Hence the low concentrations formerly found for tubular condensers (2X 10* nuclei per cubic centimeter) is made to approach the value found from plate and spherical condensers (4 10°), more nearly. Brown University, Providence, R, I. Mizxter—Determination of the Heat, ete. 347 Art. XX XVI.—On the Determination of the Heat of Dissocia- tion and of Combustion of Acetylene, Ethyleneand Methane ; by W. G. MIXTER. [Contributions from the Sheffield Laboratory of Yale University. | Tue heat of formation of an organic compound is deduced, as is well known, from its heat of combustion and that of its constituents, since it is impossible to form any hydrocarbon under conditions at which the thermal effect may be measured. The determination of the heat evolved when acetylene is | ‘| i | | il il | | | ii i i | | il | | il | | li | | iy | i J i il | i | | | | | ) i | | | | | | | | | | | | | | | | | | i | decomposed is, however, one of the easiest of calorimetric experiments, giving at once the heat of combination, as the thermal change of dissociation equals that of combination. Moreover, it is probable that the direct measurement of the change is less liable to error than the indirect method, which involves burning the gas. The thermal effect of the dissocia- tion of other hydrocarbon gases may be found by mixing them 348 Miater— Determination of the Heat of Dissociation with sufficient acetylene to effect when exploded complete decomposition. The heat found less that due to the acetylene equals the heat of dissociation of the compound. The method is open to the objection that the errors are cumulative. The apparatus used in the work is shown in fig. 1. The outer double-walled vessel, J, J, called for convenience the jacket, contains 20 liters of water, which may be stirred before an experiment by the screw propellers. The tin vessel, A, A, separates the calorimeter from the jacket by two air spaces. The calorimeter vessel is tinned iron. ‘The top of it was imperfectly covered by a thin sheet of celluloid and was exposed to tne air of the room. The steel bomb B has a capacity of 602°, weighs about 1900 grams, and is supported by a bar of wood resting on the top of the calorimeter case. It was silver-plated to prevent rusting and the silver on the interior surface was covered with a thin plating of gold. After the explosions of acetylene the inside plating showed minute perforations with projecting edges, owing doubtless to - enclosed particles of plating solution which gave off steam when heated, as eleetro-plate is known to shut in minute parti- cles of solution which are not removed by washing or drying. The explosions of a mixture of acetylene or ethylene and- oxygen dulled the surface of the gold and- gradually removed the metal in the state of fine powder. The explosions of acetylene were noiseless, while those of mixtures of oxygen gave a distinct report, threw the water against the cover and loosened the screws at the top and bottom of the bomb. Ignition was attained by electric sparks between a strip of platinum foil connected with the bomb and a platinum wire, E, insulated by a piece of clay pipe stem projecting into the bomb and a glass tube in the steel tube, D. The foil: was commonly melted and the wire, which was 13™™ thick and extended into the bomb several millimeters, was bent by the detonations of acetylene and oxygen. The mean of five accordant ealori- metric determinations of the water equivalent of the bomb and calorimeter can was 250 grams. ‘The error is probably too small to affect the results in the fourth figure, since the weight of the apparatus times the specific heat of iron is 262 with no allowance for the small amounts of other metals present. Upon changing a fitting on the bomb the number 252 was adopted. A differential thermometer graduated to hundredths of a degree was used. It was found to agree with a normal thermometer and one graduated to tenths. The zero of the differential thermometer was not changed during the work and corresponded to 16°4° of the normal. : ; In none of the experiments was any measurable quantity of acetylene found in the gas remaining after an explosion and in and of Combustion of Acetylene, Ethylene and Methane. 349 most of them not a trace was detected. The pressures used were higher than those at which the gas was exploded in earlier experiments,* when several per cent of acetylene was found after exploding. Whether the different result in this respect is due solely ‘to pressure or to other causes, the writer has not had time to investigate. In order to find if the acety- lene decomposed in the calorimetric bomb without formation of hydrocarbons, the following tests were made. The gas in the bomb after an explosion was passed through a solution of potassium hydroxide to remove hydrocyanic “acid and then burned. No carbon dioxide was revealed by lime water. 2°4 grams of carbon from the acetylene exploded in the bomb were heated in a combustion tube through which a current of dry air was passed to remove moisture. The combustion was then made in the ordinary way with oxygen and copper oxide. The result was 0°016 per cent of hydrogen. This amount may be due to the error of the analysis. These results show that in the calorimetric experiments the acetylene decomposed com- pletely into hydrogen and carbon. The composition of each sample of gas used was determined, and the amount taken was calculated from the capacity of the bomb or was found by weighing the bomb before and after filling: when hydrocyanic acid was estimated in the residual gas an equivalent quantity of acetylene was deducted from that taken. As a rule, the description of the many details involved in finding the ‘amount of gas used for an experiment are omitted. Since the accurate determination of differences in temperature is the most difficult problem in calorimetry, the observations are stated in detail. The water in the calorimeter was stirred by the small screw propellers shown in the figure, driven by a motor. ‘The observations of temperature were recorded at intervals of three minutes, and that noted imme- diately before explosion is taken for the initial temperature, no correction being made for slight changes preceding. The final temperature assumed is that observed the sixth minute after explosion plus five-sixths the average fall in temperature for a subsequent six minutes. The reason for not allowing for loss of heat during the first minute is this: the water of the calori- meter did not attain the maximum temperature observed until one, two or sometimes three minutes after the explosion, and it is less accurate to consider that as much heat is lost by the calorimeter in the six minutes immediately following explosion as in the next six. It is possible the corrections made are excessive, but any error in this respect is small as the correc- tion in but a few instances amounted to one per cent, and in most of the experiments it was less than half as much. * This Journal, ix, 1, 1900. 350 Mixter— Determination of the Heat of Dissociation The results are given for gram molecules and 1 and 12 are taken for the atomic weights of hydrogen and carbon respec- tively. No corrections are made for water vapor or nitric acid remaining after explosion, as these would only change the sixth figure of the results. Acetylene. The acetylene gas was made by thrusting lumps of calcium carbide into water in a glass gas holder. By means of ferrous hydroxide it was freed from traces of oxygen. Thus prepared the gas contained from 1 to 3 per cent of nitrogen. It was dried before pumping into the bomb by passing it through a kilo of small sticks of potassinm hydroxide. The oxygen used was made from chlorate and was kept over water containing a little sodium hydroxide. Lixperiment 1.—Acetylene, 7-24 grams, gave too great a rise of temperature to be measured by the differential thermom- eter. Accordingly the thermometer graduated to tenths of a degree was used in the calorimeter. The result was 51,900 calories for 26 grams of acetylene. Experiment 2.—Acetylene, 5488 grams; temperature of jacket 2°7° by the differential thermometer used in the ealori- meter. The residual gas was free from acetylene. Minutes. Temperature. 0 0°285 Water in calorimeter, 3000 grams. 3 0°290 Water equivalent of 6 0-295 calorimeter, 25:On ainees 9 3°740 : o 12 3°730 3250. 15 3°709 Temperature interval, 18 3°696 3°73 — 0°295 +0°027 = 3°462° 21 3°682 Multiplying this temperature interval by 3,250 and by the ratio of 26 to the mass of the substance taken we deduce for 26 vrams of acetylene 53,300°. Laperiment 3.—In this experiment the bomb leaked after charging with acetylene. Hence it is better not to record the calorimetric result although it accorded fairly with the other. Experiment 4.—Acetylene, 6-493 grams; jacket, 34°. The residual gas did not contain acetylene. Minutes. Temperature. 0 0:0 Water and water equivalent of 3 0-01 | calorimeter, 3250 grams. 6 0°02 9 0°027 Temperature interval, We 4°110 4°114 —0°027+0°018 = 4:105° 15 4-114 and of Combustion of Acetylene, Ethylene and Methane. 351 Minutes. Temperature. 18 4°100 ZA 4°090 24 4°080 27 4°070 Reducing as before the resulting value is 53,400°. Experiment 5.—Acetylene, 5°3829 grams; jacket, 4; room, 4°5°. The residual gas contained too small a quantity of acety- lene to measure. Minutes. Temperature. 0 0°45 Water and water equivalent of 3 0°47 calorimeter, 3250 grams. 6 0°49 2 0-51 12 3°892 Temperature interval, 15 3 887 . 3°887 — 0°51 +0°007 = 3°384° 18 3°880 a 3°875 24 3°875 27 3°870 30 3°865 The resulting value is 53,200°. Fixzperiment 6.—Acetylene, 4:043 grams; jacket, 3°6°. The residual gas was free from acetylene. Minutes. Temperature. ) 1°05 Water and water equivalent of 3 1°07 calorimeter, 3252 grams, 6 1°083 9 1-097 Temperature interval, 12 3°649 3°645 — 1:097 +0:004 = 27552” 15 3°645 18 3°642 21 3°640 The resulting value is 53,300°. Omitting experiments 1 and 3, the following are the results of the determinations of the heat evolved on exploding acetylene: 53,300 53,400 53,200 53,300 The average is 53,300° for 26 grams and 2050° for 1 gram of acetylene. Experiment 7.—The bomb with acapacity of 602° was filled with dry gas (97-4 per cent by volume of acetylene) at 20°7° and 756"™. The calculated weight of this volume of acetylene 352 Mixter—Determination of the Heat of Dissociation is 0°6342 gram. Rather more dry oxygen was added than required for complete combustion. : Minutes. Temperature. ; 0 0°935 Water and water equivalent of 3 0°943 calorimeter, 3252 grams. 6 0°950 9 3°270 Temperature interval, 12 3°264 3°264 — 0°95 +0°007 = 2°321° 15 3°260 . 18 3°254 21 3°250) The resulting value is 309,400°. Experiment 8.—The bomb filled as before contained 0°6527- gram of acetylene. Jacket 1°6°. Minutes. Temperature. 0 0 340 Water and water equivalent of 3 0°350 calorimeter, 3252 grams. 6 0°360 9 O50 Temperature interval, _ 12 2°762 2°762 — 0'°36+0°007 = 2:409° 15. 2°760 18 2°756 Pal 2°780 24 2°743 The resulting value is 312,100°. Lixperement 9.—The bomb was filled as for experiments 7 and 8 with dry oxygen and acetylene and contained 0°6456 gram of the latter gas. Jacket, 2°6°. Minutes. Temperature. 0 nog Water and water equivalent of 3 Talal calorimeter, 3252 grams. 6 Lis 9 3°502 Temperature interval, 12 . 3°d00 3°5 — 1:1138 +0°012 = 2°399° 15 3°490 18 3°482 my 3476 24 3°470 The resulting value is 313,700°. The mean of the three determinations of the heat of com- bustion of acetylene is 311,700° and of the last two it is 312,900° at constant volume and 313,800° at constant pressure. Before discussing the foregoing results reference must be made to the heats of combustion of hydrogen and amorphous and of Combustion of Acetylene, Hthylene and Methane. 3538 earbon. Thomsen* made three determinations of the heat of combustion of hydrogen in which 18,928 grams of water were produced, and obtained as the mean 68,357°. Berthellot and Matignont+ adopt as the results of their experiments the num- ber 69,000. Thomsent used the number 96,960 for CO, - obtained by Farve and Silbermann,$ and Berthellot| found 97,650°. Thomsen, and Farve and Silbermann made the com- bustion slowly at atmospheric pressure, while Berthellot made them almost instantly in his bomb. The former worked with large quantities and the latter with small, in case of hydrogen with about 23 milligrams. Berthellot’s results are somewhat higher than the others. Thesum of the heats of combustion of the components of acetylene is, according to Thomsen, 262,277°, to Berthellot 264,300°. The former found for the combus- tion of acetylene 310,050° and the latter** 315,700° at constant pressure. The writer’s result is 313,800°. Thomsen made the heat of formation of acetylene —47,770°%. If we calculated it from Berthellot’s figures it is —51,400°. Ethylene. The ethylene gas was prepared by the action of zine on pure ethylene bromide dissolved in alcohol. It was freed from oxygen by ferrous hydroxide but contained several per cent of hydrogen gas and doubtless a little nitrogen. It was dried by sulphurie acid. Lixperiment 10.—Gas containing 93°7 per cent by volume of ethylene, 602° at 18°5° and 748™™ pressure equal to 0°655 gram of ethylene; acetylene added 4°828 grams; jacket 1°4°. The gas after explosion was free from hydrocarbons. Minutes. Temperature. 0 1:22 Water and water equivalent of 3. 1°22 calorimeter, 3252 grams. 6 1°22 9 4°329 Temperature interval, 12 4°362 4°362 — 1'22+4+0°026 = 3°168° 15 4°343 18 4°330 21 4314 24 4°300 The thermal result is 10,300°. The heat of dissociation of 4°828 grams of acetylene is 9,900 calories; subtracting this from 10,300 we have 400 calories * Thermochemische Untersuchungen, ii, 52. +Ann. Ch. Phy. [6], xxx, 553. ¢ Thermochemische Untersuchungen, ii, 283. § Anu. Ch. Phys. [3], xxxiv, 357. | Ann. Ch. Phys. [6], xviii, 89. | §{ Thermochemische Untersuchungen, vi, 74. **eAnn. Ch. Phys. [6], xxx, 556 354 Mixter— Determination of the Heat of Dissociation due to the 0655 gram of ethylene. These data give 17,000 calories for the heat of dissociation of 28 grams of ethylene. — Experiment 11,—The mixed gases used contained 0-707 gram of ethylene and 4-091 grams of acetylene. The gas after explosion was freed from hydrocarbons. Jacket, 1°6°. Minutes. Temperature. 0 1°88 Water and water equivalent of 3 1°88 calorimeter, 3252 grams. 6 1°88 9 4°618 Temperature interval, 12 4°581 4°58) — 1°88) 0025 — ag 15 4°568 18 4°550 21 4°554 24 4°520 These data give 18,900° for the heat of dissociation of ethy- lene. Lixperiment 12.—The bomb contained 1°952 gram of ethy- lene and 5-045 grams of acetylene. The gas after explosion was free from hydrocarbons. Jacket, 0°; room, 0°35°. Minutes. Temperature. 0 0°-+70 Water. 232 eee 3 0°470 Water eq. of calorimeter, 252 6 0°470 9 3690 3452 grams. 12 3°660 Temperature interval, 14 3649 3°66 — 0:'47 +0°032 = 3:222° lef 3°628 20 3604 23 3°587 26 3069 3452 . 3°222 = 11,122 calories, 29 3°590 From these data we deduce for the heat of dissociation of 28 grams of ethylene 11,200° at constant volume and 10,600° at constant pressure. This result is probably the best of the three because about three times as much ethylene was taken as in the two preceding experiments. Lixperiment 13.—The gas used contained 93°7 per cent by volume of ethylene. Hydrogen was assumed to constitute the remainder, as it was found after removing ethylene. The bomb contained 602° of the dry gas at 15°5° and 755:5™™ pressure. The calculated amounts are 0°669 gram of ethylene and 0:0032 gram of hydrogen. 3:2 grams of oxygen were added. Jacket, 1°5°. and of Combustion of Acetylene, Hthylene and Methane. 355 Minutes. Temperature. 0 0:239 Water and water equivalent of 3 0°240 - ' calorimeter, 3252 grams, 6 0°240 9 2°801 Temperature interval, 12 2°783 2°783 — 0:°24+0024.= 2°567° 15 2°769 18 2°753 21 2°739 24 2°724 3252 . 2'567 = 8348° Deduct for hydrogen 109° 8239° and for 28 grams of ethylene, 344,800° The bomb leaked at the valve after exploding in 14th experi- ment. The result was 340,400 calories. Lixperiment 15.—602° of gas (97:2 per cent ethylene) at 21°6° and 762™™ pressure, containing 0°685 gram of ethylene, was mixed with the excess of oxygen. Jacket, 3°; room, 5°. Minutes. | Temperature. Our 1°813 Water and water equivalent of 3 1°824 calorimeter, 3252 grams. 6 1°835 9 4°410 Temperature interval, 12 4°410 4°41—. 1°835 +0°007 = 2°582° 15 4°400 18 4°400 21 4°394 24 4°393 The heat of combustion of 28 grams of ethylene from these data is 3438,400°. | Lixperiment 16.—602° of gas (97 per cent O,H,) saturated with water vapor at 19-4° and 760-4™", containing 0°67 gram of ethylene, was exploded with an excess of oxygen. Minutes. Temperature. 0 0°857 Water and water equivalent of 3 0-860 calorimeter, 3252 grams. 6 0°860 ; 9 3°398 Temperature interval, | 12 3°394 3°394 — 0'860+0°012 = 2°546° 15 3°389 18 3°380 21 3°388 24 3°370 27 3°360 30 3°351 From these data we have 345,700°. 356 Miater— Determination of the Heat of Dissociation The results of the combustion of ethylene, omitting ape ment 14, are Experiment 13 344,800 sc 15 343,400 Son ere 345,700 An average of 344,600 calories at constant volume and 345,800 at constant pressure. Although the results agree well they may be slightly in excess owing to a possible oxidation of the bomb. There was, however, no rust where the steel was exposed and the fine powder washed out of the bomb after an explosion was mostly gold and was nearly free from iron. Thomsen* obtained for the combustion of ethylene 333,350° and Berthellot+ 3841,100° at constant pressure. In deriving the heat of formation we have the difficulty already mentioned” regarding the choice of the numbers to be taken for the heats of combustion of the constituents of ethylene. According to Thomsen (loc. cit.) it is —2710°, and using Berthellot’s figures ‘ we have —7800° for the formation of ethylene from hydrogen and amorphous carbon. If, however, we deduct 341°100 (B) from 330°600, the heat of the combustion of the constituents of ethylene adopted by Thomsen, we have —10,500°% The writer’s direct determinations of the heat of dissociation of ethylene vary too widely to have much significance, but they indicate that ethylene may be more endothermic than has been assumed. Methane. The methane was made in the usual way by heating a mix- ture of sodium acetate and soda-lime, and also of the acetate and barium hydroxide, and the two products were united. The gas was purified by passing it slowly through fuming sul- phurie acid and then allowing it to stand over water containing ferrous hydroxide. A combustion of the gas gave 0°677 gram of CO, and 0572 gram of water, corresponding to 93°8 vol- umes of methane and 6-2 volumes of hydrogen. The gas was fairly free from nitrogen. Lixperiment 17.—The bomb contained 1:18 gram of methane, 4‘765 grams of acetylene and 0:061 gram of hydrogen and nitrogen. Jacket, 14°; room, 0°. The gas left after explod- ing gave no carbon dioxide when burned. Minutes. Temperature, 0 1:034 Water and water equivalent of 8 1°034 calorimeter, 3252 grams. 6 1°034 9 3°589 Temperature interval, 12 3°578 3°578 — 1-084 +.0°023 = 2°567° * Thermochemische Untersuchungen, iv, 65. + Ann. Ch. Phys. [6], xxx, 557. and of Combustion of Acetylene, Ethylene and Methane, 35% Minutes. Temperature. 16 3°559 19 3°544 22 3°530 28 3°517 The experimental result is 8348°: subtracting this number from the thermal effect of the acetylene taken, which is 9768°, we have 1420° required to dissociate 1°18 gram of methane and for 16 grams 19,300° at constant volume. Experiment 18.—Methane 1:431 gram, acetylene 5-467 grams and hydrogen and nitrogen 0:063 gram. ‘The gas after the explosion contained no acetylene, but gave a little carbon dioxide when burned. Apparently some methane was not decomposed. The jacket and room were 0-7°-according to the differential thermometer. Minutes. Temperature. 0 0°36 Water and water equivalent of 3 0°36 calorimeter, 3252 grams. 6 0°36 9 3°28 Temperature interval, 12 3°273 3°273 — 0°36 +0°023 = 2°936° 15 3°260 18 3°245 21 3°230 The result caleulated as before is 18,700°. The mean of the two results with methane is 19,000° volume constant and 18,420° pressure constant required to dissociate methane. Thomsen’s value of the formation of methane at constant volume is 21,170°. Calculated from Berthellot’s data it is 21,500°. 358 Nason— Geological Relations and the Age of Art. XXX VII.—The Geological Relations and the Age of the St. Joseph and Potosi Limestones of St. Hrangois County, Missouri; by FRANK 8S. Nason. | | DurinG the months of March and April, 1901, the writer was engaged in some engineering work for the Derby Lead Co. of St. Francois county. The nature of the work necessi- tated a study of the local geology. Incidentally this led to the exact determinations of the relations existing between the underlying St. Joseph, or Bonne Terre, and the overlying Potosi, as well as the age of the two limestones. The writer is indebted to Mr. O. M. Bilherz, superintendent of the Doe [tun mine at Flat river, for first calling his atten- tion to fossils which he had found, and to Mr. Arthur Thacher and Mr. J. T. Morrell, president and superintendent of the Central Lead Co., for the assistance which they tendered him. The rocks of St. Francois county have a general, but very slight, S.W. dip. The country is hilly, the higher points reaching from one hundred to one hundred and fifty feet above the datum level of Big and Flat rivers. The hills and ridges are not due, in general, to either monoclinal or anticlinal folds, but, so far as is now known, to erosion entirely. Both Big and Flat rivers have flood plains a mile or more in width, thus cutting through the overlying measures. Into these rivers on either side flow smaller tributary streams which have cut more or less deeply into the long divides, breaking them up into more or less hill-like domes. The gulches formed by these streams are dry and almost wholly denuded of soil, leav- ing the nearly horizontally bedded rocks exposed. As the summits of these hills are approached the mantle of residuary clay becomes thicker and on the summits of many of the hills this clay, filled with drusy, cherty quartz, is often fifty to one hundred feet thick. The limestones capping these hills and divides are more or less cherty, having cavities lined with druses of quartz, locally known as mineral blossom. In many places these same limestones have their jointing and bedding planes covered with the same quartz. In one locality the writer found a bed of sandy rock completely honeycombed with their shell-like druses. The bed was at least one foot in thickness. Immediately above and below at least fifty per cent of the rock was also drusy quartz. Underneath these strata the lime- stone becomes almost entirely free from quartz and in general appearance is hardly to be distinguished from the St. Joseph limestones. The occurrence of cherty or drusy quartz in a limestone has hitherto been the sole means of distinguishing St. Joseph and Potosi Limestones of Missouri. 359 between the Potosi and the St. Joseph. If the limestone was cherty it has been called Potosi; if not, St. Joseph, or Bonne Perre. Most cf the gulches cut by the streams expose the contacts between the cherty and non-cherty limestones. As the cherty limestones show well up the sides of the hills, it has been assumed that only the summits of the hills and divides were Potosi. In the geological map of St. Francois County, Mo. (Bull. U.S. G.S. No. 132, “ Diss. Lead Ores of 8. E. Mo.”), Mr. Arthur Wins- low has accepted this erroneous conclusion and further states (2bzd., p. 17): “ The rocks of this formation [ Potosi] are found principally west and north of the area here treated of [St. Francois Co.] and they occur within it only over the hills. The upper limits of the formation are, therefore, not reached, and it is probable that no sharp line of separation between it and the underlying St. Joseph limestone exists.” The writer discovered a positive break between the two series in the bed of Flat river about one mile up the river from Elvins, just a little above where the M. R. and B. T. Rk. R. bridges the stream. Here the river has washed bare a heavy bed of limestone conglomerate. The appearance of the conglomerate is very striking, being composed of flat, round- edged, disks of limestone lying edgewise, as is shown in the accompanying geological column. The conglomerate here appears to be about ten feet in thickness and below the layer of disk-like pebbles, five to six inches thick, it is massive. In the railroad cut above the river exposure, the conglomer- ate is seen to be overlaid with soft clay slates from a few inches to several feet in thickness. From this up to the chert- less Potosi, there isa succession of clay slates interstratified with thin beds of conglomerate. Judging by drill holes and by natural sections, the thickness of the conglomerate series is not less than fifty feet and probably in places it is one hundred and fifty to two hundred feet thick. The conglomerate is not a mere local occurrence. From its outcrop in Flat river the writer traced it continuously for one mile towards the Central Lead Co.’s office; and in isolated exposures for another mile to near the Theodora shaft of the same company. Southeast, it was found on the face of a bluff about two miles from Farm- ington, about four miles air line from the original outcrop. To the west, near Irondale, a distance of about eight miles, drill cores showed it to be present at a depth of two hundred and sixty-two feet. The pebbles ot the conglomerate, as far as seen, are all magnesian limestone. ‘The interstitial paste or cement is with- Am. Jour. Sci.—FourtH Suriss, Vou. XII, No. 71.—NovemsBer, 1901. 25 , 360 Nason—Creological Relations and the Age of | _ | aS ~ | 2s S Residuary Potosi Clay with Drusy | | co Z = Quartz. | aa a as} | eal el | ® Cae: s % Cherty Limestone. QJ g > a — = H c m4 “D Ea = 04 = = Drusy Quartz. Shaly Limestone. Bae Eee a 2 ~ Fossiliferous beds Conglomerates. 2 : 5g S of Slates and Trilobites. So Hs Lingulella. Brachiopods. mee = . ase Pteropods. (S) joanbona Tota Dividing line. Potosi. ce panna erie 2a iomabeess Sa ole 7 S E Upper Lead Zone. = HF ae S d oe co e lay. Byolvagis ai as bY =) iS) w Lower Lead Zone. Linguleila. 2 -~ i = Sandstone. Lingulella 2 3 oa ; 8 . [ — .é ta! “5 o Geological Column near Flat river, St. Francois county, Missourt. St. Joseph and Potosi Limestones of Missouri. 361 out exception a crystalline limestone, nearly, if not quite, pure. Accompanying the conglomerate is an oolitic limestone. Although not traced continuously, to the points noted, the conglomerate is assumed to be identical, for the following rea- sons: The disk-like conglomerate, the crystalline limestone paste, the oolitic limestone, and finally, the invariable occur- rence of fossils, principally trilobites and brachiopods; also beds of interstratified slate. Fossils are very scarce in rocks of both formations save as here noted: Lingulella and Obolella in or near the junction between the La Motte sandstone and the St. Joseph; in the bands of lead-bearing clay slate in the lower and upper lead- bearing zones of the St. Joseph limestone (see geological column) and in the slates of the conglomerate series; in the conglom- erate paste (trilobites, brachiopods, pteropods). The writer feels little hesitancy, therefore, in stating posi- tively that the beds of conglomerate mark a break in geological time between the St. Joseph and Potosi limestones. There is no doubt that this division will be found to be widely extended, but at present nothing can be here offered in support outside of St. Francois county and the eastern edge of [ron county. The Potosi will henceforth be recognized as including and lying above the slates and conglomerates, and will extend far down many of the streams instead of being referred to the hill tops above. As to the age, its determination rests on fossils, and these have been submitted to Professor Beecher, who has generously consented to examine them. 362 C0. E. Beecher—Cambrian Fossils of Missouri. Arr, XXXVIIL — Mote on the Cambrian Fossils of St. Francois County, Missouri ; by C. E. BEECHER. THE small collection of fossils submitted to the writer by IF. L. Nason, for identification, is interesting, especially as it determines the geological horizon of an extensive series of - limestones, sandstones, conglomerates, etc., in southeastern Missouri, the age of which has hitherto been somewhat in doubt. Also, since these strata are intimately associated with the lead-bearing rocks of this region, the identification has considerable economic value. It is stated by Arthur Winslow, ina paper on “The Dis-_ seminated Lead Ores of Southeastern Missouri”* (p. 11), that — although these rocks are placed in the Lower Silurian “ The possibility still remains that there may be a faunal break which will admit of some of the lower strata being classed as Cam- brian, though there is nothing in the stratigraphy to suggest it. This must, therefore, be left to the paleontologists, and owing to the dearth of fossils the problem is not an easy one for them to solve.” In Volume LX of the Missouri Geological Survey (Pt. LV, p. 52, Keyes, 1895) the Fredericktown dolo- mite (= St. Joseph limestone) is referred to the Upper Cam- brian on account of the presence of Lingulella Lamborné (Meek), but since this species is peculiar to the horizon, and the genus has a much wider range, this correlation is not established. A general statement is made by Keyes regarding this region (/. ¢., p. 44) that “ No strata younger than the Cam-’ brian are believed to be represented. But few fossils have been found in the rocks of the area, so that the faunal evidence as to geological age is somewhat meager.” The present collec- tion of fossils, made by Mr. Nason, indicates that the entire series is older than the Lower Silurian (Ordovician) and that at least the upper portion probably belongs to the Upper Cambrian. All but one species of the fossils were obtained from the lower members of the Potosi limestones, and, since this is the topmost formation of this region, its correlation is of the first importance. The fossils occur abundantly in the limestone and conglomerate beds and more sparsely in the sandstones. They consist chiefly of fragments of trilobites, with a few brachiopods and other forms. Lithologically there is a very close resemblance between these fossil-bearing beds and those of a similar horizon in the Black Hills of South Dakota. Limestones, limestone-conglomerates, and sandstones of the same appearance are found in both sections. T'aunally, there * Bulletin No. 132 of the United States Geological Survey, 1896. C. EL. Beecher—Cambrian Fossils of Missouri. 368 is a suggestion of affinity with the Potsdam fauna of Wiscon- sin and Texas. A careful comparison, however, reveals that these resemblances are more general than specific and that the species seem to be distinct. Nevertheless, the facies of this fauna seems to indicate Upper Cambrian, though further studies with additional material may show it to belong to the middle member. Owing to the small number of specimens in the present col- lection, the number of species is necessarily limited. It will doubtless. be considerably increased by future collections. Among the trilobites the genera Piychoparia, Ptychaspvis, Chariocephalus, and Crepicephalus, are more or less clearly identifiable. A species of Chariocephalus closely agrees with the C. onustus of Whitfield. The species of brachiopods seem to be fairly abundant, especially an orthoid shell resembling in some respects Ld- lingsella. It occurs in the shaly partings between the layers of limestone. A species of Acrotreta and a Lingulella are common both in the limestones and arenaceous beds. Hyolithes primordialis Hall and asmall species of Platyceras also occur in the limestones, together with segments of cysti- dean or crinoidal columns. Abundant remains of a linguloid shell are found on the lower, or La Motte, sandstones constituting the basal member of the clastic rocks of the section. Making allowances for differ- ent conditions of preservation, this species may be identified with the Zingulella Lamborni of Meek, which occurs in some green shales of the same age in Madison County, a little fur- ther south. In the absence of other evidence, the diagnostic value of this brachiopod is very slight, and it is impossible to say whether the Bonne Terre, or St. Joseph, limestones and the La Motte sandstones represent Lower Cambrian terranes or whether they with the Potosi all belong to the Middle or Upper Cambrian. The important point of this correlation is that, upon paleon- tological evidence which has hitherto been largely wanting, an extensive area and thickness of sedimentary rocks are definitely placed in the Cambrian. Yale University Museum, New Haven, Conn., June, 1901. 864. «CL EF. Beecher—Cambrian EHurypterid Remains. Art. XXXIX.— Discovery of Eurypterid Remains in the Cambrian of Missour.; by C. E. BEECHER. (With Plate VII.) ma THE wonderful development of merostomes in various parts of the world at about the close of the Silurian has long been recognized, and the suddenness of their appearance out of an apparently clear paleozoic sky has been a matter of considera- ble speculation. Almost at the same instant of time there appeared on the geologic horizon a marvelous assemblage of these ancient arthopods. was heated on the steam bath for 30 minutes, the results agreed with those obtained by the treatment at a volume of 25°, The results of these two sets of experiments are given in sections A and B of Table IV. After the iodine in the experiments which are recorded in section A of Table LV had been titrated by thiosulphate, sulphuric acid was added to the mixture and a blue color appeared immediately, showing, even in these ex- periments conducted at a volume of 25°", that the reduction of the persulphate had been incomplete. This fact suggested a study of the effect of the presence of sulphuric acid on the process conducted otherwise as described by Mondolfo. Accordingly experiments were made similar to those already ‘described excepting that a small amount of sulphuric acid was present. The results of these experiments are given in section C of Table IV. In a blank test made with 1°™* of sulphuric acid present, a small amount of iodine was liberated; and when, in a similar experiment the air over the liquid was replaced by carbon dioxide the amount of iodine liberated was less. The results of these experiments in blank are given in section D of Table IV. Itisseen from the results of experiments given in sec- tions A and 6b of the Table that the method as described by Mondolfo gives constant results provided the volume is small and the heating prolonged. The instantaneous appearance of the iodine color upon acidifying with sulphuric acid showed, however, that a very small amount of the persulphate had escaped reduction. Still, the presence of sulphuric acid in the * Loc. cit. 3872 Peters and Moody— Determination of Persulphates. TABLE IV. : (NH4)2S205 Ammonium Time Na.S.03 calculated persulphate HeS04 of required. from sodium solution, KI. 1:1. Volume. heating. app N/10 found. em’, orm. em?, cm*. minutes. em’. erm, A » ERG 1-0 =e ae 25 10 10°65 0°1200 12°5 1:0 Pek 2 25 10 10°74 0°1210 12°5 1:0 BYR 25 30 10°73 0°1209 B das) 1:0 2 100 10 9°89* 0°1124 ales 1:0 foe gb LOO 10 899% 01018 12°5 1°0 Banas 100 30 10°70 0°1206 12°5 1:0 Tees 100 30 10°72 0°1208 12°5 1:0 eae GOO 30 10°72 071208 | C 12°5 1:0 0 05 25 30 10°78 01214 12:5 1:0 0°05 100 30 10°81 0°1218 1255 1:0 1°0 100 _ 380 LOSE 0°1218 12°5 L:0 1°0 100 30 10°84 0°1221 D Se 120) eres 100 30 0°00 0°0000 Eee 1:0 1°0 100 30 0°07 0°0008 i ae 1:0 EO) 100 30 0°10 0°0011 pe ee 1:0 1L@ 100 30 003+ - 0°0008 process is hardly to be recommended, because of the sensitive- ness of the hydriodic acid set free to the oxygen of the air and the oxygen dissolved in the water. The value of 12°5°™ of the persulphate solution obtained by the method of Mondolfo, found by averaging the results of experiments in sections A and B (omitting the first two experiments under section B) is 01207 grm., while that found by averaging the results of ex- periments in section C and deducting the correction given in the experiments under section D is 0°1208 grm. Method of Namias. In the process for the estimation of persulphates carried out as described by Namias,t a mixture containing 12°5°™* of the persulphate solution, 1:0 grm. of potassium iodide, and water sufficient in some experiments to make a volume of 25°", and in other experiments 100°, was allowed to stand 11 hours in a stoppered bottle and the iodine set free was titrated by thio- sulphate. After the titration by thiosulphate the mixture was allowed to stand still longer, and 16 hours later the blue starch iodide color which had developed in the experiments with the * Color returned. + Air over liquid replaced by carbon dioxide. t Loe. cit. Peters and Moody—Determination of Persulphates. 373 volume of 100°™ was destroyed by thiosulphate. No more color appeared on standing. I - given in section A of Table V. In section B of Table V are recorded experiments in which the mixtures were allowed to Ammonium persulphate solution. em?, 12°5 12°5 12°85 12°5 KI. erm. (ell coe cell ee SiS eS Seesceocaogea pe py pp ee py pe py ee ee ee ee ee Se eGgeace oo ee @cto eS liquid. em?. 25 25 100 100 25 25 100 100 100 25 25 100 100 25 100 100 25 25 25 25 100 100 100 100 100 25 100 25 100 TABLE V. Time of standing. hrs. A 26 26 26 26 B 20 24 20 24 24 20 25 20 24 Atmos- phere above liquid. alr air air air (NH4)oS208 calculated from iodine found 11 hours standing. erm. 0°1203 0°1196 O:1191 0°1190 em3, The results of these experiments are After standing the number of hours indicated in the fourth column. erm. 0°1203 0°1196 0°1207 0°1209 O°1211 0°1213 O°1211 O°1211 0:0000 0°1216 0°1214 0°1212 0°1217 0°1248 0°1220 0°1226 Oss, 0°1212 O-1222 0°1223 0°1214 0°1214 0°1214 0°1219 0°1220 0'0014 0°0006 0°0011 0°0003 374 Petersand Moody—Determination of Persulphates. stand 20-25 hours, the air about the liquid in some experi- ments being replaced by carbon dioxide. Though earbon di- oxide shows no tendency to liberate iodine in the blank test, the persulphate appears to liberate in its presence a little more iodine than in its absence. This experience naturally suggests an effect of acidity. The effect of the presence of sulphuric acid in the process otherwise conducted as described by Namias was therefore tried, the air over the liquid being replaced by carbon dioxide. The results of these experiments are recorded in section C of Table V. | It is seen from the experiments conducted as described by Namias that the reduction of the persulphates in the time stated by Namias, 10-12 hours, is plainly incomplete unless the volume of the liquid is small. If the experiments are con- ducted at the greater dilution the time of standing must be increased in order that subsequent standing may not result in the return of color. In the presence of sulphuric acid more iodine is liberated than in its absence, and the amount is greater when the air above the liquid is not replaced by carbon dioxide, and when the volame is small. The amounts of iodine liberated under similar conditions in blank experiments prove to be appreciable and naturally greater when the atmosphere above the liquid is air. When these amounts, averaging the equivalent of 0°0007 grm. of the persulphate when the atmos- phere is carbon dioxide, and about 0:0010 gram when the atmosphere is air, are deducted from the actual indication, the figures agree well with those found for the same amount when sulphuric acid is not present. Plainly, the addition of the acid adds nothing to the regularity and value of the process. The value of 12°5 em® of the persulphate solution obtained by the method of Namias is 0°1208 grm. | Arseniate—Iodide Method. The estimation of chlorates is accomplished according to Gooch and Smith* by allowing a mixture containing the chlorate, a definite amount of standard potassium iodide solu- tion, an arseniate, 20 of 1:1 sulphuric acid, to boil from a volume of about 100°™* to 35™*; the difference between the amount of iodine required to oxidize the arsenious acid pro- duced and the amount of iodine in the potassium iodide origi- nally present being the measure of the chlorate taken. It is interesting to note that persulphates can be estimated in a similar manner. Mixtures containing 12°5™ of the persulphate solution, 0°5 grm. of potassium iodide, 2°38 grm. of hydrogen * This Journal xlii, 220. Peters and Moody— Determination of Persulphates. 375 potassium arseniate, and 20° of 1:1 sulphuric acid, and water enough to make the total volume about 100™*, were boiled in a trapped Erlenmeyer beaker until the volume decreased to 35°, and the arsenite present after the solution was made alkaline with potassium bicarbonate was estimated with iodine. - The results of these experiments are given in Table VII. TABLE VII. Todine required Iodine (NH4)2S20¢ Ammonium for liberated equivalent persulphate KI oxidation by to iodine solution. present. of arsenite. persulphate. liberated. em?. erm. erm. grm. grm. 12°5 0°1875 0°0514 0°1361 0°1225 12°5 0°1875 0°0522 0°1352 0°1217 12°5 0°1875 0°0514 0°1361 0°1225 12°5 0°1875 0°0516 0°1359 0°1222 These results agree closely with one another and in the aver- age, 0°1222 grm., accord well with the results of the process of Griitzner and of LeBlane and Eckhardt. To compare the values obtained for the persulphate solution the averages of the results obtained by the different methods, together with the average of all the experiments, are given in Table VI. TABLE VI. Number of Average of results. Process. experiments. grm. Mend oliGts2 Soest iy i 6 0°1207 Nanas ed Sey ens igs 8 0°1208 LeBlane and Eckhardt .-.---- 1 CAF 7 Gritzner (corrected). ==. - 0-1219 Arseniate-iodide method .-_-. 4 0°1222 Average of whole _--- 0°1213 The results obtained by the process of Mondolfo and the process of Namias, both of which involve the liberation of iodine from potassium iodide and the titration of that iodine by thiosulphate, are practically identical and lower than the results obtained by the other three methods. The results obtained by the process of LeBlane and Eckardt in which the persulphate is reduced by a ferrous salt, by the process of Gritzner in which an arsenite solution is the reducing agent, and the arseniate iodide method in which the persulphate is determined by the difference between the amount of iodine in Am. Jour. Sci.—FourtH Series, Vou. XII, No. 71.—Novemper, 1901. 376 Peters and Moody—Determination of Persulphates. an iodide added and the amount necessary to oxidize the arsen- ite remaining after boiling the solution, are all in quite close agreement and are all higher than those obtained by the pro- cesses of Namias or Mondolfo. It is readily seen that the results of the experiments obtained by the process of LeBlane and Eckardt are nearer the average of the results of all the methods than those of Namias or the similar process of Mondolfo, and this fact would show that the statements of Namias regarding the inaccuracy of the method of estimating persulphates by the reduction with a ferrous salt is without foundation. The process of LeBlane and Eckardt is simple, rapid and convenient. The method of Gritzner is advantageous in that the ordinary arsenite solution is the standard for the pro- cess, but requires the application of a correction. The method of Mondolfo is simple and fairly rapid, but tends to give low results. The method of Namias is slow because the experi- ments must stand 10 to 12 hours, and the results of these ex- periments, like those obtained by the method of Mondolfo, are lower than the results obtained by the other methods. The arseniate-iodide method introduced as a control, is accurate ‘but less simple than the other methods. We wish to thank Professor F. A. Gooch for oe given during the progress of this investigation. Wortman—Studies of Kocene Mammalia, ete. B77 Art. XLI. — Studies of -Hocene Mammalia in the Marsh Collection, Peabody Museum ; by J. L. WortTMAN. [Continued from p. 296.] The Vertebre.—With the exception of the right half of the atlas, and perhaps more than half of the caudals, the vertebral column is present in its entirety ; it is preserved, moreover, very nearly in its natural position, with the vertebre inter- locked in such a manner that there can be no doubt whatever of the presacral formula. The atlas bears a very strong resemblance to that of the felines. The transverse process is proportionally less extended and does not project so far behind the facets for the axis ; it is emarginate behind, considerably thickened, and pierced by the canal for the vertebral artery much as in the cats. The ante- rior margin of the process displays a deep notch, in which. the artery winds forward in its course to the snus atlantis. This latter has much the same relative size, position, and relation as in the felines. The cup-shaped articular facets for the con- dyles are large and roomy, the tubercle for the transverse ligament prominent, and the superior arch is of good breadth. As Scott has remarked of Jesonyx, the axis exhibits some peculiarities, the most conspicuous of which is the very large backwardly projecting spine. The odontoid is rather long and conical, the atlantal facets are large and slightly convex, the body is strongly keeled, and the transverse process long and tapering. The spine rises high above the body; it projects comparatively little forward, reaching only opposite to the root of the odontoid, but posteriorly it is produced into a_ long tapering process, which overhangs nearly the entire third cer- vical. The remaining cervicals, like the atlas and axis, are proportionally stout and heavy, evidently in relation with the large size of the head. ‘Their neural spines increase rapidly in height, that of the last almost equaling in length that of the first dorsal. The bodies of all the anterior cervicals have strong inferior keels which bifureate posteriorly and terminate in two long protuberances. The plan of arrangement of the transverse processes is similar to that of the Carnassidents. The costal processes or inferior lamellz are small in the ante- rior vertebree, but increase rapidly in size to the seventh. The | vertebra promimens has a simple transverse process, which is not perforated by the vertebrarterial canal. The dorso-lumbar formula is 19, as in the Marsupials, whereas it is very generally 20 oe ‘the Carnassidents. There is reason to believe that all the Creodonts had the marsupial 3878 Wortman—sStudies of Locene Mammalia in the rather than the carnassident formula.* The dorsals are twelve in number, and of these the neural spine of the first is broad, flat, and greatly elongated; it is the longest of the series, those of the succeeding vertebree gradually decreasing in height to the eleventh or anticlinal vertebra. The spines have a very backward slope up to this point, where they change abruptly to a forward direction. This, as far as I am aware, is the point at which the anticlinal occurs in most of the Carnassi- dents. On the eighth dorsal, metapophyses make their appear- ance and increase in size posteriorly, but there are apparently no representatives of the anapophyses developed. The lumbars are seven in number; they have short, broad, neural spines and heavy, keeled centra, increasing in size from before backward to the antepenultimate, which is the largest. The transverse processes are long and thin, with a forward and downward direction. The zygapophyses are powerful and, as Cope has observed, are closely interlocking. The articulation is effected by means of the half-cylindrical postzygapophyses fitting into half-cylindrical prezygapophyses. They differ materially from those of some other Creodonts, notably Patriofelis, in which they are sigmoid in cross-section. The sacrum is somewhat damaged, but enough remains to show that the auricular process was large and rngose, that the spines were of moderate size, and that there were very prob- ably three vertebree entering into its composition. It is rela- tively very narrow. The caudals are much weathered and many are missing. The proximal ones are short and propor- tionally but little larger than those of the dog. Those from a more posterior position are long and slender, but not much more so than those of many of the modern Carnassidents. There is no means of judging of the length of the tail, but m comparison with such a type as Patriofelzs, the tail was reduced and slender. This supposition comports well with the highly-developed cursorial habits of these forms. Ribs and Sternum.—There are twelve pairs of ribs, and for a member of the Carnivora, the anterior ones are remarkably broad and flat; they are, indeed, much more like those of an Ungulate than a Carnivore. The first is short and stout. The third is the broadest, those posterior to it becoming gradually narrower to the eighth, which is slender and subround in trans- verse section. All the ribs have a well-developed tubereulum and capitulum, with the exception of the last three, in which the tuberculum is absent. As compared with that of the dog, the capacity of the thoracic cavity was apparently considerably smaller. Some four or five of the sternebree, figure 45, are * An exception to this must be made in the case of Oxyena lupina, in which the formula is the same as in the Carnassidentia. It is probably likewise true that its successor Patriofelis had the same formula, Marsh Collection, Peabody Museum. 379 preserved with the skeleton, but not being in place, it is not absolutely certain just how they should be arranged. They are so strikingly different from the narrow elongate sternals of any of the modern Carnivora, that, were they found separately, one would searcely suspect that they belonged to an animal of this order. On comparison with the corresponding bones of the Marsupials, however, especially those of an opossum, figure 46, they are seen to bear a very decided resemblance,— so marked, in fact, that I have no hesitancy in arranging them after this species. The presternal piece is missing as well as { ; 2 FigurRE 45.—Ventral view of the sternum of Dromocyon vorax Marsh; one- fourth natural size. (Type.) Figure 46.—Venutral view of the sternum of an opossum, Didelphys virginana ; natural size. 46 the xiphisternum, but apparently all the mesosternal segments are represented. ‘The anterior one of these is long and narrow, but the succeeding ones widen rapidly and become characteris- tically broad and flat. There is evidence of five of these pieces, the last being the widest, as in the opossum. This latter seg- ment has a thickened ventral keel which is confined to the posterior end; its truncated extremity furnishes a thickened 380 Wortman—Studies of Hocene Mammalia in the base for articulation with the xiphisternum. This decided marsupial aspect of the sternum adds but another character. to the already full list of osteological features in which the two groups resemble each other. Fore Limb.—The scapula bears a general resemblance to that of the felines. A prominent spine divides the external surface into two subsequent fossee, and terminates proximally in a rather short blunt acromion. The metacromion is not a distinct process, as in the cats, but consists of a more or less distinct lamina overhanging the glenoid border. The neck of the bone is relatively longer than that of either the dog or the cat. The glenoid cavity is oval and rather shallow, and the coracoid is small. The coracoid border is much curved and is interrupted by a very wide superscapular notch. The glenoid border is straight and much thicker than the coracoid border. The area for the origin of the teres major is very distinct, being separated by a well-defined ridge from the main part of the posterior fossa. Lying immediately beneath the glenoid cavity, imbedded in the matrix, was found a small splint-like bone, which without doubt represents the clavicle; it apparently has an articular surface at one end, probably for contact with the acromion. The bone is broken so as not to display its full length, but, judging from its size, it is hardly probable that it reached the manubrium. It appears to resemble the clavicle of the Dasyure more than that of any of the living Carnivores. : The humeri are both considerably crushed in the proximal half of their extent, and, on this account, do not display the characters of this region very distinctly. The head is very convex from before backward and somewhat pointed behind. The greater tuberosity is prominent and extends above the level of the head; the postero-external border is drawn out into a broad laminate process, which reaches well backward. The lesser tuberosity is prominent, and there is a broad shallow bicipital groove. The character of the deltoid crest is much obscured on account of the crushing, but it appears to be con- siderably reduced from the more primitive Creodont condition. The distal end resembles that of the dog more closely than that of any other living mammal; it is much compressed from side to side—more so than in the dog,—the external condyle is little protuberant, and the supinator ridge is much reduced. The internal condyle projects behind and slightly below the edge of the trochlea, and is separated from it inferiorly by a deep notch. The anconeal and anticubital fossee are very deep, and are placed in communication with each other by means of a large supertrochlear foramen. There is no entepicondylar foramen. The distal articular surface is divided by a very distinct rounded ridge into capitular and trochlear portions, both of Marsh Collection, Peabody Museum. 381 which are very concave from side to side, and convex from before backward. In its transverse concavity, the capitulum differs from that of the dog, in which it is nearly flat. The humerus and scapula are of nearly equal length. The ulna, figure 47, differs from that of the dog in being shorter, heavier, and stouter in every way. The olecranon is of great proportional length, and in this respect, like so many other Creodonts, betrays its marsupial re- lationship. The proximal end is traversed by a vertical groove, the inner or radial lip of which is un- usually prominent. The sigmoid cavity is deep; its posterior wall is hook-shaped, and considerably over- hangs the articular surface. The shaft is somewhat flattened from side to side and is traversed by a deep longitudinal groove. The sty- loid process is rather short and — obtuse. The radius, figure 47, is about equal to that of the dog in size, although somewhat shorter; it is not quite as long as the humerus, and is a little longer than the fore foot. The proximal articular sur- face is flattened from before back- ward, and the proximal extremity articulates with the entire distal end of the humerus. This surface may be described as having a central concavity, with a rounded sloping edge internally, and an anteriorly beveled articular face, occupying the antero-external angle. The ar- ticular surface for the ulna is much flattened as in the dog, and is con- fined entirely to the under side, so that the power of pronation and supination was not greater than in the living Canide. The tubercle for the insertion of the biceps is placed upon the outside, just below the proximal extremity, as in the dog, and is small. The shaft is oval in transverse section, slightly curved, and has a well-marked ridge upon its 47 Figure 47.-—-Right ulna and radius of Dromocyon vorax Marsh; front view; one-half natural size. (Type.) The radius is represented as slipped down below its natural position. 382 Wortman—Studies of Eocene Mammalia, ete. inner border. The distal extremity is enlarged, rugose, flat- tened from before backwards, and presents many sulci for the passage of tendons. The articular surface is concave in both 48 4 ‘ i 3 i UU i Sut) i Tien My —_ Figure 48.—Right manus of Dromocyon vorax Marsh; front view; three- fourths natural size. (Type.) c, cuneiform; wu, unciform; m, magnum; tm, trapezium; td, trapezoid; s, scaphoid; ce, centrale; J, lunar; p, pisiform; III, phalanges of third digit. directions, and is divided by a distinct fore and aft ridge into an outer and an inner facet for contact with the scaphoid and lunar, respectively. [To be continued. | Adams— Carboniferous and Permian Age, etc. 383 Art. XLII.—TZhe Carboniferous and Permian Age of the Red Beds of Eastern Oklahoma from Stratigraphic Evi- dence ; by GEORGE I. ADAMS. [Published by permission of the Director of the U. 8. Geological Survey.] VARIOUS opinions have been expressed from time to time concerning the age of the Red Beds of Oklahoma. By some they have been called Triassic and by others Permian, but because of the general absence of fossils and the lack of strati- graphic work they have remained an uncertain group. Collec- tions of fossils made by C. N. Gould at White Horse Springs, sixteen miles west of Alva, from the Red Bluff formation of Cragin, situated one hundred feet or more above the gypsum ledges, have been determined by Schuchert and Beede as Upper Permian forms.* Vertebrate remains from Orlando in Logan County and from Hardin in Kay County, although not fully studied, are considered by 8. W. Williston as equivalent to Cope’s Lower Permian fauna from the Wichita beds of Cum. mins in northern Texas.t The writer has recently done some stratigraphic work which has a direct bearing on the problem of the Red Beds, and it is believed that it supplies a correct interpretation of the beds and furnishes a basis for future detailed study of them. In tracing the outcrops of the limestone formations of the Carboniferous of Kansas, the writer observed that in going southward there is a gradual transition in the character of the sediments to those which are more arenaceous, and that there is a thickening of the shales and sandstones and a thinning and final disappearance of some of the limestones. Moreover, in describing the shales of the higher portion of the Carboniferous and the lower portion of the Permian, the occurrence of purplish and maroon-colored shales was noted. The signifi- cance of these observations was not fully known, since the work of the Kansas survey was limited by the State line. In study- ing the oil and gas fields of Kansas and Indian Territory, the writer traced to the southwestward the extension of the out- crop of the Fort Scott limestone from the southern border of Kansas into Indian Territory. It extends from west of Chetopa in Kansas to Chelsea, Claremore and Catoosa, and thence to the Arkansas river west of Weer, when it becomes inconspicuous. The horizon as marked by the associated sand- stones was followed to a point between MHoldenville and * Amer. Geologist, vol. xx, No. 1, p. 46. + Article not yet published. ¢ Forthcoming Bulletin U. 8. Geol. Survey. 384 Adams—Carboniferous and Permian Age of the Wewoka. This made possible the correlation of the Kansas section of the Carboniferous. with the Indian Territory section thus far worked out by Mr. Joseph A. Taff and the writer.* It appears that there are in the Choctaw Nation 9,000 feet of shales and sandstones above the lowest productive coal, which are lower than the Fort Scott limestone and its equivalent, while in southeastern Kansas there are but 450 feet of shales and sandstones between the Fort Scott limestone and the Missis- sippian or Lower Carboniferous. It will be observed that the line of outcrop of the limestone as above traced diagonals the divisions of the Coal Measures as drawn by N. F. Draket and the horizon extends into the area which he erroneously called Permian. - : , 2 en ineee The results of tracing this horizon strengthened the writer’s: conviction that similar work in the higher portion of the sec- tion would determine the relation of the Red Beds to known formations in Kansas, and in June a trip was made through the Osage Nation into Oklahoma. The previous geological work which had been done in this locality by Drake and by Gould consisted of sections across the rocks and did not permit of accurate correlations.t The formation which was selected to be traced is the limestone described in the writer’s field notes as the Elk Falls. It occurs. about 700 feet below the base of the Permian as determined by Prosser. It was chosen because, from its thickness and its relation to the adjacent for- mations, it was believed that it would be found persistent for a considerable distance southwestward. The line of its outerop was followed from near Hewins, Kansas, and it was found to pass just west of Pawhuska, where it is the equivalent of the Pawhuska limestone, named by J. P. Smith,| and mentioned in the sections. by Drake and Gould. In southern Kansas, there are two heavy ledges of limestone separated py shales. In the Osage Nation, although it has not been previously so noted, there are three, all of which are persistent as far as the Arkansas river, although considerably thinner southward. The line of outcrop crosses the Arkansas river at Blackburn and continues to Ingalls, where it is the same as the limestones mentioned by Gould in his section made east of that place. Southwest of Ingalls the limestone becomes thinner. Its strike will carry it across the Cimarron river near Perkins. [rom Ingalls the route followed by the writer was to Ripley and thence to Chandler. In traveling southward, the shales and * Coalgate Folio, U. S. Geol. Survey. | + Proc. Amer. Philos. Soc., vol. xxxvi, p. 326. ¢{ This Journal, March, 1901, p.. 185. § Vide vol. iii, Kans. Univ. Geol. Surv. || Jour. Geol., vol. ii, p. 199. fed Beds of Eastern Oklahoma ho SSS sandstones associated with the limestones were seen to become gradually redder in color. South of the Arkansas river they are typical Red Beds, In going from Ingalls to Chandler and = = — ==> dP SaRES ere: ; u U mO; Qu lex) — Jee OO) G 1 i=} > o Sala: ee ae LEGEND PERMIAN UPPER LOWER CARBONIFEROUS CARBONIFEROUS 4 Stroud, one passes over formations which are much lower geo- logically, and they are likewise red in color. It appears, therefore, that rocks in Eastern Oklahoma which have been referred to the Red Beds on lithologic grounds, are in part of Upper Carboniferous or Coal Measure age. The sedimentation from the Carboniferous into the Permian is an 386 Adams—Carboniferous and Permian Age, ete. unbroken sequence. From what is known of the Permian limestones of Kansas, they will be found, when followed south- ward, to diminish in thickness, and this change will be accom- panied by a transition to more sandy beds. This is in accord- ance with the observations made by Mr. Gould. The age of that portion of the Red Beds which is in strike with the Permian of Kansas may confidently be expected to be found to be of Permian age. This is in accordance with the evidence already furnished by the vertebrate fossils. Above the Permian limestones in Kansas occur the Wellington shales, which are bluish and greenish gray in color. They are prob- ably represented southwestward by formations which are red. The succeeding formations are typical Red Beds, and have thus far yielded only Permian fossils. Upon the accompany-' ing map (p. 385) the approximate line of transition in color has been drawn with the purpose of showing that it is diagonal to the strike of the Carboniferous and Permian formations. Chemistry and Physics. 387 SCIENTIFIC INTELLIGENCE. I. CHEMISTRY AND PHYSICS. 1. A New Method for Producing Aniline and the Analogous Bases.—\t has been found by SaBaTIER and SENDERENS that free hydrogen, acting at a slightly elevated temperature in the pres- ence of certain finely-divided metals, is capable of attaching itself to the molecules of a number of unsaturated organic com- pounds, and that this method of hydrogenation is capable of very readily transforming nitro-benzol and analogous nitro-com- pounds into aniline and the corresponding bases. Copper (pre- pared by the reduction of its oxide) at a temperature of about 300° to 400° acts rapidly upon a mixture of nitro-benzol and hydrogen, the latter being in excess. ‘The reaction is prac- tically complete, the aniline is almost colorless, and the metal is not altered by continued use. Nickel acts even more readily than copper, and it may be used for this reaction at about 200° ; but it is liable to cause an excessive fixation of hydrogen, for at 250° some ammonia’and benzol are formed, while at 300°, with a large excess of hydrogen, methane and ammonia are the chief products. Platinum-blacksponge, and platinized asbestos effect the reduction of nitro-benzol and similar bodies at a temperature of about 230°-210°. Water-gas or purified illuminating-gas may be substituted for hydrogen in these reductions; in fact, in the case where copper is used, carbon monoxide takes part in the reaction, being oxidized to carbon dioxide to a considerable extent. In view of the cheapness of water-gas, it is expected that this method will find practical application in the manufac- ture of aniline, in place of the customary reduction in the wet way by means of a metal and an acid.— Comptes Rendus, cxxxiii, 321. H. L. W. 2. On the Existence of Ammonium.—An unsuccessful attempt to isolate ammonium has been made by Orro Rurr. In the first place, potassium iodide was dissolved in liquified ammonia in a cell which branched into two compartments at the bottom, and an electric current was passed through the liquid by means of platinum wires which passed through the bottoms of the two branches, the cell being kept at a temperature of —70°. Under these conditions very small drops of a liquid having a metallic appearance with the color of copper were formed at the negative pole, and these dissolved in the ammonia, giving the intense blue color which potassium-ammonium, KNH,, is known to give. Similar experiments were then conducted with ammonium iodide in place of the potassium iodide, but in this case hydrogen was continually given off, and no blue color, nor any other evidence of the formation of ammonium, was observed. The last experi- ment was repeated with a closed cell so that the evolved hydro- 388 Seientifie Intelligence. gen would produce a high pressure; but even at a calculated pressure of 60 atmospheres the appearance of the products did not change. ‘These experiments indicate that free ammonium is incapable of existence even at —70° and at a high pressure.— Berichte, xxxiv, 2604. H. L. W. 3. A Modified Gooch- Crucible—W. C. HERarus is manufac- _ turing filtering crucibles, the bottoms of which are made of spongy platinum. This device was suggested to the manufac- turer by Dr. Neubauer of Breslau and is called the Neubauer crucible. It is stated that the porous bottom is strong, and that, with reasonable care, precipitates may be scraped off ‘with a spatula. The crucible filters rapidly with a good suction-pump, and yet retains the finest precipitates, such as freshly precipitated barium sulphate and silver chloride. Such exceedingly fine pre- cipitates and gelatinous ones, however, soon form a layer through which the liquid cannot pass, so that ‘this modification has the same limitations in filtering as the ordinary Gooch-crucible where an asbestos filter is used. ‘The Neubauer crucible possesses the disadvantage that it cannot be used for filtering and igniting precipitates which cannot be dissolved out of the porous platinum. It is stated that, on this account, barium sulphate should not be weighed in these crucibles ; but, since this substance is readily soluble in hot, concentrated sulphuric acid, there is no doubt that it would present no obstacle to a chemist acquainted with the proper solvent. The advantages of neariy constant weight, and of the possibility of dissolving precipitates in hydrofluoric acid, appear to be the chief reasons for using the new device in prefer- ence to the customary one.—Zetischr. fiir angewandte Chemie, XXXVH, 923. H. L. W. 4. Radio-active Lead.—It has been mentioned previously in this department of the Journal that Hormann and Srrauss have obtained radio-active lead salts from various rare minerals. These authors have recently described further investigations upon the subject, and are still of the opinion that this radio-active sub- stance differs in many respects from ordinary lead in its chemical behavior, but they have not yet obtained it in a pure state. They find that, while the various salts of radio-active lead have nearly the same action upon the electroscope, only the sulphate is capable of acting upon the photographic plate through glass or aluminum ; moreover, the sulphate becomes particularly active in the latter respect after having been evaporated with nitric and sulphuric acids, or after having been ignited at 450° for 15 hours with access of air. It appears also, that while radio-active lead sulphate acts more energetically upon the photographic plate than certain polonium (radio-active bismuth) preparations, the latter have a much greater action in discharging the electroscope. The authors believe that it follows from this that the rays detected by photographic means are not identical with those which produce electric discharges.— Berichte, xxxiv, 3033. u. L. W. Chemistry and Phystes. 389 5. A Salt of Quadrivalent Antimony.—It has been found by We ts and Merzerr that the black, or rather very dark blue, salt, Cs,SbCl,, which was described by Setterberg in 1882 as erystallizing in short prisms, is really octahedral in form, and that it crystallizes isomorphously with Cs,PbCl,, giving crystals of various shades of green when mixed with’ this yellow salt. The octahedral form is characteristic of a large number of double chlorides of quadrivalent elements, like K,PtCl,; hence it is probable that Setterberg’s salt contains antimony ‘tetrachloride. Attempts were made to prepare SbCI,, or to find some evidence of its existence, but without success. It should be black in color, and the oxide corresponding to it should be black also ; hence it is probable that the well known oxide SbO,, being white, does not correspond to the tetrachloride.-— Amer. Chem. Jour., xxvi, 268. H. L. W. 6. A New Method for the Gravimetric Determination of Tel- lurium.—GuTBIER has found that hydrazine bydrate and the salts of this base readily precipitate tellurium quantitatively from solutions of all tellurium compounds, and he highly recommends this method for the determination of this element. The precipi- tation is made in a nearly neutral or in a hydrochloric acid solu- tion, and the precipitated metal is made flocculent before filtering by boiling the liqnid. It is remarkable to notice that the author recommends weighing the tellurium on a paper filter, since the Gooch-filter is far better adapted for the purpose. The test- analyses given are satisfactory.— Berichte, xxxiv, 2724. H. E. W. 7. Diffusion of Hydrogen through Palladium.—The depend- ence of this diffusion upon pressure of the diffusing gas is the subject of a study by A. WinkELMANN. It was found that the quantity of hydrogen which diffuses through glowing palladium is not proportional to the pressure of the hydrogen. With diminishing pressure the quantity of gas is greater than this supposition demands. If one supposes that a dissociation of hydrogen occurs,-and that the diffusing quantity is proportional to the pressure ‘ofthe dissociated molecules, then the facts can be sufficiently explained. It is evident on this supposition that only the atom and not the molecule of hydrogen passes through the glowing palladium.— Ann. der Physik., No. 9, 1901, pp. 104— 115. Seed! 8. Cathode Rays.—The emission theory of these rays appears to be the prevailing one, at least in Germany. W. Serrz has undertaken a study of the questions, what is the action between the emission particles and the material atoms or molecules in case that the latter can be considered at rest in comparison with the great velocity of the rays, whether one has to deal with fric- tional forces or conservative forces, and whether attraction or repulsion is exerted. For the study of these questions there appear to be two ways open: the study of reflection and that of the phenomena presented by the passage of the rays through thin membranes. It was found that the cathode rays are dif- 390 Scientific, Intelligence. fusively reflected from metal surfaces connected to the earth. With aluminum, zine, iron, and copper, the maximum of intensity of the reflected rays was displaced toward the mirror. While with platinum, silver, and gold, the reflection was strongest in a direction between that of the incidence ray and the normal to the metal surface. In the case of the first class of metals the reflection increases with the angle of incident ; and with the last class diminishes. At perpendicular incidénce the reflective power increases with the atomic weight. Zinc departs slightly from this law. The absorption of cathode rays in thin plates is independent of the tension. The absorption-coefficient increases with the thickness of the window through which the rays pass. Lenard’s law that thin plates of different metals, having equal masses per unit. of surface, absorb the same fraction of the inei- dent rays, is only a first approximation. The increase of the absorption-coefficient, with the thickness of the window, cannot be explained by the assumption that the electrons lose their velocity by friction with the molecules and are then absorbed, for no difference is observed between the magnetic and electric deflection of the rays in the case where they pass direct from the cathode, or in the case where they have passed through an aluminum window. Moreover, the rays emerging obliquely from the window behave in the same manner as those emerging along the normal to the window. The following values were obtained: v = 0°703.10"; -. — 0°645.10".—Ann. der Physik, No. 9, 1901, ( . 1-33, Jena an Photography of the Infra-hed Spectra of the Alkali-metals. —H. Leumawnn describes the method of sensitizing plates for the red portion of the spectra, and adopted Burbanks’ sensitizing bath, which is constituted as follows : Distilled. water .cc24e2- eae 2h Sees 16030 Cyanine solution (Burbanks) -..-.----- los Ammonia (Sp. iF, OOM 4 seattle Loy Dilivermitmate (ll ¢AO)\pes ck 42 = Flee 5 drops This enabled the author to photograph the spectra of glowing metallic paper as far as1000 pu. The arrangement of spectrum apparatus is fully described. Five new lines of rubidium and nine of cesium were discovered.—Ann. der Physik, No. 7, pp. 633-658. J. T. 10. Hither and Gravitational Matter through Infimte Space.— In the course of an article on this: subject, Lorp Kenvin thus remarks upon the old hypothesis that if we could see far enough into space the whole sky would be seen occupied with discs of stars all of perhaps the same brightness as our own sun, and that the reason why the whole of the night sky and day sky is not as bright as the sun’s disc, is that light suffers absorption in travel- ing through space. Chemistry and Physics. 391 ‘“Now we have irrefragable dynamics proving that the whole life of our sun as a luminary is a very moderate number of million of years, probably less than 50 million, possibly between 50 and 100 million! To be very liberal, let us give each of our stars a life of a hundred million years as a luminary. Thus the time taken by light to travel from the outlying stars of our sphere to the center would be about three and a quarter million times the life of a star. Hence if all the stars through our vast sphere commenced shining at the same time, three and a quarter million times the life of a star would pass before the commence- ment of light reaching the earth from the outlying stars, and at no one instant would light be reaching the earth from more than an excessively small proportion of all the stars. To make the whole sky aglow with the light of all the stars at the same time the commencement of the different stars must be timed earlier and earlier for the more and more distant ones so that the time of the arrival of the light of every one of these at the earth may fall within the durations of the lights at the earth of all the others.” — Phil. Mag., August, 1901, pp. 161-177. Jers 11. Nineteenth Century Cloud over the Dynamical Theory of Heat and Light.—Lorpv Kertvin classifies these clouds as fol- lows: Cloud I. Relative motion of ether and ponderable bodies. Lord Kelvin discusses the various theories in regard to the con- stitution of the ether, and in the course of this discussion remarks upon the suggestion of Fitzgerald and of Lorenz, that the motion of ether through matter may slightly alter its linear dimensions— “according to which if the stone slab, constituting the sole plate of Michelson and Morley’s apparatus, had in virtue of its motion through space occupied by the ether, its linear dimensions short- ened one one-hundred-millionth in the direction of motion—the result of the experiment would not disprove the free motion of ether through space occupied by the earth.” The author still regards cloud I as very dense. Cloud II arises from a proposition first enunciated by Waterston, which is as follows: “In mixed media the mean square molecular velocity is inversely proportional to the specific weight of the molecule. This is the law of the equilibrium of vis viva.” Without any knowledge of the paper of Waterston, Maxwell, in 1859, enunci- ated the following proposition: “Two systems of particles move in the same vessel : to prove that the mean wis viva of each parti- cle will become the same in the two systems.” This is also Waterston’s proposition regarding the law of partition of energy. Lord Kelvin does not regard either the proof of Waterston or that of Maxwell successful. The subsequent theorems of Boltz- mann and Maxwell in regard to the equality of mean kinetic energies, are also criticised. Lord Kelvin, after an exhaustive criticism of the Boltzmann and Maxwell law, quotes Lord Ray- leigh’s remarks on the difficulties connected with the application of the law of equal partition of energy to actual gases (Phil. Am. Jour. Sc1.—FourtH Series, Vou. XII, No. 71.—NovemsBer, 1901. 27 392 Scientific Intelligence. Mag., Jan., 1900), and remarks upon the sentence, ‘‘ What would appear to be wanted is some escape from the destructive sim- plicity of the general conclusions.” ‘The simplest way of arriving at this desired result is to deny the conclusion, and so, in the beginning of the twentieth century, to lose sight of 4 cloud which has obscured the brilliance of the molecular theory of heat and lhght during the last quarter of the nineteenth century.”—Phil. Mag., July, 1901, pp. 1-40. Bee 12. The Resistance and Electromotive forces of the Electric Arc.—W. DuppE tu has carried on a series of experiments having as their object the measurement of the resistance of the electric are under various conditions, as also of the electromotive forces present. ‘The method adopted involved the use of a steady cur- rent to which a testing current of high frequency (120,000 periods per second) was added. ‘The measurements of impedance and the power factor were made by the three voltmeter method, employing (1) an alternator for high frequency ; (2) a new, very delicate instrument, called a “‘thermo-galvanometer,” to measure the three voltages and (3) a standard resistance for comparison of the impedance of the arc, having a time constant of only 2°7 ~ 10~" second. Some of the results obtained are as follows. The true resistance of an arc 3™™ long between 11™™ solid carbons, through which a 9°91 ampere current was flowing, was 3°81 ohms. The whole dif- ference of potential was 49°8 volts, of which 37°8 was accounted for by the ohmic drop, the real EK M.F. opposing the flow of cur- rent being 12 volts. With cored carbons the resistance was found to be 2°54 ohms and the back E.M.F. 16°9 volts. With increase of the direct current the back E.M.F. first decreased and then increased, the minimum value being 11°3 volts for 6 amperes. With the direct current constant, increase of arc length increased the resistance. Further the composition of the electrodes was found to have great influence on the resistance and back E.M.F. Thus with direct current and arc length constant, soaking the © solid carbons in potassium carbonate reduced the _ resistance from 8°81 to 2°92 ohms and increased the back E.M.F. from 12 to 15°6 volts. It is suggested that perfectly pure carbons would probably make the resistance so great that the mainten- ance of the arc would be impossible, traces of impurities being essential The author concludes that the are resistance consists of three parts; thus, for the case above noted, of (1) the resistance at or near the contact of the positive electrode and the vapor column of 1°61 ohms; (2) the resistance of the vapor column of 2°5 ohms ; (3) the resistance between the latter and the negative carbon ot 1:18 ohms. The back E.M.F. consists of two parts located at or near the contact between the electrodes and the vapor column. That at the positive electrode, about 17 volts, opposes the flow of the direct current, while that at the negative electrode, about 6 volts, helps the flow of the direct current, i. e., is a forward K.M.F. Chemistry and Physics. 393 The greater part of these two E.M.F.’s are considered as due to thermo-electric forees, and special experiments are noted which support this view.— Proc. Roy. Soc., No. 450, p. 512. 13. Unsolved Problems in Low-Temperature Research.—An essay by Miss Acnres M. CrierxKeE, author of “A Popular His- tory of Astronomy during the Nineteenth Century,” upon Low- Temperature Research at the Royal Institution of Great Britain, 1893-1900, has recently been issued. It gives an admirably clear and complete summary of this most interesting subject, and closes with the following chapter. “The development of low-temperature chemistry is one of the most striking features of scientific history during the last decade of the nineteenth century. Many questions of profound interest have been answered through its means, and a partial insight has been gained into some of the most recondite secrets of nature. The unique condition attends it, that the ne plus ultra cannot recede as it advances. The absolute zero forms an irremovable landmark, a boundary line that may not be transgressed, an asymptote, as it were, to the curve of future progress. And every step nearer to it is harder to take than the previous one. Among many causes of augmenting difficulty is the circumstance that the molecular latent heats of vaporization diminish with the absolute boiling point. Hence, a continually more lavish expendi- ture of frigorific material is necessitated, and of material the price of which, in money and labor, rises rapidly with its frigorific efficacy. Still, although the bottom of the tempera- ture-scale may never be actually reached, the intervening space will surely be much abridged. But we shall never, it is safe to predict, assist at the ‘death of matter.’ At the stage arrived at, there is no sign of its being moribund. Forces still act within and upon it. Gravity and cohesion maintain their normal power. It sensibly impedes the passage of electricity in the purest and most highly conducting metals. Its minute particles can take up and modify luminous vibrations. Only chemical affinity seems to be extinct ; the various species of matter cease to react upon each other. The next cryogenic achievement, it is true, may alter the situation as we now see it. Our present standing- ground may be subverted, for the inquiry is just now in a critical phase. The liquefaction of helium, for example, may prove decisive of many things—it may set at rest some doubts, and raise unJooked-for issues. The conditions for its accomplishment were clearly set forth in the Bakerian Lecture. They may be realized by the use of methods actually available. This last fortress of gaseity cannot be regarded as impregnable, although its capture will be at a high monetary cost. Gaseous helium, to begin with, is of the utmost scarcity ; and what is scarce demands outlay to procure. Its condensation can be effected only by subjecting it to the same process that succeeds with hydrogen, substituting, how- ever, liquid hydrogen under exhaustion for liquid air as the pri- 394 _ Seventific Intellagence. mary cooling agent. As the upshot, a liquid will be at hand, boiling at about 5° absolute, or —268° C., but more expensive than liquid hydrogen, in a much higher ratio than liquid hydro- gen is more expensive than liquid air. By comparison, ‘ potable gold’ would be a cheap fluid. Nor could the precious metal, in that, or any other form, be employed for a higher intellectual purpose than in promoting and extending researches of such boundless promise and commanding interest as those conducted at the Royal Institution.” II. GroLoGy AND MINERALOGY. 1. Geological Survey of Canada. G. M. Dawson, Director. Annual Report (New Series) Vol. XI. Reports A, D, F, G, J, L, M, 8, for 1898. Pp. 853, 20 platesand figures in text, 7 maps. Ottawa, 1901.—The first of these, the Summary Report of the Director, has already been mentioned in these pages. Report D on the Yellowhead Pass Route is by James McEvoy. The region traversed extends from Hdmonton and Strathcona on, the north, Saskatchewan River westward to the Athabasca River valley, and thence up through the Yellowhead Mountain pass and into the Frazer River valley, ending at Téte Jaune Cache. The formations, their names and correlations, are as follows, viz : Tertiary Paskapoo beds hae Ciciadeous Edmonton beds Laramie. Pierre and Fox Hill. Devono-Carboniferous. Capea oe Mountain group. ow River series. Archean Shuswap series. D. B. Dowling, in F and G, reports on the Geology of the West shore and Islands of Lake Winnipeg (F), and on the Hast shore of Lake Winnipeg and adjacent parts of Manitoba and Keewatin, from personal surveys and supplemented by notes made by his predecessor, J. B. Tyrrell. The formations recognized are compared with those described in the Geology of Minnesota, Vol. II. The correlations are as follows, viz: ? Utica | Stony Mountain formation = Richmond Group. ( Upper Mottled limestone = Maclurea beds. T | Cat-head limestone = Fusispira and Nematopora renton 4 Teale [ Lower mottled limestone = Clitambonites beds. ? Black River Winnipeg sandstone and shales. Collections of fossils from some of these formations have already been described by J. F. Whiteaves in Paleozoic Fossils, Vol. ILI, Parts II and IIL. Geology and Mineralogy. 395 R. W. Ells contributes report J on the Geology of the Three Rivers Map-Sheet, Quebec. North of the St. Lawrence the Potsdam, Calciferous, Chazy, Black River and Trenton forma- tions are recognized. Lists of the fossils collected from them are reported by H. M. Ami. The “ Medina” outlier south of the St. Lawrence is defined. A. P. Low reports (L) on the South shore of Hudson Strait and Ungava Bay. The rocks there met with are ancient crystal- lines, igneous intrusions, altered shales and schists, dolomites and iron ores, the latter possibly of Cambrian age. Robert Bell contributes report M on the topography and geology of the northern side of Hudson Strait. The fossiliferous rocks are of Trenton and Niagara age, and Devonian fossils are reported from the southern side of Southampton island. A remarkable feature in the geology on the north shore of the strait are the twelve bands of white crystalline limestone alternat- ing with bands of gneiss. Dr. Bell estimates the total thickness of these limestones to be not less than 30,000 feet. Lists of the plants of Hudson Strait and of the Lepidoptera of Baftin Land are appended. The closing reports are R, The Report of the Section of Chem- istry and Mineralogy, by G. Christian Hoffmann, and 8, The Report of the Section of Mineral Statistics and Mines by E. D. Ingall. H. S. W. 2. Geological Survey of Canada. Ropert Betz, Acting- Director. Catalogue of the Marine Invertebrata of Eastern Canada ; by J. F. Wuirravzs ; pp. 1-271. Ottawa, 1901.— The Catalogue is a report on the present state of our knowledge of the marine invertebrata of the Bay of Fundy, Atlantic Coast of Nova Scotia, and gulf and mouth of the River St. Lawrence, as far as the Strait of Belle Isle. H. S. W. 3. Geological and Natural History Survey of Minnesota, 1900-1901. Vol. VI of the Final Report. Geological Atlas with Synoptical descriptions ; by N. H. WincueEtu. 88 geograph- ical and geological plates with accompanying descriptive text.— In this closing volume of the survey, begun twenty-eight years ago, the author offers the following as his final classification of the geological formations of the State of Minnesota: ( Ft. Pierre | Niobrara I't. Benton Dakota. 1 l j Hamilton Cretaceous Devonian Mareellus ? Corniferous. Hudson River Galena» Trenton. Lower Silurian 396 Scientific Intelligence. ( St. Peter | Shakopee | Richmond i Lower Magnesian Jordan St. Lawrence | Dresbach | Hinckley sandstone, passing down into ( Potsdam Red sandstone, interbedded with Manitou Lavas, and Potsdam Quartzite (at New Allen ) Middle and the Puckwunge Conglomerate Cambrian. Upper Cambrian St. Croix Series. . | Cabotian (igneous). Rarenie ‘ Red Rock Apobsidian red granite |! | Gabbro and its varieties and Lavas | Lower | l Animikie — | Cambrian. Slates and Quartzite Taconite and Quartzite J ( Granite, Syenite, etc. Laurentian. ican dae } epee schists and gneiss Coutchiching. arious fragmentals K Kawishiwin Greenstone pete. | HS. W. 4. Iowa Geological Survey. Vol. XI. Administrative Reports. SamuEL Carvin, State Geologist. Pp. 1-579. Pls. i—xii, figures 1-43. 9 maps. Des Moines, 1901.—This volume carries forward the detailed Survey of the State of Ohio by the present staff of geologists to cover the counties of Louisa, Marion, Pottawat- - tamie, Cedar, Page, Clay and O’Brien. In the report on Cedar County we note that Professor Norton transfers the Coggan formation from the Silurian, to which it was previously referred, to the Devonian system, with which it is more naturally referred by its few though diagnostic fossils. H. 8. W. 5. Dragons of the Air, an account of Extinct Flying Reptiles, by H. G. SEELEY; pp. 1-232, and 80 illustrations. , London, 1901 (Methuen & Co.).—The close kinship between a fascinating novel and a well-written popular book of science is illustrated by Professor Seeley’s story of the “Dragons of the Air.” ema pee Pepe DARIN py Le So i Sa 70 Sandstone. oe ose iee Se Se oe en re a eee crete 10 Hard brown shalewe ese eee ee eee aye Ye 20 Imvensely, salt) water tabe sae ete = 305 The lower 200 feet of this section clearly belong to the Moencopie beds. Ward—Geology of the Little Colorado Valley. 405 The Shinarump. This constitutes a vast series with a maximum observed thickness of at least 1600 feet. It presents a number of phases, some of which are so distinct that if studied in only one locality they would naturally be regarded as separate sub- divisions, but such a general survey as I have been making points to a certain homogeneity in all these beds, or at least establishes the unmistakable tendency towards the recurrence in any of the phases of features that are prominent in other phases. The Shinarump constitutes the horizon of silicified trunks and there is no part of it in which fossil wood does not occur in great abundance. It also marks the limit of the wood-bearing deposits of this region. For this reason alone, in view of the etymology of the name, I should be justified in extending the Shinarump as far as the fossil trunks oceur, and it is obvious from the language used that Major Powell had the upper portions of the formation in view as well as the lower when giving the name, although other geologists in speaking of the Shinarump usually seem to have in mind only those beds which I include: under the conglomerate. It is doubtful, however, whether the remainder of the formation has really been studied or carefully observed by others, and lL fancy that in dealing with it I am entering upon a sort of geological terra incognita. The Shinarump Conglomerate.—I am using this expression, which is the one most commonly found in works that treat of these beds, in a somewhat comprehensive sense, the necessity for which will be apparent. As thus used this part of the Shinarump occupies the lower half of that series and has a maximum thickness of 800 feet. Although perhaps the most prominent feature of it is the so-called conglomerate, which sometimes is in truth deserving of that name, and contains somewhat large but always well-worn pebbles and cobbles derived from underlying formations, still, it rarely happens that this aspect of the beds constitutes the major portion of them. In the first place the conglomerate tends to shade off into coarse gravels and then into true sandstones. These sandstones are of a light color, contrasting strongly with the dark brown sandstones of the Moencopie beds already described. They are, moreover, always more or less cross-bedded and usually exhibit lines of pebbles running through them in various directions. These are true sandstones, very hard, devoid of alumina, and scarcely affected by the winds, so that their angles are usually sharp and the ledges they form are abrupt and jagged. Although the sandstones proper generally oceur lower down, still, there is no uniformity in this arrange- ment, and sandstones are often found in the middle and con- 406 Ward— Geology of the Little Colorado Valley. glomerates more rarely at the top. But in addition to these — the Shinarump Conglomerate embraces other classes of beds. There is a well-stratified layer of thinnish sandstone shales that 1s often seen immediately under the heavy sandstone cap. Some of these shales have a grayish color and are highly argillaceous. These layers tend to thicken even within the formation itself, but especially farther out, and what is more significant, they often become transformed into a bluish white marl. This condition can be seen between the beds of con- glomerate in places where the Shinarump Conglomerate is ~ comparatively thin, as in the lower valley of the Little Colo- rado, where it is only about 300 feet in thickness. This. . feature is not very prominent, but at other places, as in the Petrified Forest region where the Shinarump attains its maxi- mum thickness of 700 or 800 feet, this tendency on the part of certain beds to become transformed into marls is the most marked feature of the formation. The marls here occupy much more than half of the beds. They are very varied in color, showing besides the white and blue tints a great variety of darker ones such as pink, purple, and buff. These heavy marl beds, of which there may be several in the same cliff, are interstratified between conglomerates, coarse gravels, and cross-bedded sandstones, all of which taken together form the beautifully banded cliffs that are seen throughout the Petrified Forest, and especially along its northern flank. It thus becomes necessary to include under one designation all of these varying beds, which often change the one into the other even at the same horizon within short distances, and rather than adopt a new name I have preferred to call them all the Shinarump Conglomerate. _ It remains to mention certain minor features, which are not universal,’ but which, nevertheless, have considerable import- ance. In the lower Little Colorado Valley there occur numer- ous somewhat calcareous clay lenses, the lime taking the form of bright white stripes, while the clay is usually purple or pink. These are very distinct objects and vary in size from lenses 10 or even 20 feet in length to small lenticular blocks or somewhat oval or even spherical clay balls or pellets. These calcareous clay inclusions are scarcely seen farther to the south- east, but on Red Butte they are well marked and here the clay becomes brilliant red and constitutes a true paint stone. Another fact to be noted in connection with the Shinarump Conglomerate is that at certain localities, and notably on Red Butte, there is at its base a clear indication of a transition to the Moencopie beds. The conglomerates proper are under- lain by argillaceous shales closely resembling those of the Moencopie beds, but beneath these is a sandstone ledge which cannot be referred to the lower division, as it is more or less Ward— Geology of the Little Colorado Valley. 407 cross-bedded, possesses considerable grit, and has small clay pellets included in it similar to those of the trne conglomerate series, in which J have for this reason included it. This con- dition of things may be somewhat puzzling from the strati- graphical point of view, but the disadvantage in this respect is much more than compensated for by the evidence that it furnishes in favor of the view that all of these beds really con- stitute one great system, and as opposed to the view which it may be inferred that certain geologists hold, that the series of beds which I have included under the name of the Moencopie beds belongs to a different system, and are in some way con- nected with the underlying Paleozoic rocks. This view, in the light of the above mentioned facts is, in my opinion, quite untenable. The Le Roux Beds.—Under the name of Le Roux beds I include the remainder of the Shinarump, deriving the name from Le Roux Wash,* which enters the Colorado Valley two miles below Holbrook, and on which some 15 miles north of Holbrook this series attains the greatest development that I have observed, probably reaching its maximum of 800 feet. These beds, too, if studied at localities where they are less developed, might be supposed to form several quite distinct subdivisions. Indeed I was of this opinion during most of my stay in the lower Little Colorado Valley, but even before leaving there the proofs of their homogeneity had become abundant. At least the lower half consists of that remarkable formation in which I found vertebrate bones in 1899 and in which alone thus far vertebrate remains have been observed. I have some- times designated it as the Variegated Marls, sometimes as the Belodont beds. The distinguishing features of these beds is the presence of great numbers of small buttes, the smaller ones appearing to be blue clay knolls, but the larger ones show- ing other colors, especially purple, and sometimes several bands of different hues. Almost everywhere at this horizon there exist plains, dotted all over with these remarkable little buttes, varying from 3 or 4 feet to 20 or 30 feet in height, usually iso- lated from one another and having a form peculiar to them. They are not couical in the true sense of the word, since they do not rise to a point atthe summit, but are always rounded off and have the form of a well made haystack, the smaller ones looking like haycocks in a field. These butte-studded plains are of course simply the remains of a plateau or mesa which has *The name “ Leroux’s Fork” was given to this wash by Lieut. Whipple’s party, who followed it down some distance and encamped at its junction with the Little Colorado on Dec. 5, 1853, this being their Camp 79. See Pacific Railroad Reports, vol. iii, parti, p. 75. The name is written in two words on the Land Office map of Arizona. 408 Ward— Geology of the Little Colorado Valley. been worn away, primarily by the action of water, but for a very long period there can be no doubt that wind has been the more potent agency. There is evidence throughout that entire region that the amount of precipitation was formerly much greater than at present, and in so speaking I do not refer to avery remote date geologically, but to a period which was probably post-Tertiary. Indeed, from the present condition of many of the regions where we know that the early Indians dwelt, and who must necessarily have had access to water, now perfectly dry, with all sources of water so remote that they can no longer be inhabited, it must be inferred that there has been a change in the climate within the period of hnman oceupancy. Certain it is that water is doing very little relatively in this region now, while the agency of wind is conspicuously marked . wherever it can produce effects. The peculiar form of these buttes is not such as water could have produced, while it is precisely the form that wind would naturally produce, acting upon the very fine and soft materials, somewhat resembling ashes, that compose these buttes. Further evidence of this, if any were needed, is found in ses fact that in approaching the general escarpment, which bounds these plains, the buttes tend to lose their isolated character and form ridges projecting out from the cliffs. It never happens that an entire valley or plain is covered by a single system of buttes. These systems are separated by wide intervals, often of nearly flat country, but through which it can be easily seen that water once flowed, at least in the form of temporary floods, and in such a manner as to have swept away every vestige of the former plateau, and in crossing which there are encountered one or several wide beds to which the term “wash” is popularly applied. In descending the Little Colorado this condition of things is not met with until within some 8 or 10 miles of the Lee’s Ferry road. A large system of buttes is then found extending some 5 or 6 miles down the river and across the plain to the first terrace, a distance of 3 to 5 miles; then occurs the first wash, 2 miles in width, followed by another system of buttes, which is nearly due east of Tanner’s Crossing, and in which most of the bones were collected by our party. There is then another wide wash, but the next system of buttes does not reach the river, but trends off in a direction nearly due north. There is still another wash before the great Moencopie Wash is reached, the direction of which is such as to be highly favorable for the preservation of these buttes, and accordingly we find their greatest development, so far as this region is con- cerned, along the Moencopie Wash. They do not however follow. the stream up in the direction of Tuba City, but con- tinue to trend northward along the wide valley that lies to the west of Willow Springs and Echo Clifts. Ward—Geology of the Little Colorado Valley. 409 The reason why these conditions are not earlier met with in the valley of the river is simply that the river does not follow the line of strike, and these beds, being common to the entire formation, must always occupy the same horizon. Above the point mentioned, therefore, they must be looked for farther in the interior. Wefound them in fact five miles east of Black Falls, or 25 miles southeast of Tanner’s Crossing. The great bend in the river culminating at Winslow keeps these beds constantly so far to the northeast, and in a region where it is so difficult to penetrate, that their exact condition for a distance of over 50 miles is little known. But farther up the river, where they approach somewhat to the region of settlement, they again admit of access, and as already remarked, they appear in great force in the valley of Le Roux Wash. Here they cover an area of nearly 100 square miles and form two great amphitheaters of veritable bad lands, but in which the great variety and symmetry in the form of these buttes and ridges, as well as the variegated and iridescent colors that prevail, ren- der them a magnificent spectacle. They can be seen from the southeast for a distance of 20 miles as a white line. Viewed from the top of the mesa ont of which they have been carved, the denudation having been arrested at a particular point, they reveal more completely than at any other place the true char- acter of this formation. In the Petrified Forest the Le Roux beds are also well developed and the variegated marls are found only half a mile east of the Lower Forest. The buttes here are quite large and well developed and bones of the Belodont occur in them. Inthe northern part of the Petrified Forest region the variegated marls he somewhat farther to the eastward. What is called the Middle Forest lies in the midst of them, and the petrified wood, as everybody has observed, differs here con- siderably in its constitution and coloration from that of the upper and lower forests, which lie in the horizon of the con- alomerate series. As was remarked when treating of the conglomerates, these variegated marls are actually found stratified’ between the sand- stones by the transformation of certain shales into marls. If these beds are carefully traced a short distance in the direction of the dip, they will be seen to thicken very rapidly and soon to take on the character of the true variegated marls. As they start from underneath a bed of sandstone which caps the con- glomerates, and which does not so readily pass into marl, the buttes that are first formed are usually topped out by a block of this sandstone, and it is necessary to proceed some distance farther in the direction of the dip to reach a point where the sandstones disappear. This however ultimately takes place and the marl beds thicken to such an extent 410 Ward—Geology of the Little Colorado Valley. that they have to be regarded as virtually overlying the con- glomerates. In fact, in the bed of the Moencopie Wash, on both sides of which these beds are so well developed, the con- glomerates can be seen distinctly passing under the marls. So much for the variegated marls, which, for the purposes of our expedition, constituted the most important subdivision of the entire.formation. But as we have seen, their maximum thickness is about 400 feet and there remain still another 400 feet before we reach the base of the painted cliffs. Throughout the whole of this fossil wood is abundant, but the character of the beds as variegated marls no longer continues. In the lower Colorado Valley, where I know it best, the varie- gated marls are sneceeded by a sandstone ledge at least 100. feet in height, yielding black logs of very fine structure. At this point these sandstone beds constitute an escarpment and form a small terrace, the summit of which is a dip plane. Upon this lie the remains of the next set of beds, which are somewhat remarkable, primarily in being essentially limestones, but they consist mainly of loose material somewhat resembling dried mortar, for which reason I have designated them mortar beds. They are, however, very irregular in structure and contain much impure flint and large flinty stones. In the midst of them there occurs a true limestone ledge, well stratified, succeeded by a continuation of the mortar beds. In the region mentioned these beds extend to the limit of what I regard as true Shina- rump, and petrified wood was found above the limestone ledge. A wider acquaintance with this part of the formation shows that the conditions above described do not hold at all points and may even be regarded as exceptional. Nowhere else except at Black Falls did I find the lower sandstone ledge, and at most other points the limestones gradually supervene npon the variegated marls. In fact, it should be remarked, that not only the variegated marls but also the shales of the conglomerate series, which become transformed into marls, are more or less calcareous; and when we find that the entire upper portion of the Shinarump consists mainly of limestones and calcareous materials, we may regard all of this, including the variegated marls, as virtually a calcareous deposit. If we were to look abroad for its homologue in the Trias of the Old World we would find itin the Muschelkalk, while the conglomerate series might well be compared with ‘the Buntersandstein, and the Painted Desert beds with the Keuper, to which the French term Marnes Irisées is only locally applicable. In the extensive exposures on Le Roux Wash these relations are brought out with great force. Overlying the true varie- gated marls which stretch ont for a distance of three miles across the broad eroded valley, the limestone series comes in grad- ually and scarcely differs except in the degree of calcareousness Ward—Geology of the Little Colorado Valley. 411 from the underlying beds, but the limestone ledge is ultimately reached and is sharp and definite. It has a thickness of about 10 feet. Over it lie very heavy beds of calcareous materials beginning as mortar beds, such as I have described, but soon taking on more symimetrical forms, closely resembling the marl buttes of the valley below. The color also changes, and many of the buttes are, in whole or in part, of a deep blue or a lively purple. These constitute here the highest beds of the Shina- rump and fossil wood isabundant throughout. Much the same conditions prevail in the Petrified Forest region, but the devel- opment is here much less extensive. The Painted Desert Beds. It remains to consider the third and highest series of the Older Mesozoic of Arizona. As already stated, these constitute the elevated cliffs that bound the valley of the Little Colorado on the northeast. Although broken through in many places, and practically wanting for long distances, they still constitute what may be regarded as a great wall separating the valley from the region of high mesas that lie in the Mogui and Navajo country. As these beds seem to contain no fossil remains, and as they are throughout the greater part of their extent practi- cally inaccessible from the absence of water, their detailed study has been neglected, and I was able to acquaint myself with them only imperfectly and at a few points. There is, however, no place where they are better developed than directly east of Tanner’s Crossing, where we remained longest, and on several occasions the attempt was made to reach them from our camp and to examine them closely. Enough was learned to justify the positive statement that they consist _ almost entirely of sandstones, perfectly stratified, the different layers differing mainly in color, thickness, and fineness of struc- ture. The great central portion constituting the escarpment and having a thickness of about 800 feet is, within these limita- tions, practically homogeneous. The series begins, however, with a bed of orange red sandstone, highly argillaceous, and soft in structure, easily eroded, and readily yielding to the influence of the wind. It has a thickness of about 100 feet and in the lower Colorado region stretches across the broad valley at the base of the escarpment and les directly upon the uppermost limestones of the Shinarump. Here it forms picturesque and fantastic buttes and chimneys standing out upon the plain. It occurs in the same position overlying the Shinarump on Le Roux Wash and forming the top of the mesa which overlooks the amphitheatres that | have described. It is also seen above. the Shinarump series to the east of the Petrified Forest. It is therefore probably safe to assume that this bed is continuous from Echo Cliffs to the boundary line of New Mexico. 412 Ward— Geology of the Little Colorado Valiey. Of the painted cliffs, considering the little that is known of them, there seems to be nothing more to say. In looking at these cliffs from a distance it is seen that they are overlain by a white formation, the nature of which it is important to con- sider. Before we had visited the region, so as to obtain a close view of them, it was natural to suppose that they might consti- tute Jurassic limestones and that the Triassic system might terminate at the line which separates them from the variegated sandstones. But upon close examination this was found not to be the case, and these white rocks were found to consist of sandstones often very pure and cross-bedded, with scarcely any admixture of marl. These without question constitute the sum- mit of the Triassic system in this region. They are, however, . not always white, or at least in some places, as for example in the vicinity of Tuba City, they are underlain by a still thicker bed of soft brown sandstone, which is somewhat argillaceous and easily worn by the wind, forming chimney buttes and ruins. This bed has a thickness along the headwaters of the Moen- copie Wash of about 200 teet and is overlain at the highest points by the white sandstones to a thickness.of 100 feet more. These sandstones are very porous and all the waters that fall in that region immediately pass through them, but as they approach the summit of the much harder and firmer beds that constitute the lower portions of the series these waters are arrested and come out in the form of springs, sometimes almost of small rivers, along the crest of the cliffs above the Moen- copie Wash. It is on one of these springs that the little Mor- mon town of Tuba City is located, and this is true also of Moa Ave, Willow Springs, and other settlements in that country. Still farther back the Cretaceous lignites and limestones lie uneonformably upon these uppermost sandstones of the Trias, and the Jurassic is wanting altogether. The following columnar section of the strata of the Little Colorado Valley will make the above descriptions more clear. DESCRIPTION OF THE SECTION. (See page 413.) Feet Feet 1. “Argillaceous: shales: i 2 soe: 100| 6. Shinarump Conglomerate .-_-_~-- 800 2 .Caleareous Ghales says s.r 100)": ‘Vianievated murlse2s2 2 a= = eee 400 3. Argillaceous shales .__.__-_.- 200.2 8.2 Sandstones 2a. Jee 2 eee 100 4; Sandstones.2). 22-10. (282 100; 9: Timestoneded poses = eee eee 20 5. Argillaceous shales ._._.-2:=_- 200)| 10. Mortar bedS)= =e se 2 4-5 eee 80 —| 11. Calcareous marls____-----_--- 200 Total thickness of Moencopie beds 700 — Total thickness of the Shinarump 1,600 Feet 12s Orange, red sandstone So) ee eee 100 13, Wariegated Sandstones=. =) 2-2 eee eee 800 14 sBrowhsand ‘stonesc +. 22 Oe a Ee ee 200 15: ) White sandstones} its (ah rep ree ek 100 Total thickness of Painted Desert beds-_. 1,200 Total thickness of Trias. .... 3,500 Ward—Geology of the Little Colorado Valley. 418 White sandstones. De ee a Brown sandstones. Painted Desert beds. Variegated sandstones, regularly | stratified and brilliantly colored; !% the well-known Painted Cliffs. Red-orange sandstones. Caleareous marls sometimes worn into buttes. Mortar beds, flint stones. * Sandstone ledge. Le Roux beds. Variegated marls argillaceous and caleareous with bones of belo- donts, labyrinthodonts, and dino- saurs. ‘dmnaeuryg Conglomerates snd coarse cross- bedded sandstones with clay lenses interstratified with gray argillaceous shales and varie- gated marls. Shinarump Conglomerate. Dark chocolate-brown argillaceous shales; saliferous. Argill, sandstones, soft, dark brown. Moencopie beds. Argillaceous shales, dark brown Calcareous shales, white. Saliferous shales (=3 and 5), Carboniferous (Upper Aubrey). Unconformabie limestone or sand- stone. * 9, Limestone ledge, definitely stratified. 414 HT. N. Stokes— Pyrite and Marcasite. Art. XLIV.—On Pyrite and Marcasite; by H. N. Sroxes.* WHILE pyrite and marcasite are usually readily distinguished by their crystalline form, there remains a residuum, consisting of massive or finely-grained concretionary material, in which this is not possible. The light brass color of pyrite and the tin- white of marcasite, which can be seen when the surfaces are freshly cleaned with acid and compared with standard speci- mens in a good white light, are not always readily made out in concretions where the surface is rough and the mineral frequently contaminated with other substances. The density too may be misleading. According to Rammelsberg and- to my own determinations, pure marcasite has a density of about 4:90,+ while that of pyrite is 5:00 to 5:04. The density of even well crystallized specimens of pyrite varies very considerably, and as a criterion of the presence or absence of marcasite is practically worthless, especially in mas- sive or concretionary material. A series of crystallized pyrites, which were shown by the method to be described below to be free from mareasite, gave densities varying from 5-04 to 4°82, while a pyrite concretion, also free from mareasite, gave only 4:56. A. A. Julient has employed the density to determine the relative amount of pyrite and marecasite in mixtures, and on this method he has based the hypothesis that most specimens of these minerals, even when well crystallized, are intimate mixtures of the two, passing into complete paramorphs. It will be shown below that this view is untenable. The greater rapidity of oxidation is sometimes used to dis- tinguish the minerals. While it is unquestionably true, that under precisely similar conditions marcasite vitriolizes more rapidly than pyrite, conclusions based on this fact have usually failed to take into account the important factor of the ratio of surface to mass; a compact, brilliant, marcasite is stable while a porous mass of pyrite vitriolizes readily. Only when we know this factor can we draw any conclusion from the fact of rapid vitriolization. Penfield has described a method for distinguishing the min- erals, based on the fact that boiling nitric acid liberates sulphur from marcasite but not from pyrite, a method which is clearly not adapted to detecting pyrite in the presence of mareasite. * This paper is a condensation of Bulletin No. 186 of the United States Geolog- ical Survey and is published by permission of the Director. + The density 4°80 given by Julien (Annals N. Y. Acad. Sci., vol. iv, 1887, pp. 177, 210) is certainly too low. tl. c. pp. 166, 213. H. N. Stokes—Pyrite and Mareasite. 415 There is no method hitherto known by which even a qualita- tive determination of the two is possible in mixtures. In connection with the study of the action of ferric salts on sulphides, I was struck by the quantitative difference of the behavior of pyrite and marcasite, and out of this has been developed the following method for distinguishing them and for determining them quantitatively in mixtures. That ferric salts oxidize the sulphur of pyrite to sulphuric acid was observed by de Koninck* and the same has since been noted by other observers, without an attempt having been made to follow the reaction quantitatively. The study which I have made of this reaction shows that under uniform and easily controllable conditions the percentage of the total sul- phur oxidized is very constant for each mineral, but differs greatly in the two, and in mixtures is an index of the amount of each constituent present: To avoid tedious gravimetric: determinations I employ a standard solution of ferric ammo- nium alum containing 1 gram Fe’ per liter, with 4 grams free sulphuric acid to prevent the formation of basic salts. An excess of the carefully prepared minerai is boiled with this solution under absolute exclusion of air until the reduction of the ferric salt is practically complete. The reaction may be regarded as taking place in two stages: (1) FeS, + Fe,(SO,), = 3FeSO, +28. (2) 28+6Fe,(SO,),+8H,O = 12FeSO,+8H,SO,. Since the active mass of the solid phase is constant, it matters not how much mineral be taken for a given volume of the solution provided it be in excess, nor is it necessary to know the absolute quantities of iron and sulphur involved in these reactions, or the strength of the permanganate solution. It suffices, in deducing an expression for the oxidized sulphur, to employ symbols expressing the permanganate equivalent of the iron, and the symbols used represent simply the volumes of permanganate consumed by a given volume of the solution. For a given volume let @ = iron in the original solution, 6 = resulting ferrous iron, = resulting total iron. Then ¢ —a@=increment of iron resulting from decomposition of FeS,, and ata) (c—a) = total sulphur in decomposed sulphide. (A) * Annals Soc. Geol. Belgique, vol. x, 1883, p. 101; Zeitschr. anorg. Chem., vol. xxvi, 1901, p. 123. 416 LH. N. Stokes—Pyrite and Marcasite. Also, 3 (c—a@) = ferrous iron produced according to equation (1) - and 6-3 (c—a) = ferrous iron produced by oxidation of sulphur. According to equation (2), 1 atom of sulphur requires for oxidation 6 atoms of ferric iron, producing 6 atoms of ferrous iron: hence, 1 31°83 6 55°60 Calling the percentage of sulphur oxidized p, we obtain from (A) and (B) (0—a(e—2)] = sulphur oxidized. (B) 100 31°83 a OEE OM ens. (By 6 55°60 . 8°3336 ee _— 25. 31°83 C—a 24 hea) 55°60 It thus appears that three titrations suffice to determine the percentage of sulphur oxidized, and that neither the amount of FeS,, the volume of the ferric solution used, nor the absolute titer of the latter or of the permanganate need be known. As, however, the proportion of sulphur oxidized varies with the strength of the ferri¢ solution, the extent of reduction, and the temperature, it is necessary, in order to obtain comparable results, to use a solution of standard composition and a standard temperature, and to continue the action to complete reduction. The value », or the percentage of sulphur oxidized, may be called the oxedation coefficient. The oxidation coefficient of pyrite, as established by the study of various specimens, varies between 60 and 61 with a mean of 60:4, while that of marca- site varies between 16°5 and 18. In duplicate determinations, properly made, it may differ about one-half unit. It is clear then that in these figures we have perfectly characteristic con- stants of the two minerals. It may be noted that the figures are independent of any contaminations which do not affect the titer of the solution, or which can be removed by previous treatment. . The explanation of the oxidation coefficient and its difference in the two cases is probably the following: the molecules of FeS, are acted on by a reagent which is capable of oxidizing the iron ‘more rapidly than the sulphur. The rate of solution of the pyrite is greater than the oxidation rate of the sulphur, hence a portion of the latter, under the given conditions about 40, per cent, escapes and when once in the free state, as I have found, is scarcely attacked; at the same time the limit of the HT, NV. Stokes—Pyrite and Marcasite. 417 oxidation rate of the iron is not reached. Marcasite dissolves much more rapidly, the result being that a still greater portion of the sulphur, about 82 per cent, escapes. The same expla- nation naturally applies to their different behavior towards nitric acid as observed by Penfield. It is not an indication of the different chemical constitution, but of a different solution tension conditioned by different crystalline structure. Doubt- less asimilar method can be applied in other cases of dimorphism to determining which form is more soluble.. I propose to apply the method to distinguishing other dimorphic com- pounds. For the details of the operation and,apparatus I must refer to the extended article, stating here that the material must be absolutely free from oxidation products, and must therefore be extracted with acid and washed and dried in carbon dioxide, and that air must be rigidly excluded during the operation and complete condensation of the steam provided for. The stand- ard solution oxidizes both pyrite and marcasite much more slowly at 20°, but the per cent of sulphur oxidized is greater, namely about 81 per cent for pyrite and 81 per cent for marcasite. Doubtless the action of ferric salts plays an important part in nature in the disintegration of pyrite and marcasite, the iron acting as a transferer of oxygen. The conditions under which the sulphur undergoes complete oxidation are as yet not clear, and this subject is now under investigation. Mixtures of pryite and marcasite.—It is clear that the oxidation coefficient of any given mixture of pyrite and marca- site cannot be deduced by any simple process from the coefficients of the pure minerals. Each mineral is here decom- posing in a solution containing the reaction products of the other, and only an extensive knowledge of the influence of con- centration, acidity and dissociation would enable us to deduce theoretically a curve for such mixtures. It has therefore been necessary to construct the curve from data obtained by experi- ments on artificial mixtures. Since the action is a surface action, the result depends, not on the relative weights, but on the relative surfaces, and uniformity in this respect was obtained by always grinding weighed portions of the minerals together. The results, with different preparations and different samples, show that perfect agreement can be obtained in this way, the results for different samples of a given composition not differing more than one-half unit. The following table gives the mean results obtained: Percent pyrite, 0. 5 “10-207 40. «60 80 90. 95: « 100 Value of p, 18°0 16°0 15°2 17°1 22°3 29°0 40°3 48°9 52°9 60°5 Am. Jour. Sci1.—FourtH Series, Vou. XII, No. 72.—DxEcEmBER, 1901. 29 418 ATO. Stokes— Pyrite and Marcasite. With these data, a curve is drawn on cross section-paper from which it is possible, after finding the oxidation coefficient. of the mixture, to read off the relative proportions of pyrite and marcasite to within 1 per cent when the mixture is quite rich in pyrite and to 2 or 38 per cent when it consists mainly of marcasite. The existence of a well marked minimum in mixtures with 10 per cent pyrite is noteworthy. I have detected similar minima and maxima in curves given by mix- tures of pyrite with other minerals. The uncertainty of duplicate values for mixtures with less than 25 per cent pyrite, caused by this minimum, is obviated by mixing the specimen with a known quantity of pyrite and determining the oxidation coefticient of the new mixture. Ifa represents the percentage of pyrite in the original sample, a mixture of 90 parts of this with 10 parts pyrite will give 0°92 + 10 = per cent pyrite in new mixture, whence w, composition of the original material, is easily deduced. Here, as before, the results express simply the percentage com- position of the iron disulphide, indifferent substances being without effect. In regard to the influence of impurities, it may be stated that substances which do not contribute iron to, or effect reduction of, the ferric salt are without influence. Pyrrhotite, limonite, siderite, and other soluble iron compounds and zinc- blende or galena in small amounts may be extracted by heating with dilute hydrochloric acid; nickel and cobalt and other iron free sulphides not extractable by acid give abnormal results. For details as to the behavior of impurities reference must be made to U.S. Geological Survey Bulletin No. 186. In this connection, it is interesting to note that in a pyrite carrying 3 | per cent copper it is possible to ascertain whether the copper be present as chalcopyrite, or as chalcocite or bornite, a deter- mination which could scarcely be made by the usual analytical methods. The values of p* for a3 per cent mixture were found _ to be: Pyrite-chaleopyrite._...-------- 62°7 Pyrite-chaleocite 11a 282 ee 75°9 Pyvite-bormite): 43.2 se eee 76°4 The difference is many times the probable error of a determi- nation. ° Small amounts of chalcopyrite mixed with pyrite or marca- site may be readily detected by exposing the sample to bromine * » is here obtained as before by substituting the permanganate values in the above equation, but in this case it does not represent strictly the percentage of Sulphur oxidized. H. 'N. Stokes—Pyrite and Marcasite. 419 vapor for half a minute and then to hydrogen sulphide gas; the ehaleopyrite is blackened, while the iron sulphide remains bright. I have used the same method to detect and establish the nature of minute grains of chalcopyrite inclosed in rocks. My examination of various samples of doubtful nature, especially of concretions, shows that the finely fibrous speci- mens passing as mareasite are very commonly pyrite. Dr. Julien has very kindly given me a number of the iden- tical specimens of supposed marcasitic pyrite and paramorphs of marcasite after pyrite which were described in his paper.* I have found that those which show regular crystallization are actually pyrite free from marcasite, nothwithstanding their density would lead to the conclusion that they were either pure mareasite or a mixture. So far as my results go, there is no evidence for the existence of such mixtures or that the sup- posed paramorphs sometimes described are anything more than replacement or incrustation pseudomorphs. I have, however, found a few specimens of marcasite which enclose pyrite which cannot be detected by a lens, and which was probably simulta- neously deposited. A. P. Brown’s hypothesis —A. P. Brownt has published experiments which consisted in heating pyrite and marcasite in sealed tubes at 200° with cupric sulphate solution, according to which mareasite gives up its iron wholly in the ferrous form, while pyrite gives a mixture containing one-fifth ferrous and four-fifths ferric iron. From this he concluded that the iron in marcasite is wholly ferrous, while in pyrite it is four-fifths ferric. | have conducted a series of similar experiments with pyrite and marcasite and cupric sulphate, under the same con- ditions, in which the greatest care was taken to eliminate oxidation and to determine both ferrous and ferric iron in the solution and the precipitate. My results all agree in indicating that the pyrite is more slowly attacked than marcasite, but that the iron is found as ferrous iron only in the solution, and in the precipitate as ferric iron mixed with cuprous oxide and cuprous sulphide. The results are 62°5— 69-9 per cent ferrous iron 34°8— 30:1 per cent ferric iron. 58°2— 66°6 per cent ferrous iron 41°8— 33:4 per cent ferric iron. For pyrite For mareasite It appears therefore, that there is no essential difference between the decomposition products of pyrite and marcasite, while Brown’s view requires 20 per cent ferrous iron for * Annals N. Y. Acad. Sci., vol. iv, 1887, pp. 176, 204. + Proc. Am. Philos. Soc., vol. xxxiii, 1894, p. 225; Chem. News, vol. lxxi, 1895, p.. P79: 420 HH, NV. Stokes—Pyrite and Marcasite. pyrite and 100 per cent for mareasite. The relative amounts of ferrous and ferric salts depend simply upon the establish- ment of equilibrium between the solution and the decomposi- tion products of the pyrite and mareasite, not upon any fundamental difference in the minerals themselves, and the hypothesis in question, while possibly true, is thus far devoid of a valid experimental basis. As part of an investigation of the reactions involved in the secondary deposition of copper by sulphides, it may be men- tioned that cupric chloride at 200° decomposes pyrite according to the equation FeS, +14CuCl, +8H,O0 = 14CuCl + FeCl, + 121101 + 2HSO,. Sulphuric acid is also formed by the action of cupric chloride on cupric sulphide and apparently from pyrite and cupric sul- phate, so that the oxidation of sulphur appears to be a necessary part of the reaction when certain sulphides are acted on by cupric salts. These reactions, with other salts and sul- phides, are still under investigation. Wortman—Studies of Locene Mammalia, ete. 421 Art. XLV. — Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum; by J. L. Worrman. With Plates VIII and IX. [Continued from p. 382.] The Manus. (Figures 48, 49.) — As compared with the pes, the bones of the manus, especially the metacarpals and phalanges, appear shorter, heavier, and more robust. The carpus con- tains the eight bones common to the Creodonts. The scaphoid is rather flattened from above downwards; its proximal sur- face is occupied by an anterior convex facet for contact with the radius, posterior to which there is a roughened area for ligamentous attachment, and a large posteriorly projecting process at the inner posterior angle. Externally there is a 49 ‘FIGURE 49.—Same foot as in Figure 48, before removal of matrix; side view; one-half natural size. facet by which it touches the lunar, and inferiorly there are three facets,—an inner lunate, a median rhomboidal, and an outer elongated oval, for articulation with the trapezium, trape- zoid, and centrale, respectively. The lunar is distinct, and of a more or less quadrate form when viewed from in front; superiorly it presents a convex facet for articulation with the radius; internally a flattened 422 Wortman—Studies of Hocene Mammalia in the area by which it is closely applied to the scaphoid and centrale; externally a similar area for contact with the cuneiform, and distally a deeply saddle-shaped articular surface, divided by a median antero-posterior ridge into two eqnal parts, for contact with the unciform and magnum. The cuneiform is flattened from above downwards and is imperfectly quadrilateral in form, with the postero-external angle much produced, when viewed from above; its anterior and external surfaces are very rugose, presumably for the attachment of ligaments. Superiorly there is a cup-shaped articular depression for contact with the ulna, while behind this is a postero-external facet by which it touches the pisiform. Inferiorly there is an excavated articular surface by means of which it rests on the unciform. 2 The pisiform, like the other bones, has an unusually rugose _ surface; it has the usual form and articulation common to the Carnivora, and does not deserve any more extended description. The unciform is by far the largest bone of the carpus; it rests almost equally upon the fourth and fifth metacarpals, abutting upon the side against the third metacarpal and the inagnum, and supports the lunar and cuneiform. The magnum is relatively small, and is remarkable for its great antero- posterior length as well as the low position of the head,—a form quite unknown elsewhere among the Carnivora. Its articulations are as follows: It rests exclusively upon the third metacarpal; laterally it develops a contact with the head of the second metacarpal and the trapezoid; externally it touches the unciform, and superiorly it supports about equally the lunar and centrale. The trapez@id is also small in comparison with the other bones of the carpus; it rests upon the second metacarpal, touches the magnum upon its external surface, the centrale and scaphoid above, and upon its internal face exhibits a dis- tinet facet for the trapezium. The trapezium is a stout ossicle, somewhat larger than in the dog, which articulates with the scaphoid above and the trape- zoid and second metacarpal externally. It has an irregular triangular shape, and at its distal extremity bears an oval con- vex facet for articulation with the metacarpal of the pollex. . The last bone of the carpus which remains to be described is the centrale. This is a small antero-posteriorly elongated ossicle which les at the junction of the scaphoid and lunar, mainly under the former, and resting about equally upon the trapezoid and magnum. ‘When seen from in front, the bone appears to lie almost exclusively upon the magnum; pos- teriorly, however, it parses over on the trapezoid to a consid- erable extent. Marsh Collection, Peabody Museum. 423 The arrangement of the metacarpals, like those of the dog, is on the paraxonic plan; the third projects a trifle beyond the fourth, but the difference is so slight as not to affect materially this order. In like manner, the distal ends of the second and fifth reach about the same level. The actual length of the third is greater than the fourth, in the same way that the second is longer than the fifth, but these differences are coun- terbalanced by the overlapping of their proximal articular extremities. The overlapping and consequent interlocking of the second with the third, and the third with the fourth, are almost as great as in the metacarpals of the felines and dogs, - but the fifth articulates with the fourth by means of a nearly plane, flat face, with little or no interlocking. In size, the second metacarpal is much the largest, having a more or less flattened shaft. The third and fourth have about an equal degree of stoutness, the shaft of the third being a trifle the thicker of the two. The sides of these two bones are consid- erably flattened laterally, especially in the proximal half of their extent, where they are more closely approximated, but the amount of this compression is much less than in the dog. The fifth is the shortest bone of the four. In the matter of the stoutness of the shaft, however, it is intermediate between the second and third. The distal extremities resemble those of the dog, having well-developed keels confined to the palmar surface. The metacarpal of the pollex is missing and there is no certain means of determining the degree of reduction which it had reached, unless we judge by the size of its articu- lation with the trapezium. This would seem to indicate that it was small and more or less vestigial, so that the fore foot was essentially tetradactyle. The phalanges of the proximal row are shorter, broader, and heavier than those of the hind foot. That of the second digit is notably shorter and stouter than the others. .The fourth is the most slender, after which follow the third and second. The phalanges of the second row are relatively short and heavy, and exhibit a considerable degree of distal asym- metry. The unguals are rather short and depressed ; they are deeply fissured at their extremities and the subungual processes are of moderate size. The ungual foramen is present in all, and is of good size. Hind Limb.—The pelvis, figure 50, is much damaged and many parts are missing, but enough remains to furnish some of the more important points of its structure. Of the ilium, the region of the sacro-iliac synchrondrosis is present with that of the acetabulum. A considerable portion of the ischium, together with important parts of the pubis, sufficient to give a clear idea of the pelvic outlet, are also preserved. From these 424 Wortman—Studies of Hocene Mammalia in the fragments it can be determined that the ilum was of good length and considerably expanded anteriorly, as in Jesonyz. The tubercle for the tendon of the rectus is large and rugose. The acetabulum is large, of moderate depth, and the acetabu- lar notch is deeply incised. The ischial spine is distinet, the obturator foramen long and narrow, and the pubis compara- tively short. This latter fact, taken in connection with the narrowness of the sacrum, figure 51, gives a remarkably small outlet to the pelvis. There is, of course, no means of deter- 50 FIGURE 50.—Pelvie outlet of Dromocyon vorax Marsh; anterior view ; show- ing dimensions; one-half natural size. 51 52 FIGURES 51, 52.—Sacrum of Dromocyon vorax Marsh, and of a Newfoundland dog ; under view; showing difference in width; one-half natural size. mining the sex of the fossil; but assuming it to be a male, if the pelvic outlet in the female were as small in proportion, the young must have been very small at birth. For the sake of comparison, I give in figure 53 an outline drawing of the Marsh Collection, Peabody Museum. 425 ‘pelvic outlet of a male Newfoundland dog, and in figure 52 an under view of the sacrum. The bones of this skeleton are but little larger than those of Dromocyon. The difference is certainly very great; this is all the more remarkable, when we remember that the skull of Dromocyon is proportionally much larger than that in the dog. If, as before remarked, the females had a proportionally small pelvic outlet, there seems to be no escape from the conclusion that the young were born in a very weak and helpless condition, like the Marsupials, a fact which may have had something to do with their extinc- tion, especially when it is remembered that they were exposed 53 FIGURE 53.—Pelvic outlet of male Newfoundland dog; anterior view; show- ing dimensions; one-half natural size. to competition with the rapidly developing contemporary Canids. The Femur. (Figures 54, 55, 56, 57, 58.) — In all of its essential features, the femur presents a very striking likeness to that of the dog; it is, however, proportionally a little shorter and the shaft has a greater backward curvature. The head is relatively larger, the neck shorter, and the trochanter major of greater fore and aft extent. There is a large second trochanter which is placed upon the internal margin of the shaft, and an elongated third trochanter which extends well down upon the outer border. The distal end of the bone is 426 Wortman—Studies of Hocene Mammalia in the 54 -<<-- om oe FIGURES 54, 55, 56.--Femur of Dromocyon vorax Marsh; anterior, posterior and side views; one-half natural size. (Type.) Figure 57.—Distal end of femur of Dromocyon vorax Marsh; one-half natural size. (Type.) FIGURE 58.—Head of femur of Dromocyon vorax Marsh; end view; one-half natural size. (Type.) Marsh Collection, Peabody Museum. 427 remarkable for the great antero-posterior diameter of the con- dyles, exceeding the dog in this respect, and quite equaling some of the Artiodactyle Ungulates. There is a distinct popliteal fossa, the inner edge of which is prominent and marked at its lower extremity by an oval facet, for a rather large popliteal fabella. The prom- inence of this part of the bone indicates unusual strength for the tendinous origin for the outer and inner heads of the gas- trocnemius, which is well in accord with the cursorial powers of the species. The condyles are subequal in size and the inner projects toa slightly lower level than the outer. Thereis a deep intercondylar notch leading forward into a well-marked rotu- lar groove. The patella is relatively larger than that of the dog; it is elongated, narrow from side to side, and rather thick from before backwards. The proximal extremity is truncated and roughened for the attach- ment of the large quadriceps tendon. It has the typical form of a running animal. The Tibia and Fibula. (Figures 59, 60.)—The tibia, as in the dog, is a trifle shorter than the femur. The head dis- plays two subequal depressions, of which the outer is slightly the larger, to receive the femoral condyles. The spine is bifid, but the separation is not so clearly evident as it is in the doy. The cnemial crest is very large and extends more than half-way down the shaft, whereas in the dog it is limited to the upper third. The upper posterior portion of the shaft is deeply excavated. -The inferior extremity exhibits a trochlear surface similar to that of the dog. The internal malleolus is a strong 59 cr Sexe EI | ado a eee > as =,” N\ \n \ NN \ S vs \\: i \ 5 i : Ae ‘ ri We A A ly aN a) E \ ay | S$ ‘N+ Aa RB 3 \\ \ ah \ M\ iG y \ ay a yi ty NN i A i Mak Why Ry AU OB. oe | 6 ! Ba es «yi bt:: Nl a VM Be BW " iA 4 AW, s rally ay I N iy id A Fy i \ | ie 4 lib NE iam |: bed Hee i i q en Bl \ q iro - j AA Bae Bill is he aH] Pao (mee ee |i) | (eee ee: 1), 1 cil " H i | ee i wiih i a 1 an y ty SIN ijl ry OM fs 8 rT aM) Bei) ‘ fe fit ” H FF I hae a) ue IR a I Uh) ih { } a "4 ie es MN | eee || 7h) Pa F } a | ) H | Ho a |i), TMHT A) i ae ||| j yy 1 in ‘ fl *« ‘ FIGURE 59. — Tibia ard fibula of Dromocyon vorax Marsh; anterior view; one-half natural size. (Type.) process and the trochlear grooves are pronounced. The fibula has a relatively stouter shaft than that of the dog, but at the same time is much reduced ; it is applied to the lower end of the tibia for a considerable distance. The shaft is triangular above and oval in cross-section below. The dis- tal end is large, and forms a stout external malleolus in the ankle joint; it articulates with the astragalus by means of a lateral facet, and does not touch the caleaneum. 428 Wortman—Studies of Hocene Mammalia in the The Pes. (Plate VIII.) — As a whole, the pes is very dog- like in general appearance; the loss of the hallux, the com- pressed, elongated, and highly interlocking character of the metatarsals, as well as the distinctive “square-cut” aspect of their distal ends, recalls at once the hind foot of the dog. There are, however, some important differences which will be pointed out in the course of this description. The astragalus, if found dissociated from the other parts of the skeleton, and if the head were not so convex from Ee before backwards, might readily be mis- taken for that of a Perissodactyle Ungu- late. Indeed, it furnishes just such a SZ Z) \ transitional stage as we ma dil (Lp. Ss o y readily € believe to have’ preceded the peculiarly Cy | specialized type of this bone in the Artio- dactyla. The trochlear surface is deeply Figure 60.—Tibia and grooved, the head is set rather obliquely fibula of Dromocyon voraz non the body by a moderately elongated Marsh; end view; three- Oe Beech ° vos me fourthe natural size. meck, and its distal end is diyaded imi (Type.) two unequal facets. Of these the larger is internal, slightly concave from side to side, and articulates with the navicular; while the outer is narrow, much prolonged upon the posterior surface, and articu- lates with the cuboid. The caleaneal facet is narrow and deeply concave from above downwards, while the sustentacu- lar facet is broad, slightly convex, and more or less pyriform in outline. There is no astragalar foramen. As compared with the astragalus of the dog, the transverse axis of the head coincides more nearly with the transverse plane of the trochlea ; the calcaneal facet is narrower, and the sustentacular facet is broader. The caleaneum has a relatively longer tuber than that of the dog, but otherwise the two agree very closely. The cuboid is proportionally larger than that of the dog. Proximally there is a distinct facet for the astragalus, but otherwise the articu- lar surfaces are much alike in the two. In the posterior aspect, however, they differ considerably ; in the dog the cuboid has a double peroneal tubercle, of which the smaller is external and just above the commencement of the peroneal groove, and the other, larger, is located near the center of the posterior sur- face. The direction of the peroneal groove, which is large, is downwards and inwards. In Dromocyon there is but a single large elongate tubercle, occupying nearly the whole of the posterior surface. Beneath the distal extremity of this tubercle is a notch, which is converted into a distinct groove upon the outer surface of the bone, marking the course of the long peroneal tendon. In the dog the tendon passes obliquely Marsh Collection, Peabody Museum. 429 down to its attachment upon the base of the second meta- tarsal, and the groove is relatively large; whereas in Dromo- cyon it is small, and the tendon passed down to the outer edge of the tubercle, thence turned abruptly inwards to be inserted in the same place. The relatively large size of the peroneal tubercle, which serves principally for the attachment of the calcaneo-cuboid ligament, may be taken to indicate unusual firmness of the tarsal joint. The navicular has much the same general shape as in the dog, and differs from it in that it is narrower, and its proximal surface more deeply saddle-shaped; it articulates with the cuboid in quite the same way, and distally exhibits the three usual facets for the cuneiformia.- Posteriorly it presents some important differences from that of the dog. In the ease of the latter, there are two more or less distinct navicu- lar tubercles, separated by a wide sulcus. In Dromocyon there is a single tubercle, which is produced into a long pointed process. The three cuneiformia differ from those of the dog, in that they are decidedly higher and narrower from side to side; their manner of articulation is, however, much the same. The internal cuneiform is large and elongate, bearing upon its distal extremity an articular facet for a rudi- ment of the first metatarsal. The metatarsals closely resemble those of the dog, not only in their elongate, compressed, and highly interlocking charac- ter, but they show a decided tendency towards lateral flatten- ing of their contiguous surfaces. Their distal ends, moreover, exhibit that characteristic “‘ square-cut”’ appearance so com- mon to the modern Canide. The foot is of the strictly paraxonic type, the third and fourth metapodials being of equal length. While the second and fifth are also of equal length, the second is much the larger and stouter bone, the disparity in size being considerably greater than that seen in the dog, and about equal to that of the fore foot. The distal metapodial keels are present and well developed ; they are, however, confined to the plantar surface and do not extend more than half-way around upon the dorsal surface. The proximal phalanges are _ proportionately — shorter, stouter, and more robust than in the dog, but the articular faces have very much the same disposition and _ extent. Proximally the plantar notch is not so deeply incised as in the phalanges of the dog, but their dorso-plantar curvature appears to be somewhat greater. The median phalanges, or those of the middle row, are proportionately still shorter than those of the first, as compared with the dog. They exhibit a slight degree of distal asymmetry. The unguals are short, rather broad, and deeply fissured. 430 Wortman—Studies of Eocene Mammalia in the | The following principal measurements are herewith given : Lenethot) skull. 0.2 te Be Sie Bee ee a 318°5™™ Width of skull at posterior peor of zygomata i_..5. 160s Heightiot anterior Wares.) ou. 2 Aaa eee ei 54: Vertiealdiameter of “orbit: - 22. bos Se epee 43° Height of occiput above base of condyles_._-_.......- 113° Widthvof :condyles.22 0 ants Stee 54: Greatest, width of occiput... ©. 42.1178 pee Be Length of superior tooth line from anterior border of INCISOTS:\2 2.2. 88D eee se eo ene Leneth.of premolar senies 42> (222 62° Leneth-ot ; molar series 5 22. 2 45° Antero-posterior diameter of superior canine-_-_-------- 20° Lene thot lowers awerce sce ae arn eee eee 248° Length of inferior dentitions 2222222525 3. ae 151° Greatest depth of jaw including coronoid_----_----_-- 93° Depth of jaw opposite first molar_-22-2-27. hie eee 43° Width ot palate between Canines! 22222) 3252 eee 33°5 Wadth-ot palate-opposite tirstimolare 22-2) 2 es ee 46° Transverse diameter between inferior canines (internal). 18° Transverse diameter between inferior canines (external)_ 44° Transverse, diameter of inferior canime__ .2-- -_-. 122 329elies Totallength of vertebral column, as articulated, including sacrum, exclusive of curves; tail not included._.. 860° Total lengthof vertebral column, including curves ; tail mot included 32.200... .30 265 ie os eel Length of tail (estimated). Jags oe advan See ee rr Length of ‘cervicals <2 20. ee es ee rr Leneth :of dorsals2 2.222228 ee ee Bength of lumbars: "2 220-029 052 nee Pee eee Length of sacrals (estimated) 222222720222 a) sea 62° Length of body of axis, including odontoid.__-------- 60° Height of axis at posterior border of centrum-.------- 63° Antero- -posterior length of spine of axis .------------- 68° Height of seventh cervical, including spime --_--..----- 113° Height of first dorsal, including spine Loe Oi a eae 128° Heicht-of ninth dorsals 2.5.2 eee ee Length of body of penultimate lumbar. ..-- .--..--.-. (36° Widthiot. sacrum) io c.ete chee os Boe ul eee 40° Length-oftirst ribo. 235. 253 See ee fl Length ‘of third. rpby... (52 G3 8 5 eee Length ottsevienth “rib oss es ee eee eee 169° Wadthsof third aio. 20.2 2 See he eo eee cone 20° Width of sementhi rib... “sn. sce ee pees eas 13° Length sot scapula. 2.2.28 22a eee ee ee IS) Antero-posterior diameter of glenoid cavity.--.-.------ 38°5 Marsh Collection, Peabody Museum. 431 Transverse diameter of glenoid-cavity.....----------- pa Length of humerus ---- -- Mae Cid oOt Sek i 3 FO Transverse diameter of distal end of humerus. ....---- 35° Antero-posterior diameter of distal end of humerus.--. 36° em OWE oe we a i eee ot eee PED) BSOr Pena of olecranon of ulna... S22 Tessie 49°68 Length of radius .-.------ Eppes tie ests Fic Ts a ye eee by Ee Transverse diameter of head of radius......-....----. 26° Antero-posterior diameter of head of radius__--------- 17°5 Transverse diameter of distal end of radius__---_----- 36°5 Antero-posterior diameter of distal end of radius ------ 22: Meeeine OMT ORC LOOG 6c ir ON es en bee 152° Bim Oe Catpus: = 5 er Se tS eS 40: Meuetyor second metacarpal s_-.. 4.24.2. 02.2 5.-2252- 65° omen Or third metacarpal. i222 42222222 lene 15°5 ean Ol-tourth metacarpal. 20-422) ee ek wae MomacheOr iirch, metacarpal i. 2-22-02 ee O° feeerh of first phalanx, dicit JL _.22 22.25... -..---.- 28°5 Seearicer cecond phalanx, digit If 22. 05-2. 2.522. 15" Pee or uneual phalanx, digit IT___-:..2-.--..-..-. 15° Bemenivor nest phalanx, digit Tf 2. 4... keels. 28°5 Pewcuy or second phalanx, digit IIE__-.-...-.-----.-. Eds Beneth of unoual phalanx, digit IIl.-_2-..---.----.2.- 16° Bewetror rst phalanx, digit: Vo... 22) 22-2 s-- 2. -- 213 eneth of second-phalanx, digit V_.-...-.---.-----.- 13°5 eaech of ungual phalanx, digit V...2...-...-....-.- tS: Ben =cho: pelvis (estimated) 2.22. .0--2 222) 2-4. 2. 240: Depth of pelvis at anterior border of pubic symphysis. 64: emetealdiameter of pelvic outlet. .-2 2.25 2.22. 2 e 58° Transverse diameter of pelvic outlet...--._-.-..-.---- 42° Antero-posterior diameter of acetabulum ___--__-.-- tage SUS. ieee ree epre eT se, ee 8 eS 227- Transverse diameter of prota end-of femur. +2. 51: Transverse diameter of distal end of femur-_..--..--- 42° Antero-posterior diameter of distal end of femur-_-.-- 49° ree Gh tibiae epee yh foe SS Sa Py > Ni Transverse diameter of head of tibia -_-_-.-...------- 44: Antero-posterior diameter of head of tibia, including HCW CREE on sara s eee nie pe ee SN 54: Transverse diameter of distal end of tibia and fibula... 39° Antero-posterior diameter of distal end of tibia....... 23: BeeeemeOrehini LO0t,, 924 Fone Oe ee a 235° Length of tarsus, including the tuber calcis.....-.-.--- 91- DEAL TC. Cee Cena eo Re Sees | ele 32°5 Merrie Ge Czleanenntl vat ee he a; 66° Meme Oe anita alus: 2222s ees oN ee 49° ee tiati re Astra AUR oe iS ae ae Sa BOF Wee nage cuuoid ir fronts - oy 5 eee ee 2 OE" 432 Wortman—Studies of Eocene Mammalia, ete. - Length of ‘second metatarsal’ 23.2 2a.) eee Lengthvot third metatarsals. 22 235 = ee eee Length of fourth metatarsal 927635) a ee ee Leneth of: fifth metatarsal: os see) eee eee Length of airst phalanx, digit Il: = {cee as ee Length of first, phalanx) digit Tl. ea see ee Length of first phalanx, dicit [V2 = 22555 22 saree Length of firstphalanx, digit; Vii. 22122 See ee Length of second phalanx,: dicit TI: ....22: 222222522 Length of ungual phalanx, disiti Tl 222- 72c2 ee aeeeae This specimen is from the voleanic-ash bed of Henry’s Fork, Bridger Basin, and was found by J. Heisey. [To be continued. | Am. Jour. Sci., Vol. XII, 1907. . Plate VIII. EXPLANATION OF PLATE VIII. Right hind foot of Dromocyon vorax Marsh; anterior view; three-fourths natural size. (Type.) c, caleaneum: a, astragalus; n, navicular; c/, c2, c3, internal, middle, and exter- nal cuneiformia; cb, cuboid. Plate IX. Am. Jour. Sci., Vol. XII, 1901. (edszq) ‘OZIS [RANJVU YAJ[OM4-OUO + MOLA OPIS + YSsVy Loca Uohoowosg JO UOJS[OYS powunoyy ‘X] @LVIg JO NOMVNVIaxg Hormell— Dielectric Constant of .Paruffins. 433 Art. XLVI.—J/%electric Constant of Parafins; by WiLL G. HORMELL. THE following paper presents some experiments in which by means of a modified form of the Blondlot oscillator the dielectric constants of four commercially different paraffins were obtained for 40, 60, and 80 centimeter waves. The paper also contains an investigation of the relative velocities of 80 centimeter waves along magnetic and. non-magnetic wires of different diameters. And finally there is given in the table, by way of comparison, the light index of refraction of the different paraffins for the D, line as obtained by means of the Abbe refractometer. Apparatus. Figure 1 is a diagram of the connections made to the vari- ous pieces of apparatus used in the experiment. The terminals T Col 4 ks ¢ G a and } are connected to a commercial line delivering an alter- nating current with a pressure of 52 volts. The resistance R is used to regulate the quantity of current flowing through the circuit. T is a transformer whose primary of about 200 turns is wound around a core of soft iron wire. The secondary of the coil contains about 2000 turns. C is a battery of four Leyden jars each having a capacity of 5900 electrostatic units. T, is a Tesla coil made after the formula given by Tesla and is capable, when fed to its full capacity, of giving a ten-inch spark. In all the following experiments a current of about 7 amperes was taken from the commercial line which would give in the Tesla coil a spark of not more than 3°". Gisa spark gap consisting of two small iridium* cylinders inserted in 2-in. brass spheres. The length of the spark gap can be nicely adjusted by means of a screw thread on the axle of one of the spheres. The constancy of the spark at this gap is of the greatest *@G. W. Pierce, this Journal, vol. ix, p. 256, 1900. Am. Jour. Sc1.—FourtH Series, Vou. XII, No. 72.—DrEcEemBER, 1901. 30 434 Hormell— Dielectric Constant of Paraffins. importance in securing the best possible results from the oscil- lator. A rapidly rotating disc presenting projections to the opposite terminal of the gap gave very indifferent results. Zinc balls, polished brass balls, and brass balls with platinum terminals inserted after the fashion of the iridium cylinders, were used, but none of these gave the constancy secured through the use of the iridium. Platinum terminals gave the next best results, but on account of their rapid deterioration. they required almost constant attention in order to keep their surfaces well polished. The iridium gap needs no attention except for occasional adjustment of length. The secondary terminals of the Tesla coil are joined to the primary terminals E’ E (figure 2) of the Blondlot oscillator. The remaining portion of figure 1 is shown in detail in figure 3. The Blondlot oscillator (figure 2) is the modified form used by W. D. Coolidge.* Some additions, however, have been made in order to bring the adjustments more completely under * Annalen der Physik, Ixvii, p. 578, 1899. Hormell—Dielectric Constant of Paraffins. 435 the control of the observer. The two semicircles comprising the primary coil are firmly fastened to two hard rubber uprights which are pivoted on the crossbar L. The screw D regulates the spark gap m._ 0:58. ,0°34==101-98 The chief point of interest lies in the presence of the graphite, distributed in thin scales. It shows on heating, after treatment with nitric acid, the phenomenon of “ sprouting ” which has been held by Moissan to be characteristic of graphite crystallized from fusion. Not only is the graphite in the main rock type but also in coarse-grained “contemporaneous veins” and in black basic lenses or dike-like bands. As the occurrence of graphite as a well-defined constituent of a rock which has been held heretofore to be of purely igneous 468 Scientific Intelligence. origin is a novel phenomenon of great interest, the author has studied with care the field relations of the rock mass. There is, however, no direct evidence to be obtained from the exposures as: _to its intrusive nature. So far as can be made out, it occurs as. lenticular masses in the schists. There are many points, how- ever, to be seen in the field which are characteristic of igneous masses, and balancing all-the evidence obtainable the writer feels. it to be distinctly in favor of considering the rock masses erup- tive in origin. ‘The foliated structure is thought to have been imposed before complete consolidation. Agreeing with this view, we are now forced to recognize graphite as a new and pos- sible constituent of igneous rocks, which may give rise to new _ types on occasion. The corundum occurs in a feldspathic rock associated with the eleolite-syenite. Some varieties contain a red garnet, others chrysoberyl. This syenite is found with the eleolite-syenite in such a way that its exact relations cannot be definitely told, but it is held to be a differentiated product of the same magma. There is thus a remarkable analogy with the occurrence in Hastings County, Ontario.* Lin Var Be 11. On a Peculiar Form of altered Peridotite; by T. H. Houtiranv. Mem..Geol. Surv. India, vol. xxxiv, pt. 1, pp. 1-9, 1901.—The auther describes a rock mass at Huliyar, Mysore State, India, consisting of a matrix of interlaced colorless tale and pale green picrolite (columnar serpentine) with numerous granules of a rhombohedral carbonate and magnetite dust in which lie well-formed, porphyritic crystals of dark gray breun- nerite, from one-half to one inch across; they form 35 per cent of the rock. From its staining with iron oxide the rock appears like a diabase with porphyritic pyroxenes or hornblendes. The | matrix and breunnerites were analyzed separately, and combining the analyses the rock is found to have the composition SiO, Fe,0, MgO 00; H.0 27°3 27 35°) 20°0 5°0 = 100°0 From this it is assumed that the rock was a dunite which has been changed by the action of water and carbonic acid. _L. V. P. 12, Perknite.—This name has been proposed by H. W. TurNER for the dark-colored, heavy, basic, holocrystalline igneous rocks composed of monoclinic amphibole or pyroxene or both. In addition to these chief components, rhombic pyroxene, olivine or feldspar may occur as secondary ones and there may be accessory biotite, iron ore, ete. ‘hese rocks have hitherto been called pyroxenite or amphibolite or hornblendite, but they have not * The occurrence of graphite as a well-marked constituent of an igneous rock is a matter of great interest and importance. It will be correlated with the diamonds in the peridotites of South Africa and mm certain meteorites and throws new light on its occurrence in pegmatite dikes described from Canada. If it can be definitely shown by sufficient evidence that carbon is a constituent of the deep-seated magmas, certain vexed questions in geology may be susceptible of a better explanation than any yet offered. L. V. P. Geology and Mineralogy. 469 been grouped under one name, as here proposed.—Jour. Greol., vol. ix, pp. 507-511, 1901. © Le Vie: 13. Outline of Elementary Lithology ; by Gro. H. Barron. Pp. 112, 12mo. (Boston, 1900.)—This little work is a digest of the main features of the subject of Lithology. The important rock-making minerals and their properties, the names, textures, and classification of rocks, are given in, synoptical form. It will be found useful by students reviewing a lecture course on this subject. ib 5 BE 14. Synchisite and Molybdophyllite.—A recent number (No. 9) of the Bulletin of the Geological Institution of the University of Upsala contains a mineralogical paper by G. Flink with notes on several known species and also descriptions of the new species, synchisite and molybdophyllite. SyNncHisiTg, from Nararsuk in southern Greenland, was earlier described by G. Nordenskidld as parisite. It is now found, how- ever, to be distinct. The crystallization is rhombohedral instead of hexagonal and it shows no distinct cleavage. Orystals have commonly the form of acute rhombohedrons. The hardness is 4°5 ; specific gravity 3°902; color wax-yellow. The composition “CeFCaC,O, is deduced from analyses, the most recent of which (by Mauzelius) is as follows: CO. ThO, Ce.03 (LaDi),02 Y,03 Fe,03 CaO FE H.O 25°99 0°30 21:98 28°67 1°18 O11 16°63 5:04 2°10=102:00 MoLyBpoOPHYLLITE is a new lead silicate from Langban, Swe- den. 1t occurs in irregular foliated masses, resembling mica, imbedded in granular limestone ; crystallization hexagonal ; cleav- age basal perfect. The hardness is 3 to 4; specific gravity 4°717. It is colorless with pearly luster on the cleavage face, elsewhere glassy. The composition R,SiO,+H,0O is given by the analysis : SO EDOM. MeO mw MsOs: NasOl K,O,, HO 18°15 61°09 OLE PALL 0°46 - 0°82 0°69 6-32=—99°24 15. Crystallization of Stannite.—L. J. SPENcER shows that stannite crystallizes not in the tetrahedral group of the isometric system as has been assumed, but in the scalenohedral group of the tetragonal system, as early suggested by Haidinger. Crystals from Bolivia are very near chalcopyrite in occurring forms and in angles, and like it are often pseudo-cubic from twinning.— Min, Mag., xiii, 54. ° 16. Conchite and Ktypteite.—H. V ater shows that the supposed new form of calcium carbonate, called conchite (p. 84) by Agnes Kelly, is in fact not distinct from aragonite. The ktypteite of Lacroix is probably also identical with the same mineral.— Leitschr. f. Kryst., xxxv, 149. 17. The Meteorite of Helix, Perry County, Alabama.—An account has recently been given by G. P. Merritt of the stony meteorite which fell near Felix, Alabama, on May 15, 1900. 470 Scientific Intelligence. The explosion of the original meteor yielded three pieces ; .of these, the main mass recovered weighed about 7 pounds; a second small piece was found but lost and a third has not been found. The material of the meteorite is soft and friable, resembling somewhat the stones of Warrenton, Missouri, and Lancé, France, though darker than the former and more choniritic than the lat- ter. It consists essentially of the minerals olivine, augite, and enstatite, with troilite and native iron, the silicates occurring in ' the form of chondrules, or associated more or less fragmental particles, embedded in a dark, opaque, or faintly translucent base, which is irresolvable so far as the microscope is concerned. ‘The non-metallic portion (including troilite and chromite) made up 97 per cent of the whole and nickeliferous iron only 3 per cent. The dark color of the stone is attributed to carbon distributed as minute flakes of graphite.—Proc. U. S. Nat. Museum, xxiv, 192. Li ZeoLoGy. Note on the Nomenclature of Bermuda Birds.—Mr. A. foe VERRILL published a short paper on the Bermuda avi- fauna in this Journal for July, 1901 (issued the Jast of June). He also printed a more detailed article in ‘The Osprey,” pp. 838-85, for June, 1901, with figures of the three following species and of the Tropic Bird, photographed from life. In these articles he described the Bermuda Cardinal Bird and the Blue Bird as new sub-species, peculiar to Bermuda. The Cardi- nal Bird he named Cardinalis cardinalis Somersii ; the Blue Bird, Stalia sialis Bermudensis ; the Ground Dove, Columbi- gallina passerina Bahamensis. Outram Bangs and Thos. 8S. Bradlee also published a paper on the Birds of Bermuda in “The Auk” for July, 1901, pp. 249- 257, in which new names are given to two of these birds and two others, which they call new species. They name the Ground Dove, Columbigallina bermudiana ; the White-eyed Vireo, Vireo ber mudianus ; the Catbird, Galeo- scoptes bermudianus ; the Cardinal, Cardinalis ber mudianus. Mr. Verrill’s article in this Journal appears to have been pub- lished a few days earlier than the latter. To me it seems quite useless to regard these very slightly dif- ferentiated forms as distinct “species.” The differences noted in the Ground Dove, Catbird, and Vireo, are trivial and scarcely sufficient to constitute varieties. To consider them as “ sub- species” is certainly a sufficient strain on the much-stretched meaning of the term “sub-species.” I should at most call these mere local varieties, scarcely differentiated. In respect to the Ground Dove, there are reasons for believing that it was introduced to Bermuda from the Bahamas, since the settlement of the islands, like many other things. None of the earlier writers mentioned it in the lists of birds that they gave. Loology. 471 This would hardly have been the case had it been present, for it is exceedingly tame and familiar. Mr. A. K. Fisher, in Bird Lore, Oct., 1901, p. 178, states that the original Motacilla sialis Linné, ed. x, p. 187, was from Ber- muda. This is not correct. Linné gave it as from ‘“ Bermudis & America calidore.” He also quoted Catesby, Hist. Carolina, etc., p. 47, pl. 47, 1731. Catesby says that he had seen it in “Carolina, Virginia, Maryland, and the Bermudas.” But he states in his preface that his birds were mostly drawn in Caro- lina and Georgia, where he spent several years in drawing them. A few were drawn in the Bahamas, where he spent about a year, mostly on the fishes and plants. He does not say that he made any drawings in Bermuda, where he probably made a mere pass- ing visit. The Bluebird does not occur in the Bahamas. His figure clearly SEisetes the common North American variety. AL BS Vi 2. Reports on the Kauna of Porto Rico.—As a result of the scientific expedition to Porto Rico on the Steamer “ Fish Hawk,”’ 1898-99, the United States Fish Commission has published, in its Bulletin for 1900, several valuable reports on the fauna of the island. . The largest and most fully illustrated is the Report on the Fishes and Fisheries, by Everman and Marsh, 1900, (350 pages, 49 colored plates, and a map). In this report 291 species of fishes are included. The plates are excellent. They were drawn and colored from life by A. H, Baldwin and C. B. Hudson. It forms an excellent manual for any part of the West Indian region. The Report on the Brachyura and Macrura, by Miss M. J. Rathbun (137 pages, 2 colored plates) is also a. very complete report on these Crustacea. It contains descriptions of all the genera and species, with analytical tables, and will serve as a standard manual. The Report on the Anomura, by J. E. Benedict (17 pages, 4 plates) is of the same excellent character. Mr. R. P. Bigelow has reported on the Stomatopoda (10 pages, with cuts). Mr. H.- F. Moore has prepared the Report on the Isopoda (14 pages, with cuts, and 5 plates), and Mr. M. A. Bigelow, a very short Report on the Cirripedia, which were evidently much neg- lected. The Report on the Echinoderms is by Professor H. L. Clark (33 pages, 3 plates). It is fairly complete for the shallow water forms, though many common West Indian starfishes and ophiuroids are lacking. The Report on the Annelids, by A. L. Treadwell (27 pages, with cuts) contains a considérable number of new species, but is evidently based on a very incomplete collection. Mr. W. k. Coe has also contributed a short report on the very few nemerteans that were obtained. 472 Scientific Intelligence. J. P. Moore has described two new leeches. We understand that several other reports, on other groups, are soon to be printed. A. eBoy 3. Papers from the Alaska Harriman Hapedition.—Several useful and valuable reports on the collections obtained by the Harriman expedition have already been printed by the Washing- ton Academy of Sciences. The Report on the Nemerteans, by W. R. Cor (March, 1901, 84 pages, with cuts and 13 plates, 6 colored), is, perhaps, the most important and complete of those hitherto published. It relates to a group that is very richly developed on the Alaskan coast, and which has hitherto been almost entirely neglected in those waters. It includes 32 species, of which 28 are new. The col- ored figures are excellent. Ao Vn 4. Zoblogy: An Klementary Text Bock, by A. HE. SaipLEy and EK. W. MacBripz. London and NewYork, 1901 (The Mac- millan Co.). 8yvo, 632 pages, 347 cuts.— Although called “ ele- mentary ” this is a rather complete text-book, suitable for the more advanced students in colleges and universities, or for a ref- erence book. The treatment of the subjects is clear and the illustrations are generally excellent. ‘The Vertebrata (or Chor- data) occupy an unusually large part.(nearly half, or 286 pages) of the book, which will probably be considered an advantage by many instr uctors. But it compels a severe cutting down of the space for some of the important groups of Invertebrata. Thus the great group of hexapod insects occupies only 23 pages, and the Mollusca 39 pages. It is an excellent addition to the list of recent text-books and manuals of zodlogy. A. Ey V. 5. An Elementary Course of Practical Zodlogy ; by T. JEFFREY Parker and W. N. Parker. London and New York (The Macmillan Co.). Small 8vo, 608 pages, 156 cuts.—This is intended mainly as a laboratory guide for a course in “ biology” or comparative anatomy, such as is commonly given in many of our colleges and scientific schools. For this purpose it seems to be admirably adapted, being more up to date than most of the similar works now in use. About one-third the book (215 pages) is devoted to the study of the frog, in detail. The balance is devoted to various other typical forms, such as the ameba, hydra, earthworm, crayfish, mussel, amphioxus, dogfish, rabbit. About 60 pages, at the end, are devoted to cell structure and embryology. AY Have lV. MiIscELLANEOUS SCIENTIFIC INTELLIGENCE. 1. National Academy of Sciences.—The following is a list of the papers entered to be read before the National Academy at the meeting in Philadelphia, held Nov. 12-14. GEORGE F. BECKER: Note on linear force exerted by growing crystals. Note ou the orogenic theory of tilted blocks. Miscellaneous Intelligence. 473 Horatio ©. Woon, JR.: On the vaso-motor supply of the lungs. GEORGE F, BARKER: Biographical memoir of Frederick Augustus Genth. The monatomic gases. The newer forms of incandescent electric lamps. JAMES M. CRAFTS: On the pseudo-catalytic action of concentrated acids. SAMUEL L. PENFIELD: On the use of the stereographic projection in making accurate maps; with criticism of some recent methods of map projection. The tendency of complex chemical radicals to control crystallization because of their mass effect; a study in isomorphism. CHARLES S. PEIRCE: On the logic of research into ancient history. Epeéar F. Smita: Observations on tungsten. S. Werk MircHELL and Simon FLEXNER: Snake venom in relation to hzmo- lysis, bacteriolysis and toxicity. IrA REMSEN: On the nature of the double halides. Cyrus B. Comstock: Biographical memoir of General John Newton. Henry F. Ossorn: Dolichocephaly and Brachycephaly as the dominant factors in cranial evolution. Cranial evolution of Titanotherium II. Latent or potential homology. . i EpwarpD W. Moruey and CHARLES F. BrusH: A new gauge for the direct measurement of small pressures. Transmission of heat through vapor of water at small pressures. CarL Barus: On quadrant electrometry with a free light needle highly charged through a conductor of ionized air. On nuclear condensation in the vapor of non-electrolytes like benzene; and on graded condensation. Henry L. Bowpitcw: The work of the International Association of Acad- emies. CASWELL GRAVE: A method of rearing marine larve. 2. Annual Report of the Board of Regents of the Smithsonian Institution, for the year ending June 30, 1900. Pp. xlv, 759. Washington, 1901.—The Annual Report of the Smithsonian Institution, recently published, includes in its preliminary part the Report of the Secretary, Professor 8. P. Langley, which was issued earlier in separate form, and has been noticed on page 400 of the Maynumber. The remainder of the volume is occupied with the general Appendix, which contains reprints of a well-selected series of papers, on topics ranging from astronomy to medicine, and all calculated to be of great interest to the ordinary reader as well as to the scientist. With its other important functions, the Smithsonian institution does a valuable work in thus present- ing to the public well-digested reviews of the progress of science in its different departments and also in republishing articles which would not otherwise be readily accessible. Of the subjects here included, one of particular interest is an account, liberally illustrated, of the Langley Aérodrome, and another on the Zep- pelin Air Ship accompanies it. A complete list of the papers printed in the volume would be required to show their range of subject and individual importance. | 3. Lhe Smithsonian Institution: Documents relative to its Origin and History, 1835-1899 ; compiled and edited by WiLt1am JonES RuxEES, in two volumes. Vol. I, 1835-1887. Twenty- fourth Congress to Forty-ninth Congress. Pp. hii, 1044. Washington, 1901.—The very interesting volumes relating to the history and work of the Smithsonian Institution, presented to the public some years since (1879, 1896) are now followed by the republication in full of all the public documents which relate to 474 Scientifie Intelligence. its inception and history. The volume now issued covers the period from 1835 to 1887. 4, Beitrdge zur Chemischen Physiologie und Pathologie— Zeitschrift fiir die gesamte Biochemie; herausgegeben von Franz Hofmeister, 0. Professor der physiologischen Chemie an der Universitit Strassburg. Braunschweig. (Fr. Vieweg und Sohn. Two volumes per year; price 15 marks for each yolume of 12 numbers.)—The progress of recent years in various branches of biology and medical science has been particularly characterized by the growing interest in the chemical problems involved, and the increased application of chemical methods for their solution. The rapid development of physiological chemistry is indicated by the beginning of a new journal devoted to original investiga- tions of bio-chemical nature. In addition to the presentation of more exhaustive contributions, it is proposed to afford a medium for the speedy publication of briefer communications regarding new discoveries, etc. Under the editorship of one of the most active workers in this field, a successful future for the new pub- — lication is assured ; and the scientific papers printed in the three numbers already published give evidence that the “ Beitrage” will form a part of the permanent literature of biology. Contributions may be directed to the editor, Wimpfeling- strasse, 2, Strassburg i. E. EBs Oe 5. Annals of Harvard College Observatory.—Recent publi- cations include the following : Vol. XXVIII, Pt. I; Spectra of Bright Southern Stars ; Photographed with a 13-inch Boyden Telescope, as a part of the Henry Draper Memorial, and discussed by Anniz J. Cannon, under the direction of Epwarp C. PickrrRiINnG, Director of the Observatory. Pp. 131-2638, with three plates. Vol. XLI, No. VIL. Comparison of Results obtained with dif- ferent forms of Apparatus in Meridian Observations ; by Arthur Searle. Pp. 189-211. 6. Royal Society of London.—The Copley Medai has recently been awarded by the Royal Society to Professor J. Willard Gibbs of Yale University, New Haven, for his contributions to mathematical physics. OBITUARY. Rupotpw Koente, the well-known investigator in acoustics and skillful maker of acoustical apparatus, died at Paris on October 2, at the age of sixty-eight years. Proressor A. F. W. Scurmprer, the eminent botanist, died on September 9 in his forty-sixth year. PS. LO; VOLUME. xTL* A Academy of Sciences, National, ° meeting at Philadelphia, 472. Adams, E. P., electro-magnetic effects of moving charged spheres, 155. Adams, G. I., Carboniferous and Permian age of the Red Beds of Oklahoma, 383. Air, separation of least volatile gases | in, Liveing and Dewar, 207. Air-tight glass stop-cocks, and Kekardt, 463. Alaska, Harriman Expedition, 472. Almy, J. E., discharge-current from a surface of large curvature, 175. Alternating current discharge, spec- trum, Wright and Downs, 66. American Philosophical Society, 400. Argentine Republic, Meteorological Atlas, Delachaux, 88. Argon, viscosity, Schultze, 76. Arkansas, Geol. Survey, vol. v, 78. Association, American, meeting at Denver, 174, 323. — British, meeting at Glasgow, 324. B Bailey, E. H. S., Laboratory Guide to Qualitative Analysis, 462. Barton, G. H., Elementary Lithology, 469. Barus, C., on temporary set, 247; effect of temperatnre and moisture on the emanation of phosphorus, 527. Beecher, C. E., Cambrian fossils of | St. Francois Co., Missouri, 362; Eurypterid remains in the Cambrian of Missouri, 364. Bermuda, additions to the Avifauna, Verrill, 64; nomenclature of birds of, Verrill, 470. — death of fishes at, in 1901, Verrill, 88. Birds of Bermudas, Verrill, nomenclature, Verrill, 470. 64 : Thiele | | Bolton, H. C., Chemical Bibliog- raphy, 89. BOTANY. African plants collected by Dr. Welwitsch in 1853-61, catalogue, 524, Cocos nucifera, anatomy of, Win- ton, 269. Milchsaft und Schleimsaft Pflanzen, Molisch, 87. Brazil, manganese deposits of Minas Geraes, Derby, 18. British Museum catalogues, 89. Browning, P. E., estimation of cesium and rubidium, 301. der Cc Cady, H. P., Laboratory Guide to Qualitative Analysis, 462. ‘Canada Geol. Survey, Annual Re- port, 79, 394. Canadian Paleozoic corals, revision of the genera, Lambe, 79. Cathode rays, Seitz, 389. Chemical Analysis, Qualitative, Per- kin, 76; Bailey and Cady, 462. Chemischen Physiologie und Pathol- ogie, Beitriige zur, Hofmeister, 474. Chemistry, Bibliography of, Bolton, Oo. CHEMISTRY. Acetylene, heat of dissociation and combustion of, Mixter, 547. Acid nitrates, Wells and Metzger, 460. — solutions, methods of standard- izing, Hopkins, 461. Alloys, Thermochemistry of, Baker, , 460 Ammonium, existence of, Ruff, 387. — cyanate, Walker and Wood, 75. Aniline and analogous bases, Saba- tier and Senderens, 387. * This Index contains the general heads, Borany, CHEMISTRY (incl. chem. physics), GEOLOGY, MINERALS, OBITUARY, ROcCKs, and under each the titles of Articles referring thereto are mentioned, 476 CHEMISTRY. Antimony pentachloride, double compounds of, Rosenheim and Stellman, 460. — salt of quadrivalent, Wells and Metzger, 389. Cesium and rubidium, etce., mation of, Browning, 301. — and thorium, double chlorides, Wells and Willis, 191. Cesium-tellurium fluoride, and Willis, 190. Calcium, strontium and barium, estimation, Peters, 216. Carolinium, a new element, 462. Caro’s acid, composition of, Baeyer and Villiger, 73. esti- Wells’ INDEX. D Davis, B., behavior of small closed cylinders in organ-pipes, 185; locating nodes and loops of sound in the open air, 263. Davis, J. W. , motion of compressible fluids, 107. Derby, 0. A. , Manganese deposits of Minas Geraes, Brazil, 18. Dewar, J., the nadir of temperature and allied problems, 168; separa- tion of the least volatile gases of atmospheric air, 207. Dielectric constant of Hormell, 483. Downs, E. S., spectrum of alternat- ing current discharge, 66. paraffins, Kuropium, a new element, Demar-| Duff, A. W., secondary undulations gay, 3821 Gooch-crucible, modified, Heraeus, 388. Hydrogen, see Hydrogen. Lead, radio-active, Hofmann and | Strauss, 388. Molybdic acid, estimation of, reduced by hydriodic acid, Gooch and Pulman, 449. Persulphates, determination Peters and Moody, 367. Phosphorus, emanation Barus, 327. — ignition temperature, 73. Quantitative analysis, new method of, Thatcher, 320. Quartz, vitrified, Shenstone, 74. Radium, radio-activity produced by salts of, Curie and Debierne, 319. Selenium in sulphuric acid, Jouve, 79. Sodium thiosulphate, solutions of metallic salts, Norton, 115. Supersaturated solutions, influence of magnetism, de Hemptinne, 79. of, from, Kydman, Tellurium, gravimetric determina- | tion, Gutbier, 389. Thorium, new element associated with, Baskerville, 462. Clerke, A. M., Unsolved Problems of Low Temperature Research, 393. Colorado Valley, geology of the Little, Ward, 401. Connecticut, river system of, Hobbs, 467. Copper voliameter, new solution for, Shepherd, 49. Cosmogony, mathematical notes to rival theories of, Fisher, 140. Cuba, bituminous deposits, Peckham, 33. action on | Spa Ethnology, publications of Bureau shown b y tide-gauges, 123. E Earthquake investigation committee, publications, 90. Electric arc, resistance and electro- motive force, Duddell, 392. — convection, magnetic effect of, Adams, 155 : Pender, 178 ; Wilson, d22. — discharge-current from surface of large curvature, Almy, 175. — flow in gases, Stark, 77. — resistance, experiments on high, Rood, 91. Electro-Magnetic Phenomena, Treatise, Lyons, 77. Electromotive force of the Clark- and the Weston-cell, Jaeger and Lindeck, 76. Ether and_ gravitational matter through infinite space, Kelvin, 390. of American, 89. F Fauna of Porto Rico, reports, 471. Field Columbian Museum, publica- tions, 89. Fisher, O., mathematical notes to rival theories of Cosmogony, 140. Fluids, motion of compressible, Da- vis, 107. Ford, W. E., calaverite, 225. G Galvanometers of high sensibility, Mendenhall and Waidner, 249. Gases of atmospheric air, separation of the least volatile, Liveing and Dewar, 207. INDEX. GEOLOGICAL REPORTS. Arkansas, vol. v, 78. Canada, 79, 394.. Indiana, 466. Towa, vol. xi, 396. Maryland, 77 Minnesota, vol. vi, 395. New Jersey, 466. United States, 21st Annual Report, 465. Geology, lectures on Economic, Bran- ner and Newsom, 467. — lectures on. EKlementary, Bran- | ner, 467. GEOLOGY. Boston Basin, Burr, 79. Bituminous deposits of Cuba, Peckham, 33. | Cambrian fossils of Missouri, Beecher, 362. Carboniferous and Permian age of | the Red Beds of Adams, 383. — terranes, time values of provin- cial, Keyes, 305. Crinoid from the Charlestown, Ind., 297. Dragons of the Air, Seeley, 396. Kocene deposits of Maryland, Clark and Martin, 77. — Mammalia in the Marsh collec- tion, studies, Wortman, 148, 193, 281, 377, 421. Etinde, volcano of the Cameroon Mts., Africa, Esch, 81. Eurypterid remains in the Missouri Cambrian, Beecher, 364. Glaciers, see Glaciers. Helgeland, geology, Vogt, 82. Ice ramparts, Buckley, 466. Igneous rocks, classification, von Federov, 83. — Laccoliths of Highwood Mts. Mon- tana, geology of, Weed and Pir- sson, 1 Limestones of Missouri, age and geological relations, Nason, 358. Little Colorado valley, geolog y; Ward, 401. Oklahoma, Hamilton of new, Wood, Loess of Iowa City and vicinity, | Shimek, 467. Manganese deposits of Minas Ge- | raes, Brazil, Derby, 18. Paleozoic corals, Canadian. Lambe, | 79. River system of Connecticut, Hobbs, 467. | Goldschmidt, ATT Vesuvius, explosive activity of, during April aud May, 1900, Matteucci, 81. Gibbs, J. Willard, receives Copley medal, 474. | Glaciers, ancient, in Tierra del Fuego and Antarctic, 464. _— in Switzerland in 1900, 465. V., Harmonie Complication, 325. Gooch, F. A., Research Papers from the Kent Chemical Laboratory, Yale Univ., 321; molybdic acid reduced by hydriodie acid, 449. und 'Gooch-crucible, modified, Heraeus, 388. | Gray, E., Nature’s Miracles, 325. Amygdaloidal melaphyre of the) H Harmonie und Complication, Gold- schmidt, 329. Harvard College Observatory, An- nals, 474. Helgeland, geology, Vogt, 82. Helium, viscosity of, Schultze, 468. Hoffmann, G. C., mineral occur- rences in Canada, 447. Hormell, W. G., dielectric constant of paraffins, 433. Hydrogen, conductivity by nega- tively charged ions, Townsend and Kirkby, 76. — diffusion through palladium, Win- kelmann, 389. — spectra of, Trowbridge, 310. I Indiana Geol. Survey, etn Annual Report, 466. Ions, conductivity pro dced in hy- drogen, etc., by the motion of neg- atively charged, Townsend and Kirkby, 76. — migration of, Gans, 463. Iowa Geol. Survey, vol. xi, 396. K Kelvin, dynamical theory of light and heat, 391; ether and gravita- tional matter in space, 390. | Kent Chemical Laboratory, Research papers from, Yale University, Gooch, 821. Keyes, C. R., time values of pro- vincial Carboniferous terranes, 305. L | Light and heat, dynamical theory of, Kelvin, 391. 478 Lithology, Outline of Elementary, | Barton, 469. Liveing, G. D., separation of the least volatile gases of atmospheric air, 207. Lloyd, M. G., magnetic effects in| tellurium, 57. Low Temperature solved Problems, Clerke, 398. Lyons, T. A., Treatise of Electro- | Magnetic Phenomena, 77. M Magnetic effects of moving electric charges, Adams, 155; Pender, 173; Wilson, 822. Marine invertebrates of Kastern Can- ada, Catalogue, Whiteaves, 599. Marsh collection, Peabody Museum, Kocene mammalia of, Wortman, 148, 198, 281, 377, 421. Maryland, Eocene deposits, Clark and Martin, 77. McNairn, W. H., phlogopite crys- tals from Ontario, 398. Mendenhall, C. E., galvanometers of high sensibility. 249. Meteorite, the Veramin, Ward, 458. — stony, of Perry, Alabama, 469. — swarms. Hogbom, 399. Meteorites, iron, classification, Co- hen, 86. Meteorological Atlas of Argentine Republic, Delachaux, 88. Mexico, Bulletin of the Geol. Insti- tute, Ordefiez, 84. Mineralogie, Handbuch der, Hintze, 399. Mineralogy and Petrography at Yale University, Penfield and Pirsson, 398. MINERALS. Atacamite crystals. Chili, 100. Barite, Kansas City, Mo., 47. Brookite, No. Carolina, 180. Calaverite, California, 225. Cal- cite, 42. Caledonite, Montana, 47. Celestite, Salina Co., Kan., 48. Ceruléite, Chili, 85. Chrysoberyl, New York, 104. Conchite, 84, 469. Corundum, in_ syenite, Madras, 467. Datolite, Quebec, 447. Eleolite-syenite, Madras, 467. Faujasite, Quebec,.448. Galena, 45. Hussakite, Brazil, 85. Ktypteite, 469. Lassallite, France, 85. Leadhillite, Research, Un-| INDEX California, 46. Linarite, Califor- nia, 46. Marcasite, 414. Mercurie iodide, New South Wales, 98. Molybdo- phyllite, Sweden, 469. Octahedrite, No. Carolina, 180. Pectolite, New Jersey, 99. Phlo- gopite crystal, Ontario, 398. Pyrite, New Jersey, 45; 414. Pyroxene, Tennessee, 105. Quartz, vitrified, Shenstone, 74. Realgar, Washington, 103. Seligmannite, Switzerland, 85. Sul- vanite, So. Australia, 85. Stan- nite, crystallization, 469. Syn- chisite, Greenland, 469. Termierite, France, 85. Torrensite, 85. Vesuvianite, New Mexico, 104. Viellaurite, 85. Violaite, 86. Minnesota, Geol. Survey, voi. vi, 395. Missouri, Cambrian limestones, Nason, 358; fossils of, Beecher, 362, 364. Mixter, W.G., heat of dissociation ane combustion of acetylene, etc., 347. . Montana, laccoliths of Highwood Mts., Weed and Pirsson, 1. Moody, S. E., determination of per- sulphates, 367. Mose A. J., mineralogical notes, 98. N Narkose, Studien tiber die, Overton, 320. Nason, F. L., age and geol. relations of the St. Joseph and Potosi lime- stones of Missouri, 399. Nature’s Miracles, Gray, 320. New Jersey, Geol. Survey, 466, Nodes and loops of sound in the open air, method of locating, Davis, 263. Norton, J. T., Jr., action of sodium thiosulphate on solutions of metal- lic salts, 115. Noyes, W. A., Elements of Qualita- tive Analysis, 462. O OBITUARY— Lacaze Duthiers, Baron de, 326. Le Conte, Joseph, 174, 248. Nordenski6ld, Adolf Erik, 326. Koenig. Rudolph, 474. Schimper, A. F. W., 474. Schott, Charles Anthony, 326. Schur, Dr. Wilhelm, 326. INDEX. Oklahoma, Carboniferous and Per- | mian age of the Red Beds of, Adams, 383. Organ-pipes, behavior of small closed cylinders in, Davis, 185. FE Paraffins, dielectric Hormell, 433. constant of, Peckham, H. E., bituminous deposits of Cuba, 33. Penfield, S. L., tributions to Mineralogy and Petro- graphy from — Scientific School, 398. Perkin, F. M. Raalitatine Chemical Analysis, 76. Peters, C. A., and barium, calcium, strontium estimation of, 216; determination of persulphates, 367. | Phasenlehre, die heterogenen Gleich- gewichte vom Standpunkte der, Roozeboom, 463. Phosphorus, effect of temperature and moisture on the emanation of, Barus, 327. Photometry of the ultra-violet rays, Kreusler, 463. Pirsson, L. V., geology of laccoliths in Highwood Mts.. Montana, 1; Contributions to Mineralogy and Petrography from Sheffield Scien- tific School, 398. Porto Rico, reports on Fauna, 471. Pulman, O. S., Jr., molybdic acid reduced by hydriodic acid, 449. Q Qualitative Analysis, Noyes, 462. — — Laboratory Guide to, Bailey and Cady, 462. Elements, R Radio-activity produced by salts of radium, Curie and Debierne, 319. — of lead, Hofmann and Strauss, 388. Resistance, effect on, of amalgamated gases, Rollins, 322. Rhees, W. J., Origin and History of the Smithsonian Institution, 473. Robinson, H. H,, octahedrite and brookite from No. Carolina, 180. ROCKS. Corundum-syenite and _ eleolite- syenite containing graphite, Ma- dras, Holland, 467. Igneous rocks, classification, von Federov, 83, . 479 — of Cameroon Mts., Esch, 81. Laccoliths of Highwood Mts., Mon- tana, Weed and Pirsson, 1. Kedabekite, 247. Kenyte, 247. Melaphyre of Boston Basin, Burr, 79 Peridotite, altered, 468. Perknite, Turner, 468. | Rogers, A. F., mineralogical notes, 42 calaeerita 2850 )bi: | Rollins, W., effect of amalgamated gases on resistance, 322. Rood, O. N., experiments on high electrical resistance, part II, 91. | Roozeboom, H. W. B., die hetero- genen Gleichgewichte vom Stand- punkte der Phasenlehre, 465. Royal Society, Copley medal, 474. S) ‘Scientia, Bohn, 89. Secondary undulations shown by tide-gauges, Duff, 128. Seeley, H. G., Dragons of the Air, 396. Shepherd, W. K., new solution for copper voltameter, 49. Smithsonian Institution, Annual Report, 473; Origin and History, 1835-1899, vol. i, Rhees, 473. Sound, experiments in, Davis, 263. Spectra, infra-red, of the alkali- metals, photography, Lehman, 390. Spectrum of the alternating current discharge, Wright and Downs, 66. Sey H.N., pyrite and marcasite, 414. 185, AY Tellurium, magnetic effects in, Lloyd, 57. Temperature, the nadir of, and allied problems, Dewar, 168. — Unsolved Problems of Low, Clerke, 398. Temporary set, Barus, 247, Thatcher, R. W., new method of quantitative analysis, 320. ., Tide-Gauge observations, Duff, 128. Trowbridge, J., spectra of hydro- gen, 310. U United States Geol. Annual Report, 465. — — Naval Observatory, publica- tions, 90. Survey, 21st 480 INDEX. i V Wood, E., new Crinoid from the Vernl ooAe Ee denth’ of ashes ait Elamulijen: of Charlestown, Ind., 297. Bermuda in 1901, 88. Verrill, A. H., additions to the Avi- fauna of the Bermudas, 64; nomen- clature of Bermuda birds, 470. Vesuvius, explosive activity i in April, May, 1900, Matteucci, 81. W Waidner, C. W., galvanometers of high sensibility, 249. Ward. H. A., Veramin meteorite, 453. Ward, L. F., geology of the Little Colorado Valley, 401. Weed, W. H., geology of laccoliths in Highwood Mts., Montana, 1 Wells, H. L., cesium-tellurium fluo- ride, 190; double chlorides of cex- sium and thorium, 191. Willis, J. M., cesium-tellurium fiuo- ride, 190; double chlorides of ce- sium and thorium, 191. Winton, A. L., anatomy of Cocos nucifera, 265. Wortman, J. L., studies of the Hocene ‘Mammalia in the Marsh collection, 143, 198, 281, 377, 421. Wright, A. W., spectrum of alter- nating current discharge, 66, Y Yale Bicentennial pubes d2l, 398. | ez, Zoologisches Addressbuch, 90. Zoology, Shipley and MacBride, 472. — Practical, Parker and Parker, 472. ZOOLOGY. Avifauna of the Bermudas, addi- tions to, Verrill, 64; nomencla- ture of, 470, Fauna of Porto Rico, a1. Fishes, death of, at Bermuda in 1901, Verrill, 88. Established by BENJAMIN SILLIMAN in 1818. AMERICAN “1 JOURNAL OF SCIENCE. EHprror::: EDWARD: 3S. DANA, ASSOCIATE EDITORS Proressors GEO. L. GQDODALE, JOHN TROWBRIDGE, > W.G. FARLOW anp WM. M. DAVIS, or CAMBRIDGE, Proressors A. E. VERRILL, HENRY S. WILLIAMS anp L..V. PIRSSON,-or NEw. HAvEN. PROFESSOR GEORGE F. BARKER, oF PHILADELPHIA, PROFESSOR JOSEPH S. AMES, or BALTIMORE, Mr. J. S.. DILLER, or WASHINGTON. FOURTH SERIES. VOL. XII—[WHOLE NUMBER, CLXII.] No. 67.—JULY; 1901. NEW HAVEN, CONNEG@PICUT. SUA ae THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 125 TEMPLE STREET. See eee . Union. Remittances should be made either by money orders, registered letters, or bank checks (preferably on New York banks). : , C ay Ad : . ava Va Re an Pee) a $e Eon r. Cyrus Adler, sy s Sea uk Librarian U. S. Nat. Museum. g OQ Das way = eage == e és = “eRe eS, Fwilee 3 cs Published monthly. Six dollars per year. $6.40 to countries in the Postal Several shipments of interesting minerals from Montana have lately Meum : eats 3 received, a number of the most notable being reviewed below. The great Kee mineral wealth of this district is well-known, and to its present activity oh league By due the increasing number of mineral species reported. CRYSTALLIZED RHODOCHROSITE. Characteristic rhombohe- drons of arich pink in aggregates, clean cut and perfect. They lack the © translucency of the Colorado find, but in other features closely resemble the material from that state. Crystals vary in size from 14 to an inch and more across with corresponding thickness. Occasionally, groups are coated with a white drusy quartz making an association both Berek and attractive. Prices 50¢. to $2.00. Massive 50c. per Ib. CRYSTALLIZED ENARGITE. 4 few neat groups of the rare cop-. per sulpharsenate, illustrating habit and modifications. Orystallized metal- lic minerals find quick sale, and it is confidently expected that at our low ~ scale of prices, the best of the lot will be aie sold. Prices 50c.. to. $3.00. Massive 7dc. per Ib. AMETHYST. Our correspondent epeit several months at the Ame- thyst locality, and after much difficulty succeeded in securing what are con- sidered the finest specimens as yet found, The types show parallel growth, interesting capping forms and single doubly terminated crystals. COclor is’ good, generally deepening at the tips. Convenient cabinet sizes—d0e. to $4.00. TOURMALINITIC QUARTZ. Two shipments of this odd occurrence permit us to describe its varied habits and strange associations. In appearance the crystals are black, due to ~ the fine network of included Tourmaline hairs, which occasionally are widely separated, reminding one of the more common “‘ fléches d’amoar.” A pecu- liar ‘‘stem’’ extends beyond the base of some crystals, around which the Quartz seems to have formed. The Tourmaline is often isolated on the ‘“stem,” appearing in bristling tufts and masses. The very fact that these specimens differ widely from the usual Quartz forms, makes them attractive to the collector, *but aside from this point they combine such interesting features scientifically, that a steady demand is ap crpavet BEE 50c. to $3.50 in 2 and 4 inch sizes. COVELLITE, CHALCOCITE and. BORNITE in pure massive lumps neatly trimmed to cabinet sizes. 50c. to $1.00 for a typical specimen. Trial orders solicited. Approval consignments submitted. Foote MIN SrA —- FORMERLY DR. A. E. FOOTE, WARREN M. FOOTE, Manager. ESTABLISHED 1876. PHILADELPHIA, PARIS. 1317 Arch Street. 24 Rue du Champ de Mars. AN OLD COLLECTION, We shall offer for sale another lot of miscellaneous specimens from an old collection, besides the several hundred advertised last month, which ineluded Roeperite, Bementite, Pa. Amethyst, French Creek= Chalcopyrite, Cubic Spinels, Japan Stibnites, Hungarian Opal, Siberian Emeralds and Beryl, Atlasite, Huchroite, Langite, Lett- Somite, Serpierite, Connellite, Libethenite, Manganite, Dawsonite, Partzite and extra fine Cumberland Mimetite - and Pyromorphite. (Many of these have been sold.) EXTRA FINE MOLYBDENITE CRYSTALS! Choice crystals of molybdenite have always been high priced and not at allcommon. Of the large Jot of unusually well defined crystals offered last month, from a new locality in the far west, we have still a good supply. Well formed crystals, one inch to four inches in diameter, with low pyramids of about 40 and 55 degrees, 10c. to $1.50... EXCELLENT FLEXIBLE SANDSTONE. A stone that will bend! What makes it bend? Buy a specimen and see if you can discover the secret, Splendid specimens, 3 x1 inches, 5c.; 5 x 2 inches, 10c.; 7x2 inches, 20c.; on up to 13x24 inches, 75c.; and even larger speci- mens. We call your attention to the fact that these prices are unprecedentedly low. Do not lose this opportunity to secure a good specimen if you have not already purchased one. CROOKESITE. A few specimens remain in stock. Prices, 50c. to $20. OTHER IMPORTANT RECENT ADDITIONS. Mohawkite, Stibiodomeykite, Domeykite from the Mohawk Mine, each piece . tested. Parisite, doubly terminated crystals in the matrix. Australian Atacamite, Cerussite, Anglesite, Pyromorphite, lIodyrite, Embolite, Cuprite, Murshite, Stolzite, etc. -New and rare species from (Greenland. Iridescent Pyrite from New Jersey. fine Milky Quarta crystals from Stop) Lead from Franklin, showy massive Bornite, Millerite in radiating crystals from Antwerp, many fine Barites, Calcites, Fluorites,, Hematite and Quartz, Sphaler- ites, etc., from England. JUST RECEIVED. Covellite, from Wyoming, a new find. Wollastonite, Rhodonite, Willemite, ete., from Franklin. OUR SPRING BULLETIN. 24 pages, 29 illustrations, will be sent free to anyone desiring it. It describes many other recent additions to our stock and gives a compiete list of Mineralogi- eal Books, 124-page Illustrated Catalogue, giving Dana Species number, crystal system, hardness, specific gravity, chemical composition and formula of every mineral, 25c. in paper. 44-page Illustrated Price-Lists, also Bulletins and Circulars free. - GEO. L. ENGLISH & CO., Mineralogists, Dealers in Scientific Minerals, 3 and 5 West 18th Street (First Door West Fifth Ave.), New York City. at ~ es ee ~ CON Eo aT Se ART. I.—Geology of the Shonkin Sag and Palisade Butte — Laccoliths in the Highwood Mountains of Toe by 3 W.-H. Wuxp and L. V. Pirsson. 5 bal II.—Manganese Ore Deposits of the Quelue (atapeney Dis- <5 3 trict, Minas Geraes, Brazil; by O. A. DERBY ___. -___~- 18 tS IlIl.— Bituminous Deposits ne at the South and East of Ne Cardenas, Cuba ; by H. KE. Peckuam .. 2,-2.2 7 ITV.—Mineralogical Notes, No. 2; by A. F. Rocurs ._.._._ 42 é V.—New Solution for one Copper Volttameter; by W. K.- ce SHEPARD (2222002 cc Lee PRE ee * VI.—Thermo-magnetic and Galvano-magnetic Effects an, eee Tellurium; by. G. Diroxp, 2300) 32 See ee 57 VIL—Additions to the Avifauna of the Bermudas with diagnoses of two new Subspecies; by A. H. Verritt_._. 64 VII —Induced Alternating Current Discharge studied with Reference to its Spectrum and especially the Ultra-Violet Spectrum; by A. W. Wricut and E. 8. Downs__-_._-_ 66 SCIENTIFIC INTELLIGENCE. ( Chemistry and Physics—lguition temperature of Phosphorus, EypMAN: Composi- tion of *‘Caro’s Acid,” BAEYER and VILLIGER, 73.—Vitrified Quartz, SHEN- Bhs STONE, 74.—Influence of Magnetism upon Supersaturated Solutions, De HEMP- Nee TINNE: Ammonium Cyanate, WALKER and Woon: Selenium in Sulphurie Acid, E JOUVE, 75.—Qualitative Chemical Analysis: Viscosity of Argon, SCHULTZE: | Conductivity in Hydrogen and Carbonic Acid Gas, TowNSEND and KiRKBY: | ~ E.M.F. of Clark- and Weston-Cell, JAEGER and LinpECK, 76.—Electrical Flow | § in Gases, STARK: Treatise on Electro-Mapnetic Phenomena, Lyons, 77. Geology and Natural History—Kocene Deposits of Maryland, and Systematic ~ Paleontology, 77.—Annual Report Geol. Survey Arkansas, 1892, 78.—Summary Report Geol. Survey Canada, 1900: Revision of Genera and Species of Cana- _ ian Paleozoic Corals, LAME, 79.—Amyedaloidal Melaphyre in Mass., BURR, . * 80. —Explosive Activity of Vesuvius, Marreccct: Der Vulcan Etinde in Kam- aoe: erun und seine Gesteine, EScu, 81.—Sdndre Helgeland, Vogt, 82.—Classification of Igneous Rocks, FEDEROW, 83.—Boletin del Instituto Geologics de Mexico: ah Brief Notices of recently described Minerals, 84.—Iron Meteorites, 86.—Mileh- | § saft und Schleimsaft der Pflanzen, Mouiscu, 87. —Remarkable Instance of Death ‘¢) es ae of Fishes at Bermuda, V ERRILL, 88. Ne Gee Miscellaneous Scientific Intelligence—Atlas Meteorologico de la Republica Argen- Bas tina, 88.—British Museum Catalogues: Select Bibliography of Chemistry, 1492-1897: Publications Bureau of American Ethnology: Field Columbian Museum: Scientia, 89.—Zoologisches Addressbuch: Publications Earthquake Investigating Committee ; Publications U. 8. Naval Observatory, 90. Dr. Cyrus Adler, ae Yous ge ee Librarian U. S. Nat. Museum. pe ee yon xii AUGUST, 1901. Established by BENJAMIN SILLIMAN in 1818. THE AMERICAN | JOURNAL OF SCIENCE Epirorn: EDWARD S. DANA. ASSOCIATE EDITORS. : Proressors GEO. L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or CAmprincE, Proressors A. FE. VERRILL, HENRY S. WILLIAMS anp L. V. PIRSSON, or NEw HAVEN. ProFessor GEORGE F. BARKER, or PHILADELPHIA, ProFressor JOSEPH S. AMES, or BALtTimorg, Mr. J. S. DILLER, or WASHINGTON. FOURTH SERIES. VOL. XII—[WHOLE NUMBER, CLXII.] No. 68.—AUGUST, 1901. NEW HAVEN, CONNECTICUT. LS OF. THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, I25 TEMPLE STREET. Published monthly. Six dollars per year. $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). es mineral wealth of this district is well pe, oe 0 its i due the increasing number of mineral species ‘reported. CRYSTALLIZED RHODOCHROSITE: Charact Prices 50c. to $2.00. Massive 50c. per Ib. . CRYSTALLIZED ENARGITE. A few neat groups of. eS per sulpharsenate, oe habit and modifications. ona interesting capping forms and single doubly terminated veyelae = good, generally deepening at the tips. Convenient cabinet sizes- $4.00. . Two shipments of this odd occurrence. permit, us. to ‘icone ee habits and strange associations. In appearance the ayee are ee ‘‘stem,” appearing in panne: tufts and masses. The very fact that’ specimens differ widely from the usual aria —— makes, them se: $3.50 in 2 and 4 inch sizes. ee COVELLITE, CHALCOCITE and - BORNITE in pure lumps neatly trimmed to cabinet sizes. 50c. to $1.00 for a ee pe Trial orders solicited. Approval consignee submitted. | = Os MINERAL oo FORMERLY DR. A. E. FOOTE, mer Sch : WARREN M. FOOTE, Manager. : ESTABLISHED 1876. eS PHILADELPHIA, ; PARIS; == 1317 Arch Street. 24 Rue du Champ de a a — _ We have just priced and now offer for sale another lot _ of miscellaneous specimens. There are few duplicates and then generally the localities are different. The prices are low, while the quality in many. cases is excel- lent Some specimens are now difficult to obtain, because the localities have long since been worked out or some similar reason. Such specimens as Pyromorphite, Beryl, Amethysts, etc, from Delaware County, Pa, are exaniples. Some specimens are valuable for their beauty, while others are valuable because they are rare. Your inspection invited. ; . IMPORTANT MINERALS. _ We still have limited supplies of the following, previously advertised, some of the lots having been replenished, but most of them balance of stock on hand. Mohawkite, Stibiodomeykite, Domeykite, from the Mohawk Mine, each piece tested. Parisite, doubly terminated crystals in the matrix. _ Australian Alacamite, Cerussiie, Anglesite, Pyromorphite, Iodyrite, Embolite, Cuprite, Marshite, Stolzite, etc. New and rare species from Greenland. fridescent Pyrite from New Jersey. fine Milky Quartz crystals from Suttrop, Lead from Franklin, showy massive Bornite, Millerite in radiating crystals from Antwerp, many fine Barites, Calcites, Fluorites, Hematite and Quartz, Sphaler- ite, ete., from England. = MOLYBDENITE. - This important mineral is generally found in rounded, ill-defined crystals, when they may be called crystals at all. Choice crystals of Molybdenite have always been high-priced and not at all common. Of the large lot of unusually well defined crystals offered last month from a new locality in the far west, we still have a good supply. Well-formed crystals, one inch to four inches in diameter, with low pyramids of about 40 and 55 degrees, 10c. to $1.50. EXCELLENT FLEXIBLE SANDSTONE. A stone that will bend! What makes it bend? Buy a Speen and see if you can discover the secret Splendid specimens, 3 x 1 inches, 5c.; 5 x 2 inches, 10c.; 7x2 inches, 20c.; on up to 13x24 inches, T5e.; and even larger speci- mens. We would call your attention to the fact that these prices are unprece- dentedly low. Do not lose this opportunity to secure a good specimen if you have not already secured one. RECENTLY RECEIVED. Covellite, from Wyoming, a new find. Wollastonite, Rhodonite, Willemite, etc., from Franklin. - FREE TO TEACHERS. On request we will send free to teachers, our pamphlet, “Suggestions to Teachers of Mineralogy.” A postal will secure it. ; OUR SPRING BULLETIN. 124-page Illustrated Catalogue, giving Dana Species number, crystal system, hardness, specific gravity, chemical composition and formula of every mineral, 25c. in paper. 44-pnage Illustrated Price-Lisis, also Bulletins and Circulars, free. GEO. L. ENGLISH & CO., Mineralogists, Dealers in Scientific Minerals, 3 and 5 West 18th Street (First Door West of Fifth Kies ) New York City. | BARG AENS Le ee ate Py Pa ta CON TEN FS: ARE DX. Experiments on High Electrical Resistance, I] 5 byO.5N. Roope soe fa ee ee ee X: —Mineralogiea Notes ; 5 by A. J. Mosns “=2- os | gauges : 52 Oe ‘ XIV. —Mathematical Notes to Rival Theories of Conmogoo ; ‘by 0. WASHER OR SS. 26s oa ee ee 14 XV.—Studies of Eocene Mammalia in the Marsh Collection, ae Peabody Museum; by J. L. Worrman..-=-___----- “14 XVI. S Wipcwonaa ec Effects of OS Charged Spheres ° oe, by HP. ADAMg: oe eo eee Bok XVIL—The Nadir of Temperature and Allied Problese te 3 J. Dewar . 2.222222 222222 eee ee eee eee | SCIENTIFIC INTELLIGENCE. Se Miscellaneous Scientific Intelligence—Magnetic effect of Electrical Concent PeNnDER, 173.—American Association for the Advancement of Science, 174, Obituary.—Dr. JosepH LeConTe: PRoressor Perer GuaTrie Tatt, 174 Pr VYius fi Gier, Librarian U. S. Nat. Museum. ae S6 Ss. aie ooo —™——C*SC#MSEPT'EMBBEER,, 19901. Established by BENJAMIN SILLIMAN in 1818. _ AMERICAN JOURNAL OF SCIENCE. _ ASSOCIATE EDITORS Proressors GEO. L. GOODALE, JOHN TROWBRIDGE, ’ W. G. FARLOW anp WM. M. DAVIS, or Camprwer, - Proressors A. E. VERRILL, HENRY S. WILLIAMS anp L. V. PIRSSON, or NEw Haven. PROFESSOR GEORGE F. BARKER, or PHILADELPHIA, PRoFEssoR JOSEPH S. AMES, or BALTIMoRE, Mr. J. S. DILLER, oF Wasurncron. FOURTH SERIES. VOL. XII—[WHOLE NUMBER, QEXTH!20 Ins), No. 69.—SEPTEMBER, 190)" “ NEW HAVEN, CONNECTICUT. 1901. THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 125 TEMPLE STREET. Published monthly. Six dollars per year. $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). | Sees Se, Ss o> Sao NOSES te Bs SRE EY 2 SS ec > Brg EM IIS ie: — = LA Pe See = Several shipments of interesting minerals from Montana have lately been 2 received, anumber of the most notable being reviewed below. The great Pe mineral wealth of this district is well-known, and to its present eS is due the increasing number of mineral species reported. ae CRYSTALLIZED RHODOCHROSITE. Characteristic rhombohe- drons of a rich pink in aggregates, clean cut and perfect. They lack the | translucency of the Colorado find, but in other features closely resemble the — . material from that state. Crystals vary in size from 14 to an inch and more across with corresponding thickness. Occasionally, groups are coated with ~ a white drusy quartz making an association both interesting and attractive. Prices 50c. to $2.00. Massive 50c. per Ib. . CRYSTALLIZED ENARGITE. A few neat groups of the rare cop- per sulpharsenate, illustrating habit and modifications. Crystallized metal- lic minerals find quick sale, and it is confidently expected that at our low scale of prices, the best of the lot will be immediately sold. Prices 50c. to. $3.00. Massive 75c. per lb. AMETHYST. Our correspondent spent several months at the Ame- thyst locality, and after much difficulty succeeded in securing what are con- sidered the finest specimens as yet found. The types show parallel growth, interesting capping forms and single doubly terminated crystals. Color is good, generally deepening at the tips. Convenient cabinet sizes—d0c. to $4.00. TOURNMALINITIC QUARTZ. Two shipments of this odd occurrence permit us to describe its varied habits and strange associations. In appearance the crystals are black, due to the fine network of included Tourmaline hairs, which occasionally are widely separated, reminding one of the more common “ fléches d’amour.” A pecu- liar ‘‘stem’’ extends beyond the base of some crystals, around which the Quartz seems to have formed. The Tourmaline is often isolated on the “stem,” appearing in bristling tufts and masses. The very fact that these specimens differ widely from the usual Quartz forms, makes them attractive to the collector, but aside from this point they combine such interesting — features scientifically, that a steady demand is anticipated. Prices 50c. to $3.50 in 2 and 4 inch sizes. . COVELLITE, CHALCOCITE and BORNITE in pure massive lumps neatly trimmed to cabinet sizes. 50c. to $1.00 for a typical specimen. Trial orders solicited. Approval consignments submitted. E'OOT EE. IMaIN See See FORMERLY DR. A. E. FOOTE, WARREN M. FOOTE, Manager. ESTABLISHED 1876. PHILADELPHIA, PARIS, 1317 Arch Street. 24 Rue du Champ de Mars. DE pep tines AT gh ae is Be Neo ea ao Pathe Mai its ch : Pan . 328 BEAUTIFULLY CRYSTALLIZED RHODOCHROSITES. The largest number of Rhodochrosite specimens we ave ever secured was recently received from Colorado. The few obtained last spring were quickly sold. The present collection contains an assortment of sizes and prices which enables us to meet the requirements of every collector, 15c. to $4.00. The association of green fluorite with the pink rhodochrosite forms a most pleas- ing combination of colors. Such specimens are higher in price, but, nevertheless, cheap at 25c. to $6.00. SHOWY COLORADO HUBNERITE. 60 excellent specimens of the Cement Creek Hibnerite (formerly known as Megabasite) are just being placed on sale. No such lot has ever before been offered for sale. The dark brown, more or less radiating blades of the Hibnerite stand out in bold contrast with the white Quartz matrix, 25c. to $3.00.. MONTANA VESUVIANITE. A small lot of the showy Montana Vesuvianites in blue calcite is just in, $1.00 and $1.25. APATITE CRYSTALS. A large lot of Canadian Apatite crystals arrived during August. Quite a num- ber of them are doubly terminated and of good size, 35c. to $2 00; others have a single good termination, 10c. fo 25c. They are most excellent illustrations not only of the mineral but of the hexagonal system. RECENT FINDS AT FRANKLIN. It seems likely that the day will soon be past when specimens of any kind will be obtainable from Franklin Furnace. We have been buying up everything really good (but no rubbish) and can offer many attractive specimens of crystal- lized Rhodonite, pink zinciferous Wollastonite, Bementite, Caswell- ite, Leucophoenicite, Hancockite, Polyadelphite Garnets, etc. Our accessions during August have included many exceptionally desirable specimens —all at most reasonable prices. AUSTRALIAN MINERALS. 200 Atacamites are now on sale at low prices, 25c. to $2.50. Many choice Embolites at 25c. to $1000; Cerussites at 25c. to $3.50; Anglesites in groups of crystals coating reticulated Cerussite, 50c. to $6.00; Cuprite, loose erystals, 5c. to 15¢.; matrix groups of Cobaltite crystals, $1.00 to $2.00; Pyro- morphite, extra good, $2.00 to $350. Also many other desirable specimens such as reticulated Copper, Azurite, Gold Quartz, Molybdenite, Cassiterite, Iodyr- ite, Stolzite, Silver, crystallized Smithsonite, crystallized Chalcopyrite, ete. We are now selling our sixth hundred of the peerless Penfield Contact Gont- ometers. the cheapest, yet the best ever on sale. No crystallographer can afford to do without them. Two models, 50c. each: $5.00 per dozen. 124-page Illustrated Catalogue, giving Dana Species number, crystal system, specific gravity, chemical composition and formula of every mineral, 25c. in paper. 44-pnage Illustrated Price-Lists, also Bulletins and Circulars, free. GEO. L. ENGLISH & CO., Mineralogists, Dealers in Scientific Minerals, 3 and 5 West 18th Street (First Door West of Fifth Ave.), New York City. netic " sae a > © ce Pee) Shoe apa ts chor ae f a XXIL —Double Chlorides of Cesium ane Thorium ; = L. WeEtis and J. M. Wiras - X XITI.—Studies of Eocene Mammalia in he Marsh Coll tion, Peabody Museum ; by J. L. Worrman _...- || XXIV.—Separation of the Least Volatile Gases of Atm . ee Air, and their Spectra ; by G Ge D. Live See Fie Se eS ee ee XXV.—Estimation of Calcium, Strontiam, and Ba the Oxalates ; oe C.. AS Prrers. SFE SO ma a SCIENTIFIC INTELLIGENCE. Miscellaneous Scientific Intelligence—Temporary Set, Ma ‘Banus: rock-types, 247. ta Obituary—JOSEPH econ 248. me Cys at” Librarian \ORESS Fear Muse rar ig ett ae Be m ‘wet =a ; estes Eo tin! “SO um. eg 5 as “VoL. ead Sto oo SOC TOREE, 1901: ae ial by BENJAMIN SILLIMAN in 1818. ai AMERICAN _ | JOURNAL OF SCIENCE. Epiror: EDWARD S. DANA. ASSOCIATE EDITORS Proressors GEO. L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or Camprincg, -Proressors A. E. VERRILL, HENRY S. WILLIAMS anp L. V. PIRSSON, or New Haven. ProFEssoR GEORGE F. BARKER, oF PHILADELPHIA, PROFESSOR JOSEPH S$. AMES, or Battimore, Mr. J. S. DILLER, or WasHINGTON. FOURTH SERIES. VOL. XIJI—[WHOLE NUMBER, CLXII.] No. 70.—OCTOBER, 1901< WITH PLATES I-VI. NEW HAVEN, CONNECTICUT. 1901. THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 125 TEMPLE STREET. Published monthly. Six sare per year. $6.40 to countries in the Postal. Union. Remittances should be made either by money orders, registered letters, or bank checks ea ay on New York banks). 2 ae ee - " ee , “ee Sea ti i es ik i ' 4 a % io RPSL, ae oe aS i grees ea ig Ror eee 8 a, Sibi OE WEE ES aD SR = ally a liana on in cel ee whee eas +e STANDARD COLLECTION OF REQUIRED FOR Osborn’ § Prospector’s Field Book and Guide, An indispensable aid and guide to users of this book. The list of 120 specimens includes all important. minerals mentioned in the S ee text, besides illustrating the Scale of Hardness and the six systems of Crystalli- — =o. zation, i ee In selecting specimens from our large stock, a collection is secured which rep-— 32 resents, in a brief way, the varieties with which the prospector or miner is mest i, So likely to meet, and it has, therefore, a thoroughly practical value. Toaes te The following sizes are kept in stock ready for shipment. With the neat ae epee durable oak cases they can be kept in small ‘space. Every specimen is acecu- rately labeled with name, composition and locality, and numbered to correspond = to list. No. 23. Prospectors’, $16.00. 120 specimens, averaging 24x2 in. | ‘Handsome oak ease, three drawers, fitted with pasteboard trays, $10.25 extra. No. 24. Prospectors’, $7.00. 120 specimens, averaging 1}x1}in. Oak compartment case, $1.60 extra. tose The following special sets of ores are put up to order: No. 25. Useful Metallic and Non-Metallic Minerals. 300 speci- mens, averaging 24x 2 in., $175.00. Includes various examples of all important : minerals possessing economic value. S No. 27. Metallurgical Collection. 200 specimens, averaging 23 x 2 in., Stes $90.00. Embracing all the more important ores of common, rare or precious aes metals. ; No. 29. Metallurgical Collection. 100 specimens, averaging 24x2 in., Saree $25.00. An abridgement of No. 27. - No. 32. Ore Associations. 60 specimens, averaging 24x 2 in, $12.00. — i Including all of the minerals most commonly found with valuable ores. Sah ice an. The above are sold in 14 inch sizes at about half price. No. 34. Gold ores. 10 specimens, averaging 17 x 17 in., $10. 00. No. 35. Silver ores. 15 specimens, averaging as 1d ji in, $7.50. Also series illustrating occurrence of Iron, LEAD, CopPEr, ZINC, NICKEL AND COBALT, AND THE RARE HLEMENTS; ROUGH GEMS AND PRECIOUS STONES. (@- Minerals purchased in quantity. Send small mail samples. FOOTH MINERAT cow FORMERLY DR. A. E. FOOTE, a WARREN M. FOOTE, Manager. ESTABLISHED 1876. PHILADELPHIA, PARIS, 1317 Arch Street. : 24 Rue du Champ de Mars. YLVANIA GARNETS. - The peculiar excellence of the Garnets from Delaware and Chester Counties, Pennsylvania, is not so well _ knowu as might be supposed, largely because local col- lectors have bought them up with such avidity as not only to absorb the entire supply but also to run the prices up to figures which outside collectors would not pay—$10.00 to $35.00 each. We have just obtained 26 specimens from a collector who has held them for many years and finally sold them out cheaply. The best of them are very good, all are above the average, but the prices are only 25c. to $2.50. They are Spessartites of large sizes, 1 inch to 3 inches, quite perfect, fairly bril- liant, dif brown or brownish-red and are mostly implanted on a quartz-feldspar Re matrix. ; PENNSYLVANIA BERYLS. The same collector sold us quite a good lot of Delaware County Beryls, both loose and on the matrix, which we are selling at low prices, 10c. to $1.00. RARE SPECIES. From an old collection we have obtained many species which we are able to - offer at much lower prices than usually prevail among European dealers. Among __ them we would mention Pyrochlore xls on matrix, Polycrase xled, Ludlamite __xled, Bergmannite, Sarcolite, Humboldtilite xled, Dewalquite, Tritomite, Webster- ite. Hielmite, Pyrrhite, Ittnerite, Praseolite, Aspasiolite, Voigtite, Romanzovite, Breithauptite, Dipyre, Polymignite, Yttrotantalite, Acmite, Tachyaphaltite, Ran- dite. There are many other equally desirable species and all are at low prices. BRAZILIAN PYRITE TWINS. Over 100 extra large and fine twinned pyritohedrons of Pyrite from Brazil have just been secured. Though much better than any ever before in stock we can sell them at but 10c. to 35c. each. No lover of crystallography should miss this opportunity. RICH, POLISHED RUSSIAN MALACHITE. Russian Malachite is famed the world over. The present consignment con- _ sists of 15 specimens, some small pieces at 35c. to 50c. each, a few elegant large pieces at $10. 00, $12.50 and $15.00. We also have in stock a few choice = polished specimens from Arizona, some of them associated with Azurite. MONTANA VESUVIANITES. _ The second recent collection of the Beautiful Montana Vesuvianites has just arrived. It is but a small lot which will doubtless all be sold quickly, but if your order comes promptly it may secure for you a better specimen than any you have yet seen. 5c. to $2.00. COLORADO RHODOCHROSITES AND HUBNERITES. The delicate pink rhodochrosites with green fluorite advertised last month are 4 worthy of another and much more extended notice than we can here give them. ae Never before has so large a lot been offered for sale, nor have the prices been so ~ reasonable, l5c. to $2.00, with a few extra fine specimens at $2.50 to $6.00. The Hitibnerites are much the best ever on sale, the rich, dark-brown color of the radiating blades contrasting most pleasingly with the .white quartz matrix. 25c. to $3.00. We are now selling our sixth hundred of the peerless Penfield Contact Goni- ometers. the cheapest, yet the best ever on sale. No crystallographer can afford to do without them. Two models, 50c. each: $5.00 per dozen. 124-page Jilusirated Catalogue, giving Dana Species number, crystal system, specific gravity, chemical composition and formula of every mineral, 25c. in paper. 44-page Illustrated Price-Lists, also Bulletins and Circulars, free. GEO. L. ENGLISH & CO., Mineralogists, Dealers in Scientific Minerals, - Band 5 West 18th Street (First Door West of Fifth Ave.), New York City. Py. ie ce, oe ey i a ce ame le Sa oe ea ae = ye a — 2 aes 2 SE es . xD pee —: penenennaes Ct CLE IE PLLA AEE DELLE CIID DA CON TEN TS; Page | § Art. XXVII.—Galvanometers of High Sensibility; by C. HK. MENDENHALL and CO. W. WAIDNER .. =. .2._-..2_..7 949 A: XXVIII.—Method of. Locating Nodes and Loops of~Sound in the Open Air with Applications; by B. Davis. .--. 263 XXIX.—Anatomy of the Fruit of Cocos nucifera; by A. ik WINTON. 82 0 a ee ee 265 XXX.—Studies of Eocene Mammalia in the Marsh Collec- tion, Peabody Museum; by J. L. Worrman. With — Plates: TAPVo S20 so ee 28] - XXXI.--Crinoid from the Hamilton of Charlestown, Indiana; by Ee Woon. With Plate Vi 2-220. = ee 297 XX XII.—KEstimation of Cesium and Rubidium as the Acid Sulphates, and of Potassium and Sodium as the Pyrosul- phates + by PvE) DROWNING J cs oo ee 301. XXXIIL—Time Values of Provincial Cochonieran: Ter- ranésis by C.K. Wayans) ee oe eee XXXIV. EeScctrd of Hydrogenand some of its compas : by J. Trowsriper,. With Plate-VI~ 202. 2 ee 310 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Induced Radio-activity produced by Salts of Radium, P. ~ Currig and A. DEBIERNE, 319—New Method of Quantitative Analysis, R. W. THATOHER, 320.—Europium, a New Element, DemMargay: Research Papers from the Kent Chemical Laboratory of Yale University, F. A, GoocH, 321.— Magnetic Effect of Electrical Convection, H. A. WILson: Effect of Amalgamated Gases on Resistance, W. ROLLINS, 322. Miscellaneous Scientific Intelligence—American Association, 323.—British Associa- tion: Catalogue of the African Plants collected by Dr. F. Welwitsch in 1853-61: Vol. II, Part II, Cryptogamia: Leitfaden der Wetterkunde, gemeinverstandlich bearbeitet, R. BORNSTEIN, 324.—Studien tber die Narkose, HK. OVERTON: Ueber | Harmonie und Complication, V. GoLDScHMIDT: Nature’s Miracles, Familiar Talks on Science, H. GRAY, 325. Obituary—CHARLES ANTHONY ScHoTT: Baron ADOLF ERIK NORDENSKIOLD: Dr. WILHELM ScHUR: BARON DE LACAZE-DUTHIERS, 326. ive Cyrus JXQHEL, ie il ae a Librarian U. S. Nat. Muse). eee _ a VOL. XU. ae NOVEMBER, 1901. Established by BENJAMIN SILLIMAN in 1818. THE |— | AMBPRICAN JOURNAL OF SCIENCE, Eprror: EDWARD S. DANA. ASSOCIATE EDITORS PROFESSORS GEO. L.' GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anv WM. M. DAVIS, or CamsBrince, Proressors A. E. VERRILL, HENRY S. WILLIAMS anp L. V. PIRSSON, or NEw HAvEN. PROFESSOR GEORGE F. BARKER, oF PHILADELPHIA, ProFressor JOSEPH S. AMES, or BALtimorg, Mr. J. S. DILLER, oF WasHINGTON. — FOURTH SERIES. VOL. XII—[WHOLE NUMBER, CLXIL] No. 71.—NOVEMBER, 1901. WITH PLATE VII. NEW HAVEN, CONNECTICUT. BOOT THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 125 TEMPLE STREET. Published monthly. Six dollars per year. $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). ; oe Se ee Ee % ‘ Te Fe ee ine: Fi pee arte a . © ae » JUST OUT! New 64-Page “Collection Catalog Numerous full-page photo-engravings, Gives prices and descriptions of : Minerals for Study and Reference arranged in Systematic Collections. Sets of Ores for Prospectors. - Detached Crystals for Measurement. Series illustrating Hardness, Color and ether Physical Charac- ters. | Laboratory Minerals sold by Weight. Sundry Supplies, ete. =i j Mailed Free to any address. Sweeping Reductions in Prices of Specimens for Museums, Private Collectors and Students. NEW ACCESSIONS From European and American localities. Over seventy boxes of fresh stock lately received. A more detailed announcement will shortly appear. - OO IMObIn See a ee FORMERLY DR. A. E. FOOTE, The Largest and Best Equipped Mineral Supply-House in the World. Highest Awards at Nine Great Expositions. | HSTABLISHED 1876. PHILADELPHIA, | PARIS, 1317 Arch Street. 24 Rue du Champ de Mars. ow ‘Leaflet ‘Fall Announcements” Published during October, will give some idea of the many recent additions to our stock. If you have not yet received a copy, drop us a card and it will be at once mailed to you. IMPROVED MINERAL COLLECTIONS. The rapid strides in the science of mineralogy and the many and important new finds, the best of which natur- ally come to us, make frequent revisions of scientific collections of minerals imperative. It has always been our policy to keep our collections right up to date. Those which we are now putting up are materially better than those which were put up a year ago, and far superior to those put up in 1894 when the last edition of our complete catalogue was published It is safe to state that our collections are 25 per cent. better than they were in 1894. OUR COLLECTIONS OF LOOSE CRYSTALS Are not only composed of better crystals, but much more care is exercised in their selection and each collection is accompanied by a catalogue enumerating the forms of the crystals. Prices, $5.00, $10.00, $25.00 and $50.00. OUR PHYSICAL COLLECTIONS Are unapproached by any others in the market. OUR BLOWPIPE COLLECTIONS Have just been revised (October, 1901) and any student buying one of them will find it of inestimable assistance in his laboratory work. The leading works on determinative mineralogy have been studied from beginning to end in order to make it certain that everything essential for a student of any one of these books - to have is included in our collection of 200 specimens at $8.00, while the smaller collections, 100 for $3.50, 50 for $1.50, 25 for 75c., contain the most important of the species. All of these collections are put up in handsome quar- tered oak compartment boxes without extra charge. OUR SYSTEMATIC COLLECTIONS OF MINERALS Range in price from 75c. to $1,000, and each collection is so carefully selected that its scientific worth is much more than its price. Our ‘“‘ Dana” Collection, No. 1, 125 specimens averaging 2 x 2 inches, $10.00, has attained an international reputation, no ten dollar collection ever before on the market being comparable to it. Our “ Manhattan” School Collections at $1.00 and $2.00 are meeting with suecess which would be wonderful were it not so well deserved, over 500 of them having been sold to one school. Our 12-page illustrated leaflet, ‘‘ Suggestions to Teachers of Mineralogy,” quite fully describes them. a GEO. L. ENGLISH & CO., Mineralogists, Dealers in Educational and Scientific Minerals, 3 AND 5 WEST 18th STREET, NEW YORK CITY. ao 5 gs Se Te ASS ae ee ee Boas, 5 et ~ tas ei = Ee re on sana) CONTENTS. Art. XXXV.—Effect of Temperature and at Moistaeet on : the Emanation of Phosphorus, and ona Distinction 1 An the Behavior of Nuclei and of Ions; : by _C. Banusete2. 3 XXX VI.—Determination of the Heat of Dissociation ‘and of ay Combustion of Acetylene, Ethylene and Methane ; PE W .. G.MExr er 2 oe Se SO ee ee XXXVII.—Geological Relations and the Ae of abe St. =a Joseph and Potosi Limestones of St. Frangois County, Missouri; by FE. S.(NASON. 222.322 a XXX VHUI.—Cambrian Fossils of St. Frangois County, Mis- cas sourl; by C.-H.. BEECHER. —-:.~.-7.12. 3) 22) XX XIX. E Distieee of Eurypterid Remains in the Cam- brian of Missouri; by C. E. Bescuzr. With Plate ae 364 XL.—Determination of Persulphates ; by C. A; PETERS and 8. E. Moony. 22:3 @25 55 ee ee XLI.—Studies of Eocene Mammalia in the Mar sh Collection, Peabody Museum; by J. Ll. Worrman.____-.--- 2-2 as XLITI.—Carboniferous ‘and Permian Age of the Red Beds or Eastern Oklahoma from Stratigraphic Evidence ; by G. T> A@AMS oo 002 aoe Ce Gee Bases, “SABATIBR and SENDERENS: Existence of Ammonium, O. "Hea ‘381 Modified Gooch-Crucible, W. C. Herakus: Radio-active Lead, HOFMANN and STRAuSS, 388.—Salt of Quadrivalent Antimony, WELLS and Mrerzcer: New | Method for the eran Determination of eed GUTBIER: Diffusion: 0: —Photography of the Infra- ‘Red Spectra of the Alkali- metals, H. LEHMANN Y Ether and Gravitational Matter through Ha Space, Lorp KELVIN, 390. ~ KELVIN, 391. —Resistance and Blectromotive: forces of the Rectie Mae DUDDELL, 392.—Unsolved Problems in Low- = Pen StHenee Bescon Ane CLERKE, 393. J Geology and Mineralogy— Geological Sakiey of Canada, G.-M. ‘pines 394. —_— Geological Survey of Canada, R. Bru and J. F. WHITEAVES : Geological and Natural History Survey of Minnesota, 1900-1901, N. H. WincHeELt, 395. —Tlowa Geological Survey, 8. CALVIN: Dragons of the Air, an account. of Extinct Flying Reptiles, H. G. SreLey, 396.—Contributions to Mineralogy and Petr graphy from the Laboratories of the Sheffield Scientific School of Yale Uni. versity, 8S. L. Penrietp and L. V. Pirsson: Large Phlogopite Crystal, W. H. -MoNairn, 398.—Handbuch der Mineralogie, C. HINTZE, 399. Miscellaneous Scientific Intelligence—Meteorite Swarms, A. G. Héarom, 599 American Philosophical Society of Philadelphia, 400. ve hn Stee he ie et TS fee WHYS in Lariit-) Smithsonian Institution. a Yor. XI. oe eee DECEMBER, 1901. ‘Established Re BENJAMIN SILLIMAN in 1818. THE AMERICAN JOURNAL OF SCIENCE. Eprror: EDWARD S. DANA. ASSOCIATE EDITORS PROFESSORS GEO° 4. GOODALE, JOHN. TROWBRIDGE, W. G. FARLOW anv WM. M. DAVIS, or Camsrince, Proressors A. E. VERRILL, HENRY S. WILLIAMS anp L. V. PIRSSON, or New Haven. ProFEssoR GEORGE F. BARKER, oF PHILADELPHIA, ProFessor JOSEPH S. AMES, or BALTIMoRE, Me. J. S. DILLER, or WasHINGTON. FOURTH SERIES. VOL. XII—[WHOLE NUMBER, CLXIL] No. 72.—DECEMBER, 1901. oy : ma a" Padi Pr wean = OW ree oF, Pe PD ee, mS Py ‘ tao at Te 2a be : . eae, | wee aes F é jg Mh Ri 4 Bes ry PO gee RE ite aren. Ps ; ry a | vere ie»! eo WITH PLATES VIII-IX. NEW HAVEN, CONNECTICUL....>~ +1901. meme, THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 125 T LE STREET. Bublished monthly. Six dollars per year. $6.40 to countries in the Postal Union. Remittances should be made either by money orders, registered letters, or bank checks sepecrerahiy on New York banks). a ea Aer Se Ee NE r te P oe =~ as. ra <2 re 2 ahah’, ne © Son ae stock of. polished sp things. They are suitable for the face, leaving the natural fracture on bg | aos love an odd formation, combined with | e fe) the following of high decorative value. The ryst natural polish equal to the best lapidary work, ike Amazon-stone crystal. Jasperized Wood. Amber with insects. Labradorite. Amethyst crystals. Malachite. lal Aventurine (Sunstone). Meteoric tron, "5 = sulp Barite crystal. - Moss Agate. ey Fluorite crystals. Opal in Matrix. Garnet. Onyx Marble. Hematite crystals. - Opalized Wood. Iceland Spar. Pyrite erystals. Prices, 50c. to $3.00 and $4.00 for the beet owes CG E MS ae = The brief list below is but a suggestion of our eae L good stones eut for mounting. Prices average one- third le than charged by jewelers. 3 ae Almandite. _ Olivine. z Speasar ates ey Amethyst. ~ Opal. ace Topaz. : Aquamarine. : Peridot. = Tourmaline. Cairngorm. Rhodolite. — ors oe Hiddenite. — x Rubellite. . NEW CATALOG Numerous full-page photo-engravings. aoe Gives prices and descriptions of : ee Ste Minerals for Study and Reference arranged ink §) Collections. see ee Sets of Ores for Prospectors. _ Peek ee Detached Crystals for Measurement. Ee : Series illustrating Hardness, Color and other Physical ters. Laboratory Minerals sold 2 Weight. Sundry Supplies, ete. Mailed Free to any address. Sweeping Reductions in Prices of Specimens fr mM se uns, Private Collectors and Students. | NEW ACCESSIONS ; From European and American localities. Over seventy | boxes c -fresh stock lately received. A more detailed announcement shortly appear. . FORMERLY DR. A. E. FOOTE, The Largest and Best Equipped Mineral Supply-House in the Wor! Highest Awards at Nine Great Expositions. “5! ESTABLISHED 1876, ieee PHILADELPHIA, PARIS, a 1317 Arch Street. 24 Rue du Champ. Have been obtained recently as the result of the purchase of a large collection, the first portion of which (to and including the oxides), embracing some 1,500 specimens, is now on sale. Many specimens from the-Spang and Bement collections are included. Among species worthy of special mention are: Brucite. A score of superb groups from Pa. Hematite. Many magnificent Swiss groups and ‘Tron Roses.” Alexandrite. Several Jarge, well formed and gem- my crystals. ; : Amalgam. Two crystallized specimens. Acanthite. Several exceedingly good groups. Argentite. Splendid groups. Arquerite. One good specimen. Proustite. Extra choice loose crystals and groups. Pyrargyrite, Polybasite and other silver minerals in fine specimens. Corundum in brilliant, gemmy crystals from North Carolina. Cassiterite. Exceptionally large twins and most brilliant groups—a truly grand assortment. Fluorite. Many rare forms and fine specimens, including a three-inch pink octahedron from Switzerland. _ Pyrite. Rare forms and combinations including trapezohedrons and diploids. Chalcocite. Splendid old-time groups from Bristol and Cornwall. Copper, Groups of remarkably sharp and well-developed crystals. Silver. Excellent wire and crystallized specimens. Sulphur. Very perfect loose crystals and beautiful groups. Platinum. Good-sized nuggets and vials of little grains. Iridosmine from several localities, some of it in distinct though very minute erystals. Platiniridium. A very little. Palladium. One expensive specimen. Chilenite. Two specimens. Argentopyrite. One fine specimen. Manganite. Several extra fine groups. Cobaltite. Excellent crystals. Hundreds of other very desirable minerals, of some of which there are twenty or more specimens, of others but one. Among them are fine specimens of Tetra- hedrite, Nagyagite, Bournonite, Octahedrite, Aikinite, Binnite, Jordanite, Sartor- ite, Dufrenoysite, Antimony, Bismuth, Diamonds, Petzite, Hessite, Dyscrasite, Tetradymite, Freieslebenite, Matlockite, Rutile, Brookite, many varieties of Quartz and Opal, Ilmenite crystals. Franklinite in rare forms, Magnetite, etc, etc. It is, of course, impossible to do justice in these few words to so important an accession to our stock. DEKALB DIOPSIDE CRYSTALS. The season’s output includes few large crystals but an unusual number of beautifully transparent and perfect crystals, 4 to 14 inches in length. The usual unattractive character of pyroxene specimens finds no counterpart in the flashy beauty of these crystals whose popularity is so well deserved. WONDERFUL TITANITE TWINS. A new locality in the Province of Quebec has just yielded forty startlingly fine and large Titanite twins. They much resemble the Renfrew County twins, but we have never seen any so good from that old locality. 124-page ILLUSTRATED CATALOGUE, giving Dana Species number, crystal sys- tem, hardness, specific gravity, chemical composition and formula of every mineral, 25c. in paper. 44-page ILLUSTRATED PRIceE-LIsTs, also BULLETINS and CIRCULARS, FREE. GEO. L. ENGLISH & CO., Mineralogists, Dealers in Scientific Minerals, 3 AND 5 WEST 18th STREET, NEW YORK CITY. et rt Re HoLuAND: Perknite, H. W. TURNER, 468.—Outline of Elementary Lithology, — ROSENHEIM and STELLMANN: Thermochemistry of Alloys, T, J. Baker: Acid Nitrates, WELLS and Metzger, 460.—Methods of Standardizing Acid Solutions, C2Ge HOPEINS, 46].—New Element Associated with Thorium, C. BASKERVILLE: ee Laboratory Guide to the Study of Qualitative Analysis, E.H.S Barney and H. P. Capy: Elements of Qualitative Analysis, W. A. Noyus, 462.—Hetero- = gene Gleichgewichte von Standpunkte der Phasenlehre, H. W. B. RoozEBoom: Viscosity of Helium, H. ScauttzE: Air-tight glass stop-cocks, H. THIELE and 2s Violet rays, H. KREUSLER, 464 Be Geology and Mineralogy—Glaciation in Tierra del Fuego and in the Antarctic, H. .ARCTOWSKI, 464.—Glaciers of Switzerland in 1900, S. FINSTERWALDER and HE. MurRET: United States Geological Survey, C. D. Watcott, 465.—Indiana: Department of Geology and Natural Resources, W. 8. BLATCHLEY: Geological 466.—Syllabus of a Course of Lectures on Elementary and Economic Geology, ne J. C. BRANNER and J. F. Newsom: Loess of Iowa City and Vicinity, B. } Re Survey of New Jersey, Annual Report, 1900: Ice Ramparts, EH. R. BucKLEY, 4 hah : SHIMEK: River System of Connecticut, W. H. Hopes: Sivamalai Series of | Eleolite-Syemites and Corundum-Syenites in the Coimbatore Distr, Madras Presidency, T. H. HOLLAND, 467.—Peculiar Form of altered Peridotite, 1. A. is G. H. BARTON: Synchisite and Molybdophillite: Crystailization of Stannite, L. — County, Alabama, G. P. MERRILL, 469. Bed. Zoology—Nomenclature of Bermuda Birds, A. H. VERRILL, 470. —Reports on the | as Fauna of Porto Rico, 471.—Papers fromm the Alaska Harriman Expedition, nm * W. R: CoE: Zoology : An Elementary Text Book, A. EH. Saretey and EH. W. MacBripE: Elementary Course of Practical Zodlogy, T. J. PARKER and W. N. PARKER, 472. Miscellaneous Scientific Intelligence— National Academy of Sciences, 472—Annual Report of the Board of Regents of the Smithsonian Institution, for the year ~ ending June 30. 1900: Smithsonian Institution: Documents rclative to its Origin and History, 1835-1899, W. J. Ruexs, 473.—Beitrage zur chemischen Physiologie und Pathologie: Annals of Harvard College Observatory, H. C. PICKERING: Royal Society of London, Copley Medal, ATA, Obituary—Rupotpu KorexiG: Proressor A; F. W. Scurmper, 474. 4\ ep | VN} =, i‘. M. EcK4rpT: Migration of the-ions, R. Gans, 463.—Photometry of the Ulira- — eS J. SPENCER: Conchite and Ktypteite, H. VATER: Meteorite of Felis, Perry — CONTENTS. ae Page . | b Bt Art. XLITL.—Geology of the Little Colorado Valley ; 5 DY = Sean a ; li; WOWaRD ioe Se 401 |f- 4 XLIV.—Pyrite and Marcasite ; by H. N. Sroxss __-- og 414 = R = XLV.—Studies of Eocene Mammalia in the Marsh Collection, ce ee Peabody Museum; by J. L. Wortman. With Plates 3 ( i, Vill and [X*)) wt a i as XLVI.— Dielectric Constant of Paraffins ; by W.G. Horm 488 an | ‘a 3 XLVII.—New Mineral Occurrences in Canada; by G. C. ee HOFFMANN (22222 so ae ce ee ' Sas XLVIII.—Estimation of Molybdic Acid reduced by Hydriodie ae i Acid; by F. A. Goocnu and O. 8S. Putman, Jr..__-._-- 449 {ff a : XLIX.--Veramin Meteorite, by H. A. Warp __..._.___-- 453 ag | | SCIENTIFIC INTELLIGENCE ) ve Chemistry and Physices—Double Compounds of Antimony Pentachloride, ete, | *: \ ; 2 Nw, i 1 & } ial i¢ TIM