SSeS SII Wop SPO Se ree enna ee ee ee eee ee ee eee en a te ng at Ma le gtk Mal tim ae et td a Ng thea le Ml A a Ft AR mt EU TE Ll De te RAT om ye Hit, 6 LNA A es nm Sind nah rem rhe ad Ah ee ele Ne he eee ee Ret Mgt eat ate ig etl anh ak Yn the gi Rit Ke at te Of W tana Sieh Me RS ae. te Tl A Ee elt BN elit Et tH tte hl th PBR ALA IVA he Ref SN ett aS Fe eth oe ‘ ee a a ie A te te pee oath Sit att NI Pl ak he te Atle Ms a lt ac ee ee el 7 en an he ON I Hn ate Ingen, | nth ig Reg ert te et LL tn BL Rade te OR Mic se ee Pete hin ete lage Mat ee es ee Ce eer ee oe Se ee ee Wee eer or ee ee eee oe ee or ere Net thee eh oH ot Bath Hegde Maa te ll Hl lg ee Rtn ES Rt NEI i it AR AL ek nal "al at i et atin tg teehee ante Ne ete the aed TN te RL Lt A AE Hin ec ee Al teeth Sener dietite., thay yh. Win Ne Byte ee re ee ae ew in nian em ee a Nie Atl AP adie Mba A tee Phe the fat Nh i i il ah a Me esi” Rant nndbnd’, Ji ee ee ee MR ee RF RA te Aa, ft he RR et Sete RL kt ce eet Si el cdo tne Raaiieriae Te Vint Rye a Rt tote Qh iu Se Oe Na Nd A th Rat Ange RN le eMart Sint et the Din eel Ste Re eT he Lee beta tate be Meter mie ew eee ee es We ee Ca MR PO eee eee ee eo Se ee en eee ee Oe ee ee a ae ce RA BR Beh Sets Pe Tn he MaMa ReaD ERs uni Oe” wad Tete tad atenet a Me ek at ee ee oe eee (ee eT eae eee he eA, Net etn Seether Ria m iltet le Ste ba Meh ail oan iy 5 oe heel Mein Tee Merete vain Ss eer et en Wee ee We Me a ee ee, ee ee ee pth atte aie a batine™ Ye Uw biol ate te oe Ta Ra Ee ie ale tte a fs Neg ae ew he he a he OE Mea Degas (tat ng linha ste he Megdeine O ee ee eee ee I eet, i ee ete teh Aes. L a etd Phe ele Mg MO An Re tet IR eRe ttt I ttl A tie ete Qi ytd PM tet SATA, Aner t/a Me ty be a meg ttl Sethe Flot atte aU its ona, tenth ole ll omelet Rare ae thm Mg tg Pah ee me ae A Mtl ae ta A OR ee ne i AA ee Rt Noh aN Mt Aa NTI Kc Mee tien tet STN hte Hin A Lat Ry Bethe ot Raith ees ty nla cle tn Rae UA ate tn ede emt the Ne tnt Te ate De te Ra Teck ne ap te ee Pate Rabin teil LF a Fade te Tetaite Tetfetigwae 1" Din Paar © ah) All Palate ig emit A tht, Ci Bie ne erty male nie Atle hehe Ha tig eR ant Me wth ee wn on Le I Ri ER re eet rend MH tad gh Meath oe este le Mah Bho e tan taed MB TUN TO a Le Mette Tag Me tho bee 5 Nagle Me Le teil a a hee : het AT hi aN a ak i ORT tl th a tt th TN it hie I le TN GM eR a Se ml a A hl Ee dt nih ait Toth Pig Nay either teats Nore tgs ie ake ea aS Bee ER ie oe ae tM RL ke dt he ee Me ite Le aed OL Maye Rete Mel ot Paha ° a eT |e eee an re Te a ee ee, ee ee eee eee ee Lee Dee (athens en ee te tei Na hee tt ak Bm ee elke Ma Be te net, ally Te i Rhea Ea MM Ma A TRIE Ce Sa ea ll ST AN i Ft mat AU eR a me HRM at i ee hl Re A LT EE eae MT ee Fay Rann Ree = HF nde Dasa Hi in Dagtt~ itee mn tt Rint 4 Pen Rosin Hearte te hehe == nwt ten Meee nite bee teats ea mete ol Mata RP Fett ete St Fee ee ee Fe ee ee ee ee ee ee ee ee ee Be ee Oe Peete ee eee ee Se ee Oe ee cc ee OSS gh Se a ‘ eV a athe tting Hg ete ne la ba Hee Carta Reta re ae te Oe te tad Petal om ers | EL He Ran a Me Ne Ln te Riley hee oars Meine Ho Men nS a tn Nethm min he the the nett? mgt ht ee eM OM ee Sig oti & ol ke ee BO ttn ee ee kt kt eis el At et ne + —* CMP nate ee A le tee tha I a EL IIT A Meese MN eh TM ae te ig RR Ath i, Beilin Me, Me Mtn Me He teeth etter Ma Gi he Tate Barts Rn Pie ee omit FP = meets Mae a Bne Ce ee ee ~—A* DB all Arce Ae he ee Se Te LR Te it met ne te Mt RS eRe LT be Te ee MaMa Beat aha Me ery Made. eV oP ete Gaere Le vote. ee OR A Oe te By oy. - NR Ae the 0 te aya eh am 2 are, See ore Be peta ah lw Bay fe ORLA IR Ne el thet hat aR, oe Me tt Met ee | Mati Btg te el we Fe Me POE ey ee ree eee ener eine ae ear = ~ tah ee No pe gM et Pe et Le Bi a, PI oe OR ncrO t,he Rel dh hcl re th AW, te ee ta tle hn Shei eet etter Me teeth a Ne te athe a han A tte gh a i 7 a re een ee ee eee ee Hag RS Serine re Th CRA pa ATR ae ee TL Le i LL Stains Ape eet RS tm SIS tree ate Au lFate Rath reer bad Eytan Pe pr Rg ete Si Re ety mt ene Oe ee ee ee on, ee Par A eK met A Re LN ERE A Bil eh i ee Ae at MI he a EDM, Mette Te Pet IL Re hg fede 7 a Cathe Me Mra eg hah heresy Onan 4 ee een array ew eee © ee ere mre PIN I et WN See te Tg tee AL RL Ue sree hae EME et MaMa Bantctetn ene oa tam sbeinemtette ty Ne telhe ih \ a — A eee er Ake Pe oe a Pn Nee ers AAA LN eot, ee a ek P ° Le Se RRR OE Le Re i et IL Aa te ee ee ee ee ee oe oe ft a ee Tears ee eee ere te. a othe MR Neg A = 7 _*> ee See ee es nn LL ee tN et et ee le HMO He gta te Me ate a + ee Ree sede te te Diy te ee, RE ee MO AE RL te ee ee ee en ee ere me ar eee ee ee ee ee Sr ee ee ee eee eee ee ee ee ee ee A meA = Aas —% te . ae wee -_* sen ‘ ~ ~% edt Deter RR DL Te Re, bee tnt eg Ue ee i te ett a tette De Rete Me Arts oh ER eh Am ee ee es rt i ei eee OL iether ay Tarte. ee a LL ML Qe ethene, toe 8 te ee om oo % Kn wah ee . “ e ta ds ke TM 8 Ly = 7 i ele Sete te Reg tee Meee oe te tee alten eg cate tntetetor Re I ee Ne eel i A Re le RR A Re SN me Rd I eR HOR Crs I tn ite oe Di Ag Et es Be rh ee he et Dee te tel Wet wen ate Vie, he as thy ates oe See eee eee ke Oe ee A et ee ete 2. aS ne eh, a ee Lee Pe See te TK meth Rit, te Ne teigt Ror tPF igP a ta th Ne, dace *s,* c De A ee ete ~*~ Fo a RARE AIR ter, a ae a ee ee en oe ee eee ee, eee eT Bee Oe es a ' a : . 7 . > = nr At oe shal gts - --s thn SoA ree ee * ne ee ee i a er a pores ~~ e Me ky ann * 5. et -~* -tow& we RN te et Re Rea Cee a tear -* oe ae — lee ~ ~~ « — ~ ‘ Ah ~~ re re " ou —— ee og ee a See etn ie eg te aN ene i Mgnt He Re detete -sehette lets ihe de a tm ee ~_* “e Var “ee ~ = - ee aie EIN ok te FAIS he OR > gt A i 8 ee tng hah MMs Aa ante Cath ity, ten - tied 7 ~ ~ Rote a « ~~ ~ ” = a A . - Pee Oe a oe eH ti he te te + > —~ ae Teck A Pe te m “uta ee, ae ae ee ae Md tn, ee eee, ae Baty A} eu tm A he tte a Le te! ei Ne “ LALA be we - a a a “ . et es A a ee ee er aT an haivtnntin ta? uae be = ee ee he att 4 mete ee ee ‘ a en ve « ee, . + Ree th ee ee ee ee ee “ » = - as tet ee a ae a a ee a we et, 7 n « a ~ - li ~~ eo a er a ey a ~% Pe a A et Se Pod we a= ote i - hm _ oe, -* Oe ee ey ™ .6. ‘ ~ « 7 9 fe Phas Od ME g's eames ee, ee ate mete = .« -< Po res . ee os, x e's —*& u* it Ped Rede ge Mette? — tame te ag Rare, hele “ ~ “ ~.* - . I Pe rh os ee Le o.t - ‘ “ thet > wah “A ath art Mee Pew elt) OM, o tate & Sot > re ™ _ = “ os = = 4 Vette - - ~ ~ . - Ay AN & 2 tw ae eR Om e Ota Trt at > = = a « ~ - - “ ~ Nei we ale ae as ere ena ~ em eR Se 7 ot ‘ Ae eat Me coaiied ~ “ ~ ae J _— * ea Fa _* » ~~ 48 en +. 7 - “ ~ oe i « ~ - oa a te ~~ oo aa Ae ~ me “ * ~~ —— an 8. Cette eRe D. Ln we wits ae ee eee Se re .. aos > a -* tn a - ' ~ % ~~ sattgihs inn -* ¢ “grata Ss - - < na tg Bax hw ate =o. ~ - ~ 4 Va eee cy . a aN Mg! Rae, eM ee ewe He 8 ee ae oy rears 7 - te a m ~* ie ete : ; A a > . + ss Pp o a 7 ~ ee es ee weed | ry ae St ee eva, 4 or » - os a ~ $ r tte rs - —* : s - — - . at, = *ate = - 7 a = ee ee ae mA ~. - a = s ~ at a ea PS, atk 2s ma ote ee te S ~ “ re a oo oe és = 2 x n- Ls wos “ - =" - ” “ - i . Sn ~ .- = ~ ro - * Nae 7 ~o s ~ a a . a 1 a = sed x ” - ” ae he 4 a ~ bs ~ ? = ay 4 ~ “9 ™ . > *. ~~ s - - a 3 THE AMERICAN JOURNAL OF SCIENCE. Epirorn: EDWARD S. DANA. ASSOCIATE EDITORS Proressors GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or Camprince, Beneecsons ADDISON E. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT E. GREGORY AnD HORACE S. UHLER, or New Haven, Proressorn HENRY S. WILLIAMS, or Irwaca, -Proressor JOSEPH S. AMES, or Battimore, Mr. J. S. DILLER, or Wasurineton. FOURTH SERIES VOL. XLIV—[WHOLE NUMBER, CXCIYV]. WITH PLATE If. NEW HAVEN, CONNECTICUT. boOrhte. HOUSE CONTENTS TO VOLUME XLIV. Number 259. Page Art. I.—The Motion of Ions and_ Electrons through Gases; rae) aN DHIISOPy io oe ee ech et lp <1 JI.—Correlation of the Devonian Shales of Ohio and Penn- Seivatige OCW cA. NRW ERE Ge oo ee 33 IJJ.— Evidence of Uplift on the Coast of New South Wales, BeUSianas bye te Be LLARP ER SoS Salo Po oo el 48 I1V.—The Use of the Platinized Anode of Glass in the Elec- trolytic Determination of Manganese; by F. A. Goocu INO ONL IMOMANCA STR EMU UAT ek D/O Le et ho BD V.—Preliminary Note on the Occurrence of Vertebrate Foot- prints in the Pennsylvanian of Oklahoma; by W. R. PC arene nie a es al as Se ei okay 86 VI.—New Evidence of a.Recent Volcanic Eruption on Mt. St. Helens, Washington; by W. R. Jrinison .---.----- 59 VII—Some Notes on Japanese Minerals; by 8. Icnikawa 63 VIII.—The Retardation of Alpha Particles by Metals; by ire eI Se ee os O89 PMG NER MW NGG Bic mane Sty AN a ae eed oe AR SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Analysis of Pyrolusite and other Oxidized Manga- nese Ores, O. L. BARNEBY and G. M. BisHop: The Life of Robert Hare, an American Chemist, E. fF. Surry, 76.—A Course in Food Analysis, A. I. Winton, 77.—A Text-Book of Sanitary and Applied Chemistry, KE. H. S. Baitey: Nature of Solution, H. C. Jonzs, 78.—Theory of Measurements, L. Tuttte: Laws of Physical Science, E. F. Norturup, 79. Geology and Natural History—United States Bureau of Mines, V. H. Man- NING, 80.—Canada, Department of Mines, R. W. Brock and &. HAANEL, 81.—Pennsylvania Glaciation, First Phase, E. H. Win.iams, Jr.: Nebraska Pumicite, HX. H. Barsour: Guide to the Insects of Connecticut, Part ILI, H. L. Visreck, etc., 83.—The Biology of Twins (Mammals), H. H. New- MAN: The Theory of Evolution, W. B. Scotr: A Chemical Sign of Life, S. Tasuiro, 84.—Fundamentals of Botany, C. S. GacmrR: A Laboratory Guide for General Botany, C. S. Gacer: Laboratory Manual of Agricul- tural Chemistry, C. C. HupGres and W. T. Bryant, 85.—Manuring for Higher Crop Production. E. J. Russetu: A Manual of Organic Materia Medica and Pharmacognosy, L. E. Sayre, 86. Obituary—G. H. Srone: H. F. E. JuncEersen, 86. 1V : CONTENTS. Number: 260. Page Art. [X.—Physiographic Development of the Tarumai Dome in Japan; by HipEz6 Simoromatl (TANAKADATE)_ 87 X.—Lavas of Morro Hill and Vicinity, Southern California; by G. A. Warine and C. A. Warine_-_--- ee 98 XJ.—On Tri-Jodide and Tri-Bromide Equilibria, especially in Cadmium Solutions; by R. G. Van Name and W. G. BROWN 225. S223 22 ee re’ XII.—The Environment of the Amphibian Fauna at Linton, Ohio; -by : He G@. Cus oie ee ee 124 XIIi.—Some Fossil Beetles from the Sangamon Peat ; by Hi. FL Wick nam. 00 2 Se ee 137 XIV .— Granite in Kansas; by S. Powmns-=_"" 2 = ae 146 XV.—A New Method for the Determination of Hydrogen Peroxide; “by G. S. JAMIESON: 225-2.) 210-9) 5 a 150 SCIENTIFIC INTELLIGENCE. Geology—The Coral Reef Problem and Isostasy, G. A. F. MOLENGRAAF, 153.—A Study of the Magmatic Sulphide Ores, C. F. Totman, Jr., and A. F. Rogers: Origin of Massive Serpentine and Chrysotile-Asbestos, Black Lake-Thetford Area, Quebec, R. P. D. Granam, 156.—Contribu- tions to the Knowledge of Richthofenia in the Permian of West Texas, E. B6sE: Contributions to Geology, 157.—Geological Survey of Alabama, EK. A. Smita: Bibliography of the Geology and Mining Interests of the Black Hills Region, C. C. O’Harra: Story of the Grand Canyon of Ari- zona, N. H. Darton: Bulletin of the University of Texas, 1916, No. 66, ' J. A. UDDEN, 158. Miscellaneous Scientific Intelligence—Food Poisoning, E. O. Jorpan, 108.— Principles of Agricultural Chemistry, G. S. Fraps: The Secretion oi the Urine, A. R. CusnHny, 159—Field Museum of Natural History, Annual Report of the Director, F. J. V. Skirr: Chemical and Biological Survey of the Waters of Illinois, E. Barrow: British Museum Publications, 160. Obituary—T. McK. Hucues: H. T. KEenneDy, 160. CONTENTS. V Numiber? 261. Page Art. XVI.—Voleanologic Investigations at Kilauea, with Platenl(frontispiece)s) by T..A. Jacear, Jrc fs. ---. 161 XViI.—On the Qualitative Separation and Detection of Gallium ; by P. E. Browntine and L. E. Porter.--- .- 221 XVIII.—On the Calibration and the Constants of Emanation pelecmoncopes. by OC) Lusrun (222 2-2 22 22.5 XIX.—Measurements of the Radioactivity of Meteorites ; io he Ounce and Lro VINKELSTEIN .2...03...0..4 237 XX.—Occurrence of Euxenite in South Sherbrooke Town- ship, Ontario ; by W. G. Mizrer and C. W. Knieur_.. 243 XXI.—A Remarkable Crystal of Apatite from Mt. Apatite, Emepienn Miainie: a by, Won Korps ie 2.2L. ee 2 Pine EnOCK OLARK 22.08 200 eee a ogy vi CONTENTS. Number 262. Page Arr. XXII.—Block Mountains in New Zealand; by C. A. COTTON | 22 Sse ce i ee X XIII.—Dinosaur Tracks in the Glen Rose Limestone near Glen Rose, Texas; by E. W. Sauer .._.. 22 52 3eeee 294 XXIV.—Outline of the Geological History of Venetia dur- ing the Neogene; by G.STEFANINI __)22___ 2 2 eee 299 XX V.—On the Qualitative Detection of Germanium and its Separation from Arsenic; by P. E. Brownine and S. CH Scorn fe 2 les a eae Es a ed XXVL—A Peculiar Type of Clay; by Fi. Rrus—-_- 322 316° XX VII.—Marine Terraces in Southeastern Connecticut ; by Lavma: Havrore: 225 oS Oe See ee 319 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Recovery of Sulphur from the Sulphur Dioxide of Smelter Gases, A. EK. WELLS, 330.—Use of Large Glass-Stoppered Con- tainers in Autoclaving, R. B. Krauss: A Cryoscopic Method for the Determination of Added Water in Milk, J. T. Kester, 351.—Inadequacy of the Ferric Basic Acetate Tests for Acetates, CURTMAN and HARRIS: The Priestley Memorial Committee of the American Chemical Society, 382.—The Electron, R. A. MCE: Failure of Poisson’s Equation, G. PRASAD, 333. —Composition of X-Rays from Certain Metals, G. W. Kaye, 334. Relations between the Spectra of X-Rays, J. IsHIWARA, 390. Geology—Grundziige der Palaontologie (Palaozoologie, F. BRorL1: Palaeozoic Crustacea, the publications and notes on the genera and species during the past twenty years, 1895-1917, A. W. Vogpres: The Lower Cambrian Hol- mia fauna at Tgmten in Norway, J. Kime, 336.—Recurrent tetrahedral deformations and intercontinental torsions, B. K. Emerson : On the crinoid genus Scyphocrinus and its bulbous root Camarocrinus, F.. SPRINGER, 3987.— On a new hydrozoan fossil from the Torinosu-limestone of Japan, I. HAYASAKA, 308. Obituary—A. von BAnyER: E. BucHner: M. E. Sarasin: R. Bzti: D. D CAIRNES: C. W. DRYSDALE. CONTENTS. Vil Nua ber 263: ; Page Arr. XXVIII.—The Great Barrier Reef of Australia; by Ree iD eva tr ha ee eA ee ie Sl 2 BOO XXIX.—Wave Work as a Measure of Time: A Study of time Ontario basin; by A. P.-CoLEMAN 220202 22.22442- 351 5X Arthropods in Burmese Amber; by T. D. A. CoCr RIE Eo Sat ae ta ee ela es od 360 XX XI.—A Calcium Carbonate Concretionary Growth in Cape Ponce ss CV. MEA UY sce tm Sle S60 fox X11.—On the Preparation and Hydrolysis of Esters Derived from tbe Substituted Aliphatic Alcohols; by Wee Drousnnn and GIR. BANCROPT.22 220020 22.2}. a 7al XXXUI.—The Perchlorate Method for the Determination of the Alkali Metals; by F. A. Goocu and G. R. BiaxeE 381 XX XIV.—Protichnites and Climactichnites; A Critical Study of Some Cambrian Trails; by L. D. Buruine __------- 387 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—A New Method for the Recovery of Salts of Potas- sium and Aluminium from Mineral Silicates, J. C. W. Frazer, W. W. HoLuanpd and EK. MinuErR, 398.—Electrochemical Equivalents, C. Hprine and F. H. German: A Laboratory Manual of General Chemistry, W. J. Hare, 899.—A Short Manual of Analytical Chemistry, J. Muter: Allen’s Commercial Organic Analysis, W. A. Davis, 400.—The Ionizing Potential of Sodium Vapor, R. W. Woop and S. Oxano: Penetrating Power of X-Rays from a Coolidge Tube, RutHsRrorp, 401.—Problems in General Physics, M. Mastus, 404. Geology—A monograph of Japanese Ophiuroidea, arranged according to a new Classification, H. Matsumoto, 404.— Publications of the United States Geological Survey, G. O. Smrtx, 405. Miscellaneous Scientifie Intelligence—Eleventh Annual Report of the Presi- dent, H. S. Prircnett, and Treasurer, R. A. FRANKS, of the Carnegie Foundation for the Advancement of Teaching, 407.—Publications of the Carnegie Institution of Washington: Publications of the British Museum of Natural History, 408. Vill CONTENTS. Number 264. Page Art. XXX V.—Origin of the Chert in the Burlington Lime- stone; by, W. A. Tame 2. sie. 2 409 XXX VI.—Ionization and Polymerization in Cadmium Iodide Solutions; by R. G. Van Name and W. G. Brown___- 453 XXX VII.—Famatinite from Goldfield, Nevada; by E. V. DHANNON esciesties! Et qe een Ee oe 469 XXXVITI.—On the Functions of the “Sacral Brain” in Dinosatrs; by Re S8.. LULL: 23.2 2-2 47] SCIENTIFIC INTELLIGENCE. Chemistry and Physics— Colorimetric Determination of Manganese by Oxidation with Pericdate, H. H. WinLtarp and L. H. GREATHOUSE: Preparation of Cyanamide from Calcium Cyanamide, E. A. WERNER, 478.—New Method of Separating Tin and Tungsten, M. TRavers: Yellow Mercurie Oxide as a Standard in Alkalimetry, G. Inczr, 479.—New Oxy- chloride of Tin, H.. F. KeLuer: Equilibrium Temperature of a Body Exposed to Radiation, C. Fasry, 480.—Numerical Application; Solar Radiation: Colored Flames of High Luminesity, Ged: HeMs ae 482.— X-Ray Band Spectra, DE BROGLIE, 484. Mineralogy and Geology—New Mineral Names, W. E. Forp, 484.—Deserip- tive Mineralogy, W. 8. Bartey, 486.— Wave Work as a Measure of Time: A Study of the Ontario Basin, A. P. CoLEMAN, 487. Miscellaneous Scientific Intelligence—National Academy of Sciences: Ameri- can Association for the Advancement of Science: Negro Education, T. J. Jones, 487.—Science and Learning in France, J. H. WicmMorg, 488. VOL. XLIV. SULLY, 1017, Established by BENJAMIN SILLIMAN in 1818. ANSOHIAL INSEE, © a c JOURNAL OF SCIENCE. | Epiror: EDWARD S. DANA. | ASSOCIATE EDITORS Proressorss GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW ann WM. M. DAVIS, or Camsrince, PROFESSORS ADDISON E. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT E. GREGORY anD HORACE S. UHLER, or New Haven, Proressor HENRY S. WILLIAMS, or Irnaca, Proressor JOSEPH S. AMES, or Bartmore, Mr. J. S. DILLER, or Wasuinerton. « FOURTH SERIES No, 259--JULY, 1917. NEW HAVEN, CONNECTICUT. VOL. XLIV—[ WHOLE NUMBER, CXCIV ]}. 2 19 f:6- a TE tA as _ THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. = Published monthly. Six dollars per year, in advance. $6.40 to countries in the ‘Postal Union ; $6.25 to Canada. Single numbers 50 cents. Entered as nen class matter at the Post Office at New Haven, Conn., under the Act ‘. of March 3, 1879. LIST OF CHOICE SPECIMENS. Calcite, Joplin, Mo.; violet twin, beautiful color. 5” long, 5” wide, 2%" thick. $10. Aragonite, sulphur, Girgenti, Sicily, twinned crystals. 34x 34". $2.50. Calcite, Rossie, New York. 3x38". $1. Apatite crystal. Rossie, New York, blue, fine termination, 24” long, ¢" in diameter. $2. Blende on calcite, Rossie, New York. 3x 22". $2.50. Zircon in matrix, Rossie, New York. Specimen 22” x 14" ; erystal 7" long, +" diameter. $2. Spodumene crystal, Norwich, Mass. 34" long, 2” wide, 1" thick, nicely terminated. $8. Spodumene, Chesterfield, Mass. Specimen 8 x 24": crystal 23" long, 8" diameter. $2. Iolite, Guilford, Conn.; very gemmy, beautiful color. $1.50 to $8. Malachite on hone! Cheshire, Conn. 24x 24", 14" high; partly covered with malachite. $2. Selenite, Cayuga Lake and Lockport, Nev 0c. to $2. Sulphur crystals on gypsum, Cayuga Lake, Ny Yo 432: Azurite, Mineral Point, Wisconsin. $2.50. Azurite, malachite, Chessy, near Lyons, France. 24x 1%". $2. Uranium mica, Schneeberg, Saxony. 24x14". 5c. ; Chrysoberyl, Greenfield, near Saratoga, New York. $1 to $4. BerylLin matrix, Haddam Neck, Conn. ‘dc., $2.50, $4. Calamine, Sterling Hill, N. J. Blue. 50c.-75c. Red wulfenite on pyromorphite, a Mine, Phoenixville, Pa. 24x14". $3. Atacamite, Atacamas, Chili. 2x13". dc. Millerite, Antwerp, New York. 3x2". $2. Cobaltite, xlized, Schneeberg, Saxony. 2x1}. $2. 30 Quartz gem Crystals, Middleville, New York. Atelier clear crystals from 4" to 24"; rare forms and ‘planes ; some with enclosures and some with bubbles : a few groups 7odc. to $3. Celestite, Strontian Island. ‘5c. Adularia, Haddam Neck, Conn. 50c. Tremolite, Gouverneur, St. Lawrence Co., N. Y. ‘dc. Fluorite, White, Muscalonge Lake, St. Lawrence Co., N. Y. 50c. Gypsum, Salt Vats, Syracuse, N. Y. 5c. Topaz in matrix, Adun Tchelon, Siberia. 2 x 14", good many crystals embedded. $3. Pitchblende, Johanngeorgenstadt, Saxony. $1. Tremolite, Lake Superior, Michigan. 50c. Sphene, Orange Co., N. Y. 5c. Galena, group of crystals, Rossie, New York. $2. ALBERT H. PETEREIT 81-83 Fulton St, New York City \ tet ee og Or $ : THE lal Mi Us il AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES.] a es Arr. I1.—The Motion of Ions and Electrons through Gases ; by E. M. Waututscu, Lecturer in Applied Mathematics at the University of Sydney. 1. INTRODUCTION. THE experiments described in the present paper were carried out in the Sloane Laboratory of Yale University and are a con- tinuation of those which have already been described in this Journal (May, 1915). In determining the mobility (%) of the ion as a function of the pressure (p) of the gas, previous investigators had found that the product pk showed an abnormal increase as the pressure of the gas was reduced. ‘This result had been interpreted as indicating a diminution in the size and mass of the ion at relatively low pressures; for the nega- tive ion in air this diminution appeared to set in at pressures below 10° while for the positive ion it did not occur till the pressure was reduced below 1™”. The investigation to which reference has already been made provided experimental and theoretical indications which were entirely different from the foregoing. For the positive ion in air no anomalous results were found ; the law pk = const. held good to the lowest pressure employed (-05™"). The negative carriers were found to consist of two distinct kinds, electrons and ions, the former coming more and more into evidence as the pressure of the gas was reduced. When once this separa- tion had been effected all the preceding anomalies disappeared ; the law pk = const. was verified for the negative ion in air from 1 atmosphere down to *15™", indicating that the ion remains unaltered in character over this range of pressures. The electrons appeared to travel freely through the gas with- out attaching themselves to molecules. No indication was found of any intermediate stage in the nature of the negative Am. Jour. Sci.—FourtH SErizs, Vou. XLIV, No. 259.—Juty, 1917. 1 2 Wellisch—Motion of [ons and Electrons through Gases. carrier, the separation between the ions and the electrons remaining throughout clearly marked. In the present experiments these results have been extended to other gases; in accordance with expectation the abnormal mobility values found by previous investigators for the nega- tive ions in hydrogen and carbon dioxide were shown to be capable of a similar explanation, all anomalies disappearing as soon as the resolution of the carriers into ions and electrons was effected. A brief study has been made of the motion of free electrons through carbon dioxide at relatively high pressures; in addi- rg ol. -\V, Kil | awe ee He ee eS eS tion, the motion of ions through a number of vapors has been investigated. A few discussions bearing upon the physical interpretation of the results have been included; in particular, certain out- standing problems of ionic theory have been specially considered. 2. EXPERIMENTAL METHOD AND ARRANGEMENT. A description of the experimental method and apparatus has already been published; on this account it seems advisable to repeat here only the essential features, reference being made to the previous paper for further details. Moreover advantage will be taken here to enter into greater detail in connection with certain features of the method to which only a brief allu- sion was previously made. The method employed in the determination of the mobilities was that devised by Franck and Pohl.* The ionization vessel * Franck and Pohl, Verh. Deutsch. Phys. Ges., ix, p. 69, 1907. Wellisch— Motion of Ions and Electrons through Gases. 3 (v. fig. 2) consisted of a brass cylinder divided into two com- partments by a brass partition containing a circular aperture. In the upper compartment was a copper plug on which a layer of polonium had been deposited ; great care was taken that the radiation from the polonium was confined to the upper com- partment. A circular electrode A was situated about 3°™ above the aperture and was in metallic communication with the case of the vessel. The lower compartment contained a gauze electrode insulated by a thin ebonite ring from the par- ~ tition. Two cm. below the gauze was the electrode e connected to the electrometer: this electrode was surrounded by a guard screen (W) connected to earth by means of a guard tube. Fig. 1 illustrates the method employed to effect the commu- tation of potential.. The commutating discs were of brass with a number of fiber segments of equal width placed at regular intervals along the periphery. The two potentials V, and —V, were connected across the terminals of a large metal resistance R in series with the commutator ; it was not in general con- venient to alter the potential V, except in steps of 40 volts each, and on this account the potentiometer device (v, 7, p) was employed to effect finer gradations of potential. When the commutator is in action the potential of K (fig. 1) should alternate between « and —V, where pen Rveilen peta t NONE V3) (1) ‘ pr + Rr — p’* Owing, however, to the time involved in the establishment of potential this formula will be sufficiently valid only if care be taken to maintain a satisfactory relation between the frequency of commutation and the resistance R. This was effected by an experimental method described later. We shall assume here that the potential of the ganze is given by # and —V, alter- nately, the former potential lasting for a fraction 7 of the total time; this fraction can be determined experimentally. Under these conditions the mobility & of the ion under consideration dn on nations per second, d is the distance between the gauze and the electrode ¢, and V, is the critical potential, 2. ¢. the value of « which is just sufficient to enable the ions to reach the electrode e before the field is reversed. | The diagram of connections is exhibited in fig. 2. Asin the previous experiments two commutating discs were employed ; one of these had 20 fiber segments while the circumference of the other was half fiber and half metal. The motor was is given by k= where » is the number of complete alter- 4. Welliseh—Motion of Ions and Electrons through Gases. worked generally on 110 volts which afforded approximately 42 revolutions per second. The double-pole, double-throw switch S, when thrown to the right, completed the connections as exhibited graphically in fig. 1. When thrown to the left, connection was made with a subsidiary potentiometer system (b); ; In this position the quad- rants of the electrometer could be commutated in potential between zero and any convenient potential read off on the poten- TGs: | ea Tian Bh i or ; 3 /TX tiometer. The use of this device in testing the contact at the brushes, in estimating the value of f, the fractional duration of contact, and in adjusting the position of the electrometer needle for observations, has been described in the previous paper. For large current values readin gs were taken with the capaci- ties B and C added to the electrometer system ; the capacity of the system was then increased 174% times. 3. EXPERIMENTAL PROCEDURE. For convenience in manipulation a table was prepared of the potentials assumed by the gauze for different values of p, Wellisch—Motion of Ions and Electrons through Gases. 5 V,, V, and R. This was effected by means of formula (1) which for the purpose was put in the following form :— ( py (vy 4v.4 0 p ("~p) : a v,+{& ( ev ge =) a ae er = V,+e (say). (2) v was always chosen equal to 40 volts and 7 was always 15,000 ohms. The calculated values of c for various values of p and V,+ V, were then tabulated and the value of « under any desired conditions could be guickly obtained. Establishment of Potential. It was important to ascertain that the experimental con- ditions admitted of an effectively instantaneous establishment Hie. 3. of the withdrawing potential —V, through the resistance R; in other words, the ions must eommence to retire as soon as the commutator brushes make contact with the fiber segments. This point was tested experimentally in the following manner: the commutator and resistance R were put in series with a battery V (fig. 3) of which one terminal was earthed ; K represents a Kelvin multicellular electrostatic voltmeter which was meluded in the manner shown in the dia- eram. The commutator was set in motion at its highest speed and readings were taken on the voltmeter corresponding to different values of R. If the values of R were excessively large there would not be sufficient time during an alternation to admit of the earth connection with the gauze being fully established and in consequence the steady reading of the volt- meter would be too large. It was found that when the small frequency comutator was employed this steady reading remained constant for values of R up to 1,000,000 ohms; for the high 6 Wellisch—Motion of Ions and Electrons through Gases. frequency commutator the value R = 200,000 afforded a read- ing greater than the normal by less than 2 per cent. Inasmuch as the potential (V,) was always chosen considerably greater numerically than the advancing potential (a) the value R= 200,000 was sufficient to ensure the realization of the desired conditions. In the present series of experiments this value of R was chosen in preference to a smaller value because in the determination of electron velocities V, is often small and it is advisable to have ¢ in formula (2) small compared with V.. Manipulation of Switches. In general, when the gauze is raised to any potential, the electrode é is raised by induction to a potential which has to be taken into consideration when the electric field is estimated. It was found possible, however, by a suitable manipulation of the switches 8S, f and g to arrange that the electrode e was practically at zero potential when the potential (#) had been established on the gauze, so that no correction for induction was necessary. The series of operations involved in taking a single reading was as follows :— (i) potentiometer (6) fixed at a convenient value so that the elec- trometer needle should have a suitable range of deflection : & closed : S closed on 6 side: f and g both closed: earth key K open: motor and commutator running but not operating on account of the short-circuit at g : capacities B and C included in the system. (ii) & opened : S switched to the right. (111) g opened, if it is desired to work with added capacity. or (ili) capacities B and C cut out and g then opened, if it is desired to work without added capacity. It will be seen from the foregoing that the effect of induction was to alter only the reversed field whose value had not to be known at all accurately. The electrometer needle experienced always a small kick when the switch g was opened but this quickly subsided and the current was measured with the needle jn steady motion, the midpoint of the range of deflections being so chosen as to coincide with the zero of the instrument. 4, ExprrimentraL Resvtts. (A). lectrons in Gases. In fig. 4 of the previous paper typical curves were given showing the relation between the current due either to positive or negative ions and the potential (v) for various pressures ; from such curves the critical potential V, could be deduced and the ionic mobility determined. In figs. 5 and 7 of the Wellisch— Motion of fons and Electrons through Gases. 7 same paper there were given the curves s corresponding to the negative carriers in air at relatively low pressures ; the charac- teristic feature of these curves is their compound nature result- ing from the independent passage through the gas of electrons and ions. It is convenient to designate as I curves the former type which is due solely to ions, while the latter type may be referred to as EI curves ; moreover, those curves or parts of eurves which arise solely from the motion of electrons will be called E curves. On resuming the experiment an investigation was made of the gases CO, and H,. The CO, was prepared in a Kipp’s apparatus by means of the action of dilute HCl on marble and was passed through NaHCO, Aq. in order to remove acid fumes; the H, was obtained by the action of dilute HCl on zine and was passed through KOH Aq. In each case the gas was passed through a series of tubes of CaCl], and P,O, in order to remove traces of moisture. A series of I and EI curves was obtained for these gases under various conditions, a few examples of the latter type being given in fig. 4. The free electrons were more numerous in each of these gases than in air at the corresponding pressure; this point is brought out by the fact that with the same frequency of commutation the electrons appeared at much higher pressures than in air, e. g. it was just possible to detect electrons in air at 8°™ pressure whereas in OO, they appeared in large numbers at a pressure of 14°™ and in H, they were readily observable at atmospheric pressure (v. Curve A fig. 4 which was obtamed with a fre- quency of only 42°6; also curve in fig. 5). _ This result was to be expected from the conclusions of previous experimenters who had found that the abnormal increase in the ionic mobility set in for these gases at higher pressures than for air. It should be remembered that we cannot form any definite infer- ence as to the relative number of electrons by comparing the ionization currents in the E curves for different gases at the same pressure because these currents are due to the electrons which have passed through the meshes of the gauze electrode and the fraction of electrons which accomplish this depends upon the gas concerned. When the pressure of the CO, or H, was relatively high the free electrons appeared to be extremely sensitive to the pres- ence of impurities in the gas under consideration ; the number of free electrons was greatly decreased if the gas were allowed to stand undisturbed for a few hours in the measuring vessel which was presumably air-tight. This effect is illustrated in the curves of fig. 5; curve A refers to CO, at 79™™ pressure, the readings being taken quickly after the introduction of the gas ; curve B exhibits the values after the gas had been allowed to remain 24 hours in the closed vessel. 8 Wellasech—Motion of Lons and Electrons through Gases. For lower pressures of the gas this effect practically van- ishes ; with CO, at a pressure of 44™" the EI curve obtained after the gas had remained undisturbed in the vessel for 2 days was identical with that obtained immediately after the intro- duction of the gas. It is probable that the above effect arose from a very slow leak of oxygen into the vessel from the outside atmosphere ; actual experiments were performed to test this point and it was found that traces of air added to CO, or H, at relatively high pressures resulted in a marked decrease of the number of free (higek zs ALT A eal Ze ed 2s ate i ekees ee TTS Saas miekae ee Bere Ase 0 20 40 60 80 100 120 Volts. electrons whereas when these gases were at low pressures the number of electrons was not appreciably affected by the admixture. 7 It should, however, be mentioned that a similar though much more intense effect was found in experimenting with the free electrons in the vapor of petroleum ether (w. sec. 4 D); in this instance the diminution in the number of electrons was very rapid and could not reasonably be ascribed to a small leak of air into the apparatus. All the indications pointed to the appearance in the vapor of a constituent capable of absorbing electrons at ordinary temperatures. It is convenient to refer to nuclei, whether molecules or aggregations, which possess this property, as ‘electron sinks’; the electrons cannot remain in the free state during their motion through a gas which con- tains these sinks other than in excessively small quantity. All _ Wellisch— Motion of Ions and Electrons through Gases. 9 the experimental evidence indicates that the molecules of oxy- gen do not belong to this class of impurities and that the larger electron velocities attendant upon the act of ionization are necessary for the formation of negative oxygen ions. It is of course possible that the decay of the electrons in CO, and H, does not arise from an air leak but is due to an ageing effect similar to that in petroleum ether. In this connection several unsuccessful attempts were made to remove possible nuclei from CO, which had been allowed to remain for several ue at a pressure of 81™™ in the measuring vessel. In one experiment the gauze electrode was maintained for several hours at a potential of —160 volts in the hope that the electrons which were being continually produced would ultimately remove the nuclei from the gas; however the current meas- urements failed to indicate any tendency to restore the original condition of the gas under which permanently free electrons were in evidence. The same gas was subsequently passed several times through P,O, by means of a mercury reservoir attachment in order to remove any trace of water vapor which might have arisen from the metal walls: the free electrons, however, did not reappear and the possibility of the existence of nuclei consisting of molecules of water vapor was thus excluded. A few experiments were made to ascertain whether free electrons are present in carbon monoxide. This gas was liber- ated by the action of concentrated sulphuric acid on potassium ferrocyanide and was passed through solid caustic potash, cal- cium chloride and phosphorus pentoxide before admission into the measuring vessel. A typical EI curve was obtained for CO at a pressure of 13™", demonstrating thus the existence of free electrons ; these were, however, not nearly so numerous as in air at the same pressure and, as the manipulation with this gas presented difficulties, it was not considered expedient to extend the investigation. It seems fitting to refer here to an apparent difficulty in connection with the existence of free electrons in gases. The electrons were shown to appear in measurable amount in dry air at pressures as high as 8™ and yet it has been mentioned in this section that a trace of oxygen is sufficient to cause them to disappear from CO, or H, at relatively high pressures. Refer- ence is made later (sec. 5) to this apparent “discrepancy ; ; the dif- ficulty is in large measure removed by a consideration of the experimental fact that the sensitivity to oxygen decreases rapidly as the pressure of the original gas is reduced. To take actual filoures, it was found that a trace of air would rob H, at 1 atmosphere of its free electrons and yet in a mixture of ‘H, at 825™ and air at 237 the electrons appeared in consider- able numbers. 10 Wellisch— Motion of Ions and Electrons through Gases. (B) Motion of Free Electrons. A number of experiments were undertaken to determine the velocity with which the free electrons moved in an electric field through CO, and H,. Mobility values have already been assioned by Franck* for the electrons in argon, helium and nitrogen at atmospheric pressure; the values given were respectively 209, ca. 500, and 120°™ per sec. per volt per em. The mobility values were found to be extremely sensitive to the presence of impurities in the gas under consideration, the slightest trace of oxygen, for example, causing a considerable reduction in the value. Recently Hainest has investigated the HiGs 5: TPE poe SHEP dun7 anne Saba) 4607 eeeen Nin dea CoP COVA CPC kg Ye mee Pet Te are Current. motion of free electrons in pure nitrogen at atmospheric pres- sure and has obtained a mean value of 367 for the mobility. Carbon dioxide appeared especially suitable for experiments in this connection because the electrons were relatively numerous in it and at the same time the density of the gas was sufficiently great to justify the belief that the velocities would not be inor- dinately large and thus incapable of measurement with the apparatus at disposal. Even with the high frequency of 800 alternations per second and at the highest practicable pressures of the CO, it was found that the values of the critical potential (V,) were considerably less than 10 volts, so that the observa- * Franck, Verh. Deutsch. Phys. Ges., xii, pp. 291, 613, 1910. + Haines, Phil. Mag., vol. xxx, p. 508, 1915. Wellisch— Motion of Lons and Electrons through Gases. 11 tion error in the determination of the electron mobility was of necessity considerable. Moreover there was also the difficulty connected with the presence of the ageing effect which, as men- tioned above, occurs at the higher pressures ; it was of course not feasible to attempt determinations at the lower pressures where this effect is absent because the electron velocities become excessively large. It was in every instance found that the effect of age (i. e. of allowing the CO, to remain for any length of time in the ‘apparatus) was to reduce considerably the velocity of the elec- trons. On this account great care was taken to exclude impuri- ties, the gas being in all cases swept several times through the measuring vessel, and the observatious quickly made after the final introduction. In figs. 5 and 6 there are given a few typical E curves which were obtained in the determination of V, for the free electrons ; the ions do not make their appearance until much higher poten- tials are employed. Reference will be made later to the fact that the experimental results rendered doubtful the assumption that the velocity of the electron is proportional to the applied field,so that the use of the term ‘mobility’ is not certainly justified ; however it was thought useful to make the calcula- tions on the assumption that there exists a distinct mobility for the electron just as for the ion. In the following table there are given the results of the mobility determinations for freshly prepared CO, together with some of the results for CO, in various degrees of impurity; the symbol K denotes the mobility reduced to atmospheric pressure on the assumption of the validity of the law pk = const. Freq. f p We k K Remarks. 807°9 | °545 35°D 1°4 4232 197°6 Fresh eZ0°6")°°518 54 Qh Fl Zoe 183°8 + 824°5 | *527 87 2°8 2233 255°6 |, MIG ee 7530 || N37 4°5 1336 241°0 833°3 | °d40 79 3°0 2056 214°0 830°1 | °545 54 8 ols, 54°] 2 days old 834°2 | °550 25 2°2 Qi OiT) 83°4 do. 833°3 | °540 79 4°6 1341 139°4 24 hrs. old 828°2 | 540; 81 7°35 834 (oe CO, aye Qmm air In fig. 5 there is given an E curve for freshly-prepared hydro- gen at atmospheric pressure ; this curve was 1700™ per sec. the value of K deduced from 12 Wellisch—Motion of Lons and Electrons through Gases. The values of K for freshly prepared CO, are scarcely in sufficient agreement to justify the assignment of an average value. It is evident from the table that the electron velocities are very sensitive to the presence of impurities; the highest value of K obtained was 255°™ per sec., but even this cannot be regarded as a maximum as a greater degree of purifieation would probably result in still higher values. There is some evidence of an indirect character that the elec- tron does not move with a velocity which is strictly propor- tional to the applied field but traverses with an accelerated Fic. 6. pes SEY SUE SET eS 2 RT AOD a Sal ae mY MM | | ERERERERSERABE EMRE BAS ee as i ee ae [freshly prepared COzt | yr RB -|- “ee EES a wa Se SiN me Oi i RE i ae I cca pet eh EERE ici EY em Zee nea era rae a A a A =4r-0e. fe oy ic A 72 MANE en AR es Gn | HB bt 0 a ak B= shee nee motion distances comparable with the distance between the electrodes. The close approach for small potentials of the E curves in fig. 6 which refer to CO, at the same pressure (187"™) but with different alternation frequencies suggests very large values for the velocities of the electrons; this is more ay kV where 7’ is the current for potential V when the alternating field is employed and 7 is the current for potential V directly applied. If we take the velocity calculated from V, at the higher frequency, viz. k=1336 at 187"™", the above formula (with d=2) gives 7’ ,/7’,=4:1 for V=5-78 volts where the suf- fixes 1 and 2 refer to the low and high frequency respectively ; the value obtained from the experimental curves is only 1:3. Similarly for V=6'96 volts we obtain a calculated ratio of 2°6 whereas that obtained experimentally is 1:2. These considera- tions would seem to imply that the value k=1836 is too small readily understood if we apply the formula @=7 (f— Wellisch— Motion of Tons and Electrons through Gases. 13 and that the critical potential is really smaller than the value (V ,—4'5) apparently obtained. , In this connection it is a significant fact that several of the E curves, especially those obtained at the highest pressures, showed a distinct curvature in the neighborhood of the poten- tial axis, the tendency being to shift the point of intersection towards the origin. This shape of the current-potential curves in the vicinity of the origin suggests accelerated motion of the electron or a slow acquisition of a terminal velocity. Further experimental data are of course necessary before the nature of the motion of the electron is definitely ascertained ; the suggestion here given is that the electron may traverse a considerable distance with accelerated motion before its ter- minal velocity is acquired. It should be remembered that Franck and Hertz* have already shown that the collisions of electrons with the molecules of the inert gases are practically perfectly elastic so that the drift motion of the electron would under these circumstances be accelerated. The experiments with regard to the effect of impurities upon the number of free electrons in CO, or H, strongly suggest that the collisions - of electrons with the molecules of these gases have a high degree of elasticity, although naturally not so high as with the inert gases. The effect of this high but imperfect elasticity would be to cause the electrons when moving under an electric field in CO, or H, to move with an accelerated motion until their terminal velocity is acquired. On this view the effect of traces of impurities in the gas in diminishing the velocity of the electron is readily explained ;, the impact of the electron with the molecule of the impurity is. in all probability either inelastic or considerably less elastic than the collision with the gas molecule, and, in consequence, the electron is unable to acquire as great a velocity as in the pure gas. (C) lectrons in Vapors. The demonstration of the existence of free electrons in air, CO, and H,, at relatively high pressures rendered it fairly obvious that all permanent gases were able to contain electrons in the free state. [ranck’s experiments had shown previously that the inert gases were especially conspicuous in this respect, the negative carriers appearing to consist entirely of free elec- trons. It became of interest to extend the investigation to the case of vapors, especially as these are liable to occur as impuri- ties in gases. It was thought extremely improbable that the electrons, if they were present in the free state, would occur in large numbers except at very low pressures; preliminary *Franck and Hertz, Verh. Deutsch. Phys. Ges., xv, pp. 373, 618, 1913. 14. Welltseh—Motion of Ions and Electrons through Gases. - trials with a few vapors justified this conclusion. With the high frequency commutator there was not the slightest indica- tion of the presence of free electrons either in dry SO, ata pressure of 7™™ or in CH,I at a pressure of 28™™. It was how- ever quite possible that lower pressures would bring the elec- trons into evidence, but as the apparatus did not readily lend itself to securing low vapor pressures the investigation was resumed in a slightly different manner. A small quantity of the vapor under consideration was mixed with a permanent gas and experiments were made to ascertain whether free electrons could continue to exist in this mixture; if the vapor molecules behaved as electron sinks and were present in appreciable Hic. 7. ocd A Lye, Lea lee a ia| en 7a SMT MRI er ce HEEREEEEE EP Heft ie he sr dsnesey.©. der ost ce Current. 20 40 60 80 100 120 140 160 Volts. amount then the number of collisions and subsequent attach- ments between electrons and vapor molecules would be suf- ficiently great to prevent the existence of free electrons. This information was of importance in view of the experimental results with regard to the effect of impurities on the number of free electrons in a gas. Three vapors were tried in this con- nection, viz.: ether, aleohol and water; these were chosen because they were deemed to be the most probable absorbers of electrons. In each of these instances hydrogen at a reduced pressure was chosen as the gas with which the vapor was mixed because of the copious supply of free electrons which it affords. An EI curve was first obtained for dry hydrogen at a pres- sure of 36™™ (fig. 7); ether vapor was then admitted until the pressure of the mixture was 38", and the readings were again taken. It was found that even in the presence of 2™™ of ether Wellisch—Motion of Lons and Electrons through Gases. 15 vapor a considerable number of free electrons were able to traverse the distance between the electrodes. The number was less than in the pure hydrogen, but the EI curve for the mix- ture (fig. 7) was sufficiently definite to justify the conclusion that the molecules of ether vapor do not behave as electron sinks. The experiments with alcohol vapor were conducted in a similar manner; an EI curve was obtained for a mixture con- sisting of hydrogen at 35™™ and alcohol at a pressure slightly less than 1™™. The number of electrons was again distinctly smaller than in the pure gas, but was sufficiently great to make it evident that the molecules of alcohol were unable to absorb the free electrons. In order to experiment with traces of water vapor present in the gas, the tubes containing the drying agents were removed so that the hydrogen passed into the measuring vessel directly after generation in the Kipp’s apparatus. The moist hydrogen was introduced at a pressure of 37™™ and a current-potential curve (v. fig. 7) was obtained in the usual manner ; the presence of the moisture caused a reduction in the number of free elec- trons, but these were in sufficient evidence to show that the water molecules do not behave as electron sinks. It is of course quite possible that in all these instances a loose attachment may occasionally exist between the electron and the vapor molecule; the experimental results indicate, however, that such an attachment, if it occur at all, persists only for a time which is small in comparison with that during which the electron remains free. Vapor of Petroleum Ether and the ageing effect. The previous experiments with CO, and H, suggested that the atoms of carbon and hydrogen were in great measure responsi- ble for the relatively large number of free electrons in these gases as compared with air. It became of interest to make a special study of some member of the parattin series whose molecules contain only atoms of carbon and hydrogen or indeed of any vapor which does not contain electro-negative atoms such as those of oxygen or iodine. It was originally proposed to make the experiment with pentane, but as this was not immediately available the vapor of petroleum ether was em- ployed instead. Petroleum ether (sp. gr. ca. ‘67) consists of a mixture of pentane (C,H,,) and hexane (O,H,,); its molecules contain, therefore, only atoms of carbon and hydrogen. A number of determinations of ionic mobilities were also made for this vapor ; reference is, however, made to these only as far as they concern the motion of electrons, the actual values obtained for the mobilities of the ions being deferred to a later section (4D). 16 Wellésch—Motion of Ions and Electrons through Gases. The first experiments with this vapor, which was introduced at a pressure of 95™™, gave a normal value (Kk = :41) for the mobility of the positive ion; the current-potential curve for the negative carriers (fig. 8) was distinctly abnormal, as it afforded evidence of two types of carriers; in addition ‘to the normal negative ion (K = :44) there appeared a carrier (qd, fig. 8) for which K had the value 1° 692, which is about four times as great as one might have reasonably expected for the nega- tive ion and, on the other hand, considerably less than the value corresponding to a free electron. On attempting to repeat the experiment after the vapor had been allowed to SHTGzaS: SCC co HB TP eee Peiroleum Etter |__| | Haee 20 16 Current, —_ Li) Kya [7 | pane Fe eA BA er ae cent ew Om A a Le 100 120 140 Volts. remain for about two hours in the vessel, only the normal value was obtained for the mobility of the negative ion. In the next experiment, after a preliminary evacuation of the vessel, streams of vapor were swept through repeatedly in the hope of removing traces of impurities ; the vapor was finally admitted at a pressure of 76™™ and the readings quickly taken. The curve obtained is given in fig. 8; the direction of the arrow signifies that the current measurements were made in descend- ing order of potential. It will be seen that this curve shows the presence both of ions and of free electrons; the ions enter at 80 volts and possess a normal mobility, viz. K = -480. The curve marked 1 was obtained only a few minutes before that marked 2 and it will be observed that the electrons have decayed appreciably during this short interval of time. The third curve was obtained 24 hours after the introduction of the vapor ; there is now only the slightest indication of free electrons while the negative ion has still a normal mobility (K=-428). Wellisch— Motion of Ions and Electrons through Gases. 17 Subsequent experiments were made with freshly introdueed vapor at a pressure of 20™" and gave evidence of a very large percentage of electrons; however, the ageing effect was very pronounced, the free electrons decreasing in number so rapidly that no regular curve was obtained. The general indications seem to be that in the pure vapor of petroleum ether a large fraction of the negative carriers are free electrons, the negative ions if present at all appearing only in small numbers; the free electrons are, however, extremely sensitive to the presence of some constituent which arises gradually in the vapor, with the result that at the expiration of a few hours the electrons have disappeared and the current of negative electricity is due entirely to ions. The nature of the constituent which occasions the ageing effect in the vapor can at present only be conjectured ; system- atic experiments are necessary before a definite conclusion can be reached. There is distinct evidence, however, that we are dealing here with a true electron sink; in other words, this constituent, whatever be its nature, is capable of absorbing an electron during its drift motion through the vapor and in this respect must be carefully distinguished from impurities such as oxygen, which seem to require for the absorption of electrons velocities considerably higher than those which are afforded by thermal agitation at ordinary temperatures. The effect of the latter type of impurity is to reduce the number of free elec- trons in a gas and at the same time to diminish appreciably the velocity of the electron through the gas; this diminution in velocity has been ascribed (v. sec. 4B) to the comparatively inelastic impact between the electron and the molecule of the impurity. In the experiments with the vapor of petroleum ether the effect of the impurity is to cause likewise a reduction in the number of free electrons and a diminution in the electron velocity ; the electron, however, appears now to be capable of acquiring all velocities intermediate between that of a free electron in the pure vapor and that of a negative ion. We seem, therefore, to be dealing with a carrier which changes continuously and progressively from a free electron to a nega- tive ion; the most feasible hypothesis is that the electron as it drifts through the vapor is for part of the time in the free state and for the remainder in attachment with the molecule of the impurity. It is highly probable that this attachment, occurring as it does as a result of ordinary thermal motion, is of a very loose nature and is liable to be broken at molecular encounters; we would thus expect continual alternations of the electron between the free and combined states. Am. Jour. Sct.—FourtH Serizs, Vou. XLIV, No. 259—Juty, 1917. 2 18 Wellisch—Motion of Ions and Electrons through Gases. With regard to the nature of the sink which is gradually formed in the vapor of petroleum ether, nothing at all definite can be said. We may imagine that polymers or small aggre- gates of pentane or hexane are formed gradually under the influence of the radiation from the polonium; such systems would probably be able to form stable negative ions for large electron velocities and unstable ions for small velocities. Initi- ally, when the vapor is pure, the negative carriers are for the most part electrons; as the sinks appear, the velocity of the electrons would be reduced through the formation of unstable ions. The fact that ultimately the carriers consist entirely of negative ions may be explained by ascribing to a polymer the property of being able to effect occasionally a union between an electron and a molecule of the vapor. (D) tons in Gases and Vapors. Gases.—The law pk=const. was verified for both the posi- tive and the negative ions in dry air over a wide range of pressures. Some of the values obtained experimentally for K at the lower pressures have been given in the previous paper and should be sufficient to illustrate the unchanging nature of the negative ion. , hl i fim +> © Vz Fig. 2. Well records showing the thickness and character of the Devo- nian shales from central Ohio into western Pennsylvania. Explanation of Figure 2. _ In figure 2 are represented ix carefully selected well records. © It is not an easy matter to secure well records which penetrate the entire thickness of the shales and also give their character in detail; and those chosen are not as complete as desired. Still they will indicate in a general way the nature and _ thick- ness of the Devonian shales from the longitude of Sandusky, Ohio east to that of Erie, Pennsylvania. Section A is from Norwalk, Ohio and thus near the western outcrop of the shales 46 W. A. Verwiebe—Oorrelation of the in Ohio. It shows 203 feet of black (Huron) shale underlain by 10 feet of limestone (Prout) and 120 feet of shale (Olentangy). The base of the Berea is caleulated from other wells in the vicinity. Section B from Wellington shows 96 feet of red shale (Bedford) beneath which are 145 feet of black shale (Cleveland). The remainder, 658 feet, is given as alternating blue and black shale. In section C from Akron there are three zones of black shale, the thickest being 150 feet. The rest consists of alter- nating dark and light blue and erey shales. This section was discussed above (p. 387). The next section (D) constitutes the record of a well drilled in the early part of 1916 three miles southwest of Niles in Trumbull County. It shows ten feet of red shale (top of Bedford) and below that 2589 feet of “blue, white, and cinnamon” shale. An interesting detail is the 35 feet of sandy shale and sandstone occurring 612 feet above the base. This showed a trace of oil and may represent one of the lower sands found in western Pennsylvania. Section E takes us into Crawford County, Pennsylvania. Unfortunately the record was not well kept and the sands indicated must there- fore be considered as generalized. The Chemung is about 1200 feet thick, the Portage, Hamilton and Marcellus about 1500 feet, making a total of 2700 feet for the Devonian shales in Crawford County. The last section (Erie, Pa.) was intro- duced partly to show that the thickness of the shales increases toward the south and partly because this record has been kept with such perfect detail that the formations may be delimited with considerable confidence. At the base are shown 25 feet of black shale (Marcellus), above this 170 feet of Hamilton and finally 1105 feet of Portage and Genesee. The top of the sec- tion has been extended by the writer, from calculations based on a study of outcrops in the vicinity, to include the top of the Devonian. An addition of 170 feet of Portage is indicated. Above that 550 feet of Chemung (lower 225 feet-Girard) and 450 feet Bradfordian (Riceville and Venango group). In all the sections the top of the Devonian limestone is used as the lower datum plane and the base of the Berea sandstone as the upper. Summary. The correlation of the Devonian shales of Ohio and Penn- sylvania with those of New York isa difficult problem because the ordinary criteria, lithology, and paleontology are unsatis- factory guides. A careful consideration of all the available data at hand seem to justify the following conclusions: 1. The 750 feet of shales in central Ohio expand into and are stratigraphically equivalent to the 2700 feet of shale in western Pennsylvania. Devonian Shales of Ohio and Pennsylvania. 47 2. The shales in Ohio should be subdivided into two parts rather than three, an upper predominantly black colored divis- ion and a lower predominantly light colored division. 3. The detailed correlation of the various subdivisions in Ohio and Pennsylvania with those of New York is as follows: Ohio Pennsylvania New York northcentral northeast northwest westcentral western | Riceville | (Riceville) (Riceville) ) Bedford Bedford K Knapp napp ; Venango i eae o| § Oswayo 8 Cattaragus | Huron (upper) (Cleveland Chemung (upper)* Huron (lower) Chemung (lower) Prout Chagrin Portage Olentangy Hamilton (upper) ( Hamilton (lower) Drone ) Marcellus Index to Sections Used in Figures. The following were taken from Bull. 15 of the Geol. Survey of Ohio (C. S. Prosser):-No. 1—p. 80; 2—p. 54; 3—p. 185; 4—p. 242; 5—p. 300; 6—p. 316; section F.—p. 413. The following from vol. vi, Geol. Surv. Ohio, 1888 (Edw. Orton) A—p. 441; B—p. 348; C—p. 367. Section D by courtesy of J. A. Bownocker. Section E by courtesy of Frank Mossinger. *It is probable that the Chemung of western New York includes some of the Bradfordian also, “UBIPLOFPVIg 48 L. £. Harper—Lvidence of Uplift. Art. III.—Hwidence of Uplift on the Coast of New South Wales, Australia ; by L. F. Harprr. Durine a geological survey of the Southern Coal Field of New South Wales, evidences of an uplift of a portion of the coastal belt of that State were studied. | The coastal plateau of the Illawarra district is formed of Permo-Carboniferous and Triassic strata capped by the resist- ant Hawkesbury sandstone. Contemporaneous lava flows are included in the series, and sediments aud lavas alike are intruded by .dikes of basaltic rock, chiefly monchiquite and camptonite. The dip of the strata is northerly, parallel with the coast line, and vigorous wave erosion on the plateau edge has formed long lines of sheer cliffs in Triassic sandstone and cliffed headlands interspersed with shingly beaches 1 in Permian beds and lavas.* In the [lawarra district each rock headland is faced by a flat rock shelf, the surface of which is developed generally on a stratification plane. These shelves vary.in width from 70 to — 250 feet and stand 2 to 4 feet above high-water. They are most pronounced on headlands where a dense lava flow over-_ lies sedimentary stratae Vertical dikes crossing the shelf are less resistant to weathering than their bordering walls and are usually represented by open channels at sea-level or by fissures in the cliff face. (Hig. 1.) At three localities in the neighborhood of Kiama, so-called blowholes are found. These are openings into the cliff face at sea-level and.consist either of a horizontal tunnel with an out- let vertical to the land surface, or of. a tunnel only.. In the case of the former, the water propelled by waves passes through the tunnel and escapes by the funnel-like, vertical opening ag a shower of spray. Two blowholes with vertical outlets were noted. Although both occur in the same contem- poraneous lava flow, each has a separate origin. The principal Kiama blowhole was produced by the action of the sea on a decomposed dike in the lava. A tunnel about 60 feet long resulted. Owing to a depression of the land sur- face, the sea formed a vertical opening at the landward end of the tunnel, up which the spray is driven to a considerable height. (Figs. 2 and 3.) In places, the lava flow is columnar and rests on a bed of voleanic tuff—factors leading to the formation of a second type * For a description of the geology of this region see:—Harper, L. F., Geology and mineral resources of the Southern Coal Field, Memoirs Geol. Survey New South Wales, Geology No. 7, Sydney, New South Wales, 1915. L. F. Harper—Lfvidence of Uplift. 49 of blowhole. Wherever the underlying bed of tuff has been subjected to maximum wave action, the lava sheet has been undermined. In one instance, at the landward end of a cave so formed, a basaltic column dropped out and left an opening about 15 feet deep and 14 feet in diameter. The column was probably loosened both by the chemical action of the sea Fie. 1. Fic. 1. Rock shelf crossed by a weathered dike. Illawarra district, New South Wales. water forced up the joint faces and by the mechanical action of the waves on the unsupported prism of basalt. In the ease of the horizontal blowhole, which has no vertical opening, an incoming wave compresses the air in the rock chamber, and as the wave subsides the water is forced out in a horizontal shower of spray. In the event of a heavy sea and favorable tidal conditions, the effect produced by blowholes is awe-inspiring as well as spectacular. At such times, a subter- Am. JouR. Sci.—FourtH Serizs, VoL. XLIV, No. 259.—Juty, 1917. 4 50 L. F. Harper— Evidence of Uplift. ~ ranean roar is accompanied by a tremor in the roof of the cavern and is followed by a jet of spray shot from 50 to 100 feet into the air. (Fig. 3.) Caves and channels now 16 feet above high-water mark and similar in appearance to the blowholes deseribed indicate a similar origin. As blowholes are necessarily formed at sea- level, the conclusion to be drawn is that the coast has been Hares Fic. 2, Entrance to tunnel leading to blowhole. Kaima, New South Wales. . raised. Caves developed along dike channels extend under cliffs for distances ranging from a few feet to 200 feet. The dike material at the end and in the roof of these caves is from 2 to 4 feet wide, but caving of the walls has enlarged the openings and produced caverns from 6 to 30 feet wide. The largest of these caves—Hole-in-Wall—(formed in sediments of the Narrabeen stage, Triassic) is about 10 miles north of Syd- ney. The floor of this cave is strewn with sand and sandstone blocks accumulated long after it was elevated above the sea; and the present floor level does not, therefore, indicate the amount of uplift. Caves in Permo-Carboniferous strata about L. FE. Harper—Evidence of Uplift. 51 60 miles south of Sydney, at Kiama, record an uplift of about 16 feet. Accumulations of coarse shingle at the heads of many small coves in the Illawarra district also indicate an uplift of the land. This shingle is from 10 to 15 feet above a similar Inine, By Fic. 8. Spray issuing from Kaima blowhole. New South Wales. accumulation which extends from below low-water mark to just above high-water mark, and is composed mainly of bowlders of basalt and rocks from the headlands. No modern storm could have deposited shingle at this elevation. The higher beach is separated from the present beach by a well- marked declivity and in many places maintains a stunted growth of vegetation. A third proof of uplift is the presence of flat shelves of rock extending out from the cliff faces about 3 feet above high- water mark. Under present conditions, these shelves are 52 L. EF. Harper—Evidence of Uplift. reached by spray only during storms. When they were planed off, however, their surface must have been subjected to the maximum erosive action of the waves, which, judging by the depth of water, is probably about 10 feet below high-water mark. A fourth proof of uplift is based on the presence of shallow lagoons or lakes which occur at intervals along the coast and are open to the sea only during exceptionally stormy weather, or when they are artificially opened to allow flood-waters to escape. Along other parts of the coast are extensive tracts of level, swampy land, containing fresh water—areas useful for grazing purposes only during very dry seasons. They are thought to represent ancient lagoons which have been raised above sea-level. In favor of this assumption, it may be added that fragments of recent marine shells are found along the margin of these fresh-water swamps. That this uplift is comparatively recent is borne out by the geological evidence available, for there is little doubt that the intrusive dikes are of post-Tertiary age, and yet sufficient time elapsed ‘prior to the uplift to permit of marine excavation along their courses to a maximum extent of 200 feet. A review of the evidence leads the writer to estimate the amount of uplift as follows: 1. Coastal shelves, formed 10 feet below high-water mark. now, 4:1eeCt apOviess = = .5 32s = eee 14 feet. 2. Ancient blowholes formed at sea-level, now 16 feet above. After taking off for the shelf elevation 12 feet The total uplift is, therefores about --2222222 22254 26 feet. Geological Survey, New South Wales, Australia. Gooch and Kobayasha—Platinized Anode of Glass. 58 Art. I1V.—The Use of the Platinized Anode of Glass in the Electrolytic Determination of Manganese ; by F. A. Goocu and Marsusuke Kosayasal. {Contributions from the Kent Chemical Laboratory of Yale Univ.—cclxxxix. | Ty a recent paper* we have shown that the use of very small rotating electrodes of platinum in solutions of usual volume (100°™*) is perfectly feasible, although the time required for the complete formation of the electrolytic deposit increases with the volume of the solution from which deposition is made. The process was illustrated by the deposition of copper and nickel upon the cathode, and of lead dioxide upon the anode. We have attempted, also, to effect the deposition of hydrated manganese dioxide upon the very small rotating anode from a solution of manganous sulphate, but in this case the special difficulty is presented that the superoxidation of the manganese to the condition of permanganic acid takes place under the action of the higher current density implied in the use of the usual strength of current and the same anode surface. Chrome alum, alcohol, and formic acid, used as deoxidizing agents, were introduced to effect the reduction of the permanganic acid, but the results which are summarized in the following statement show that the deposition of the manganese dioxide upon the small anode, under the ordinary current strength, is incomplete and unsatisfactory. Preliminary Tests. (Volume of solution, 100°™*: anode surface, 1°5°™3 (approx.): time, 90 min.) Manganese weighed as Manganese __ dioxide, taken as dried at Reagents added to Electrolytic sulphate 200° Initial current Solution erm, germ, amp. volt 0°1085 0°0337 75) 16 Acetic acid, 20°™* Chrome alum, 1 grm. Ammonium acetate, 5 grm. 0°1085 0°0389 2° i Alcohol, '5°™* Ammonium sulphate, 1 grm. Sulphuric acid, 15 drops 0°1085 0:0275 1° 30 Alcohol, 10°" Formic acid, 5°™* 12 Acetic acid, 3°™ Chrome alum, 2 grm. Ammonium acetate, 10 grm. Alcohol, 10°™* “Chis Journals xii, 3015 1917. 0°1085 0°0438 Lo Or 54 eee. and Kobayashi—Platinized Anode of Glass. These results point to the conclusion that the high current density which is a necessary consequence of the use of the very small anode with currents large enough to effect an electro- lytic determination of manganese within a reasonable time is impracticable. In the succeeding experiments, therefore, the use was made of the rotating electrode of platinized glass, described by Gooch and Burdick,* since this device affords a large surface for the deposition with small expenditure of plati- num. The anode used in these experiments was made by heating, to a temperature sufficient to volatilize glycerine, a tube of lead glass, shaped like a test tube, painting upon it a viscous emulsion of dry chloroplatinic acid in glycerine, and burning the film of deposited platinum into the glass at the softening point of the latter. Connection of this platinum film with the rotating shaft was made by platinum wire bound about the tube, reaching over the edge of the latter, and pressed by a rubber stopper into contact with a strip of platinum foil in Fie. 1. electrical contact with the metal shaft. An important modification of the original form of this type of electrodet is the attachment of the binding platinum wire at a point so low that the wire will be kept cool by immersion in the electrolyte and thus avoid the possibility of cracking the glass electrode by the over-heating of the wire when carrying a high current. This electrode is shown in the accom- panying figure. With the anode of platinized glass two sets of experiments were carried out. In one set the cathode was a piece of plati- num foil measuring 0°5 cm. x5cm. In the other set of experiments three ¢a- thode toils of twice this size were em- ployed, thus permitting the passage of the same strength of current under a lower potential. At the end of the electrolysis the solution was drawn off by means of the filtering tube (fig. 2) made by fusing the flared end of a lead glass tube to a disc of platinum gauze and coating the disc with a filtering mat of asbestos by dipping it in an emulsion of asbestos and applying suction. The hydrated manganese dioxide deposited upon the anode * This Journal, xxxiv, 107, 1912. + Loe. cit. HIG; 2; Gooch and Kobayashi—Platinized Anode of Glass. 55 and collected upon the asbestos was dissolved in a cold mix- ture of sulphurous and sulphuric acids, and the solution, filtered on paper from the asbestos, was evaporated. The residue was dried over the radiator, at about 450°, and weighed as the anhydrous sulphate. The material taken for analysis was a solution of manganese sulphate standardized by evaporating portions of the solution to dryness, drying over the radiator, at about 450°, and weigh- ing as the anhydrous sulphate, MnSO.,,. Details of these experiments are given in the following table. Electrolysis of Manganous Sulphate. (Volume 100°™°: anode surface of platinized glass, 25°™? (approx.): anode revolu- tions, 150.) Manganese deposited as hydrated di- Manganese oxide and taken aS weighed as sulphate sulphate Hrror Initial current Time Reagents added grm. grm. grm. amp. volt min. Large cathode: 3 (1cm. x 5 cm.) 0°1080 01057 —0-0028 ili 12°5 120 ( Chrome alum, 0°5 grm. Acetic acid, 5°™3 0-1080 0:0922 —0-0058 1°6 13° 120 0°1080 0:1080 0:0000 16 12°5 150 es 0°1080 0:1073 —0-0007 Ly 14:8 150 ng 0°1080 0:1086 +0-0006 1°6 10°3 150 a 0-1098 01094 —0-0004 1°5 16° 150 oe 0:1098 01095 —0-0008 1°6 15° 150 ap 0:1098 0:1074 —0-0024 1°6 Ie oe Alcohol, 10°™? Acetic acid, 5°™* 0°1098 0:1040 —0-0058 1°5 18°3 180 a 01098 01097 —0-:0001 16 18° 240 { Alcohol, 5°"? Acetic acid, o°™? 0°1098 01095 —0-0003 15 18° 240 se 0:1098 0:1098 0-0000 1°5 18°3 240 0:1098 0°1093 —0-0005 1°5 Small eathode: 0°5 em. x 5 em. 01082 00-0875 —0-0207 1°2 27 150 ( Alcohol, 5°™3 Acetic acid, 5°™? Ammonium sulphate, 2 grm, 0°1082 0:1081 —0-0001 eZ 26 180 co - 0°1082 0:1080 —0-0002 1°2 25 210 a 01082 0°1081 —0-0001 le 20 210 0-1082 01084 +0-°0002 1-2 26 210 ne 0°1082 071080 —0-0002 1°4 20 240 ss 0°1082 01079 —0-0003 1°2 33 240 ( Alcohol, Homs Acetic acid, 5°™3 eaearnoni i sulphate, 1 grm. 0°1082 0°1085 +0-0003 Lea 30 ae Alcohol, 5°™% Acetic acid, 5¢™% 56 Gooch and Kobayashi—Platinized Anode of Glass. These results show plainly that the estimation of manganese by electro-deposition of the hydrated dioxide upon the rotating anode of platinized glass, and subsequent conversion of the manganese to the anhydrous sulphate, is entirely feasible. From the electrolyte containing in a volume of 100° approxi- mately 0:1 grm. of manganese in the form of sulphate, acetic acid (5°*), and chrome alum (0°5 grm.) the time required for the complete deposition of the hydrated manganese dioxide was two and a half hours; and when alcohol (5™’), preferably with ammonium sulphate (2 grm.), was substituted for the chrome alum (0°5 grm.) the time required was extended to a safe minimum of three and a half to four hours. Art. V.—Preliminary Note on the Occurrence of Verte- brate Footprints in the Pennsylvanian of Oklahoma; by Witrarp Rouse JItison. Durixe the summer of 1916, while mapping the structural geology of a portion of the Osage Nation, Oklahoma, the writer had the good fortune to discover a series of casts of ver- tebrate footprints in one of the sandstone members of the Middle or Lower Pennsylvanian. The location of the bed con- taining these fossil trails is in Township 27 North, Range 10 East, Section 31 in Elm Creek about six miles northeast of Pawhuska. Stratigraphically the clay-sandstone member in ~ which the casts of the animal tracks are preserved is calculated to be about 200 feet below the top of the Elgin sandstone which Adams, Girty, and White* regard as the Oklahoma equivalent of the Kanwaka shales of Kansas. As the Elgin sandstone is somewhat thicker in Oklahoma than in Kansas this would correlate the track-bearing horizon of the formation with the uppermost part of the Le Roy shales of Kansas.t This correlation fixes these beds as middle or lower Pennsy]- vanian, which is probably as close a determination as can be made until the Oklahoma Carboniferous and Permian in their relation to the Kansas equivalents are better understood. . Fossil footprints, or casts, from this undoubted marine hori- zon of the Oklahoma—Pennsylvanian series have never before been described, and because of their rarity such occurrences are to be regarded with interest. The slab on which the casts * Upper Carboniferous Rocks of the Kansas Section, U. S. Geol. Surv. Bull. No. 211, p. 45, 1908 + Idem, pp. 65-66. W. R. Sillson— Vertebrate Footprints. 57 occur is about one foot thick. It is overturned from its orig- inal position above a soft gray clay which has recently been removed at this point by the waters of Elm Creek. Both the impressed clay and the casting sandstone are small and unim- portant members of a long series of interbedded layers of sand- stone and shales. Although a diligent search was made, no exposure of the original upper surface of the clay could be found. Due to the nature of the clay deposit, it is not thought TEES Ibs Fig. 1. Detailed study showing two series of tracks. Note the lack of heel impressions (horizontal surface 2). that it will ever be possible to collect anything of greater sig- nificance from this locality than these casts. The specimen shows a double trail of casts across the slab. In the upper trail the movement was from left to right, and just below one may note the probable return or “ back track- ing” of the same individual (fig. 1). Whether this creature was amphibian or reptilian there is no absolute proof, though many small factors point toward the former. Being adapted to a strandline habitat it undoubtedly possessed aquatic ten- dencies, which statement finds considerable support in the size and physical character of the footimpression. A five-toed ani- imal, its feet were apparently almost as broad as long, and this taken into consideration with the close proximity of the foot- prints suggests: (1) a poor adaptation to land locomotion or crawling ; (2) probably a more efficient adjustment to swim- ming or paddling; and (8) a compromising of these two in a 58 W. Lf. Jillson— Vertebrate Footprints. bottom crawling. Detailed anatomical studies of these tracks are planned, and the results with a series of measurements and outline diagrams will be presented in a later paper. One of the striking features of this slab of sandstone casts is the absence of the impressions of the feet of one side of the animal in both series of tracks. This peculiarity is generally to be accounted for in one of two ways. First, the animal may “back track” over one of the lines of fresh impressions with the result that the superimposed body weight flattens out and destroys the new undried tracks. In this case, however, the double track of the last movement should be left undisturbed. The specimen, however, does not show this double track. Second, an explanation for the single series of impressions is frequently found in the steeply dipping surface of muds on which the animal crawled along. Such a condition would of course give a series of good impressions on the lower side of the body, but on the upper side of the animal there would be very poor impressions or none at all. Tracks made under such conditions might be expected to show pronounced evidence of the fact in their increasing outward impression. A strong heel impression ought also to occur. Neither of these character- istics can be said to occur in these tracks, and as a result pres- ent speculations as to the reason for the absence of the complementary series of tracks have led to no definite conelu- sions. Only the very slightest and most occasional tail groov- ings could be detected, indicating very possibly the comparative physical insignificance of the caudal appendage. The block herein described has been presented by the writer, at the suggestion of Drs. M. G. Mehl and R. L. Moodie, to the oO Department of Geology of the University of Oklahoma. Yale University, New Haven, Conn. W. LI. Jillson—Recent Voleanie Hruption. 59 Arr. VI—WNew Evidence of a Recent Volcanic Eruption on Mt. St. Helens, Washington; by Wittarp Rovse JILLSON. Every intelligent person returning from the wooded foot- hills or snow-capped summit of Mt. St. Helens, Washington (9671 feet—U.S.G.8.) brings back stories of recent volcanic action. The tourist, though he keep to the well-beaten, government trails, sees the evidence on every side, and if he will but listen, may still hear from the lips of a few old pio- neering guides very interesting, though perfectly unreliable reports of “the last eruption of the mountain.” During the summer of 1915 the writer spent several weeks examining the geology and physiography of the Mt. St. Helens region, making the climb to the top on August 5th. Many recent flows were noted at elevations of 6000 and 7000 feet, but by far the greater masses of extrusive material lie below 4500 feet. In September while investigating the surficial character of the great recent flow on the southwest side of Mt. St. Helens, between Big and Cougar Creeks, the block of lava shown in the accompanying illustration (fig. 1) was collected. The specimen, which is now in the Yale University Geologi- eal Collections of Peabody Museum, in New Haven, was taken from the surface of the flow at the end of a spur of old meta- morphic rock which rises out of the floor of the flood, like an island im the sea. The spur is, in effect, the lower end of the sharp divide between two old lava-fiiled valleys. At this point two long lava ribbons stretching down the side of Mt. St. Helens unite before cascading into the bottom of the Lewis River Canyon two hundred feet below. In spreading out laterally at the lower end of this spur to join each other, the flows evidently became slower, and formed, if the comparison may be allowed, a “ lava back-water.” From the standpoint of the physical principle involved, the condi- tion found here must have been essentially the same as that which is seen to be operative at the foot of any stream island. The slackening of the lava flow is obvious, for the surface con- sists of a series of roughly outlined terraces leading down from both main flows into a considerable depression which finds its greatest depth close to the spur-end. As far as the preserva- tion of the tree moulds or casts is concerned, this slowing up and thinning out of the lava has been a fortunate thing. In places where the lava is much thicker and the flows were more rapid and more powerful, very few traces of tree casts are to be found. At the place where the specimen shown in the illustration was procured, are to be found the lava casts of a 60 W. &. Sillson—Recent Volcanic Eruption. score or more of standing and fallen logs. The abundance of casts is a measure of the protection from the destructive force of the two main flows. The occurrence of this small locality has not been mentioned before in the literature of the subject, although a number of other more conspicuous places adjacent to the main trails have been described. The vertical casts appear like man-holes in the lava floor, “the wells in lava” of Elliott.* At depths of from nine to twelve feet these regularly outlined cylindrical carbon- ized casts of the trees extend outwards and downwards into giant root casts. The horizontal casts appear as long, regu- larly cylindrical lava-moulds extending back considerable but undetermined distances into the rock. The horizontal casts are in general of a uniform size and from two to three feet in diameter. The vertical trunk or stump casts seem to be larger (which we should logically expect), some of them attaining five feet in diameter. The museum specimen (fig. 1) was taken from one of the horizontal casts of this area. It shows in fine relief the longi- tudinal breaking or splitting of the wood, and the transverse or circumferential check rings, both of which are due to the special kind of carbonization and rapid contraction of the wood cells which the tree trunk underwent in its nearly-closed lava kiln. The fine longitudinal striee-casts show the character of the annual growth layers of the wood. The size and rough character of the largest ridges on the cast indicate that the log had been greatly charred before the lava struck it. Studies of the carbonization of the cast have led to the placing of the cast-producing tree in the conifer group. This statement finds strong support in the great diameter of the casts themselves. The size of these trees is indicated by the flatness of the arch of the section of the lava cast shown in fig. 1. The important bearing of these tree casts upon the determi- nation of the dates of various flows is pointed out by Diller,t who also gives a résumé of the better known literature of the subject. He publishes a letter from Mr. F. V. Coville of the Department of Agriculture, who while studying the flora of Mt. St. Helens found some interesting charred trees under recent gravels of the Kalama River. The following statements occur in the second paragraph of Mr. Coville’s letter: ‘‘ The character of the charcoal, which need not be described in detail here, is such as at first to suggest that it was made in a very carefully prepared kiln. There are, however, no char- coal pits in the region. . . . It is evident from the peculiarities * Elliott, C. P., Nat. Geog. Mag., vol. viii, p. 227, 1897. + Diller, J. S., Latest Volcanic Eruptions of the Pacific Coast, Science, N.5S., vol. ix, pp. 689-40, 1899. W. R&R. Sillson— Recent Voleanie Hruption. 61 of the flora of Mt. St. Helens, and from its limited erosion, that it is a mountain of very recent volcanic origin.” Anda little further on, speaking of the tree casts, he says, “‘ Though I was unable to visit the places where these tree moulds occur, I talked with . .. men who had seen these casts, but none of them had seen. charred bark or wood in the holes.” His con- Fie. 1. eo Fic. 1. Base view of lava block. Shows the strongly carbon-casted sur- face of the lava where it found contact with the previously charred log. The lava is distinctly vesicular. clusions are, however, that these casts are the source of the peculiarly charred logs of the Kalama. Diller agrees with Coville, and says these charcoal trees are probably at least 100 years old, and, “If this be true it is probable that some of the charred loos are not the result of the last eruption of Mt. St. Helens, but an earlier one.” This statement virtually amounts to a recognition by Diller of volcanic activity on Mt. St. Helens within the last century. 62 W. f. Sillson— Recent Volcanic Eruption. Of first significance then, and in line with the arguments of Diller and suggestions of Coville, was the discovery in the same locality from which the Yale cast was taken, of the decayed and disintegrating remains of asmall tree trunk. The material lay in the bottom of one of the horizontal casts and extended back into it for several feet at least. A portion of this residue, which consisted mainly of broken, powdered chareoal and a small amount of the decayed wood, was col- lected but, unfortunately, was lost in packing out of the mountains. In consideration of the fact that the temperatures of most fluid lavas are greatly in excess of the ignition point of wood, it is realized that any carbonized log producing a lava-cast would have a very slight chance of being preserved, even to the end of the period of volcanic activity, unless completely imbedded within the lava. In such an event the log would be redueed to absolute charcoal, but the chances of its discovery in recent, uneroded lavas would be slight. It is thought that the only possible means by which original spontaneous combus- tion eould have been stopped, thus preserving the charred logs in the moulds, would be by the introduction of a completely encasing water jacket immediately following the contact of the lava with the wood. It may be noted again that the casts herein described occur in a considerable depression which might well have been a small collecting basin for the hot and rapidly vaporizing surficial waters which commonly accom- pany volcanic disturbances of this kind. The rough terracing of the lava seems to indicate the existence of a rising and widening water barrier and the spur described above is favor- ably situated for directing surface water into a depression at its base. The woody material taken from the lava casts has received only superficial examination. The writer, therefore, considers it inadvisable at this time to enter into further abstract con- siderations. If new collections and detailed studies show that this decaying woody residue was a part of the original cast- producing log, and not, as has been suggested, foreign material introduced in some way into the case, the views of Coville and Diller regarding the recency of volcanic activity in the Cas- cades receive direct support. These facts, exclusive of much existing corroborative documentary evidence, would be suffi- cient to establish the occurrence of mild extrusive volcanics at Mt. St. Helens well within the last century. Yale University, New Haven, Conn. S. Ichikawa—Some Notes on Japanese Minerals. 63 Arr. VII.—Some Notes on Japanese Minerals; by Suim- matsu IcnikKaAwa.* V. Natural Etching of Garnet Crystals. AurnoueH erystals of garnet are found abundantly at many localities, the natural etchings of the crystals have not yet been discussed by mineralogists. In 1908, I collected garnet crystals from Wadatoge, Sinano Province, and at this time observed erystals with natural etchings. The results of the study of these specimens are shown in the accompanying figures (1). Garnet crystals from Wadatoge occur in cavities of a vitreous andesite. ‘Their color is black, luster vitreous, and crystal faces show the combination of the dodecahedron (d) and icositetra- hedron (nm); the habit varies according to which of these forms predominates (see figs. land 2). Crystals are mostly imperfect and model crystals very rare: the etched crystals were collected from earth produced by the decomposition of a white, tufaceous rock accompanying obsidian; they measured 3 to 9™™ in diameter ; etched figures were more frequent in imperfect than perfect crystals. Some of the crystals were so much etched as to be round like balls; in general etched crystals are distin- guished by their stronger luster. The rounded edges of the etched crystals can be barely observed by the naked eye, and the pits, elevations, etc., can only be investigated minutely under a magnification of 75 to 140 diameters. Fig. 1 shows the natural etching of an icositetrahedral crys- tal; details are given in figs. 3 and 4. Fig. 2 shews a crystal of dodecahedral type. Jig. 3 is a horizontal projection en the principal axis of fig. 1; the solid angles at the extremities of the axes are rounded by etching and elevations having the form of an octagonal pyramid are found on their surfaces. The cruciform edges on the extremities of the same axes are rounded or hollowed and similar elevations formed on the ridges or grooves. The edges formed by the combination of d and 1 are transformed by faces of the hexoctahedron, mOn, by solu- tion. Fig. 4 is a horizontal projection on the trigonal inter-axis of fig. 1; the triplane solid-angles on the extremities of the axes are leveled as faces of the octahedron by etching and pits of the form of a hexagonal pyramid formed on their surface ; the pits appear as negative-crystals on the hexaplane solid angles on the trigonal inter-axes of the hexoctahedron. The trifureate edges on the extremities of the same axes are cut vertically, as — +f, by etching, and the new faces show neither pits nor ele- vations, but sometimes striations are noted. * For an earlier paper, see vol. xlii, pp. 111-119, August, 1916. S. Ichikawa—Some Notes on Japanese Minerals. 64 S. Ichikawa, del. Natural etchings of garnet crystals. The figure is. Fig. 5 shows examples of octagonal pyramidal elevations and a front view and the black parts show the shading of the eleva- these groups formed on the rounded edges of n. S. Ichikawa—Some Notes on Japanese Minerals. 65 tions. Figs. 6 and 7 show the relation between the outlines of the different pits and the edges of the dodecahedron (d); the parallel lines in the figure give the zonal structure of the erys- tals. The former are observed abundantly but the latter very rarely. Figs. 8 to 11 show the relation between the outlines of the various forms of natural pits of very rare occurrence and the edges of the icositetrahedron (7); the parallel lines in the figures show the zonal structure of the crystal. Fig. 12 shows a single rectangular pyramidal pit; also a group of these as commonly observed on the cubic faces. By the above study, it is proved that in the direction of the three principal axes of the crystal, elevations of an octagonal pyramid are formed; in those of the four trigonal inter-axes, pits of a hexagonal pyramid are formed; also that the edges through the three principal planes of symmetry are rounded or grooved. The resulting form of the etching is supposed to be a hexoctahedron. VI. Hlongated Gypsum Crystals. In 1909 I visited Udo, Usagi-mura, Hikawa-gun, Izumo Province, and observed gypsum crystals of unusual length; notes on this gypsum and its use have already been published in Japanese.* The following is an abridged translation with only slight changes. The gypsum of Udo occurs associated with pyrite in a clay in massive (radial-fibrous) and lamellar (parallel-fibrons) forms. The massive gypsum appears as spots in a hard clay and the lamellar forms in a very soft clay easily. pierced by the fingers; the elongated crystals were collected from the latter. Crystal faces observed are the « P, 0 P &,and «© Po,ete. The prism is elongated in the direction of the axis c; individuals measure 5 to 25™™ on the axis 6, and 60 to 2007" on the axis ¢. The erystals are colorless and trans- parent. Inclusions are observed of powdery gypsum, pyrite, diatoms (Cyclotella, ete.), etc.; the diatoms were studied with a magnification of 3850 to 650 diameters. Fig. 1 (see IL) shows a familiar swallow-tail twin ; A isa front view and B a horizontal projection on the vertical axis of A. Fig. 2 shows a specimen in which at one end the same twin separates into three individual. Fig. 3 is a regularly developed twin. A isa front view; B a horizontal projection on the vertical axis; C a cleavage frag- ment. Fig. 4isa hoop of a flat and long swallow-tail twin, arti- ficially bent along the face oP o:; Aisafrontview. If the hoop in A is bent in the direction opposite along the face a Px a hoop like B is the result; in this case thin polished *See my notes in Jour. Geol. Tokyé, vol. xv, p. 509, 1908; vol. xvi, p. 92, 909. Am. Jour. Sci.—FourtH Sreriges, Vou. XLIV, No. 259.—Juty, 1917. 5 66 8S. Ichikawa—Some Notes on Japanese Minerals. i. Elongated gypsum crystals. S. Ichikawa, del. folia (ex. ab, etc.) are sometimes separated. Fig. 5 shows the same twin artificially twisted; the parallel lines in the figure are cleavage fissures parallel to the face P. ; S. Ichikawa—Some Notes on Japanese Minerals. 67 In the above study it is proved that the cleavages of yypsum erystals are in the direction of the faces ~P &, P, o Pa, etc., and the latter are less perfect than the first-named. The flexibility of the crystals is more conspicuous along the face co poo than in other directions. VIL. Dendrites of Manganese Oxide. In 1912 I visited mineral localities in Echizen Province and collected some interesting specimens of manganese oxide in dendritic and circular forms; some of these are illustrated in the accompanying plate (III). Fig. 1 (mat. size) shows dendrites formed in the fissure of a Tertiary siliceous sandstone from Yamanokoshi, Kamiyama- mura, Nanjo-gun. The outlines of the dendrites, as is often the case, much resemble natural views with mountains, trees, etc. | Fig. 2 shows manganese oxide in concentric zonal rings, formed in the fissures of the platy joints of an andesite from Kurashita, Tani, Kitatani-mura, Ono-gun. When fresh the rings are black and yellow, but these colors are gradually bleached on exposure to the air. B and C show a group of concentric zonal rings; the rings marked @ in each figure are yellow and the others black ; b shows plagioclase crystals which form phenocrysts in the rock. A is reduced to one-third ; B and O are natural size. A plate of the dendrites was pre- sented to the Twelfth International Geological Congress in Toronto, Canada, in August, 1913. Fig. 3 shows dendrites formed on calcite erystal from Shimo- shinjo, Shinyokoe-mura, Imatate-gun (magnified 10 times). A, front view ; B, a horizontal projection on the vertical axis. The calcite crystal ‘is colorless and transparent, but its surface is somewhat weathered. In the specimens a group of dendritic forms is sometimes changed to yellow. It is shown here that dendrites of manganese oxide, besides the usual forms, form concentric zonal rings, yellow and black, alternately. Errata.—The following errata are to be noted in my papers in this Journal: Vol. xxxix, April, 1915, p. 459, footnote: for vol. xvi, p. 197, read vol. xvi, p. 1.—P. 463, second footnote: for Jour. Geogr. Toky6, read Jour. Geol. Tokyo.—P. 468, second footnote: for vol. in, No. 13, read vol. ni, No. 12. Vol. xl, August, 1916, p. 115, line 14 from top and also in the second and fourth footnotes: for Jour. Geogr. Toky6, reud S. Ichikawa—Some Notes on Japanese Minerals. 68 ~ Tie S. Ichikawa. del. P. 115. fourth footnote, for 1904 read 1914. , for Imitate-gun, reed Imitate-gun. Dendrites of manganese oxide. oOo Jour. Geol. Toky Pp. 119, last line Kitashinjo-mura, Imitate-gun, Fukui-ken, Japan, December, 1916. \ Vennes— Retardation of Alpha Particles by Metals. 69 Art. VIII.— The Retardation of Alpha Particles by Metats,; by H. J. Vennes. Tue retardation of alpha particles by metals has been inves- tigated by several experimenters, and among the most recent is a series of experiments carried out by Marsden and Rich- ardson.* They found that the amount by which a metal foil reduces the range of alpha particles depends on the part of the range in which the foil is placed. The air equivalent in the foil was shown to be considerably greater when placed directly over the source than when placed near the end of the range. in an earlier experiment, Taylort+ showed that when a layer of hydrogen was used in place of a metal foil, the ionization near the end of the range of alpha particles was greater when the layer of hydrogen was placed directly over the source than when placed near the end of the range. When a metal foil was used, ionization was greatest when the foil was placed near the end of the range. In the experiments carried out by Marsden and Richardson, the scintillation method was used for determining the end of the range. Taylor, in his experi- ments, did not actually determine the range of the alpha parti- cles after passing through the foils and layers of hydrogen, but made experiments on the relative amounts of ionization when they were placed in different parts of the range. The purpose of the experiments carried out by the writer was primarily to test the point discharge method in determin- ing the range of alpha particles after passing through matter. This was done by carrying out experiments, firstly, using the scintillation method, the procedure being almost identical to that followed by Marsden and Richardson, and secondly, using the point discharge method of counting the alpha particles in - place of the scintillation method. The apparatus used in connection with this experiment is shown in the accompanying diagram, and consists of an ordi- nary microscope from which the stage has been removed. The zine sulphide screen is held in place by means of a brass frame, the upper part of which is clamped to the objective of the microscope. The metal foils and source of alpha particles are held in position by two small arms which can be clamped in any position on the vertical rod fastened to the base of the instrument. The point discharge chamber is constructed as shown in the diagram at “5,” and is provided with an arm * Phil. Mag., xxv, p. 184, 1913. + Ibid., xviii, p. 604, 1909. 70 Vennes—fetardation of Alpha Particles by Metals. and clamp so that when in use it can be supported by the ver- ticle rod of the instrument directly above the center of the supports for the foils and the alpha ray source. The micro- scope has a pivot joint near the base, so that it may be swung out of the way when the discharge chamber is used. The dis- charge point is connected to a string electrometer and a poten- Pre. 1: To Elechromerer- (S00 volts Yel tial of about 1500 volts is used on the chamber. Polonium which had been deposited on a small copper plate was used as a source of alpha particles. This source was made very strong so there would be a large number of particles even at the end of the range entering the counting chamber. Measurements were made with aluminum, Dutch metal, silver, and gold foils. The results obtained with these foils by the scintillation method appear to be almost identical with those obtained by Marsden and Richardson, and especially in the case of gold and silver, (ve Vennes— Retardation of Alpha Particles by Metals. x 3 SS R Q ) x JAVON ILE Alf Lyrergerr! rarige rr Cv. 72 Vennes—fetardation of Alpha Partacles by Metals. the air equivalent drops off quite rapidly near the end of the range. The same results were also obtained when the point. discharge method was used. In order that the range may be accurately determined by the discharge point method, it is essential that the point be sensitive enough to respond to the small ionization produced in the chamber when the range of the alpha particle ends imme- diately after entering the chamber. In determining the end of the range by this method, the distance was measured from the lower end of the chamber. 2 The results obtained are shown by the curves plotted in figure 2, the emergent range being plotted as the abscissa, and the air equivalent of the foil as the ordinate. The points obtained by the scintillation method are given by circles, while those obtained by the point discharge method are given by crosses. As shown by the curves, the change in air equivalent near the end of the range is more pronounced in the case of the thicker foils and especially in the case of those of higher atomic weight. It is quite evident that in any investigation where the range of alpha particles must be determined, the point discharge method is as equally well adapted as the scintillation method. In observing scintillations, there is always a certain amount of strain on the eye, and besides a considerable length of time is required for getting the eye adapted to darkness. This disad- vantave is altogether eliminated in the point discharge method, for the deflection of an electrometer fiber indicating a discharge is very definite and can be seen without the aid of a dark room. In conclusion the writer wishes to state that this work has been carried out under the direction of Professor Alois F. Kovarik, and much eredit isdue him for the success which was attained. The work was done at the Physical Laboratory of the University of Minnesota. Arnold Hague. 73 ARNOLD HAGUE. Arnoitp Hague, the able geologist and a man of rare per- - sonal gifts, was born in Boston December 3, 1840, and died at his home in Washington May 14, 1917. The immediate ‘cause of his death was cerebral hemorrhage and was undoubt- edly hastened by his recent fall in Albany while attending a meeting of the Geological Society of America. For nearly fifty years he was prominent in the geological affairs of the country. His parents, the Rev. Dr. William Hague, a noted clergyman and writer, and Mary Bowditch (Moriarty) Hague, lived in Boston during his youth. There his education began but later he attended the Albany Academy. James D. Hague, his elder brother, studied mining engi- neering at the Lawrence Scientific School of Harvard, and Arno!d may have acquired from him his taste for geology. At the Shetteld Scientific School of Yale where Arnold Hague graduated (Ph.B.) in 1863 he met as classmate Clarence King, who had much to do with his career. Three years in succes- sion Hague studied in Europe at the Universities of Gottingen and Heidelberg, and the Freiberg School of Mines. While in Bunsen’s laboratory he devoted himself chiefly to chemistry and mineralogy. The spring of 1865 found him in Freiberg, where he met 8. F. Emmons. They were especially congenial, and with the same bent they soon became and continued through life devoted friends and colleagues. Hague in his excellent memoir of Emmons tells much of himself. Indeed, much of that loving tribute to his friend reads like an anto- biography. He writes “I was always ready to lay aside metal- lurgical studies for field geology. Together we took all the week-end excursions with dear old Bernhard von Cotta, visit- ing many parts of Saxony and studying petrology as laid down in that now antiquated text-book, Cotta’s ‘ Die Gesteinlehre ’ (Zweite Auflage, 1862). Many an evening Emmons and I spent together over the map of Saxony, acquiring our initiative experience in geological cartography which later stood us in good service. Both came to realize the influence of Cotta upon our future careers, as he gave us much of his time. In this way, during these few months of German student life, was formed a friendship that always endured.” Hague returned to his home in Boston in December, 1866, and soon received from his friend, Clarence King, an offer to join in the Geological Exploration of the 40th Parallel which King was just succeeding in carrying through Congress with- out the customary delay. Hague lost no time in bringing Emmons to the attention of King, who secured him, at first as 74 Arnold Hague. a volunteer, for the 40th Parallel Survey. Work began on the Pacific Coast in 1867, and the party went thither by way of Panama. The only other available route was by Wells, Fargo and Company’s overland stage, a tedious not to say dangerous, journey. Hague and Emmons had separate parties in the field, and King with his own camping outfit and greater freedom of motion conducted special investigations over the whole region, all parties meeting frequently for conference. The topographic and geologic surveys of a belt 100 miles in width along the proposed route of the Central and Union Pacific Railroad pro- ceeded together from the Humboldt country of western Nevada to the Great Plains east of the Rockies. Field work was finally completed late in the autumn of 1872, but it should be borne in mind that finished topographic maps on which the areal geology was to be shown, as Hague remarks, were seldom in the hands of the geologists till a year after completing the field work. After the completion of the field work the final preparation of the report with its accompanying atlas was accomplished in New York, where Mr. King and his two colleagues worked together and lived in ties of closest friendship. Hague’s first scientific publications, “Chemistry of the Washoe Process” and the “Geology of the White Pine Dis- trict,’ occurred in 1870 when he was 30 years of age. They grew out of his 40th Parallel work and appeared in Volume III of that organization. The great work, Descriptive Geology, of which Hague and Emmons were joint authors, appeared as Volume IJ, in 1877. King published Systematic Geology, Wol tan, 18s: For a comparative study, the 40th Parallel geologists in 1870 visited the Cascade Range. King climbed Mt. Shasta, Hague climbed Mt. Hood and Emmons Mt. Rainier. They observed about these lofty volcanoes the first active glaciers noted in the United States, and, using the lavas collected, Hague and Iddings made a comparative study of the volcanic rocks of the Cascade Range and the Great Basin. In 1877 Hague received the appointment as government geol- ogist of Guatemala and traveled extensively over the republic visiting mines and active volcanic centers. The following year he was engaged by the Chinese government to examine gold, silver, and lead mines in Northern Chine. Congress created the bureau of the U. 8S. Geological Survey in 1879, thus withdrawing Congressional authorization from existing surveys and exploration parties and accomplishing a complete reorganization. Clarence King was appointed the first director, and took the oath of office May 24. Arnold Arnold [Haque. 75 Hague, who had returned to the United States, was appointed a geologist in the U. 8S. Geological Survey July 8, 1879, but did not take the oath of office until April 10 of the following ear. 4 Under the new organization he was sent to Nevada to study the geology of the Eureka district. His report, published in 1893, is Monograph 22 of the U. 8. Geological Survey. In 1883 he was made geologist of the Yellowstone National Park. With the aid of a number of able assistants and specialists the general study of the Yellowstone National Park was completed some years ago and the results published as Monograph 32, part 2, leaving part 1 to be prepared as a final report by Mr. Hague. It is to include a special study of the geysers which engaged his attention for a number of years. This work, his last and greatest, Hague leaves practically complete. Mr. Hague in addition to his larger reports has contributed papers to a number of scientific periodicals, especially to this Journal. Among these may be mentioned the “ Early Terti- ary Voleanoes of the Absaroka Range,” delivered as his presi- dential address before the Geological Society of Washington and “The origin of the thermal waters in the Yellowstone National Park,” his presidential address before the Geological Society of America. His bibliography of scientific papers includes 39 titles, the last. of which is the memoir to his life- long and devoted friend, 8. F. Emmons, published in 1913 by the National Academy of Sciences. Hague was a fellow of the Geological Society of America of which he was president in 1910, of the Geological Society of London, and a member of numerous other scientific societies. In 1885 he was elected to the National Academy of Sciences, of which he was an active member and officer. As a member of the Commission appointed at the request of the U. S. Gov- ernment by the National Academy of Sciences, he had much to do with the plan for our National Forest reserves. Columbia University honored him with the degree Sc.D. in 1901, and in 1906 he received the degree of LL.D. of the University of Aberdeen. He was vice-president of the Inter- national Geological Congress at Paris 1900, Stockholm 1910, and Toronto 1918. Nov. 14, 18938, he married Mary Bruce Howe, of New York. Mr. Hague was not a ready writer nor voluminous, but exact. He aimed more to write well and truly than much. He was a charming host, and there are but few scientific men in America who have had so wide a circle of devoted friends as Arnold Hague. J. 8. DILLER, U.S. Geological Society, Washington, -D. C., May 26, 1917. 76 Scientific Intelligence. SCIENTIFIC INTELLIGENCE. I. CxHEmiIstry AND PHysIcs. 1. Zhe Analysis of Pyrolusite and other Oxidized Manganese Ores.—O. L. Barnesy and Geo. M. Bisyor have devised a very simple and convenient iodometric method for the determination of active oxygen in these ores. A sample of about 0°2 g. of the very finely powdered ore is placed in an Erlenmeyer flask with 10°¢ of a normal solution of potassium or sodium iodide, followed by 5°° of concentrated hydrochloric acid. The flask, covered with a watch glass, is allowed to stand with frequent shaking until the reaction is complete. Under these circumstances iodine is liberated, not only by the higher oxides of manganese, but also by the ferric oxide that is always present in the ores. Then from 0-2 to 0°5 g. of powdered sodium tartrate is added and the solution is diluted to about 150°. This reagent prevents the subsequent precipitation of iron by sodium bicarbonate, which is now added in small portions until a considerable excess is present. In the resulting solution the iron is changed back by the free iodine to the ferric condition. The remaining free iodine, corre- sponding to the higher oxides of manganese, is now titrated with standard arsenite solution, using starch as an indicator.—/our. Amer. Chem. Soc., xxxix, 1235. H. L. W. The Life of Robert Hare, an American Chemist; by Epear Faus SmituH. 8vo, pp.508. Philadelphia, 1917 (J. B. Lippincott Company).—The present-day student of physics and chemistry may probably know that Robert Hare of the University of Penusylvania invented that very important piece of apparatus, the oxy-hydrogen blowpipe, and by its aid succeeded in melting considerable quantities of platinum, and showed that practically all of the hitherto refractory substances could be melted by the use of this powerful means of heating. The student may have learned also that Hare’s ‘“ deflagrator ” was a galvanic battery of much importance for currents of great heating power in the times preceding the modern development of electric generators. While these well-known facts indicate the remarkable experimental ability of this early American chemist, the present biography shows him to have been a’'man of varied and high attainments in science and of admirable personal character. Provost Smith has produced a very attractive and interesting biography of his eminent subject. He admits that he has become an enthusiast in regard to him and consequently, while he has allowed Hare himself to tell much of his story, largely through previously unpublished letters and other documents which were buried in forgotten journals and pamphlets, he has evidently devoted a vast amount of research to his task, and his comments show much admiration for his hero. Chemistry and Physics. 1G One of the most interesting features of the book is the presen- tation of a large number of letters between Hare and Benjamin Silliman, the elder, the founder of this Journal. This correspond- ence records a close and sincere friendship between these two early American seientists, which began in the time of their youth when they studied and lived together in Philadelphia in 1802, and continued until the time of Hare’s death in 1858. They con- sulted each other in the most intimate way in regard to their _work, and the letters have an interest similar to those of Liebig and Wohler. It should be mentioned also that the greater part of Hare’s scientific publications appeared in this Journal under Silliman’s editorship. Hare was born in 1781, and was about two years younger than Silliman. The historical position of these men may be shown by the statement that Silliman, Davy and Berzelius were of almost exactly the same age. Only afew of Hare’s achievements can be alluded to here. It is interesting to find that his best known invention, the oxy-hydrogen blowpipe, was made when he was only 20 years of age. It was on account of this invention that the Rumford medal, granted for the first time, was conferred on him in 1839. It appears that he was the first experimenter to convert charcoal] into graphite by heat, that he was the first to isOlate metallic calcium, and that he produced calcium carbide and obtained acetylene from it, although the latter was not recog- nized by him. He was a wonderful experimenter, particularly interested in electricity, and his experiments before his classes were performed usually on a large scale and in a most impressive manner. | The book contains much of Hare’s theoretical discussions. Many of the theories advocated by him: have not survived, but. others are still important. He was fond of argument, wrote long letters to such celebrated men as Berzelius and Faraday criticiz- ing their views, and was regarded highly enough by them to receive their elaborate replies, which add much interest to the book. Three portraits of Hare, here presented, show him to have been a man of imposing appearance. H. L. W. 38. A Course in Food Analysis; by AnpDREw L. Winton. 8v0, pp. 252. New York, 1917 (John Wiley & Sons, Inc.).—The purpose of this book is to provide a laboratory course compris- ing about 40 periods of work. The author suggests that the course may be used as a substitute for the usual training in inor- ganic analysis, since it presents more variety in methods and is perhaps of more general interest and of greater practical impor- tance to many students. As the author has had a vast amount of experience and is one of the highest authorities in this line of _work, it is found, as would be expected, that the best methods and their most important applications have been presented. It may be added that the operations are very clearly and fully described. The examination of a great variety of products is presented, with due attention to the detection of various preserva- 78 Scientific Intelligence. tives, substitutes and adulterations. An important feature is an excellently illustrated chapter on the microscopic examination of vegetable foods, a subject of which the author has made a special study. H. L. W. 4. A Text-Book of Sanitary and Applied Chemistry ; b K. H. 8. Barney. 12mo, pp. 394. New York, 1917 (The Mac- millan Company).—This is the revised fourth edition of a book dealing very satisfactorily, in a general way, with the chemistry of water, air and food. While the book is intended primarily for the use of students, the scope is rather popular, so that it may be highly recommended for the use of general readers who desire information in regard to the chemistry of the necessities of life, especially of foods. The composition of all the important articles of food is given, and much good advice is imparted concerning the proper balancing of rations. The book makes no attempt to describe the methods of quantitative analysis, but many experi- ments are supplied for the use of students. Most of these experiments are qualitative and simple in their character, but they are well selected and instructive. H. L. W. 5. The Nature of Solution ; by Harry C. Jones. Pp. xxiii, 380. New York, 1917 (D. Van Nostrand Co.).—The present volume was written by Jones during the last summer of his life and put into the hands of the printer, but later he withdrew it from publication. After his death his friends and colleagues decided to issue it as a memorial volume. The text proper is pre- ceded by a full and accurate bibliographical sketch written by E. Emmet Reid who also surpervised the bringing out of the volume. There are also brief tributes by Professors Arrhenius, Ostwald, and Woodward. The frontispiece is a reproduction of an excellent autograph photograph of the author. In his preface Jones says: “The present work is not a text- book, but a general discussion of some of the more important properties of solutions, true and colloidal. It is therefore written in a nonmathematical, indeed, largely in a semi-popular style.” The first chapters deal with the importance of solution and the historical development of the earlier views as to the nature of solution. These are followed by chapters on osmotic pressure, on the relations between solutions and gases as demonstrated by Van’t Hoff, on Arrhenius’ theory of electrolytic dissociation, on freezing-point depressien, etc. The twelfth chapter comprises a lucid and extended account of the phenomena presented by col- loidal solutions. The last two chapters deal with the newer hydrate theory and the solvate theory of solutions, for the final development of which Jones is almost entirely responsible. The author and subject indexes are immediately preceded by a com- plete bibliography of articles and books written by Jones and his coworkers. In the opinion of the reviewer this is the best of Jones’ lite- rary efforts, since the entire text forms an extremely well- balanced whole, since the ideas and arguments succeed one another Chemistry and Physics. iS) with perfect smoothness, and since the perspective is unusually broad. If it be not inappropriate in this place, the writer of this inadequate notice desires to add emphasis to the remarks made by Reid concerning the cordiality always shown by Jones to his students and friends, since he belonged to the latter group from early boyhood and since he later had the honor of being one of Jones’ students and assistants on the Carnegie Foundation. Ey St, 6. The Theory of Measurements ; by Luctus Tutrie. Pp. xiv, 303, with 66 figures. Philadelphia, 1916 (The author).—In writ- ing this text the author has kept in mind the needs of the student of mathematics as well as those of the student of physics. No knowledge of trigonometry, however, is presupposed, and none is imposed upon the reader of the book. In addition to the statements of facts and theory each of the chapters of the book includes directions for actual experimental work to be performed by the student, and the amount of this work has been so planned that each lesson will require about three hours for the pupil of average skill and ability. Although the field covered is comparatively small the subject matter is taken up in elaborate detail. ‘This may be seen at once from the following list of topics: weights and measures, angles and circular functions, significant figures, logarithms, small magnitudes, the slide rule, graphical representation, curves and equations, graphic analysis, interpolation and extrapolation, codrdinates in three dimensions, accuracy, the principle of coinci- dence, measurements and errors, statistical methods, “ deviation ” and “ dispersion,” the weighting of observations, criteria of rejec- tion, least squares, indirect measurements, and systematic and constant errors. ‘The index follows an appendix of physical and mathematical tables which have been prepared with extreme care. Notwithstanding the fact that the material is presented in great detail the text has not been “ padded” or overexpanded. It has been carefully graded and many parts may be omitted if the student is already conversant with them. This book undoubt- edly merits the attention of all earnest teachers of elementary physics and mathematics since it contains a wealth of valuable pedagogical material and since it is admirably designed to cause the student to think for himself ina clear, concise, logical manner. Ba Ss, U. 7. Laws of Physical Science ; by Enwin F. Norturup. Pp. vii, 210. Philadelphia, 1917 (J. B. Lippincott Co.).—This vol- ume is designed as a reference book on the general propositions or laws of physical science. ‘The material is systematically arranged in six parts pertaining respectively to I Mechanics, II Hydrostaties, Hydrodynamics and Capillarity, III Sound, IV Heat and Physical Chemistry, V Electricity and Magnetism, and VI Light. Whenever doubt arose as to whether an important fact could be classified as a law “a policy of inclusion has been fol- lowed in preference to one of exclusion.” The manner of pre- 80 Scientific Intelligence. ~ senting a law may be readily seen from the following typical illustration: ‘“Coulomb’s Law. The electric intensity of a point p close to the surface of a conductor surrounded by air is at right angles to the surface. It is equal to 470 where o is the surface density of the electrifica- tion. If the surface of the conductor is in contact with a dia- electric of specific inductive capacity K, then the electric intensity at the point p is, tei ee (Thomson, Elements of Electricity and Magnetism, pp. 36, 122.)” The text proper is followed by a bibliographical list of authors, reference books, and journals, and by an index. The publishers have taken pains to make the volume as convenient as possible by using clear type, matt paper, and limp leather binding. Although the book is useful and much can be said in its favor, nevertheless it seems desirable and fair to point out a few of its general and specific defects. In the first place, a reference book of this kind should be as full and complete as possible. It ‘ails in this respect since no mention is made of the laws and funda- mental phenomena of radio-activity (save only the “heat pro- duced by radium’”’), spectroscopy, and X-rays. Again, with regard to details, there is room for improvement. (a) The formula given (p. 187) for Newton’s rings is a purely mathematical relation between the sagitta, the associated semi- chord, and the radius of a circle, when the square of the sagitta is negligible. The formule for the bright and dark rings are not even suggested. It would be just as fair to imply that 12 Y : “ie eee “4 — —_” js the formula for a thin lens, since it is often used as 2R a lemma in deriving the lens equation by the wave-front method. (5) The statement of Fermat’s principle of least time (p. 167) refers only to the minimum and thus gives no clue to the cases involving maxima of time. (c) There are two unfortunate typographical errors on page 51. The mass per unit length of a string or wire is specified as “grains per cm.” instead of grams per cm. The formula for the frequency of a transversely vibrat- ing cord has 21 omitted under the symbol 1. H. 8. U. II. Grotogy anp Naturaut History. 1. United States Bureau of Mines ; Vax H. Mannine, Direc- tor.—Recent publications from the Bureau of Mines (see earlier, vol. xlili, pp. 86, 87) include the bulletins whose titles are given below ; also a series of Technical Papers and Miners’ Circulars. It is announced that owing to the expense involved in the pre- paration and publication of the bulletins and the limited printing funds available, it has been necessary to place a definite Geology and Natural History. 81 price on each bulletin (usually 25 or 30 cents). Orders should be addressed to the Superintendent of Documents, Government Printing Office, Washington, D. C. Buuietins: No. 107. Prospecting and mining of copper ore at Santa Rita, New Mexico, by D. F. MacDonaup and CHAR zs Enzian. Pp. 122; 10 pls., 20 figs. No. 109. Operating details of gas producers, by R. H. Frernaup. Pp. 74. No. 111. Molybdenum ; its ores and their concentration, with a discussion of markets, prices, and uses, by EF. W. Llorron. Pp. 132; 18 pls., 2 figs. No. 119. Analyses of coals purchased by the Government . during: the fiscal years 1908-1915; by G. S. Pops. Pp. 118. No. 121. The history and development of gold dredging in Montana; by Hennren JENNINGS; with a chapter on placer min- ing methods and operating costs, by Cuartes Janry. Pp. 63; No. 122. The principles and practice of sampling metallic metallurgical materials, with special reference to the sampling of copper bullion; by Epwarp KELLER. Pp. 102; 13 pls., 31 figs. No. 124. Sandstone quarrying in the United States, by OLIVER Bowes. Pp. 1x, 143, 6 pls. No. 125. The analytical distillation of petroleum, by W. F. Ritrman and HK. W. Dean. Pp. 79; 1 pl., 16 figs. No. 126. Abstracts of current decisions on mines and mining reported from January to April, 1916; by J. W. THompson. Pp. xi, 900. No. 128. Refining and utilization of Georgia kaolins, by Ira EK. Sproat. Pp. 59; 5 pls., 11 figs. No. 143. Abstracts of current decisions on mines and mining, reported from May to August, 1916; by J. W. TsHompson. Pp 7-2. An address delivered by Dr. Manning in Washington, May 25, 1917, before the editorial conference of the Business Publishers Association, gives an interesting and instructive account of the present petroleum and gasoline situation in this country, due chiefly to the very large increase in the number of motor vehicles. 2. Canadu, Department of Mines.—Of the many publications issued by the Canadian Department of Mines in recent months, the following should be specially mentioned (see vol. xli, pp. 467— 469, and vol. xlil, p. 84) : (1.) Geological Survey Branch. R. W. Brock, Director. Memorrs.—No. 51. Geology of the Nanaimo Map-Area; by Cuarues H. Crapp. Pp. vii, 135; 13 pls., 10 figs. No. 73. The Pleistocene and Recent Deposits of the Island of Montreal; by J. SransFiELD. Pp. iv, 80; 2 maps, 2 pls., 10 figs. No. 83. Upper Ordovician Formations in Ontario and Quebec ; by A. F. Forrsrr. Pp. viii, 277, vii';.1 colored map, 8. figs. Noticed on p. 438, vol. xlii. Am. Jour. Sci,—Fourts Series, Vou. XLIV, No. 259.—Juty, 1917. 6 82 Scientific Intelligence. No. 84. An Exploration of the Tazin and Taltson Rivers, Northwest Territories; by CHarLtEes CamMseLn. 124 pp., 18 pls., 1 map. No. 85. Road Material Surveys in 1914; by L. Reinecke. Pp. vii, 244 ; 5 maps, 10 pls., 2 figs. No. 86. Iroquois Foods and Food Preparation; by F. W. W AUGH. No. 87. Geology of a Portion of the Flathead Coal Area, British Columbia ; by J. D. MacK enzin. No. 88... Geology of Graham Island, British Columbia ; by J. D. MacKenziz. Pp. viii, 221; 2 maps, 16 pls., 23 figs. No. 89. Wood Mountain-Willowbunch Coal Area, Saskatch- ewan ; by Bruce Rosz. Pp. 103; 1 map, 7 pls., 1 fig. No. 90. Time Perspective in Aboriginal American Culture, a Study in Method ; by E. Sapir. No. 91. The Labrador Eskimo ; by E. W. Hawkes. No. 92. Part of the District of Lake St. John, Quebec; by Joun A. Dresser. Pp. 88; 1 map, 5 pls., 2 figs. No. 93. The Southern Plains of Alberta; by D. B. Dowxrne. Pp. 200; 3 maps, 35 pls., 3 figs. No. 94. Ymir Mining Camp, British Columbia ; by CHartes Waters DryspaLe. Pp. vii, 185; 1 map, 15 pls., 16 figs. No. 95. Onaping Map-area; by W. H. Coxuins. Includes Map 153A. | No. 97. Scroggie, Barker, Thistle and Kirkman Creeks, Yukon Territory; by D. D. Carrnzes. Pp. 46; 1 map, 6 pls., 2 figs. In addition, a considerable number of maps have been issued, some of them in connection with the above memoirs. Museum Burwetins.—No. 23. The Trent Valley Outlet of Lake Algonquin and the Deformation of the Algonquin Water- Plane in Lake Simcoe District, Ontario; by W. A. JounstTon. Pp. 22; 3 pls., 1 map. Bay: No. 24. Late Pleistocene Oscillations of Sea-Level in the Ottawa Valley ; by W. A. Jounston. Pp. 14; 1 fig. No. 25. Recent and Fossil Ripple-mark ; by E. W. Kinp1ez. Noticed on p. 491, vol. xl. No. 26. The Flora of Canada; by J. M. Macoun and M. O. MAtre. ; (2.) Mines Branch. KEuaenr Haanet, Director. Summary Report for the Calendar Year ending December 31, 1915. Pp. Vill, 213; 12 pls., 3 figs. Also numerous separate reports on the production for 1915 of the metals (copper, gold, lead, etc.); iron and steel; cement, lime, clay, etc.; coal and coke. Preliminary Report of the Mineral Production of Canada dur- ing the calendar year 1916. Prepared by Jonn McLuisu, Chief of the Division of Mineral Resources and Statistics. Pp. 25. Annual Report ‘on the Mineral Production of Canada, during the calendar year 1915; Joun McLetsu. Pp. 364. Report on the Building and Ornamental Stones of Canada, vol. iv; Wm. A. Parks. Pp. 333; 56 pls., 7 figs. Geology and Natural [Mstory. 83 Bulletin No. 11. Investigation of the Peat Bogs and Peat Industry of Canada, 1913-14; by AterH AnRep. Pp. xii, 185 ; 92 pls., 66 figs., 69 maps. Feldspar in Canada; by Hucu 8. pe Scumip. Pp. vill, 125, XXill, 22 pls., 12 figs.,2 maps. The output of feldspar has increased from 700 tons in 1890 to 19,166 tons in 1916. 38. Pennsylvania Glaciation. Lirst Phase: by EK. H. Wit- LiAMs, Jr. Pp. x, 101, 56 figs., Woodstock, Vt., 1917.—The glacial deposits over a strip 300 miles long in Central Pennsyl- vania are unlike normal glacial drift. They more nearly resem- ble overturned local soil and the disturbed shell of bed rock. Most of the exceedingly rare erratics were furnished by floating ice. Dr. Williams emphasizes the conclusion that such drift is the natural result of a first phase of glaciation, since the first gla- cier was of necessity burdenless and passed over an aged, soft, and deep surficial mantle, with no frontal moraine. The aged appearance of the drift is “inherent and not acquired” and may not be used to determine the date of glacial advance. An inter- esting feature of the region is the presence of unaltered anthra- cite immediately beneath glacial gravels. - HEE: G: 4. Nebraska Pumicite; by E. H. Barsour. Nebraska Geol. Surv.; vol. iv, pp. 357-401, 1916.—This paper gives an illumi- nating idea of the vast extent of the Great Plains which have been covered by volcanic ash deposits. It has been found in most of the counties of the state and undoubtedly occurs in all of them. The beds run from 6-10 feet where exploited, but some are 25-30 and even 50-100 are known. ‘The reviewer pauses to note that if a deposit of one foot in thickness covered the state it would equal nearly 15 cubic miles of rock which gives some notion of the enormous amount of material carried out over the plains from the western volcanoes during their period of activity. The beds range in age from the Oligocene into the Pleistocene. The characters of this material and the chemical analyses which have been made of it show it to be a very pure rhyolite tuff, to which the author gives the name of pumicite. Considerable use for it has been found commercially, as an abrasive, a non-conductor and for constructive purposes, some 27,000 tons of it having been mined annually for the past two or three years. Pe NAGE 5. Guide to the Insects of Connecticut; Part III, The Hymenoptera, or Wasp-like Insects ; by Henry Lorenz ViERECK, with the collaboration of A. D. MacGririvray, C. T. Brurs, W. M. WuHeEeEveErR and §. A. Ronwer. Pp. 824, with 10 plates. Bulletin 22. State Geological and Natural History Survey, Hartford, 1916.— This extensive work consists of systematic descriptive keys to all the families, genera and species of hymenoptera at present known from the State of Connecticut and the adjacent regions. A total of 2411 species, of which 126 are new to science, have been included. More than eleven hundred of these have actually been collected within the state. They are represented by 634 genera and 86 families. 84. Scientifie Intelligence. ~ The codperation of the group of widely known experts repre- sented in the authorship of this work has produced an authorita- tive monograph of the greatest importance to the science of entomology. For, in spite of the fact that this order of insects in- cludes some of our most destructive pests as well as many of the most beneficial forms, it has hitherto been impossible for even the trained entomologist to identify many of the species. This report now makes the identification of genus and species possible to the general student of insects. The publication of such a volume as this should be a source of much gratification not only to the authors and to other entomologists but to the citizens of the State under whose auspices it has appeared. WRG, 6. The Biology of Twins (Mammals); by Horatio Hackrrr Newman. Pp. xiv, 185. Chicago, 1917 (University of Chicago Press). —The writer bases his discussion of this interesting sub- ject on his own researches on the process of twinning in arma- dillos. He presents evidence to support the generally accepted view that twirs in sheep, cattle and man are sometimes produced by the fertilization of two distinct eggs, while in other cases they may result from the division of a single egg or embryo. The latter are the so-called identical twins. The conditions found in twins help to elucidate some of the important biological problems connected with heredity, sex, and general development. WR. iC! 7. The Theory of Evolution, with Special Reference to the Evidence upon which it is Hounded; by Witit1AM BERRyYMAN scoTr. © Pp. ‘xiv, 183." New York, 1917 (The Macmillan Com- pany).— This volume consists of six lectures designed for presenta- tion before a popular audience. The principal evolutionary doc- trines are explained and as critically examined as the brief course of lectures will permit. The evidences of evolution as supported by comparative anatomy, embryology, blood tests, paleontology, geographical distribution and experimental work are logically presented and in sufficient detail to give the general reader a good idea of what the theory of evolution stands for at the present time. WwW. R. C. 8. A Chemical Sign of Life; by Surro Tasutro. Pp. ix, 142. Chicago, 1917 (University of Chicago Press).—In this little volume the author discusses irritability as a sign of life, and explains the relation between irritability and metabolism resulting in the production of carbon dioxide. By means of an ingenious apparatus sufficiently small quantities of this gas can be detected to determine whether a single seed, a nerve fiber, or any plant or animal tissue, still possesses the irritability characteristic of life. Tence the test for life is the capability of.c :rbon dioxide formation, and the quantity of life present can be measured by the relative amount of this gas produced in a given time. The “biometer,” by means of which these tests are made, is fully described in an appendix. W. RB. C, Geology and Natural History. 85 9. Fundamentals of Botany; by C. Stuart GAcER. Pp. xix, 640, with frontispiece and 434 text-figures. Philadelphia, 1916 (P. Blakiston’s Son & Co.).—Dr. Gager’s text book presents so many interesting features that it is difficult to call attention to them in a brief review. In his opinion the purpose of an intro- ductory course in any subject is not so much to prepare the stu- dent for advanced courses as to introduce him into a new realm of thought. With this end in view he emphasizes certain phases of botany which are not usually taken up in introductory works. After deseribing the nature of the science, and the various fields of botanical activity, he gives a short account of plant organs and of plant cells. He then proceeds at once to a consideration of the functions of plants, taking up such subjects as the loss and absorption of water, nutrition, respiration and growth. He then discusses the structure and life histories of typical plant-forms, beginning with the fern and taking up in order mosses, liver- worts, algae, and fungi. Then, in an ascending series, he con- tinues with the horsetails, lycopods, cycads, conifers, and angio- sperms. Into these discussions he introduces many collateral topics. Some of these illustrate theories or generalizations and others call attention to matters of economic importance. In con- nection with the ferns, for example, alternation of generations, reduction, inheritance, variation, and adjustment to environment are clearly presented ; ‘while, in connection with the fungi, the use of these plants as food, the diseases which they cause, the nature of fermentation, and the significance of bacteria to ‘the human race are among the subjects ‘considered. The concluding chapters of the book are devoted to such subjects as evolution, Darwinism, and heredity. Throughout the volume the author lays especial stress on the historical development of botany and introduces portraits of eminent workers, calling attention to the definite ser- vices which they have rendered. In many places he illustrates his text with original outlines or diagrams ; these and the excel- lent text-figures, many of which are new, deserve high commen- dation. The book on the whole represents a distinct contribution to botanical pedagogy. KW Ht 10. A Laboratory Guide for General Botany ; by C. StuaRtT GaGER; pp. vill, 191. Philadelphia, 1916 (P. Blakiston’s Son & Co.).—The directions given in this helpful work are unusually explicit, their purpose being not only to help acquaint the student with botanical facts but also to teach him how to observe and how to record his observations. In many cases his knowledge is tested by suitable questions. ‘The order of topics is the same as in the author’s Fundamentals of Botany, although the book could easily be used in connection with other texts. AS Wok, 11. Laboratory Manual of Agricultural Chemistry ; by CHARLES CLEVELAND Hepexs and Witiiam THoreAv Bryant. Pp. x, 94, with frontispiece and 8 text-figures. New York, 1916 (Ginn & Company).— This little book is designed to meet the special needs of students in agricultural chemistry and presupposes a 86 Serentifie Intelligence. ~ knowledge of general chemistry. It describes a long series of experiments, the significance of which is brought out by appro- priate questions. The intr oductory experiments deal with methods of quantitative analysis and these are followed by definite appli- cations to the analysis of feedstuffs, of soil, of insecticides and fungicides, of milk and of water. In the hands of a capable teacher the book should yield excellent service. aD 12. Manuring for Higher Crop Production; by KE. J. Rus- sELL. Pp. iv, 69, with 16 text-figures. 1916 (Cambridge Uni- versity Press).—Although written principally for the use of farmers in the British Isles the present work gives information of much value to farmers in general. At the same time emphasis is laid on the impossibility of giving advice which will hold good under all circumstances. After an introductory chapter on the improvement of the soil, natural and artificial manures are de- scribed, and the methods of manuring arable and grass lands are discussed at length. The book is based largely on actual experi- ments carried out at the Rothamsted Experimental Station at Harpenden, of whieh the anthor is director. A. W. E. 13. A Manual of Organie Materia Medica and Pharmacog- nosy ; by. Lucius E. Sayre. Fourth edition, revised. Pp. xviii, 606; 4 pls., 302 figs. Philadelphia, 1917 (P. Blakiston’s Son & Co.).—This work deals concisely and thoroughly with the sources, characteristics and constituents of drugs of vegetable and of animal origin. It has been extensively recast. Not only is it brought into conformity with the new U. 8. Pharmacopeia IX, but the newer botanical classification (leading from Cryptogams to Composit) is followed. Besides the elaboration of the mat- ter relating to inorganic drugs, chapters on therapeutic action and serotherapy have been added. These contain valuable matter but a good deal of antique therapeutic superstition is fostered (e. g. the employment of gold salts, sarsaparilla, etc.). Plant histology has been largely omitted, the reader being referred to Stevens’ “ Plant Anatomy.” Sayre’s new edition will attract and be of high service to those who are interested in the study of drugs. H. G. BARBOUR. OBITUARY. Prorgessor Greorce Harcoop Strong, formerly a member of the Faculty of Colorado College, died on February 20 at the age of seventy-five years. He early studied the glacial geology of Maine and in 1881 came to Colorado Springs where he resided for most of the remainder of his life. He was active as a mining geologist and was especially interested in the geology of the Pike’s Peak region. Proressor H. F. E. Juncrersen, the Danish zoologist who made important contributions to the knowledge of the fauna of Greenland, died recently at the age of sixty-three years. He was particularly interested in the Danish expeditions concerned with the investigation of the North Atlantic and the Polar seas. na! ih. * oe “ ~~ Warns Naturat Science EstasisHMENt A Supply-House for Scientific Material. Founded 1862. Incorporated 1890. A few of our recent circulars in the various departments: Geology: J-3. Genetic Collection of Rocks and Rock- forming Minerals. J-148. Price List of Rocks. Mineralogy: J-109. Blowpipe Collections. J-74. Meteor- ites, J-150. Collections. J-160. Fine specimens. Paleontology: J-184. Complete Trilobites. J-115. Collec- tions. J-140. Restorations of Extinct Arthropods. Entomology: J-30. Supplies. J-125. Life Histories. J-128. Live Pupae. Zoology: J-116. Material for Dissection. J-26. Compara- tive Osteology. J-94. Casts of Reptiles, etc. Microscope Slides: J-185. Bacteria Slides. Taxidermy: J-138. Bird Skins. J-189. Mammal Skins. Human Anatomy: J-16. Skeletons and Models. General: J-155. List of Catalogues and Circulars. i ‘ ; ! : Ward’s Natural Science Establishment 3 84-102 College Ave., Rochester, N. Y., U.S. A. a Publishers: WILLIAMS & NORGATE, 14 Henrietta Street, Covent Garden, London, W. C. 4 66 99 ‘ SCIENTIA 4 INTERNATIONAL REVIEW OF SCIENTIFIC SYNTHESIS. Jsswed monthly (each x number consisting of 100 to 120 pages). Editor: EUGENIO RIGNANO. . ‘““SCIENTIA’’ continues to realise its program of synthesis. It publishes articles which relate to the various branches of theoretic research, and are all of general in- terest; it thus enables its readers to keep themselves informed of the general course of the contemporary scientific movement. *“SCIENTIA’’ appeals to the codperation of the most eminent scientific men of all countries. It has published articles by Messrs. Abbot (Washing/‘on) —- Arrhenius (Stockholm)- Ashley (Birmingham) -Bechterew (Petrograd) - Bohlin (Stockholm)- Bonnesen (Kopenhagen) - Borel (Paris) - Bottazzi (Vapoli) - Bragg (Leeds) -Bril- louin (Paris) - Bruni (Padova) - Castelnuovo (Roma) -Caullery (Paris) -Chamberlin (Chicago)-Ciamician (Bologna) - Clark (New York) - Costantin (Paris) -Crommelin (Greenwich) - Daly (Cambridge, U. 8S. A.) - Darwin (Cambridge) - Delage (Paris) - De Martonne (Paris) - De Vries (Amsterdam) - Durkheim (Paris) - Eddington (Greenwich) - Edgeworth (Ozford) - Emery (Bologna) - Enriques (Bologna) - Fabry (Marseille)-Fisher (Vew-Haven, U. S. A.)=Foa (Torino)-Fowler (London) - Fredericq (Liége)- Galeotti (Vapolt)- Golgi (Pavia) - Gregory (Glasgow)- Guignebert (Paris) - Janet (Paris) - Jespersen (Gentojfte) - Kapteyn (Groningen) - Kidd (Oxford)-Langevin (Paris)-—Lebedew (Moscou)-Lodge (Birmingham)-Loisy (Paris)- Lorentz (Haarlem) - Loria (Torino)-Lowell (Flagstaff, U. S. A.)-Maunder (Green- wich) - Meillet (Paris) - Pareto (Lausanne) - Peano (Torino) — Picard (Paris) - Poincare (Paris) - Puiseux (Paris) - Rabaud (Paris) —- Righi (Bologna) - Rignano (Milano)-Russell (Cambridge)- Rutherford (JZanchester)- Sayce (Oxford) -Schiapa= relli (Milano) - Seligman (Vew York) —Sherrington (ZLiverpool)- Soddy (Glasgow)- Svedberg (Upsala) -— Tannery (Paris) - Turner (Ozford) - Vinogradoff (oscou)-Vol= terra (Roma)-Von Zeipel (Upsala) -Westermarck (Helsing/fors)-Willey (Montreal, Canada)-Zeeman (Amsterdam) -Zeuthen (Kopenhagen), and more than a hundred _others. ‘“*SCIENTIA’’ publishes, at present, in the section dedicated to sociological articles, a series of studies on the present questions of an international character raised by the war. ** SCIENTIA’’ publishes its articles in the language of its authors, and joins to the principal text.a supplement containing the French translations of all the articles that are not in French. (Write fora specimen number.) Annual Subscription: 24 sh. post free. Office: Via Aurelio Saffi, 11- MILAN (Italy). Bes ¥2 % D1 AOS eRe Ee en Ee ei > Ren Tt “> mre : rs CONTENTS. Arr. I.—The Motion of Ions and Electrons through Gases; by. 3i.. MW ELUISOH (22 ee ee ee Ii.—Correlation of the Devonian Shales of Ohio and Penn- sylvania; by W. A. VERWIEBE III.—Evidence of Uplift on the Coast of New South Wales, Australia; byl, HW. HaRPMwR oo os 22-248, oe eee IV.—The Use of the Platinized Anode of Glass in the Elec- trolytic Determination of Manganese; by F. A. Goocu and M. KospayasHi1 V.—Preliminary Note on the Occurrence of Vertebrate Foot- prints in the Pennsylvanian of Oklahoma; by W. R. JILLSON SoU ae Se ee eee e- Ww Me Me ew ee eB ewe VI.—New Evidence of a Recent Volcanic Eruption on Mt. St. Helens, Washington; by W. R. JILison --.-.----- VII.—Some Notes on Japanese Minerals; by 8. Icurmawa VIII.—The Retardation of Alpha Particles by Metals; by H. J. VENNES ARNOLD HAGUE SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Analysis of Pyrolusite and other Oxidized Manga- nese Ores, O. L. BARNEBY and G. M. Bisnop: The Life of Robert Hare, an American Chemist, E. F. Smiru, 76.—A Course in Food Analysis, A. I. Winton, 77.—A Text-Book.of Sanitary and Applied Chemistry, E. H. S. BatLEey: Nature of Solution, H. C. Jonses, 78.—Theory of Measurements, - L. Tutrte: Laws of Physical Science, E. F. Norturvup, 79. Geology and Natural History—United States Bureau of Mines, V. H. Man- NING, 80.—Canada, Department of Mines, R. W. Brock and £. HaANnsEt, 81.—Pennsylvania Glaciation, First Phase, E. H. Wixxiams, Jr.: Nebraska Pumicite, EK. H. BarBour: Guide to the Insects of Connecticut, Part IIT, H. L. Viereck, etc., 83.—The Biology of Twins (Mammals), H. H. New- MAN: The Theory of Evolution, W. B. Scotr: A Chemical Sign of Life, S. TasHtro, 84.—Fundamentals of Botany, C. S. GacEr: A Laboratory - Guide for General Botany, C. S. GaGer: Laboratory Manual of Agrieul- tural Chemistry, C. C. Hepcres and W. T. Bryant, 85.—Manuring for Higher Crop Production, E. J. Russextu: A Manual of Organic Materia Medica and Pharmacognosy, L. E. Sayre, 86. Obituary—G. H. Stone: H. F. E. Juncersen, 86. Se AUGUST, 1917. Established by BENJAMIN SILLIMAN in 1818. THE | AMERICAN | JOURNAL OF SCIENCE _Eprion; EDWARD 8. DANA. ASSOCIATE EDITORS ee PROFESSORS GEORGE L. GOODALE, J OHN TROWBRIDGE, W. G. FARLOW anv WM. M. DAVIS, or CamsBringz, Proressors ADDISON E.. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT E. GREGORY anp HORACE S. UHLER, or New Haven, | | | | PROFESSOR HENRY S. WILLIAMS, or IrHaca, : PROFESSOR J OSEPH S. AMES, or Bautrwors, Mr. J. 8. DILLER, or Wasuineron, FOURTH SERIES ; re a Fk HAeE Noe And 2 vi Lax J ~ ~ ro vA VOL. XLIV—-[WHOLE NUMBER, ON GIN. = {G]7 No. 2602 AUGUST: 5 Ce. ar Pr NEW HAVEN, CONNECTICUT. 1917 | THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. faned nae. Six dollars per year, in advance. $6.40 to countries in the Union ; ; $6.20 to Canada. Single numbers 50 cents. as se one nd-class matter at the Post Office at New Haven, Cons ., under the Act \ NOTICE TO COLLECTORS _ I am selling a choice collection of minerals and gems contain- ing 548 specimens. ‘The specimens are of the best quality, very well crystallized and correctly labeled; some specimens are now unattainable. The majority of the specimens are matrix speci- mens. ° They are arranged in a fine mahogany cabinet 64” high, 36'’ wide, 14” deep, with crystal plate glass shelves and mirror back. It makes a beautiful display and is suitable for any home or museum. ‘here are a number of duplicates showing species from various localities. | The following minerals are included: ; Diamonds Kunzites Silver Quartz with en- Rubies Garnets Copper closures. Sapphires Opals: 75: Galena Fluorites Emeralds Tourmalines Pyrargyrite Calcites Beryls ~ Turquoise Proustite Vanadinite Aquamarines Phenacite Spessartite Wulfenite Topaz Diopside Pyrite Petrified wood Spinels Vesuvianite Cyanite Zeolites, only. Zircon Amethyst Amber the finest - Agates Amazonstone | Crocoite Azurite Apatite Epidote Chrysoprase Chalcedony Malachite Chiastolite Labradorite Chrondrodite Carnelian Jaspers Rhodocrosite Lapis lazuli Hematite Rhodolite Jade Sphalerite of Peridote Dioptase Rhodonite different colors Hiddenites Gold Willemite And many other specimens too numerous to mention. The collection cost over $2000; will be sold for $1200. If not interested in the entire collection, I am willing to sell in- dividual specimens. A complete list will be sent upon applica- tion with photograph of the cabinet and the collection. This is a rare opportunity. If you are interested, kindly write at once or you will be disappointed. ALBERT H. PETEREIT 81-83 Fulton St,, | New York City uth Seat a ainda ae wit, Hah a an > et i heat © eA -ib THE pe peotla C AUG as v< AMERICAN JOURNAL OF SCIEN \ \ [FOURTH SERIES.] \4i a ete ' Y —____ $4) _—_—_ Art. [X.—Phystographic Development of the Tarumai Dome in Japan; by HrpEez0 Sroromar (Tanakapate), Tohoku University, Japan. TuHE rise of a new dome on the top of Tarumai volcano in Japan attracted much attention among the naturalists of the world. Though it was not an unusual type of lava eruption, itsphysiographic development is observed better than others of the same kind. As the structure of the voleano and the history of its recent eruption in the year 1909 have been described by many authors, they will be touched upon only briefly here. Tarumai is a flat voleanic cone formed principally of andesitic scorias and ejectas and its site is in the volcanic region of Hokkaido in lat. 42° 41’ 30” N. and long. 141° 21’ 40” E. Towards the south it slopes gradually to the Pacific coast, while to the north a short ridge connects it with an extinct voleano called Hu-uppusi-nuppuri. Both cones rise from the deep water of Sikots Lake, of which the area is 78 sq. km., with surface 298™ above sea-level and bottom about 60™ below sea-level. The volcano has a double crater on its summit and the Somma-wall is much dissected on the southwest and southeast. The eastern side culminates in Higasiyama at the height of 1016™ above sea-level and before the eruption this was the highest point of the volcano. According to Prof. Oinoue, the inner crater, which is now full of the recent lava, was formerly oval in shape, with major axis of 670™ running N.10°W. and minor axis of 550™; its depth was about 80". The circular bottom of the crater had a diameter of 60™, and from some pits in it sulphurous gas continuously issued. Am. Jour. Sci.—Fourts SERIES, Vou. XLIV, No. 260.—Avueust, 1917. G 88 H. Simotomai—Tarumai Dome im Japan. Several eruptions are recorded in history, but they have few direct bearings on the present discussion. From 1896 until the beginning of 1909 the voleano was quiet. In the latter year, between January and March, small activities were noted eight times by the inhabitants of the territory in the form of ash-rain, fire phenomena, detonation, and earthquake. On the 30th of March a tremendous explosion took place, throwing dives; 1, Fie. 1. The location of Tarumai Volcano. out ashes, bombs, lapilli, ete., and therewith a smoke column about 7 km. high rose from the voleano. On April 4th Oinoue visited the mountain and saw a new, deep, fuming pit on the floor of the crater bottom. On April 12th another explosion occurred in full violence, accompanied by ashes, bombs, lapilli of several sizes, and an enormous smoke column. The sur- rounding inhabitants did not notice any change on the moun- tain until April 19th; when, the cloud clearing up, a dark elevation on the summit came in sight. Indeed, a dome of augite- andesitic lava was rising in the inner crater. On April 23d LH. Simotomai—Tarumat Dome in Japan. 89 Oinoue again visited the mountain and the photographs taken on that date show a smooth, round-headed dome which filled the inner crater and rose more than 100™ above it. The pic- tures taken on May Ist show the dome to have become larger and flatter at the top. The trigonometrical measurement of the Oinoue party on May 1st gives the dimensions of the dome. The base of the dome was about circular, covering an area of 152,000sq. m. It was about 200" in height above the floor of the crater bottom and 134" above the lowest part of the inner crater-wall; its whole volume was computed at 20,000,000 cubic meters. The dome was 1046" high above sea-level and 30™ above Higasiyama, so that the voleano had become a little higher than before. On May 15th an explosion occurred on the south- ern foot of the dome and thereby the inner crater-wall was cracked. The new fissure was 3” to 8™ wide, about 18" in visi- ble depth, and 150™ long in the direction of N.60°W., ending directly on the side of the dome. From a point in the middle of the crack, sulphurous gas was issuing in great volume. In the winter of the same year snow covered the dome, though it was still fuming on the surface. After that time seven years elapsed without any further information about the dome. The writer had occasion to visit it three times in the year 1916, and will give here some short observations about the development of the dome, adding some new data to that of my former publication.* Since 1909 the dome has remained without any important change in form, conserving all the main elevations and angles, and most of the lar ge lava blocks scattered in the vicinity had kept their original shape. Talus had developed on all sides, covering the lower three-fifths of the dome. Its maximum development is at the southern half, where the prevailing wind attacks it directly, while the minimum is at the opposite side, where the dome is protected from the wind through the inner erater-wall. The southeastern side of the mountain near the termination of the large fissure shows the least steepness caused by the falling of lava, and it is now not difficult to reach the summit, which was never visited by anybody before and it was doubted whether there existed any crater or not. From out of the talus arises the dome proper, showing steep cliffs of lava in its sides which are very instructive. The superficial lava layer does not form a continued crust but is broken by vertical fissures into many irregular parts, some of which are fallen down between others as wedges. Each lava-mass is divided into numerous layers, feldspathic, iron-rich, etc. These layers must originally have been parallel * Zeitschrift, Gesellschaft fiir Erdkunde, Berlin, 1912, page 433. 90 H. Simotomai— Tarumai Dome in Japan. to the cooling surface. The thickness of the layers varies from a tew centimeters to half a meter. The sulphurous gas is still. issuing from numerous parts around the lava cliff, though it is now in minimum activity. The upper surface of the dome is practically flat, but it is so rough and uneven with sharp edges, spears, and spines of yel- lowish- gray lava that walking on its surface 1 is almost impossi- ble. Such slagey, porous lava was seen everywhere around Fig. 2 Fic. 2. Tarumai before the eruption of 1909. Seen from south from the coast of the Sikots Lake with the old pier in front. the dome at the beginning of the eruption, but now one can scarcely find any on the sides, for there this phase of the lava has fallen away and become buried in the talus. The maximum elevations are at the southwest and north- east sectors on the upper surface, and between these two heights a fissured zone runs from southeast to northwest. Many cracks of different sizes run nearly in radial directions from the center. In some cases the intervening mass between pairs of the fissures is dropped down as a wedge, forming a trench-depression (Graben). Many cracks are stillopen. ‘These features of the upper surface are like those seen on the sides H. Simotomai—Tarumaz Dome in Japan. 91 of the dome. One of the large trenches begins at the south- eastern periphery of the top-area, a little northward from the termination of the fissure on the inner crater-wall, and runs to N.60°W., in the same direction. | Towards the center of the dome the trench gradually loses its character, but at the other end it is 50™ wide and is bounded on both sides by steep slickensided cliffs which are locally about 20" high. Within this depressed zone, many minor fissures run in the same direction; some of them are open and IPGL, Bis * Fie. 3. Tarumai after the eruption of 1909. Seen from about the same place as fig. 2, with the new pier in front. about 10" deep. Another large depression runs from the cen- ter of the top surface and goes to N.N.W., and its breadth is about 60™. Its maximum depth lies a little distant from the center and forms a long, oval basin about 20" wide, and its bottom is 80™ under the highest point of the dome. Moderate amounts: of stlphurous gas are issuing through these numerous fissures. Prof. Kusakabe in Sendai, who visited Tarumai in August with us, reported to me that the photographie films which I took on the dome for his party 92 Il. Siemotomai— Tarumai Dome in Jupan. gave positive instead of negative images when developed by the normal process. This curious phenomenon was afterwards experimentally proved by him to have been caused by a momentary attack of the sulphur dioxide on the films. The exact date of the formation of these fissures is unknown, WML Map and section of Tarumai, 1916. Fic. 4. In the map the meridian is marked by a lateral edge. The sche- matic section passes from east to west, including the highest point in the somma wall (1016 m.). Scale, 1: 150,000 approx. but in photographs which were taken on the 23d of April such depressions are not visible, while in those taken on the 1st of May, 1909, we can recognize, on the eastern side of the dome, some traces of the termination of the trench just as we now see it. LH. Simotomai—Tarumai Dome in Sapan. 93 HIG. D. Fic. 5. Slaggy lava surface of the dome with cracks in front. Photo- graph by the author, October 20th, 1916. Fic. 6. Fie. 6. Tarumai dome from the S.E., ata point in the atrio of the vol- cano. Photograph by the author, October 20th, 1916. 94 H. Simotomai— Tarumat Dome in Sapan. In October, 1916, we carried a small transit and a mountain barometer to determine the exact height of the dome. For the -measurement of the heights, I took one point common with the former survey as a base, that is, the triangular point of Higasiyama, 1016" high above sea-level. From this base, I took the readings of the several characteristic points on the dome, upon which the former surveyor and I had agreed. This method gave the same result as that furnished by the mountain barometer. The exact height of the dome is 1035". Comparing this measurement with that made by Oinoue on the Ist of May, 1909, my result is 11™ less, so that we can be sure that the dome has sunk a little since the 1st of May, 1909. But this sinking of the dome surface ended probably before the summer of 1909, for since that time it has remained practically the same in size and the shape, so far as I remem- ber it. On the slaggy surface of the dome, much fragmental material, similar to that seen all around the base, is scattered. This material might be regarded as having been thrown up by the explosion on the 15th ‘of May, 1909. In August, 1916, I found many dead cicadas (Cicadina bi- hamata Motsch.) on the dome, which perhaps were traversing the mountain and were killed by the sulphurous gas. It is of further interest that three species of plants are found on the dome surface; these are: 1, Pentstemon frutenscens Lam.; 2, one species of moss resembling Deeranum sp.; 3, another species of moss resembling Pognatum sp. The two latter have not yet been classified by species. The first of them was determined by Prof. Miyabe in Sapporo. It is very common on the high mountains of northern Honsyt and Hok- kaido, and it was “first found in Japan on Tarumai; so the name “ Tarmai- So,” or grass of Tarumai, is applied by ‘him. It is not strange that this plant, whose dry seeds weigh only about 0:002 gram, has migrated to this dome from the outer slope of Tarumai, but it is remarkable that the develop- iment succeeded in only seven years on the still fuming lava dome. The smoke is continuously issuing with much violence from the same point of the main fissure on the inner crater wall as before. In January of this year, under the smoke column, the dome stood covered with snow except in the middle belt, where the cliffs show their maximum steepness, and also at several fum- ing spots on the top. Summary. 1. The activities preliminary to the 1909 lava eruption con- tinued about four months, from the middle of January to the il, Simotomai— Turumar Dome in Japan. 95 Hieoe 7, Fie. 7. View from a pointin the large trench on the top surface of the dome. Inthe background, the somma-wallis seen. The large fissure, trav- ersing the inner crater-wall, was formed on May 15th, 1909. In front is seen the steep cliff of the south wall of the trench. Photograph by the author, October 20th, 1916. Fic. 8. Fic. 8. Slickened surface on the north wall of the large trench on the top of the dome. Photograph by the author, October 20th, 1916. 96 H. Simotomai— Tarumai Dome in Japan. middle of April, and it ended with the great explosion on the 12th of the same month. 2. The lava followed this, and on the 19th of April it accumulated in the crater so much as to be seen from a great distance, and on the 23d of April the growth of the dome was not yet completed and it was smooth and round-headed. 3. On the Ist of May, 1909, the maximum growth of the dome had already been reached, though we know neither the exact time of its completion nor the exact shape at that time. Fic. 9. C Fic. 9. Dome seen from the trigonometrical point on Higasiyama (1016 m.) ; tracings from photographs to the same scale. A, view taken on the 23d of April, 1909, by Prof. Oinoue; B, view, on the 11th of May, 1909, by Prof. D. Sato; C, view, on 19th of October, 1916, by the author. The surface of the dome became very uneven through lateral expansion, which was caused by the further accumulation of lava. On the other hand, the sinking of the dome began through the settling of the lava mass, and it probably ended soon after the above date. Such depression after the maxi- mum growth of the dome must have produced the change from a small, round shape to the large, flat-headed one. 4. As the parallel layers in the lava masses show, the mode of growth of the dome was concentric, as suggested in fig. 4, and the upper crust is split into irregular sections by the sinking of the central part of the dome. LH. Simotomai—Tarumai Dome in Japan. 97 5. The vent through which the lava rose is probably at the most 60™ in diameter, because the lower part of the crater pit measured that width before the eruption. According to this supposition, the lava must have consolidated more rapidly in the narrow vent than the nucleus of the dome, so that the connection of the viscous lava between the dome and the sub- terranean deep chamber was cut off. This accounts for the fact that the settling of the lava-mass of the dome (20,000,000 cubic meters), due to consolidation, was comparatively less than in the case of Usu after its eruption in 1910, when connection with the subterranean chamber was probably not so quickly cut off. It is impossible to calculate exactly the diminution of the lava mass by settling, but a rough estimate gives about 5-10 per cent of the whole volume. 6. The principal fissure formed on May 15th, 1909, on the inner crater-wall just at the foot of the dome, and it opened a way for the gas which issued from the consolidating lava in the central part, thus tending to remove the danger of the dome’s destruction through gas explosion. The endogenetic development of the dome was completed in about one month. 7. The formation of the rock talus continues slowly and the principal epigenetic cause is wind action. 8. During seven years, some plants have migrated to the dome on the still-fuming surface of lava. 9. Tarumai is the representative type of this kind of dome. Literature of the recent eruption of the Tarumai volcano. Immanuel Friedlinder: Ueber einige japanischen Vulkane, II Teil, Mitteilungen der Deutschen Gesellschaft fiir Natur- und Voelkerkunde Ostasiens. Bd, XII, Tokyo, 1910. — Ueber den Usu in Hokkaid6é und ueber einige andere Vulkane mit Quellkuppenbuldungen, Petermann’s Geo- graphische Mitteilungen, Juni Heft, 1912. Bunjiro Koté: On the volcanoes of Japan II, Journal of the Geological Society of Téky6, vol. xxiii, No. 269, 1916. Synkusuke Kozu: Preliminary notes on some igneous rocks of Japan, Journal of Geology, xix, No. 7. Yositika Oinoue: Report on the Tarumai eruption (Japanese), Publications of Imp. Earthquake Investigation Comm., No. 64. Sidney Powers: Volcanic Domes in the Pacific, this Journal, xlii, No. 249, Sept., 1916. Denzo Saté: Report on the Tarumai eruption (Japanese), Publi- cation of the Imp. Geological Survey of Japan, No. 14, Toky6, 1909. Hidezo Simotomai: Der Tarumai Ausbruch in Japan 1909, Zeit- schrift der Gesellschaft fiir Erdkunde zu Berlin, Heft 6, 1912. $8 ~ Waring—Lavas of Morro Hill and Vicinity. Art. X.—Lavas of Morro Hill and Vicinity, Southern California; by Grratp A. Warine and Ciarence A. W arine.* Morro Hill is in northern San Diego County, in the foot- hills region of southern California. It is about three miles Tevet 1 RIVERSIDE 850 Pam SPRINGS ake Elsinore 7220 S WiLDOMAR t ve g MARTIWEZ MT, Sces39 g oFALLBROOK 732 BN sence soc cuca cHoaaSenenece Stee bsqgnccce SStess \cogsseaasesccsssosesc ss es N SCpaLomar MT : ‘ E. “S26 : y a eN Co S 340 a MoRRS WARNER SPs / DB /72 | S Hitt ) BoNsALL aR 2 Ne iF iS x > :< se > a > 6 Esconpibo OsF e pria Yst° Gi, 78s, JULIAN J © ¢200+ DIEGO 6428». CUYaMaca Kes. XP 4500 RY MAP ig aS 6S/S o 979. (000213) 2EVOOS. 079.4 578 0:00151 3°591 11°57 578 0°00142 1°632 3:171 578 0:00147 fas 4°88 578 0°00141 0°643 1°222. 578 0:00144 0:975 3°164 578 0°00145 0°3451 0°672 578 0°00148 0°4104 1-352 578 > 0700147 2573 Ool> 9078 “000152 0°2052 0°830 578 (0:00181) mean 0°00145, mean 0'00140, TABLE VI. TaBLe VII. Lanthanum Iodide. Zine Potassium Iodide, ZnI,.KI. ei a = 0°1289 molar ZnI,+ a= 0 02179 molar Lals. 0°1239 molar KI. (21) + (213) = 3a = 0:0654 molar. (31) + (21) = 0°3717 molar. b g Cc ky b g ‘ c Ky 28°79 6600 650 0:00138 | 45:4 . 105°3- 585 0°00130 18:94. 333°0 610 0:00139 24:10 586 580 0:00146 10°29 151°5 580 0:00144 10°04 27°45 580 (0°00171) 5°98 80-7578) 50-00142 742 16°42 578 0:00141 3°648 47°9 578 0°00144 S345 clr 578 0-00137 cig tee eee EERE SD Meola terse 2518.3 0.00140 : 0933 2:°031 578 0:00141 07605 1°331 578 0:00142 mean 0°00139 6 Discussion of Results. Considering first the results obtained with iodides, as given in Table III to VII, we find two different types of behavior. Strontium iodide, nickel iodide, zine iodide, and lanthanum jodide behave “normally,” that is, they give a value of the equilibrium constant A, which agrees with the value given by the iodides of the alkali metals. The nickel iodide, zine iodide, and lanthanum iodide solutions contained a small amount of hydrogen iodide, but since the calculations in the case of both of these solutions were based upon the total iodide concentra- tions, and since hydrogen iodide itself falls in the normal class, it is reasonable to conelude that the presence of this hydrogen iodide can not have had any appreciable effect upon the results. This inference is supported by the fact, shown in Table VII, that when zinc iodide and potassium iodide are present together 114 3. *8°16 2°540 1°397 0°646 0°3104 0°1411 ite Van Name and Brown—Tri-lodide and TasuE WAitk Cadmium Iodide. a = 0°5 molar Cdl. (21) + (21s) = 2a = 1:0 molar. g 384°1 196°8 95°9 70°3 40°2 21°86 11°54 a = 0°25 molar Cdl. g 440°4 231°5 122°8 65°2 34°40 18°76 9°74 a = 0°125 molar Cdl,. g 243°4 216.0 108°7 93°3 61°9 29°55 28°04 11°31 5°36 a = 0-01 molar Cdi.. g 1751 91:0 41°35 19°88 9°76 Cc 612 590 582 580 580 578 578 c 620 590 585 580 580 578 578 Cc 607 590 585 583 582 580 580 578 578 c 590 580 580 580 580 a (312) (31) 1°32 54:0 946 0°627 2651 974 0°3335 14°09 986 0°1649 6°98 993 0°1212 4°92 995 0°0693 2°932 997 0:03782 1584 998 0:01996 0'864 999 mean x (X13) (1) 1°32 43°52 456 0°710 26°43 474 0°3924 13°23 487 0°2099 722 493 0°1124 3°935 496 0°0593 2-168 498 0°03245 St 499 0°01686 0°645 499 mean c (I;) (31) 132 34°35 ig WSN 0°566 15°66 234°3 0°3661 10°38 239°6 O°1857 dro yt be Ee 0°1600 4°83 94.5°2 0°L064 3°266 PA GeT 0°0510 LS] 948°4 0°0483 1°509 248°d 0°01958 0°604 249°4 0°00928 0'°28380 249°7 mean x Gi. Jes I-32 6°84 13716 0°2967 haha 3 LT76 0°1569 1°240 18°76 O-0713 0°575 19°42 0°03428 O°2761 “19°72 0°01682 0°1243 19°88 mean K, 0:0231 0°0230 0°0233 00234 0°0245 0°0236 0:0238 0-0230 0-0234, (21) + (213) = 2a = 0°5 molar. ky 0°01388 0°0127 0°0144 0°0143 0°0141 0°0136 0°0138 0°0130 0-0137, (21) + (213) = 2a = 0°25 molar. Ky 0°0083 0°0085 0°0085 0°0086 0°0081 00080 0°0080 0°0080 0°0081 0°0082 0-0082, (=I) + 213) = 2a = 0°02 molar. K, 000254 0°00235 000237 0°00241 0°00245 000269 000246, * These values were obtained from solubility determinations. Tri-Bromide Equilibria. 115 in the ratio ZnI,.KI, the constant A, calculated from the total iodide concentration, that is, upon the assumption that the power of each iodide to combine with iodine is unaltered by the presence of the other, has the normal value. It is worth noting that this is a case where some slight tendency toward complex salt formation might possibly have been anticipated, but is in no way indicated by the results. Cadmium iodide, on the other hand, furnishes an example of abnormal behavior. (See Table VIII.) This salt gives, as compared with normal cases, a much larger value of A, which, moreover, increases rapidly with the iodide concentration, but tends at high dilution to approach the normal value. This is evident in the following tabular comparison of the values of XK, for potassium and cadmium. The iodide concentrations are expressed in gram-equivalents per liter. Iodide concentrations 0°02 0°25 0:5 1:0 ES) for KI = 0°0014 0°0014 0°0013 0°0013 3G for Cdl, = 00023 0:0082 0°0137 0°0234 For constant iodide and varying iodine concentration, however, the constancy of A, may be fairly good, even in an abnormal case. TABLE IX. TABEE Xi: Cadmium Potassium Iodide. Mercuric Potassium Iodide. a = 0°23 molar CdI..2KI. a = 0125 molar Hgl,.2KI. (21) + (213) = 4a = 0°92 molar. (31) + (2153) = 4a = 0°5 molar. b g c ky b g € Ky, 35°78 226°3 590 0°0096 14°34 Ae | 85°7 0°0185 17253 103°4 583 0°0092 8°66 DAS Con bene tele 0:0178 Sit 46°4 580) 00091 5°83 16°58 85 0:0171 6°16 34°79 580 0°0090 a o1l2 10°05 85 0:0173 2°308 elo 578 00091 -2°284 6°03 85 0°0160 0°839 52a 78- 070100 py pte! coro mean 0:0173 mean 0'0093, In the case of the double iodides CdI,.2KI, and HgI,.2KI (Tables IX and X) abnormality of the same general nature is apparent. Pure mercuric iodide could not be studied by this method on account of its insolubility in water. The experiments conducted with bromine in equilibrium with bromides were confined to cases in which abnormal behavior was to be expected from analogy with the correspond- ing iodides. The results are given in Tables XI to XIV. For bromides the normal value of 4, at 25°, as shown by Jakow- 116 Van Name and Brown— Tri-lodide and kin’s work with bromides of the alkali metals, is 0°0623. Evidently cadmium bromide, cadmium potassium bromide, CdBr.2KB,r, and mercuric potassium bromide, HgBr,.2K Br,, like the corresponding iodides, must be classed as abnormal. The same thing is no doubt true of mercuric bromide, though the low solubility of this salt made the experiments difficult and the results inaccurate.* TABLE XI, Cadmium Bromide. 1. a=0'1556 molar CdBro. 2. a=Q0:0610 molar CdBrs. (Br) +(2Brs) = 2a = 0°3112 molar. (Br) +(2Brs) = 2a = 0:122 molar. b g c Ky b g c Ky, 83°6 965 30°6 0°157 Sicy 883 30°4 07128 41:0 437 28°8 0°168 27°94 432°0 28°8 0°12 22°12 229°8 28°0 O°175 14°80 218°0 28°04 is 10°85 105°7 27°6 0°166 CCT 108°7 216 O lee 5:07 481° 974 0°163 3°673 46°6 27°4 (0-104) 2°533 24°90 27-3 0°167 1°8238 25°05 27-3 O 122 1-319 14°08 27-2 (071389) 0°934 13°94 27:2: 0° 126 0°598 5°Gay 2S 0°165 0°469 6°56 2772.) (OF E29 mean 0°165 0°2239 3°198. 27s20GaiSs mean = 0°127, TaBLE XJ. (continued) TABLE XII. Mercurie Bromide. a = 0°01447 molar HgBry. 3. a@=0:038827 molar CdBro. (2Br)+ (Brs) = 2a = 0°0765 molar. b g c Ke (=Br) +(2Brs) = 2a = 0:02894 molar. 61°3 1268 Sy ers 0°104 b g c ce 13°45 230°6 28°0 0°118 26°97 797 30°0 1¢8 13°08 223°4 28°0 0'11838 10°71 300°4 28°5 1°8 6°15 102°0 27°6 Gs 6°53 178°7 28°0 1:2 3°408 56°0 27°4 07112 4°16 113°5 27°6 2°4 1°655 97°04 27°3 0°113 2°664 1225 2'7°6 ash mean @°111, mean 1°8, *Herz and Paul (Zeitschr. anorg. Chem., Ixxxv, 214, 1914) have also studied the equilibrium between bromine and mercuric bromide at 20°. Their results, recalculated on the same basis as our own, are as follows: (=Br) + (=Brs) (2Brs) (Bre) e Br) Ke 39°98 0-9 8-1 39°08 0°352 42°4 0:8 76 41°6 0°395 39°46 0-7 70 38:76 0388 39°46 06 6:5 38°88 0°421 42°4 10°6 212°5 31°8 0°638 The last determination was not included in the original table of Herz and Paul, but has been calculated from data, given elsewhere in their article, for the solubility of bromine and of mercuric bromide in water saturated with both. Though their values of K, are smaller throughout than ours, they also show great abnormality in the behavior of mercuric bromide. Tri-Bromide Equilibria. t17 Tapue XIII. Cadmium Potassium Bromide. CdBr..2KBr. ia = 0'137l-molamCdbry. 2k pr: 2. a =0°0686 molar CdBr,.2KBr. (2Br)+(2Br3) = 4a = 0°548 molar. (2Br)+(ZBr3) = 4a = 0'2745 molar. b g (6) K, b g Cc Ky Tata ANS 0) 28°8 0°119 82°4 763 29°7 0°099 a2, 926-6 2S Ont ose ahs AGN 28°5 0'097 23°52 122°7 27°5 0°124 17°62 135°4 27°8 0'100 13°67 10°9 27°0- 07128 9°23 68°2 27°0 0°101 8°03 41°4 27-0 0°128 5°10 37°34 27°0 0°101 mean. 0:124, mean 0099, TABLE XIV. Mercuric Potassium Bromide. HgBr2.2KBr, 1. a=0°'125 molar HgBr..2KBr. 2. a=0:05 molar HgBr2.2KBr. (2Br) +(2Br3) = 4a = 0°5 molar. (2Br)+(2Br3) = 4a = 0°2 molar. b g c K, b g (G Ky, 46°8 686 29-5 0°470 32°53 613 29°6 0°330 26°94 378°3 28°6 0°469 14°89 263°5 28°3 0°324 20°07 2733 284 0°451 9:90 19053 28:0 — 0.434 1223-61 27°8. 0444 6°09 10474 27°4 0°330 7°42 ST 27,0" ' 0°47 594K Ot Ie 27-0 0300 mean 0460, mean 0°353, It is very noticeable that the only halides which give abnor- mal results are those of cadmium and of mercury, salts which are conspicuous for abnormal behavior in many other ways. All other iodides and bromides, so far as yet tested, give in dilute solution normal values of K,. Before taking up the discussion of the abnormal cases we must examine a little more closely the nature and meaning of the results in the normal ones. It is evident that the equili- brium in normal eases is practically independent of the nature and the valence of the positive ion. This indicates that the equilibrium is primarily an ionic one, which, for an iodide, can be written — (’) (1,) K = V ay, Me Comparison with the expression for our equilibrium constant mGubih) je ET) pi aibtioal shows that A = K, a where y and 9’ are the degrees of ioniza- tion of the iodide and tri-iodide, respectively, in the equilibrium mixture. Assuming the constancy of “A, i.e. the validity of Equation V, the observed constancy of A, requires that the 118 Van Name and Brown—Tri-lodide and ratio y/y’ shall be constant even when the ratio (2I)/(SI,) varies widely. This is at first sight an unexpected result, but the correct explanation has probably been given by Bray and MacKay* who show that it is in agreement with the principle governing ionization in a mixture of two largely ionized salts having an ion in common. The same reasoning and corre- sponding mathematical relations must also hold in the case of bromides. A number of direct determinations of the value of £ for iodides have been made by Bray and MacKay,t+ who obtained them by measurements of the conductivities of potassium iodide solutions saturated with iodine, and independently from the conductivities of iodine solutions saturated with cuprous iodide. These experiments were all made in rather dilute solutions, (£1) + (2I,) = 0:1 molar and below. Values for K for potassium iodide solutions from 0-1 to 1:0 molar, saturated with iodine, have also been calculated by Bray and MacKay from experimental data given by Laurie.t Comparison of the observed values of A and A, for iodides shows that these two constants are in close agreement in dilute solution, and probably agree fairly well in solutions up to normal strength. For con- centrations in the neighborhood of 0:1 normal iodide both constants are nearly independent of the iodine concentration, but in the stronger solutions the values of the constants decrease very materially, as saturation with iodine is approached. The identity of A and A, means that the ratio y/y’ in the dilute solutions is not only constant but approximately unity, or in other words, that the iodide and tri-lodide in the equilibrium mixture are ionized to practically the same extent, irrespective of the value of the ratio (2I)/(2I,). An important consequence of this, from a practical standpoint, is that a determination of A, is in effect an approximate determination of A. This, however, is only true in “ normal” cases. Finally, reasoning by analogy, we are naturally led to the conclusion that for bromides also, the values of A and A, in dilute solutions are probably identical, but until independent determinations of A for bromides are available, this inference can not be directly tested. Nature and Probable Cause of Abnormal Results. What we designate as “abnormality” with respect to the tri-halide equilibrium actually consists, as a reference to the above tables will show, in a lower capacity to unite with the halogen, and one which varies greatly with the concentration * Jour. Am. Chem. Soc., xxxii, 916, 1910. + Jour. Am. Chem. Soc., xxxii, 914, also xxxii, 1207, 1910. t Zeitschr. phys. Chem., Ixvii, 627, 1909. Tri-Bromide Kyuilibria. Byles of the halide. It occurs, so far as yet proved, only with cadmium and mercury. Cadmium iodide and mercuric chloride are conspicuous among inorganic salts for their low ionization, and for a number of other peculiarities of behavior which have long been ascribed to the presence in their solutions of complex molecules and ions. These peculiarities are known to be shared to a greater or less extent by the other halide salts of the same metals. Putting these facts together, the most obvious explanation for the abnormality which these salts display toward the trihalide equilibrium is to assume that the power to unite with the halogen is possessed only by the normal molecules and ions and not by the complex molecules and ions.* This hypothesis seems to afford a satisfactory explanation of the observed facts, and will be assumed hereafter to be approximately true. An important result of this point of view is that it permits the calculation of the extent to which the simple molecules have united to form complexes. We shall hereafter designate the fraction of the total halide concentration which is in the form of simple molecules, whether ionized or not, as the “active fraction.” Let us consider, for example, the case of cadmium iodide. The apparent value of A, as calculated from the equation A,=(2J) (I,)/(£1,) isabnormally high, but plainly the observed values of (I,) and (2I,) are independent of the presence of complexes, and only subject to the normal experi- mental error. Assuming the correctness of (I,) and (£I,) we can calculate the actual value of (21) in the equilibrium mix- ture from the equation (31,) A, (1,) in which (I,) and (I,) have their observed values in the given cadmium iodide solution, and A, = 0°0014, the value of that constant for a normal iodide in moderately dilute solution. The ‘active fraction” is then obtained by dividing (S31) + (£1,) by the equivalent concentration of cadmium iodide, the value of (21) used here being, of course, the one calculated from equation VI. The same method applies in the case of any other iodide or mixture of iodides, and also for bromides, though the value of A, at 25° in the latter case must be taken as 0°0623 instead of 0-0014. *It might, at first sight, seem possible to explain the variation in K, by assuming that the degree of ionization of the tri-halide in the abnormal cases was very much greater than that of the halide; or in other words, by ascrib- ing this variation to the failure of the relationship. y/y’ = const., on which the constancy of K, depends. This, however, is contradicted by the comparative constancy of A, in such cases when the iodine concentration is varied, and fails to explain the rapid rise in K, when the halide concentration is increased. en — (VI) Am. Jour. Sci.—Fovurts Seriss, Vou. XLIV. No. 260.—Aveust, 1917. 9 120 Van Name and Brown—Tri-lodide and TABLE XV. (Cdl,)" > (21,) (Is) (I) (1) +(21,) (calculated) 500 54°0 1°32 Dike be Oy 26°51 0°628 59°] 85°6 14°09 0°3335 59°1 Tore 6°97 0°1649 99°3 66'2 4°99 O21 56°9 61°8 2°932 0°0693 59°2 Gie4 1°584 0°03782 58°6 60°2 0°864 0°01996 60°6 61°4 0-0 0:0 60°0 60°0* 250 43°52 1-32 46°2 89°7 26°43 0°710 50°1 76°5 13°23 0°3924 47°2 60°4 ee Te, 0°2099 48°29 55°4 3°934 00-1124 49°0 eee} 2°168 0°0593 1-2 53°4 aS: 0°03245 91°0 521 0°645 0°01686 53°6 54°3 0:0 0°0 53°0 53-0r 125 34°35 1°32 36°4 70°8 15°66 0°566 38°42 54°4 10°39 0°3661 39°72 50°1 5°41] 0°1857 40°82 46°2 4°82 0-160) 42°20 47°0 3°266 0°1064 * 43:0 46°3 Gb 5 0°0510 43°25 44°89 1°508 0°0483 43°75 45'3 0°605 0°01958 43°25 43°85 0°283 0°00928 41°15 41°43 0:0 0°0 42°0 AD Os 10 6°84 1°32 (EDD 14°09 2°243 0°2967 10°58 12°82 1°240 0°1569 1a ON 12°31 0°574 0:'0713 10°28 10°85 “OF 2.6 a 0°03428 Leluet he ®) 11°46 0°1243 0°01682 10°35 10°47 0:0 0°0 110 Lio Active Fraction @. ited GRHABWH ~E wo OrorD a re dO a 10°6 Or or Ot Ot GD SO =7T on aT pre He SOS OP © DO OO OT Table XV gives the results of applying this method of cal- culation to the experimental data for cadmium iodide previously recorded in Table VIII. It is evident that the magnitude of the active fraction decreases with the iodine concentration, which is entirely natural since the transformation of a part of the simple molecules into a new substance (tri-iodide) would * Found by extrapolation. Tri- Bromide Equilibria. 121 necessarily shift the equilibrium between simple molecules and complex molecules in favor of the former. What is of the greatest interest and importance, however, is not the value of the active fraction in solutions containing free iodine, but its magnitude in a pure solution of cadmium iodide. This value can easily be found by plotting the values of the active fraction against the iodine concentrations, and extrapolating graphically back to zero iodine concentration. When several concordant determinations are available, especi- ally in the region of low iodine concentration, this extrapola- tion can be carried out with a good deal of certainty, and furnishes what, so far as known to the writers, is an entirely new point of attack, for solving the problem of the composi- tion of iodide and bromide solutions which contain complex molecules. It should be noted that this process, unlike the E.M.F. method and the catalysis method of Walton,* does not give the concentration of the iodide ions, but the total con- centration of simple iodide molecules, ionized and un-ionized. By this method we find, as shown in Table XV, that the active fraction in a pure solution of cadmium iodide increases with the dilution as would naturally be expected, its values in 0:5, 0°25, 0-125, and 0-01 molar Cdl, being 6-0 per cent, 10°6 per cent, 16-8 per cent, and 55-0 per cent, respectively. TABLE XVI. (CdBrz) (2Brs) (Br) (2Br) ({Br)+(2Br3;) Active (calculated) Fraction @. 15526. 521 S1e5 102°9 155°0 49°8 0°39 0°208 Ms ge. LLG 37°8 0°0 0°0 119°5 Unies; 38°4 61°0 22°68 29°05 48°6 (Ae 58°4 0°2275 0°241 58°8 59°0 48°4 0:0 0°0 59°0 590+ 48:4 38°3 Did epenyl 40°01 oo 14 54:4 Tale 0°665 0°:991 Asai, 5 42°4 55°4 0°0 0:0 492°0 42°0+ 54°9 Similar calculations, using the data in Table XI, give the values recorded in Table X VI, which indicate that in cadmium bromide solutions the active fraction is much larger. The active fraction in 0°125 molar CdBr,, as found by interpola tion from the data in Table XVI, is about 40 per cent, as com- pared with 16°8 per cent for Cdl, of the same concentration. For the other abnormal iodides and bromides our experi- - mental data are less complete and less accurate, making the * Zeitschr. phys. Chem., xlvii, 185, 1904. + Found by extrapolation. 122 Van Name and Brown—Tri-lodide and extrapolation rather uncertain. The resulting values for the active fraction, as summarized in Table X VII, are therefore to be regarded only as rough approximations. TABLE XVII. Total Halide Concentration. Active © Salt. grm. equiv. per liter.) Fraction %. CdI,.2KI 0-92 15 CdBr,.2K Br 0°55 47 = = 0°27 61°5 fic! 2KI 0°5 8°5 HgBr,.2K Br 0°5 13-2 ese a4 0-3 ay HeBr, 0-029 + In considering these results for solutions of double or mixed halides it should be remembered that in the case of a mixture of zine and potassium iodides, in the proportions ZnI,.KI, the value of K, was found to be normal, and the active frac- tion, therefore, 100 per cent. In other words, both iodides, which are normal when alone, act entirely independently in the mixture, a result which is probably true in general. This, however, is not the case when one of the halides is abnormal. Let us take, for example, the case of CdI,.2KI, where the active fraction was found to be 15 per cent. Since the potassium iodide made up 50 per cent of the total iodide, it follows that even assuming that the cadmium iodide united with no iodine, only 30 per cent of the potassium iodide was active. It is evident that a large part of the latter must have been taken up in the formation of complexes. The same effect is visible with CdBr,.2K Br to a shght extent, while with Hel,.2KI and HgBr,. OK Br it is even larger hea in the first example. As already stated, it is very difficult to obtain any trust- worthy results with pure mercuric bromide. Considering the low concentration of the mercury salt, the result in the table indicates that the tendency to complex formation in mercuric bromide solutions is very great. An entirely different hypothesis to account for the proper- ties of the iodine-iodide mixture has been proposed by Parsons, who ascribes the phenomena to simple solubility of the iodine in the dissolved iodide—“ solution in a dissolved solid.” The acceptance of this view, however, would leave the constancy of A, entirely without explanation, and, indeed, would involve the complete rejection of evidence of ‘the kind we have been considering. Parsons* likens the case to that of iodine (or camphor) dis- *Jour. phys. Chem., xi, 664, 1907; also Parsons and Corliss, Jour. Am. Chem. Soc., xxxii, 1374, 1910. Tri-Bromide Equilibria. 123 solving in a mixture of acetic acid and water, but this would only be a parallel if it could be shown, for example, that a number of different esters of acetic acid, in equivalent concen- tration, increased the solubility of iodine (or camphor) in water by exactly the same amount, or if the same thing were _ proved for a series of substituted acetic acids. The experimentally established fact which calls for explana- tion is that in dilute water solution equivalent concentrations of all metallic iodides, excepting only, so far as yet known, those of cadmium and mercury, take up, at the same tempera- ture, the same concentration of titrable iodine, when brought into equilibrium with solid iodine or with a carbon bisulphide layer containing iodine in a fixed and constant concentration. That this relation should become inaccurate with increasing concentration and finally fail altogether, is only what would be expected from the deviations from the law of mass action which are normally observed in concentrated solutions. Its failure for cadmium and mercury is wholly in keeping with the other abnormalities of the halides of these two metals. The existence of such clear-cut stoichiometric relations between the concentrations of iodide and of iodine is the strongest possible proof that the phenomenon is not due to solubility in the ordinary sense, but rather, to the formation of a definite compound. Application of the law of mass action leaves no doubt that this compound is the tri-iodide. Summary. 1. The iodine-iodide equilibrium has been studied at 25° in - number of cases not previously investigated, and the value of = (XJ) (1) / (21,) determined. a For the i didee of strontium, zine, nickel, and lanthanum, the value of A, in dilute solution is normal. For cadmium iodide, and double iodides containing cadmium or mercury, the value of A, is mnch higher, and rises rapidly with increasing iodide concentration. 3. Similar abnormality in the corresponding constant for the bromine-bromide equilibrium was found in the case of cadmium bromide, mercuric bromide, and double salts containing either _of these. 4. The measurement of A, in abnormal cases furnishes a useful method for determining the percentage of complex molecules and ions, not only in the equilibrium mixture, but also in solutions containing no free halogen. This method is based on the very probable assumption that the complex mole- cules and ions do not combine with the halogen. 124 ELC. Case—Amphibian Fauna at Linton, Ohio. Art. XI1.—TZhe Environment of the Amphibian fauna at Linton, Ohio ;* by EH. C. Case. THE group of vertebrates including amphibians, fishes, and the probable reptile Hosawravus cope, from the Lower Free- port Coal at the old station of Linton in eastern Ohio has been known since 1856, when the first description of Pelzon lyelli was published by Wy man in this Journal. Since that time papers have been. published by Newberry, Cope, Williston, and Moodie, all dealing with the morphology of the animals. A complete bibliography of the subject appears in Moodie’s Monograph, The Coal Measures Amphibia of North America,t aud need not be repeated here. As our knowledge of the taxonomy and morphology of the fauna is now fairly complete, it is possible to turn to a con- sideration of the various factors which influenced their life and development. Fortunately the presence of workable coal in the region has led to a large amount of exploration and exploita- tion of the beds in which the animals occur and it is possible to gain a clear idea of the conditions in which they lived. The key to an understanding of this fauna was grasped by Newberry as is shown by the following quotations : “In the descriptive portion of this volume, quite a number of species of the fossil fishes from the Coal Measures of Ohio are figured and described. A large part of these are from a single locality, which has already become somewhat celebrated for the number and interest of the fossil forms it has furnished. I refer to Linton, on the Ohio river, at the mouth of Yellow Creek. The fossils are found there in a thin stratum of cannel which underlies a thick seam of bituminous coal, that we have called Number 6, because it is the sixth workable seam from the base of the productive Coal Measures. Already about twenty species of fishes have been obtained from this deposit, and at least as many Amphibians ; and all found here for the first time, although two or three species have since been met with in other localities, in this or adjoining States. On tracing Coal-seam No. 6, in various directions from Linton, the cannel at its base is found to thin out and soon disappear. We learn, from a careful study of the deposit, that there was in this locality at the time when the coal was forming, an open lagoon, densely populated with fishes and salamanders ; and that after a time this lagoon was choked up with growing vegetables ; and peat (which afterward changed to cubical coal) succeeded to the carbonaceous mud (now cannel) * This paleogeographic study is published by permission of the President of the Carnegie Institution of Washington. + Moodie, R. L., Publication 238, Washington, 1916. i. C. Case—Amphibian Fauna at Linton, Ohio. 125 that had previously accumulated at the bottom of the water. The fishes of this pool were mostly small, tile-scaled Ganoids, belonging to the genus Hurylepis. Though here extremely abundant, they have not been found elsewhere. I have enumer- ated nine species of this genus, but possibly some of them should be considered as mere varieties. There were also in this lagoon two, or perhaps three, species of Coelacanthus (one of which is so closely allied to C. lepturus of the Coal Measures of Europe that they should not be separated, and yet this genus has been nowhere else recognized on the American continent). There are also found here the thin seales, from one to’two inches in diame- ter, some ornamented and some plain, and also the lance-head teeth of RAizodus, and the teeth and spines of Diplodus. On the whole, this must be looked upon as one of the most interest- ing localities of vertebrate fossils known on this continent ; and it is even doubtful whether any other equals it in the number of new species or in their zoological and geological interest.””* “ The large number of species of fishes and amphibians (about fifty) found in one single coal mine at Linton indicates that the vertebrate fauna of the Coal Measures was much richer than has heretofore been supposed. The cannel coal of this locality was undoubtedly deposited in a lagoon of open water in the marsh where Coal No. 6 was formed. How extensive this lagoon was, we have not as yet learned; but all the fossils found there have been taken from an area of a few hundred feet in diameter. We have probably now obtained representatives of most of the fishes and salamanders that inhabited this body of water, but cer- tainly not all, for every considerable collection made there has contained something new ; and the fauna of the epoch in which this deposit was made must certainly have been very varied, since from this one spot have been taken the remains of fifty dis- tinct species, less than a half dozen of which have been found elsewhere. This coal mine at Linton may be regarded, therefore, as a kind of loophole through which we see, in all its details, the life of one locality in the great world of the Carboniferous age. Looking through that, we have before our eyes a little pool of water swarm- ing with fishes of various kinds, some of them very large, clad in mail and provided with most formidable sets of trenchant teeth ; others, small but exceedingly numerous, covered with enameled and highly ornamented scales and plates. These latter, as we learn by coprolitic masses, were the prey of the larger ones. With the fishes were a large number of aquatic, carnivorous salamanders, some of which must have been eight or ten feet in length, and as formidably armed as the larger fishes. Others were snake-like in form, yet several feet in length, bristling with spines, or protected by thick and bony scales. Others still were afew inches in length, very slender and delicate, and, as we ere J. ion Geol. Survey, Ohio, vol. i, Paleontology, pp. 284-5, 126 £. C. Case—Amphibian Fauna at Linton, Ohio. know by their mutilated fragments, served as food for the more powerful. A remarkable circumstance connected with the Linton deposit is this: that in working up some hundreds of tons of the cannel coal which contains the fishes and amphibians, we have obtained not a fragment of an insect, and only a few small and imperfect remains of crustaceans. Mollusks, too, are entirely absent, no shell of any kind being found there, except those of Spirorbis, which is thought to have been au annelid. ‘These occur, how- ever, In millions, and we may infer from the multitudes of these delicate organisms that the water they inhabited was guiet, warm, and almost stagnant. Whether salt or fresh, we do not know, but it seems to me most probable that it was fresh. Very few remains of plants have been found in the Linton can- nel, and these, if leaves, are skeletonized, sbowing their long maceration in water. In this, as in many other respects, the Linton deposit is strikingly different from that of Mazon Creek, Illinois, which has yielded a large number of insects, crustaceans, and plants, and very few fishes and amphibians.”* The first step in the study of the Linton fauna was an at- tempt to determine whether Newberry was correct or not in his assumption that the Linton fauna was isolated in a pool of open water in the midst of a great swamp and that this pool was finally closed by the growth of vegetation causing the death of the animals. The attempt was made along two lines of evidence; first, the stratigraphy of the region around Linton and second, the method of formation of cannel coal. 1. The evidence from the stratigraphy. The Linton fauna occurs in the Lower Freeport Coal of the Allegheny series of the Pennsylvanian. To test Newberry’s idea an area of considerable size was selected surrounding the location of the fauna. This ineluded Beaver, Allegheny, Washington, and Greene Counties in Pennsylvania ; Hancock, Brooke, Ohio, and Marshall Counties in West Vir ginia ; and Col umbiana, Jeffer son, Carroll, Harrison, and Belmont Counties in Ohio, as shown on the accompanying map, fig. 1. Many sections of the Allegheny series in this area were plotted and a few of the most detailed are shown in the columnar sections, fig. 2. he location of the sections is indicated by numerals en the map. A study of sections 4, 3, 2, and 1, north and east of Linton, show a strong tendency for the Lower Freeport Coal and its accompanying beds to break up, indicating. the edge of the local swamp. From Sprucevale, Columbiana County, Ohio (4 on the map), south of the mouth of the Little Beaver liver, * Newberry, J. 8., Geol. Survey of Ohio, vol. ii, pp. 179-180, 1874. E. QO. Case—Amphibian Fauna at Linton, Ohio. 127 in Beaver County, Pennsylvania (3), then east to near the mouth of the Beaver River in Beaver County (2) and southeast to Sewickley in Allegheny County, Pennsyl- vania (1), the Upper Freeport Coal maintains a nearly uniform thickness of 2 to 3 feet underlain by a vary- ing thickness of fire clay with a thin limestone at Fig, 1. =——2 === Fic. 1. Map showing the location of Linton, Ohio and the stations from which sections are quoted. stations 4 and 2. Beneath this is a considerable thickness of sandy shales broken at station 1 by thin beds of sandstone and fireclay. The Lower Freeport Coal is very thin at these stations, represented by a carbonaceous slate only at station 1. Below the Lower Freeport Coal lie from 10 to 20 feet of sandy shales and then a heavy bed of massive sandstone, 50 to 75 feet, which at station 1 becomes a micaceous sandstone broken by the Upper Kittanning Coal. South from Sprucevale (4) the Upper Freeport Coal is 3 feet thick at Linton (5), underlain by fire clay but the limestone is 128 E. 0. Case—Amphibian Fuuna at Linton, Ohio. absent, then follows 10 feet of shale and 5 feet of limestone, obviously not the upper bed at (4), and then 50 feet of shale and sandstone. The Lower Freeport is 7 feet thick, the lower part cannel coal, with 5 feet of fire clay and below this 7 feet of shale and sandstone with possibly thin layers of coal. At Wellsburg, Brooke County, West Virginia (6), the Upper Freeport Coal has been removed by erosion, but there are 40 feet reported as fire clay, 5 feet of Lower Freeport Coal, no Hires: rim 4 vo Oo iS Ol (or) Pe Fic. 2. Sections taken from the stations indicated on the map, fig. 1. Black, coal, Upper and Lower Freeport, Upper and Middle Kittanning. White, fire clay and unknown. Dotted, sandstone. Lined, shale and shaly sandstone. Obliquely lined, ‘‘ slate.” Vertically lined, limestone. cannel, and one foot of fire clay underlain by 60 feet of shale and slate. At Wheeling, Ohio County, West Virginia, the Upper Freeport is removed by erosion, but there are 50 feet of blue shale, then 7 feet of Lower Freeport Coal, without cannel or fire clay, lying above 96 feet of shale and sandstone. West of Linton the Lower Freeport Coal shows, in the adjacent counties at least, a decided thinning, with a corresponding in- crease in the accompanying layers. “The Lower Freeport Coal becomes locally important as the Whan Coal bed within E. C. Case—Amphibian Fauna in Linton, Ohio. 129 a small area in central Columbiana, but for the most part it is worthless.”’* - “The Freeport coals both thin. ... Eastward the Upper Freeport is important, it can be followed from Yellow Creek in Columbiana Oounty somewhat decreased in thickness, but in the southern part of the county is often 4 feet 6 inches and yields good coal. The chief drawback is the frequency with which it is cut out by the overlying sandstone. ... The Lower Freeport is persistent, usually too thin to be utilized. It is rarely more than thirty feet below the upper.’’+ In Carroll County, Ohio, the following section has been given :{ Ft. Wmpenslieeport; Coals sss .2 2 obs ee ee 2 Shale, sandstone, and conglomerate _-_.--.------- 40 Lower Freeport sandstone, massive...-.--...-.-- 30 Concealed ...-.:- MEY De UGeA TSO IRT SS NS LM rate ak Seed AS Fs BIO) Coals Middle Kattanmine 5.2 2 4nl< pseu ers oles 3 ENTS GIB. 2 EN) Oa ae ee a vr ar er 3 Fee clita em gs stb ss Se IL OOS a) eI UN Oo.) OBIE Ls Se Se Pea a cet A red Gen ae eee 1 SHANG) Ga 8) Dele a SS EY Ce lr Ney et) le Near Sherodsville in Carroll County. ID iipermnecport Coals eh veo ee a lo yitetelaice Geert e eee Pe we ele 8) 6 Ee Ss Homer haa ere aE eb tal) 1 Concealed 920452 J. tun. Bera iyoeioen, Fee D Ay: dats toneCrts 40 black shale 30/2) PEEVES ey Rie RUN RU I UND Homer emceport Come wae sik Bee ek el (8 Pinerclay and modular won) ores. 3/2424 5044042. 12 Newberry’s generalized section for Stark County. Ft. Conley 7. (Upper Preeport) 210022. 222. 222. toy 3 TE'TPS GIG SOUS i Le URL 2 CR Ae Ue aoe 1 Shale and sandstone with thin coal near the TDRGAUISS SO ATE CS OEE he a age eh ene 75 to 110 Coackie. (Middle Kittanning ye 22. 2.2. 22222. 2to 6 *Stevenson, J. J., Bull. Geol. Soc. Amer., vol. xvii, p. 111, 1907. + Ibid., p. 118. ¢t Geol. Survey, Ohio, vol. v, p. 74. S Includes Lower Freeport ‘‘a mere blossom.” 130) EF. C. Case—Amphibian Fauna at Linton, Ohio. Generalized section for Tuscawaras County. In North In South Ft. Ft. Coal. (Upper Freeport) 22227 22% Ake eas 3 4 Interval. i. fa seeeee ears ee ae ae 35 Coal, 6a (Lower Freeport) . ‘ipere tasters thin 2 Conglomerate, s sandstone, and shale (Iree- DOT) 5) ate ty SR en a ee eee ae 50 52 Coal, 6 (Middle Kittanning) __.--._- ._.- 4 4 Orton’s section of northwest Guernsey County. Upper Hreeport: Coal bed, Cambriadaew2=2 =. = thin Clay, Upper Ereeport limestone... -_ 22... 222525 10 {nterval os tee. oe ee ee Lower Freeport Coal ‘bed :2.3\.0 52 bapa thin Middle Wittannin 3.26 25 32a eee 3 Stevenson remarks in the article quoted above, page 71, that the *“* Lower Freeport shows abrupt and extreme variations in thickness (in eastern Ohio) as well as in quality and occasion- ally carries on top a thick deposit of impure cannel.” A study of these sections taken in all directions from Linton shows that the Freeport Coal either becomes thinner, loses its fire clay (the leached ground soil of the marsh in which it was formed) or the accompanying layers become disturbed by inter- calated beds. While it is admitted that the Lower Freeport Coal is variable in thickness and quantity, in almost every place where it is known, the peculiarity of the thickening at Linton, the presence of underlying cannel coal, and the undis- turbed deposition of the accompanying shales and sandstones is at least strong confirmatory evidence of the presence of an open pool in the center of a great swamp. Of the Lower Freeport sandstone, I. C. White says,* “It is one of the most persistent sandstone horizons in the Allegheny Sores, | . The rock is usually quite hard, micaceous, and often pebbly, but does not split evenly.” In the western part it is very uniform in character, running from 75 to 100 feet thick, but toward the east (station 1 on the map), it is broken by shale and a thin bed of coal (Upper Kittanning). In the time before the deposition of the Lower Freeport Coal, the Linton area was evidently for a considerable time a region of deposit from moving water bearing the debris from a region undergoing rapid denndation, as is shown by the pres- ence of undecomposed mica and the pebbles included in the matrix. This was one of the longer periods of depression in the region and the depression was extended over a wide area. * White, I. C., Geol. Surv. West Va., vol. xi, page 473, 1913. E. C. Case—Amphibian Fauna at Linton, Ohio. 1381 2. The evidence from the presence of cannel coal. Without going into the history of the discussion of the origin of the various kinds of coal, it is sufticient to state that itis now the generally accepted opinion that boghead, boghead-cannel, and cannel coals have been formed by the accumulation of alge, spore exines, bits of resin, and other light, wind-blown materials which have grown in place in open water or have accumulated on the surface of open pools. Details of the dis- cussion and the general conclusions may easily be found by following the references given by David White in his treat- ment of the origin of coal.* Itis also of interest to note that the theory of the origin of cannel coal recognizes the necessity, or at least the advantage, of the presence of decomposing ani- mal matter in its formation. Accepting as a working basis that the stratigraphy and the presence of the cannel coal demonstrate the presence of an open pool in the vicinity of Linton, we may turn to the history of this pool and the other factors which influenced the develop- ment of the fauna. East of Linton the Lower Freeport Coal lies directly upon the sandstone, but at Linton the section of the old Diamond Coal Mine shows several feet of fire clay beneath it. Fire clay is the result of the action of peaty waters, containing much CO, in solution upon an underlying soil. It is apparent that succeeding the long period of submergence during which the Lower Freeport Sandstone was formed there was a slight ele- vation, but a local depression in the vicinity of Linton retained a considerable body of fresh water in which accumulated the sediments of the adjacent land. The water at the edges of this pool was not deep enough to prevent the rapid develop- ment of an abundant flora. The roots penetrated the soil and the decomposition of the vegetal material furnished an abund- ance. of CO, which deprived the soil of its alkalis, alkaline earths, and iron, reducing it to a fire clay. The bordering veg- etation did not for a long time spread out over the surface of the pool in the middle of the swamp and during this interval there accumulated the material of the cannel coal and the amphibian and fish fauna developed large numbers both of indi- viduals and species. The presence of a single form which must be placed with the reptilia indicates either a land at no great distance, from which the body of the animal was trans- ported after death, or an aquatic habit for the reptile. The physiographic environment of the fuuna. I. C. White has insisted upon the fact that though the beds of the Allegheny series are subject to frequent changes in the * White, David, The Origin of Coal, Bull. 38, Department of the Interior, Bureau of Mines, page 198, 1918. 132 E. 0. Case—Amphibian Fauna at Linton, Ohio. material, due to minor changes induced by slight but repeated and rapid fluctuations of level, they maintain on the whole a homogeneity which speaks of ‘widespread and long continued uniformity in the general aspect of the land and water. Per- haps the best picture that has been presented of the region is that given us by David White in Bulletin 38 of the Bureau of | Mines, page! 63: “SuMMARY OF TERRESTRIAL CONDITIONS. Coal formed on subsiding areas. On the whole the criteria relating to the terrestrial condi- tions of deposition show that the formation of widely extended coals in series were regions of. subsiding base-level coastal plains or filled basins. That the rate of subsidence was vari- able is shown by the varying character of the rocks of the coal measures ; the presence of marine faunas in places immediately above the coal; the occurrence of shallow water limestones, of ripple-marked, rain-marked, or sun-cracked layers or of con- glomerates or local unconformities: and in particular the occurrence of great thicknesses of coal covering large areas. The deposition of the great thicknesses of peat necessary for the production of a thick bed of coal, probably 10 to 20 feet of peat being required to produce a single foot of high-grade bituminous or semi-bituminonus coal, could hardly have taken place except under such close adjustment of the rate of sub- sidence to rate of peat accumulation as to maintain a depth of water cover within limits that would permit the growth of peat-forming vegetation for an exceedingly long time. Too rapid a subsidence would have flooded the swamps so deeply as to kill the principal peat-forming vegetation, produce open water conditions, and allow the invasion of sediment-bearing water with its oversweep of mineral matter, or of oxygenated water which would have permitted the advance of decay (biochemical process), to the complete destrnetion of the super- ficial organic matter, unless the deposition of sand or mud were sufficiently rapid quickly to arrest the decay by exclusion of the sources of oxygen supply. The roofs of many coal beds bear evidence of the latter conditions. In most cases, how- ever, the oversweep of clays and sands appears to have been so abrupt as to seal the underlying more or less aseptic organic mass from access of oxygen and to prevent its continued decomposition. On the other hand, if the subsidence of the region was too slow or there was warping of the region, the surface of the peat may have reached the upper limit of its formation and entered the zone of increasing exposure (insuf- ficient water), in which the organic matter was reduced to E. ©. Case—Amphibian Fauna at Linton, Ohio. 133 ‘humus’ soil, and even destroyed, so that the formation of a thick bed of coal was impossible, unless the rise of the water level, usually by accelerated submergence, was brought about. The formation of a thick bed of coal is therefore seen to indi- cate in general the maintenance for a long period of an approx- imate balance between the rate of peat accumulation aud the rise of the water, so as to maintain a depth of water favorable for the growth of the vegetation and its preservation as peat. In confirmation of the views of many geologists the writer’s observations of the horizontally extensive coal beds in the American fields lead him to conelude that the peat-forming vegetation, which was probably largely vascular, grew in place over nearly all of the areas of these coal beds and that it occu- pied these areas almost continuously during the deposition of the peat except at times marked by the inundation inwashes represented by the clay or shale partings in the coal. In other words, most of our commercial coal has been formed from plants that grew above the surface of the peat, and is of autoc- thonous origin. Coal that may be attribnted to the mere accumulation of drift vegetation is, according to the author’s observations, very restricted in area and variable in thickness, and much of it is too high in ash to be of value. In certain regions in which the water was quiescent, but of too great depth for occupation by vegetal growth, bituminous shales and black carbonaceous muds, many containing marine - or brackish water shells, seem to have been deposited in many places, the state of the organic material—that is, its stage of decay—being dependent largely upon the rate of accumulation of the vegetal debris and the supply of oxygen. At many points small areas of open water, temporary in duration, occur- ring in the midst of the swamps were marked by the concen- tration of spores, resins, waxes, etc., forming cannel layers or lenses, the more destructible matter being lost by decomposi- tion, which here again is dependent on the oxygen supply and rate of delivery of the plant debris. In other cases very restricted areas of open water (not occupied by vascular plant — growth) whose stagnant depths were more toxic seem to have favored the accumulation, without decay to the point of destruc- tion, of plankton of various types, as well as of wind-borne spore materials forming boghead, and boghead-cannel coal. In none of the important and widely extended coal beds examined by the writer has he observed any lenses or intercalated bodies of coal that may be interpreted as masses, floating islands, or rafts of vegetation somewhat abruptly submerged, in accord- ance with the hypothesis proposed by numerous writers.” Elsewhere David White speaks of the region as “‘ one vast peneplain.”’ 1384 £1 C. Case—Amphibian Fuuna at Linton, Ohio. The climatic environment of the fauna. The flora of the region around Linton has been reported upon by David White. His list* of the plants of the Freeport group contains no forms differing especially from those of the whole Allegheny series, and all indicate the existence of a ‘singularly equable and humid but not tropical or even semi- tropical climate.” There is no evidence either in the woody growth foliage, florescence, or fruition of any seasonal changes, either of temperature or of humidity. In other words, the animals lived in a period characterized by the extreme monot- ony of the climatic environment. The organic environment. The organic environment of any animal or group of animals may be defined as the group of contacts of that animal with other forms of life. Normally, the organic environment com- prises both the flora and fauna, but in this instance the animals were not, so far as we can see, influenced by the vegetation more than that they profited by the shade of the umbrageous growths, sought refuge in the interstices of submerged roots, or possibly fed upon some forms of the algze in the pool. None of these factors would have left any readable record in the morphology of the animals. The list of the flora occurring in the shales accompanying the coals of the Freeport group has been cited above. The character of the contacts within the fauna. The list of known amphibians from Linton as given by Moodiet includes 51 species. The genera are as follows : Brachydectes Cercariomorphus Cocytinus Ctenerpetom Diceratosaurus Eoserpeton Erpetosaurus EHurythorax Hyphasma Ichthycanthus Leptophractus Macrerpeton Molgophis * White, David, Bull. Geol. Soc. Am., vol. i, page 154, 1900. + Moodie, Roy L., Publication 288, Carnegie Institution of Washington page 18, 1916. E. C. Case—Amphibian Fauna at Linton, Ohio. 135 Odonterpeton stocephalus Pelion Phlegethontia Pleuroptyx Piyonius Saurerpeton Sauropleura Stegops Thyrsidium Tuditanus If we examine the animals as described and illustrated in Moodie’s excellent monograph, we find that they were, one and all, provided with sharp, “conical teeth, suitable only for a car- nivorous or an insectivorous diet. This eliminates the vegeta- tion of the period from consideration as a possible source, at least as an immediate source, of food, but introduces a most effective element of stress in the competition between the ani- mals themselves, on the one hand to capture prey and on the other to escape the attack of predatory forms. The possible sources of food were the fishes, the amphibia, and very probably the abundant arthropods, ’ molluses, and insects, though practically no traces of invertebrates have been found with the remains of the amphibians, except the casts of spirorbis-like forms. While there can be little doubt that some of the amphibians were carrion eaters and scavengers, the ulti- mate food supply must have been the invertebrate fauna of the waters and banks, and the very meagerness of the remains of such a fauna speaks eloquently.of the crowded habitat and the eager search for every edible particle. Beyond this the diet was of flesh and the fauna was self-devouring. From the description given it seems fairly certain that the amphibian fauna was isolated in a pool of clear water sur- rounded by a great stretch of swamp. ‘The ordinary factors of environment which influence the development of a fauna were absent or ineffective, the physiography and the climate were monotonous in the extreme ; the vegetation had only an indirect effect. The main stress upon the life was competition within the fauna. This stress became very high with the crowding of the pool, but as the monotonous environment afforded but lim- ited possibilities for the formation of new habits, adoption of new habitats or the assumption of a new group of contacts in any form, it was not relieved by any over-specialization either in structure or habit. A study of the amphibia reveals only a very normal group of animals. They varied in size from ten feet to six inches in length, some were squat and sluggish, others lithe and serpentiform, some even so snake-like that Am. Jour. Sc1.—FourtH Srriss, Vou. XLIV, No. 260.—Aveust, 1917. 136 E. C. Case—Amphibian Fauna at Linton, Ohio. they had lost their limbs. Some hid for safety in dark holes and corners, others lurked in the slime, feeding on carrion or the less active and well protected forms; still others flashed through the water in active pursuit of prey and dared give battle in their conscious strength. It was a fauna whose ele- ments occupied all the possibilities of the pool to preserve their lives and propagate their kind, but there is an almost total lack of bizarre and overspecialized forms, none heavily armored and none with an excessive development of tusk or talon or spine, and none that could be called giants of their kind. There was a full occupation of all the reasonable possibilities of life but nothing that would indicate an extreme adaptation, either for offense or defense, to limited paths of lifesuch as occur in other places and in other geological formations where the members of the faunas were very perfectly adjusted to each other. There was only the healthy growth induced by competition in a fauna which still retained all the resilience of its juvenile stage. Such an assemblage existing under very powerful stress, if even from a single source, was full of the possibilities of devel- opment; ripe for the rapid and wide radiation in habits and structures long denied them by the monotony of their environ- ment. For the animals in such a pool there were but two possible endings. Either the pool would become choked by the growing vegetation of the surrounding swamp, or in the many fiuctuations of the land, channels would open whereby the animals could escape into other habitats and encounter a new environment. It was apparently the first of these fates which came to the Linton fauna. It was overcome in its full vigor before the ultimate adjustments of life to life had pro- duced the extreme development of armor and weapons of attack seen in more mature or in senile faunas. Elsewhere in the same region similar faunas were released to expend in morphological advances and various adaptations to new condi- tions the stored up stresses of similar periods of isolation. University of Michigan, Ann Arbor, Mich. Wickham—Fossil Beetles from the Sangamon Peat. 187 Art. XIII.—Some Fossil Beetles from the Sangamon Peat ; by H. F. WickHam. AutnouesH the insect life of the North American Tertiaries has received considerable attention, that of our Pleistocene is relatively little known. Our ignorance of the exact distribu- tion of insects in this latter period is due in part to the appar- ent scarcity of deposits containmg reasonably well-preserved remains and, in almost equal degree, to the disinclination among entomologists to give the more or less fragmentary material the close study necessary for determination. Most of the known North American Pleistocene fossil insects are from the clays, peats and asphalts. Nearly all of them, so far as recognized, are beetles, the hard exoskeleton of this group resisting destructive agencies much better than the comparatively delicate integuments of other orders. It is also evident, from the nature of the remains, that ground beetles and water beetles are much more likely to be preserved than those living upon plants, the result being that collections made in different sections of the countr y may have a similar physi- ognomy. While this renders the identification of new finds more difficult, it really gives a much better basis for compara- tive work than if the same number of species were scattered through many families. We are able, for example, to com- pare the Carabidz, Dytiscidee and Staphylinidze of one loca- tion with species of the same families from other places. During the past two years, I have received from Professor T. E. Savage sendings of Coleopterous remains which he col- lected in an exposed peat seam on the north bank of the San- gamon River near Mahomet, Champaign County, Illinois. This peat lies above the [linoisan and below the Wisconsin drift. There is a slight development of loess or loess-like silt above the peat and below the Wisconsin, and Professor Savage considers that the reference of the bed to the Sangamon stage is rather definitely proven. This collection allows us to make a comparison of two fairly widely separated North American faunz which have been assigned to the same interglacial stage, since Doctor 8. H. Scudder has reported quite fully upon a series of Coleoptera from the Scarborough beds near Toronto, Canada,* considered as belonging to the Sangamon interval. He recognized 76 species of 338 genera and 8 families. From these he reached the conclusion that the climate of Ontario, at the time of their deposition, was very similar to that of to-day or perhaps slightly * Contributions to Canadian Paleontology, vol. ii, part II, Ottawa, 1900. 188 Wickham—Fossil Beetles from the Sangamon Peat. colder, a considerable number of the recent allies of the fossils being known from a more northern habitat. On the whole, the fauna had a boreal aspect though by no means so decidedly boreal as one would anticipate under the circumstances. Examination of the Illinois collection indicates the presence therein of ten determinable species belonging to seven genera and four families, the Carabide, Dytiscide, Staphylinide and Chrysomelidee. These families contain, as well, the bulk of Scudder’s Scarborough species, in the proportion of 36, 8, 19 and 2 respectively—that is, 65 out of the 76 which he has described. Five of the genera are common to both collections but all of the species appear to be quite certainly different. The basis for deductions as to climate is not very broad but, judging from the presence of Carabus meander sangamon and Chlenius plicatipennis, the general northern flavor of the remaining species and the entire absence of any without fairly close recent boreal allies, I think we are quite justified in assuming that conditions were, at any rate, more rigorous than in sonthern Illinois at present. Probably they were at least as severe as in Ontario at the date of formation of the Scarbor- ough beds. It is true that all of the genera are now living in Illinois but they also occur very far to the north, extending in part to the shores of the Arctic Ocean and we must take into account the entire absence of anything characteristically southern. A glance at the notes following the descriptions will show that the near relatives of all the fossils in this col- lection are of northern range. One might hope that the beetles would throw some light upon the identification of beds containing their remains and allow us to decide with some certainty whether or no the Scarborough deposit and the one now under investigation really belong to the same interval. The matter is complicated, however, by our ignorance of Pleistocene insects. Aside from the two collections noted above, we are acquainted, in this country, only with the probably more ancient Port Kennedy fauna, that of the widely distant Rancho la Brea asphalt deposit and an occasional scattering species from some other point. It has already been brought out that the species of the Scudder report are all different from those of the present paper, though in general closely allied. The differences are not great enough to indicate any wide dissimilarity in ecolog- ical conditions nor separation by a long period of time. On the other hand, the likenesses do not prove the deposits to be synchronous. It is apparently recognized that the Sangamon interval was of long curation (20,000 to 100,000 years*) and even if both Scarborough and Mahomet beds were laid down * Osborn, The Age of Mammals, p. 447, 1910. Wickham—Fossil Beetles from the Sangamon Peat. 139 during this stage their formation may be sufficiently remote in time to allow of some specific differentiation. We must take into account also the rather wide separation in space of the two places—but I very strongly question if ten species of the Carabidee, Dytiscidee, Staphylinide and Chrysomelide taken at random in a recent Lllinois bog would all be different from 65 species of the same families collected during the same year and in similar surroundings at Toronto. I doubt if season has much to do with the divergence in character of the fossils since peat deposits would continue forming all through the warmer parts of the year and insect remains might readily be preserved at any time. The fact that there is no evidence of intrusion of southern types in the Mahomet collection would suggest that the deposit was perhaps formed when the Illi- noisan glacial movement was well advanced on its southward route or at any rate previous to a far northward recession. The species and varieties described in this paper are all new to science and are arranged by families as follows: CaRABID& DyTiscip & Carabus meander sangamon Agabus savagel Patrobus henshawi prelugens Platynus pleistocenicus Oli engi. ae subgelidus ealvini Olophrum interglaciale Chleenius plicatipennis CHRYSOMELID.& Donacia styrioides All of the types are to be found in the Museum of the Uni- versity of Illinois. CARABUS MZANDER Fisch., var. SANGAMON new variety (fig. 1). Represented by part of a wing cover, evidently the inner basal area of the left elytron, about 6-90" in length by 2°35™™ in width, more or less damaged on all margins. The sutural bead is like that of recent specimens; immediately exterior to it is a very fine, scarcely visible carina which corresponds to a similar line (of great variability in distinctness) on the living insect. The first row of tubercles has the basal one elongate, as usual in modern individuals, while in the second row those near the base are short as in most of the specimens in my cabinet. The third row is not well preserved.. The carina between the first and second series of tubercles is interrupted instead of being entire and this carina, as well as the tubercles, is more irregular and less smooth than in any of my recent examples. The carina between the next two rows ef eleva- 140 Wickham—Fossi! Beetles from the Sangamon Peat. tions is finer. Between all these raised sculpturings, the sur- face is roughened similarly to that of recent specimens but a little more coarsely. The separation of this variety is based upon the rougher surface and par cularly upon the breaking up of the carina. This character, in itself, is of small import- Fig. 1. ance and does not indicate any radical change 2 since the fossil was laid down in the Sanga- mon stage. To-day, Carabus meander oc- curs in the north, from Maine and Labrador to Manitoba, sonth through the Rocky [ERs Mountains to Colorado and, probably as a see relict from one of the interglacial stages, in EER « the Chicago district of Illinois. Specimens pies trom all of these places have been compared RS. directly with the fossil. Several species of Sa Carabus, part of them presumed to be identi- Ke. cal with recent forms, others believed to rep- eat & oS resent extinct varieties or species, have been Re fered | described from the Pleistocene deposits of Eee BS Switzerland, Belgium and Galicia, one of wee which, C. meandroides Lomnicki, from the Se last named locality, probably closely resem- ESS bles the one above characterized. PATROBUS HENSHAWI New species. A single piece of peat carries the head, prothoracic epister- num, pronotum and elytron, the last broken off behind the middle. Color black, shining. Head minutely punctulate (as viewed under a 9x hand lens), posterior transverse impression deep, roughened a little at bottom, frontal grooves strong, rugose and punctulate, the intervening convexity somewhat wrinkled transversely. Pronotum not very well preserved, the front portion being broken off. The disk is apparently very finely and sparsely punctulate, the median groove strong, wide and deep behind (as in the recent P. septentrionis), but fine at middle, a trifle stronger again anteriorly, basal foveze mod- erately deep (less so than in P. septentrionas), strongly, closely and confluently punctured, connected by a punctate flattened area across the pronotal base. Prothoracic episternum quite strongly and closely but irregularly punctured anteriorly and posteriorly, the smoother submedian area less marked than in any of the four modern species of this genus (P. aterrimus, P. longicornis, P. septentrionis and P. californicus), with which I have been able to compare it. In general, the punctuation of this sclerite is rougher in appearance than in the recent form Wickham —Fossil Beetles from the Sangamon Peat. 141 cited. Elytra moderately deeply striate anteriorly, about as in P. septentrionis, the striz distinctly but not strongly punctate, both striz and punctures becoming finer posteriorly, scutellar stria short and very oblique, interspaces convex near elytral base, flatter on the disk, finely wrinkled and minutely sparsely punctulate, these characters of intimate sculpture being possi- bly due to accidents of preservation. Leugth of elytral fragment, > oem > Width (entire), 2-257". Probably this species is most nearly allied to the recent P. septentrionis which it approaches in size and general sculpture, differing in the points brought out in the preceding diagnosis. It is not referable to any of the three species described by Sendder from the Scarbrough clays, being larger than P. deces- sus and with short scutellar stria ; having a different prothoracic median line and basal foveee from P. gelatus; and with unspot- ted elytra in place of the profusion of pallid dots seen in P. Srigedus. I take pleasure in naming this insect after Samuel Henshaw, Director of the Museum of Comparative Zoology at Cambridge, in recognition of his numerous and varied services to recent and fossil entomology. PLATYNUS PLEISTOCENICUS new species. Represented by a single practically complete elytron of shin- ing black color without metallic luster. Striee fairly deep and strong but not punctured, all attaining the raised basal elytral margin, the scutellar slightly interfering with the course of the first discal, throwing its base over towards the origin of the second. The scutellar and first discal are practically con- fluent at the apex of the former. The ocellate punctures of the outer edge are indicated towards the apical third but are not strong. None of the dorsal serial punctures can be made out with certainty, probably on account of the peculiar trans- verse, fine, apparently adventitious, wrinkling of the integu- ment which gives the interstitial spaces the appearance of minute corrugation and, in places, of punctulation. The fifth and sixth strigz unite much farther from the elytral tip than in P. subgelidus. Length of elytron, 5°60™. None of Scudder’s descriptions and figures agree with this insect, but it would probably go nearest his P. enteritus by the form of the scutellar and adjacent strize, which, however, do not reach the elytral base in that species. The color is also different and the present insect is somewhat larger. Compared with recent North American species, the arrangement of the strie in the neighborhood of the scutellum is very similar to that seen in some Colorado specimens standing in my collection 142 Wickham—Fossil Beetles from the Sangamon Peat. under the name P. propinguus G. & H., but the surface gloss is entirely different. ' According to Scudder’s identifications, the genus Platynus was very abundant in the Scarborough beds, whence he has described eleven species. Six others, under the generic names of Platynus, Agonum and Anchomenus, have been character- ized from the Pleistocene deposits of France, Germany, Schles- wig-Holstein and Galicia. PLATYNUS SUBGELIDUS new species. Represented by a single practically entire elytron which is of a deep shining black color, the outer margin rather strongly curved, apparently muchas in the recent P. ovipennis, extreme edge sharply and narrowly reflexed, humeral angle not fully exposed and, judging from the outline of the side, probably not very prominent. Striz moderately deep and coarse basally and discally, finer apically, their punctures strong but of medium size, rounded or very little transverse and extremely close together. These punctures become much finer towards the apex, following the reduction of the striz. Interstitial areas not visibly punctate but in some lights appearing very minutely transversely corrugate, this wrinkling becoming very strong apically (through folding of the integument in fossilization) so as to obscure the tips of the strie. In consequence, it is not possible to say just where the fifth and sixth are joined, but this point is not far from the elytral apex. The scutellar stria is short, strongly punctured and not interfering with the first discal. The ocellate punctures of the outer margin, if present, are obscured by wrinklmg. Length of elytron as exposed, 6-00", in life possibly a trifle longer; greatest width, a very little over 2°00"™. A very careful comparison of this specimen with descriptions and figures of all of Scudder’s species of Platynus from the Scarborough beds has convinced me that it is not referable to any of them though approaching (but exceeding) his P. desue- tus in size. If reliance is to be placed upon the figure, P. subgelidus has much more closely punctate strie. In com- parison with recent species, P. subgelidus has elytral sculpture a good deal like that of P. crenistriatus Lec., but is larger. I do not venture to express an opinion as to which group it should enter. PLATYNUS CALVINI new species. Represented by a pair of partly overlapping elytra, shining black and moderately heavy in texture, the strize impunctate, strongly impressed but not coarse, reaching the raised elytral Wickham—Fossil Beetles from the Sangamon Peat. 143 basal margin. Scutellar stria a little less than half as long as the elytral breadth, free at apex. First discal strongly inflexed to base and almost joining the second atthat point. Fifth and sixth confluent not far from the apical fifth of the elytra. Ocellate punctures outside of the eighth stria strong behind the middle of the elytral length but not reaching the apex. On the third stria, about one fifth from the base, is a strong puncture, while between the second and third strie are two others, one a little antemedian, the other only a short distance from the apex. Interstitial areas fiat or nearly so, extremely minutely punctulate and alutaceous. Length of elytron, 5-00™™. Very closely related in size and arrangement of strize in the scutellar region to P. wnterglacialis Scudd., from the Scear- borough beds, but evidently differs in color and texture. The closest resemblance that I can find among the modern species seems to lie with the melanarvus group of Platynus, but here I do not find an exact duplication of the arrangement of the dorsal punctures. In giving the specific name, I have in mind my late friend and preceptor, Professor Samuel Calvin, who did so much to advance our knowledge of the Pleistocene formations in Iowa. CHLZNIUS PLICATIPENNIS new species. Represented by a considerable portion of one elytron, of full width but with the base and apex broken off. Color black, surface moderately shining, no signs of pubescence remaining. The striz are pretty deep but irregular, being composed of short, longitudinal, impressed dashes, much as in the recent C. interruptus. The intervals are alternating in height, as de- scribed for the modern C. alternatus, the punctures fairly deep, strong, so generally confluent as to form transverse irregular rugee, occasionally broken up into granules. Length of fragment, 6°00", greatest width, 3:00™". Apparently this elytron represents the remains of a Chleenius allied to the modern (. interruptus Horn, occurring in Ore- gon, Manitoba and the Rocky Mountains of Colorado, and C. alternatus Horn, from the Saskatchewan district. It will be noted that both of these recent forms are decidedly northern ty pes. AGABUS SAVAGEI new species. The type shows the upper surface of the head and prothorax with the two elytra detached and lying a few millimeters dis- tant. All of these fragments are black, feebly shining. The head is finely but distinctly alutaceous, with a few small, scat- tered punctures, the lines forming the posterior and inner 144 Weekham—Fossil Beetles from the Sangamon Peat. margins of the eye meeting nearly at right angles. Prosternum alutaceous, about like the head in minute sculpture, a narrow band of small punctures just behind the front margin becoming transverse rugosities in the angles, a similar but less pronounced band just in front of the prothoracic hind margin and a few scattered points on the disk. On the side which shows best, the lateral pronotal edge is beaded and very slightly reflexed. The pronotum as a whole is evidently rather strongly narrowed anteriorly, the margins nearly regularly but feebly arcuate. The front angles are sharp, strongly advanced, as shown by one which is nearly entire and the other which is somewhat more broken. The hind angles are not well uncovered but seem to be obtuse and perhaps a little rounded. LElytra alu- taceous, with moderately strong punctures arranged in irreg- ular longitudinal double series and a few inconspicuous scattered punctures in addition, marginal bead strong. Length of pro- notum 1°45™"; of elytron (not quite entire at tip) 4°60™ ; width of pronotum at broadest part, 3°45™"; of elytron, not determi- nable on account of curling. Six specimens are referred to this species, which I have named atter Professor T. E. Savage. In the features shown, A. savagei is very much like the recent A. seriatus Say, com- mon in the northern United States and in Canada. However, comparison of the present species with specimens of A. serzatus from the White Mountains and Newfoundland shows the fossil to be smaller, more strongly alutaceous and with deeper elytral serial punctures. Scudder has described A. perditus, fossil in the Scarborough beds, but calls particular attention to the lack of serial punctuation. Species of this genus are known from the Tertiary deposits of both continents and seven have been recorded from the European Pleistocene in addition to one from the Cambridge Peat. Today, Agabus is found commonly in swamp land, often burrowing in damp spots out- side of the pools themselves. | AGABUS PRELUGENS new species. The type is an elytron very similar to that of Colorado spec- imens of the recent A. dwgens Lec., in my collection. It is of a deep black color, moderately shining, finely but very dis- tinctly alutaceous, the rows of serial punctures double, quite deep but not large. The extreme apex of the elytron is broken off, but the remaining fragment is 6°40™™ in length. It differs from modern A. lugens in the entire lack of brassy reflections and in the texture of the surface sculpture. Four specimens are assigned here, all poor except the type. Wickham—Ffossil Beetles from the Sangamon Peat. 145 OLOPHRUM INTERGLACIALE new species. Represented by several elytra, 2°25"™ long, 1:10" wide, black, rather shining, subtruncate apically with the outer angle rounded off, punctuation confused, strong, moderately coarse, much of it confluent so as to form poorly defined transverse rugee, no striz visible, but the sutural bead shows faintly in some specimens. The outer margin is deflexed as in Oloph- rum, the line of flexure with a sharp edge. While the generic reference cannot be made with any cer- tainty, these elytra are apparently staphylinidous, judging from their form, size and sculpture. In all these respects they approach more nearly to Olophrum obtectum than to any other insect known to me, but are darker in color and even more strongly and closely punctate. Scudder has described three species of this genus from the interglacial clays of Scarborough, Ontario, but O. interglaciale appears to differ, by descriptions and figures, in being more strongly and closely punctured than any of these. In general, Olophrum may be considered rather boreal than otherwise in distribution. I find O. obtectwm chiefly under bits of wood in damp places and have met with QO. rotundicolle in swamp land near Leadville, Colorado. DoNAcIA STIRIOIDES new species. An elytral fragment, belongs to Donacia and resembles in fine strial sculpture a recent specimen in my cabinet collected at Coeur d’ Alene, Idaho, labelled D. pusilla Say, var. cuprea Kirby. The fossil is flattened, the strize fine but quite sharp, punctures small, not very well defined, interstitial spaces much wider than the striz, relatively coarsely transversely rugose. The color is metallic blue or purple. As exposed, the piece measures 3°85™™" in length by 1°50 in breadth. Two other small fragments are associated with the foregoing but may perhaps not be specifically identical. I cannot refer this fossil to either of the species described by Scudder from the Scarborough beds since his D. pompatica has deep strize with larger punctures and PD. stiria is said to have an exces- sively fine transverse rugulation. In North America, Donacia is much more abundant northward and D. pusilla, with which the present species has been compared, is more particularly characteristic of the country from Hudson Bay to Vancouver Island, southward to Oregon, California, Idaho, Colorado, and the Lake Superior district. The genus frequents swamp land and the shores of lakes, breeding in the vegetation common to such localities. State University of Towa, Iowa City, Iowa. 146 S. Powers— Granite in Kansas. Art. XIV.—Granite in Kansas; by Stpney Powers. Durine the last three years granite has been occasionally encountered in wells drilled in Nebraska and in the east-central portion of Kansas over an area extending from near Eldorado, Kansas, to Pawnee County, Nebraska.* These wells, starting in the same Pennsylvanian horizon, encounter the igneous rock at depths of 550 to 2500 feet, while other wells within a short distance may be drilled to a greater depth and find only sedi- mentary rocks of the normal Paleozoic succession. The granite, a medium-grained, pink, biotite type, is not intrnsive into the Pennsylvanian and must be of Karly Paleozoic or of pre-Cambrian age. In a recent paper, Twenhofel has called attention to granite porphyry, chert, and quartzite bowlders in (?) the Pennsyl- vanian strata near Rose, Woodson County, southeastern Kansas,t suggesting that the origin of the bowlders may be similar to the origin of the granite encountered in the wells. Twenhofel presents strong arguments to show that these bowlders were deposited contemporaneously with the LeRoy shales and sandstones of Pennsylvanian age. He also believes that the bowlders reached the positions where now found through the agency of ice, because “the sediments with which they appear to be associated were deposited in quiet waters— waters absolutely unable to transport bowlders of the size of those which are present. apa If the bowlders are in Pennsylvanian strata it does not seem impossible that they were derived from a buried knob of igne- ous rock such as postulated below. The bowlders are on a low anticline but Twenhofel does nut believe that they can have been derived from a granite mass in this region in Pennsy]l- vanian time, because “the strata of the region are almost horizontal and if the granite mass projecting above the present level of the bowlders were once present, it seems that some- where in the region it should still project through the sedi- ments which lie at the same level as the bowlders. There is absolutely no evidence that such is the case.”§ However, if granite is encountered in wells in Kansas at a depth of only 550 feet, it is quite probable that at places granite occurs still nearer the surface and it might have been undergoing erosion = Ke _aMOrED, On Crystalline Rocks in Kansas, Univ. Geol. Sury., Kansas, Bull. 2, 1915. He describes some of the occurrences, but denies the presence of eens in the wells. LW. H. Twenhofel, Granite Bowlders in (?) the Pennsylvanian Strata of Kansas, this Journal, xliii, pp. 363-380, May, 1917. { Idem, p. 372. S Idem, p. 372. S. Powers—Granite in Kansas. 147 somewhere in the vicinity of Woodson County during the deposition of the LeRoy shales. A list of the wells which are known to have encountered granite follows, the first well being in Nebraska and all the others in Kansas. The depth at which the granite was first found is given and, where known, the total depth of the weli is also given. It has been reported that the two wells in Nowata County, Oklahoma, three wells in Washington County, Oklahoma, and a well five miles east of Inola, Rogers County, Oklahoma, have also found granite, but it is possible that these wells were drilled through the Paleozoic series into the same pre-Cambrian granite which is found in the Ozark Mountains in Missouri. Sec. 25, T 1N, R12EK Pawnee Co., Nebraska, near Dubois, depth 550-652 feet SW+# Sec. 27, T 28, R12EK Nemaha Co., near Seneca. Two wells, depth about 600 feet NW3 Sec. 34, T 68, R11E Pottawatomie Co., near Onaga, Em- pire Gas & Fuel Co., No. 1 Albert Rokes, set 920-1680 T108, R 9E Riley Co., 4+ mile south of Zeandale, depth 958 1093 T108, R 9E Riley Co., 14 miles southeast of Zeandale, depth 945-1200 NW34 Sec. 26, T10S, R 9H Wabaunsee Co., near Zeandale, depth | 991-1093 SW Sec. 1, T118, R 9E Wabaunsee Co., near Zeandale, Km- pire Gas & Fuel Co., No. 1G. A. Root, depth 1169-1950 SW Sec. 24, T15S, R 7E Morris Co., near Kelso, Echo Oil Co., No. 1 Whiting, depth about 1970-2552 SW Sec. 18, T16S8, R12E Lyon Co., Kansas Natural Gas Co., No. 1 Miller Ranch, granite re- ported 1360-1450 SE Sec. 34, T178, R 7E Morris Co., Empire Gas & Fuel Co., No. 1 Moffitt, depth 1900-2500 , L208, R 7E Chase County, near Elmdale, Em- pire Gas & Fuel Co., No. 1 Kauf- man, depth 1875-3100 T208, R 7E Chase Co., near Elmdale, No. 1 Chase County Poor Farm, depth 1707-2501 NW34 Sec. 14, T2388, R 5E Butler Co., near Burns, Hoyt et al., No. 1 Libby, depth 2312-2502 With the exception of the well in T16S, R12E, all the granite wells listed fall into a line about 140 miles in length extending in a N.18°H. direction on the prolongation of a line A 4 Hl 7p) G bo 148 S. Powers— Granite in Kansas. of petroleum producing anticlines which includes Billings, Blackwell, Arkansas City, Augusta, and Eldorado. The prin- cipal petroleum production in Oklahoma—Kansas comes from an area of elliptical outline with the longer axis in a north-south direction 75 miles east of the southern extension of the granite axis. There is no doubt but that granite has been encountered in wells in Kansas, and possibly in Nebraska, other than those above listed, but this list will furnish some idea of the size of the area underlain by it. Anticlinal structure has determined the location of most of the wells which have struck granite, yet two of the Zeandale Hic. Wichita Sa Arbuckle Fic. 1. Outline map of Kansas, Oklahoma and portions of adjoining states showing location of granite wells in Kansas ard Nebraska with respect to the Ozark, Ouachita, Arbuckle, and Wichita mountains. wells and the Kelso well are not on anticlines. The large anticline at Eldorado, south of the Burns granite well, shows no granite at a depth of 3600 feet, nor does a well of the same depth 24 miles from Burns. Therefore, the surface of this granite mass must have arelief of about 1300 feet in 6 miles. The total lack of metamorphism in the Pennsylvanian rocks around the granite knobs proves that the latter stood as islands in the Pennsylvanian sea and were gradually buried beneath the limestones and shales. In some of the wells red shale has been noted immediately above the granite, while in others, granite bowlders, sand, chert pebbles, and weathered granite above the fresh rock show that the knobs suffered erosion in Pennsylvanian time. In the first two Zeandale wells Professor S. Powers— Granite in Kansas. 149 -Haworth secured two varieties of dark-colored schists (evidently inclusions or dikes) and in the first well he found a one-foot bed of shale 32 feet within the granite (a fissure or cave filling or a dike).* Above the granite in both these wells, there is a bed of fossiliferous shale. Pre-Cambrian granite is known in the Wichita, Arbuckle, and Ozark Mountains. The only Paleozoic granite in the region is the dike near Spavinaw, Mayes County, Oklahoma, on the edge of the Ozark Mountains,+ the known extent of which is 1200 by 50 feet. It runs along the axis of a gentle anticlinal fold in a N.80°E. direction, cutting Ordovician strata, and yet these strata are described as being free from any special metamorphic action due to the dike rock. No similar dikes have been reported. The local extent of the buried granite knobs and their lineal arrangement suggest that in the beginning of Pennsylvanian time they presented an appearance similar to that of the Wichita Mountains now or at the beginning of Permian de- position. These knobs must belong to a mountain system of pre-Pennsylvanian age and yet they are situated in a region which is supposed to have been covered at intervals and for varying lengths of time by Ordovician, Silurian, Devonian, and Lower Mississippian seas. The mountain system to which the granite belongs must have been formed either in the pre- Cambrian, the Lower Paleozoic (pre-Silurian), or in the Mississippian and is probably a part either of the Lake Superior or of the Appalachian folding—further drilling, especially in Nebraska and Iowa, should show which. A study of variations in the intensity of gravity in the region of the granite should also conclusively prove that the granite knobs are part of a large mountain structure and should delimit this structure. Up to the present time only one station in the eastern half of Kansas furnishes any data and the anomaly of this station according to the Hayford 1916 method is positive, agreeing with the anomalies of the stations in the United States situated on pre-Cambrian formations. During the Lower Paleozoic, uplifts took place west and north of the present Appalachian system in two regions: the Cincinnati axis upon which rise the Cincinnati and Nashville domes; and the Ozark axis. If it be granted that the Ouachita, Arbuckle, and Wichita mountains represent the prolongation of the Appalachian system west of the Mississippi (see fig. 1), running in a N.75°W. direction, or * Op. cit., pp. 23-26. +N. F. Drake, Proc. Amer. Phil. Soc., xxxvi, 338-348, 1698. { Wm. Bowie, Investigations of gravity and isostasy, U. S. Coast and Geodetic Survey, Special Pub. No., p. 78, fig. 12, 1917. 150 S. Powers—Granite in Kansas. making an angle of 125° with the main Appalachian trend, there is a rather striking arrangement of the Cincinnati, Nash- ville, and Ozark areas, and the Kansas granite, in an inner are with a general axis from 100 to 300 miles west and north of the inner boundary of the main Appalachian axis. The Kansas granite may therefore represent either an out- lier of pre-Cambrian structure connected with the Lake Superior region, or a Lower Paleozoic uplift from the top of which the sediments were removed in pre-Pennsylvanian time, leaving only peaks of pre-Cambrian granite to be buried by Pennsylvanian sediments just as the peaks of the Wichita Mountains at the present time represent a formerly extensive mountain range now almost completely buried by horizontally bedded sediments. Art. XV.—A New Method for the Determination of Hydrogen Peroxide ; by GEoRGE 8. JAMIESON. Tue method to be described is based upon adding a meas- ured volume of hydrogen peroxide solution to an alkaline solution containing an excess of standard sodium arsenite. When the reaction is completed concentrated hydrochloric acid is added and the unaltered arsenite is titrated with a standard solution of potassium iodate* using a chloroform indicator. The amount of arsenite found by titration is deducted from the amount taken, giving that oxidized by the hydrogen peroxide. In order to obtain a quantitative reaction with the hydrogen peroxide and the sodium arsenite it was found necessary to add sodium hydroxide in excess as directed below. It should be observed that this method is not influ- enced by the presence of organic preservatives as is the case with the well known permanganate method.t Also it has the advantage over the excellent Kingzett methodt in that both the sodium arsenite and the potassium iodate solutions are remarkably stable. These solutions can be made of known strength without standardization and used immediately which is in marked contrast with the sodium thiosulphate solution employed in the Kingzett method. Furthermore, it has been found that the iodate method gives accurate results. * J. Ind. and Eng. Chem., iii, 250, 1911. + Analyst, vili, 36. tJ. Chem. Soc., 1880, 792. Jamieson— Determination of Hydrogen Perovide. 151 In order to test the method, a solution containing 3°567 ¢. of normal potassium iodate in 1000° was used. The tenth normal sodium arsenite solution which had been made for another purpose over a year ago, was prepared by dissolving 4°948 e. of pure arsenious oxide in about 50° of water which contained 4g. of sodium hydroxide. When the oxide had dissolved, 200° of a saturated solution of sodium bicarbonate were added along with enough water to make 1000°. The relationship between the arsenite and iodate solutions was obtained by titrating a measured volume of arsenite solution acidified with two-thirds the volume of concentrated hydrochloric acid.* 5° of sodium arsenite were found equivalent to 7°5° of iodate solution or 1° of KIO, = 0°667% of As,O,. If desired, the rela- tionship of the two solutions may be calculated as follows: 1°° of KIO, = -003297 g. of As,O,+ 1° of As,O, = :004948 ge. of As,O, = 0°667° which is identical with the result obtained by titration. The first hydrogen peroxide solution used to test the method _ was made by diluting 50° of ordinary commercial peroxide to 500°. Measured quantities of the arsenite solution which must he in excess of that required by the hydrogen peroxide taken for analysis, were placed in 500% glass stoppered bottles along with 10° of a 10 per cent solution of sodium hydroxide. A measured volume of the hydrogen peroxide solution was added from a burette while the contents of the bottle were gently agitated. After the solution had stood for 2 minutes, 40° of concentrated hydrochloric acid were cautiously added. The stopper was inserted and while holding it firmly in place the bottle was violently shaken in order to separate as much carbon dioxide as possible from the solution. Then the stop- per was carefully released so as to allow the excess pressure of gas to escape without losing any solution. After adding 6-7° of chloroform the unoxidized arsenite was titrated with the potassium iodate solution with thorough shaking of the closed bottle after each addition of iodate until the end point, which is the disappearance of the iodine color from the chloroform, is reached. The amount of iodate used for the titration was converted into its equivalent of arsenite solution which was deducted from that originally taken, leaving that oxidized by the hydrogen peroxide. The following equations may be writ- ten to represent the reactions which take place : As,O, + 2H,O,= AsO, + 2H,O 2 As,O, + KIO, + 2HCl = As,O, + IC] + KCl + H,O The following results were obtained : 1° of AsO, sol. = -001701 g. H,O.. *J, Ind. and Eng. Chem., iii, 250, 1911. Am. JouR. ot eee SERIES, VoL. XLIV, No. 260.—Aveust, 1917. 1 152 JSamieson—Determination of Hydrogen Peroxide. ecH.,02. ccN/l0As203 ccK103 ccAs203 H.02 No, sol. sol. sol. sol. found Kingzett taken taken used used by H2O2. grams method 1 oe 34°9 4°9 31°6 "0537 2 15°0 34°9 5°02 31°55 °0536 "0536 © 3 20°0 46:0 6°0 42°0 0714 "0710 + 20°0 45°5 5°45 41°9 0712 5 22°0 49°9 5°90 46°0 ‘0782 0781 6 22°0 49°9 5°90 46:0 “0782 > omen These titrations were made during a period of three hours after preparing the diluted hydrogen peroxide. It is important to titrate the hydrogen peroxide solution which has been diluted with ordinary distilled water of the laboratory soon after the dilution because it was observed that the hydrogen peroxide slowly decomposed. A diluted solution of hydrogen peroxide was prepared and analyzed. It was found to contain 003639 @. of H,O, per cc. After standing 24 hours it con- tained :008551 @. of H,O, per cc. and a week later it was found to contain only : a trace of peroxide. Another solution was prepared by diluting 55° of the com- mercial hydrogen peroxide to 500° which gave the following results upon analysis: cecAs,03 ecH.0, fen ere ecKIO; sol. H.O. Kingzett No. sol. sol. used used by H.O, found det’s. ] 15°2 40°0 120 ooo "0600 "0602 2 15:0 39°9 7°35 35°0 "0595 °0595 3 P5205 39°9 FON) 30°0 "0595 "0597 4 15°20 39°8 6°70 30°3 "0600 5 has 39°8 TOO 35°13 "0597 "0598 The results obtained in each series of experiments show that the method gives accurate results. In practice, it would be recommended that a fifth normal] sodium arsenite solution along with an equivalent potassium iodate solution (10°700 g. of KIO, per 1000°) should be employed. Yale University, New Haven, Ct. Geology. 153 SOLE NPE RCOUN Tb bre WN'On. I. -Gro.Lcey. 1. The Coral Keef Problem and TIsostasy; by G. A. F. MoxtencraaFF. Kon. Akad. Wettensch. Amsterdam, vol. xix, No. 4, 1916.—Professor Molengraaff gives in this paper an interesting and ingenious hypothesis to account for the apparent submerg- ence of oceanic islands, which has given rise to barrier reefs and atolls, without having recourse to those extensive subsidences of | the ocean floor, which the Darwin—Dana theory postulates, and which has been held to be an objection to it. He gives full weight to the idea of changes in the ocean levels during Pleistocene time by the piling of ice on the land and its subsequent melting, as recently urged by Daly, but following the views of Davis, he believes that to explain the topography of the oceanic islands surrounded by barriers, greater changes of level than could be ascribed to this cause must be accepted. This tends to strengthen the Darwin—Dana theory in its demand for actual subsidence. To permit this without recourse to a general sunallenee of the ocean bottom, he supposes it to be individual in each case, accord- ing to the following hypothesis. He classifies oceanic islands and considers only those rising from abyssal depths as volcanic structures composed chiefly or entirely of basaltic material true oceanic islands. Other islands are to be regarded as those occurring in shelf seas, connected with conti- nental masses, either wholly or partly submerged, and the coral islands found in these shelf seas are to be explained by the “ gla- cial-control” hypothesis of Daly. The rocks composing these islands may be of diverse characters. The material composing the earth’s shell under the oceanic abysses is that of a basaltic substratum, called by Suess the sima, or barysphere, and upon this rests in relative flotation the conti- nental masses of a more siliceous nature and lower specific gravity, or the lithosphere. He notes that isostatic equilibrium is general over the earth, but, since there are mountain masses where anom- alies of gravity exist and isostatic compensation is not complete on the continents, he draws attention to the anomalies of gravity found on true oceanic islands and infers that none of these are isostatically compensated. From this he draws the conclusion that, whereas on the conti- nental masses anomalous mountain projections may be able to sustain themselves in virtue of a thicker, stiffer substratum, those rising from the ocean floor rest upon and are rooted in the more plastic basaltic substratum, or sima, and in the long process of time they must gradually sink down and be again welded into it. This gradual sinking down under the influence of gravity is regarded as the cause of the movement of large amount and lovg 154 Scientific Intelligence. duration which is held to be necessary to explain barrier reefs and atolls in true oceanic regions. The author suggests that a test of this hypothesis would be the finding of reef-crowned islands which had sunk to considerable depths, but admits that so far the evidence in this direction is scanty. He presents one case, however, in the Ceram sea which is held to be in the nature of the desired proof. It is obvious also that the hypothesis should apply to all volcanic islands that rise from the oceanic abysses wherever they may be found. In the Atlantic only Bermuda is held to fill this definition, and to be in the area of reef-building corals, and the fact that reef limestone was found in a boring on it to extend to the depth of 245 feet below present sea-level is viewed as agreeing with the hypothesis. The reviewer, who desires to say at the outset that he has no bias toward any particular theory as to the formation of coral islands, offers the following comments on this hypothesis. In the first place the work which has been done in recent years upon the strength of rocks in resisting deformation precludes the idea that a basaltic cone could flatten out by its own weight. There might be some lowering by the compacting of fragmental or vesicular material, but this effect should take place largely in measure as the cone grew and would soon cease. A volcanic cone of itself must be regarded as a competent structure. If the sinking takes place it must be by a yielding of the foundation on which it is placed and of the cone as a whole. It cannot be supposed that a yielding basaltic stratum can immediately underlie the sea. Moreover, a distinction must be made between material that is rock, that is a rigid solid, and that which is in any sense hquid or initially plastic. The floor under- - lying the sea must be of rock and it cannot be plastic until a depth has been reached where the resistance to deformation has been overcome. ‘The experiments of Adams show that under labora- tory conditions the crust increases in strength with depth. Or, on the other hand, if plasticity is to be referred to a change of rock from the solid to the liquid condition, this also can only occur at a depth where increasing heat is sufficient to overcome the effect of pressure, and this can be no relatively shght one. It is clear from this, therefore, that if the volcano sinks bodily into the sima, large underlying masses of the crust on which it stood must also be involved in the movement. It is also questionable whether a mass of tightly fitted earth blocks resting either on a liquid substratum, or one rendered plastic by deformative stress, would not be competent to sustain the volcanic load under the static conditions posed in the hypothesis. In order to have sub- sidence, it would seem as if differential movements in the yielding substratum must occur. But if these took place on a small scale they would be attended by outflows and quick readjustment, while the hypothesis does not permit us to assume undertow move- ments on a large scale leading to orogenic processes. It is dif- ficult to see how such subsidence as is demanded could occur without diastrophism. Geology. 155 In regard to Bermuda it may be said that since the coral rock only extends about 250 feet (nearly 75 meters) below the present sea-level, this is practically within the limit of the glacial control theory, or not allowing more than 50-60 meters in change of level according to that theory, would indicate a very minute sub- sidence from Eocene time. Molengraaff noticing the total thick- ness of the coral limestone at Bermuda of 380 feet (110 meters nearly) and that 135 feet of this is now above water, assumes both up and down movements of these dimensions but this is unneces- sary, for that part of the limestone now above sea-level is every- where remarkably and very strongly cross-bedded, indicating a pronounced dune structure. It is difficult to imagine such strongly inclined cross-bedding occurring on a small isolated island except by atmospheric action. No upward movement seems, therefore, needed and to obtain a just thickness of the limestone deposited as a marine formation the part now above water should be sub- tracted from the whole, leaving 245 feet. Considering the situa- tion of the drill hole on the outer edge of the island it would seem to the reviewer that the atmospherically weathered deposits encountered below the coral rock are most naturally explained by their being the washed-down products of land-waste lying on the outer slopes of the island below sea-level, that is a wave-built terrace, rather than as indicating soil in place, carried under by subsidence. Also the rounded forms of the pebbles in this deposit indicate much wear and transport of the material. It is difficult to see how over 300 feet in depth of soil could accumu- late on a small island without being washed off. It would seem to the reviewer, therefore, that Bermuda, stand- ing as it has since, at least, Eocene time, and no one knows how much longer, is an example of the stability of true oceanic islands rather than the reverse. One cannot avoid the impression on reading the paper and noting expressions used by the writer like these: ‘From this it follows that the continents must be considered as flows of salic composition, floating in the sima in the same way as ice bergs do in water, being submerged with about 85-95 per cent of their mass,” and “the whole sima has been called by Daly basaltic sub- stratum,” along with similar ones, that the writer conceives of the shell immediately underlying the ocean floor as being in a liquid state or possessing at all events a much higher degree of plasticity than the continental masses. To be sure he quotes from others to support his view and seems to feel that its assumption is demanded by recent researches on isostasy, but it would seem as if this were pushing the idea of isostasy to a rather extreme limit. It would at all events be of interest to test the hypothesis which the writer has so interestingly presented by the results of what recent investigations in the fields of seismology and astrophysics have taught us regarding the physical properties of the outer shell of the earth. Bs vi Pt 156 Scientific Intelligence. 2. A Study of the Magmatic Sulphide Ores ; by C. F. Tot- MAN, Jr.,and A. F. Rogers. Leland Stanford Junior University Publications, 1916, 76 pp., 20 pls., 7 figs.—The authors define “magmatic deposits” as those segregations of ore-minerals that take place under the influence of, or closely connected with, the molten stage of the parentrock. Ore accumulations accompanied by destructive pneumatolytic action, or those formed by hydro- thermal solutions, are not to be classed as magmatic deposits. Typical magmatic deposits are confined to the basic rocks. Their study has led to the hypothesis that ‘ the magmatic ores in gen- eral have been introduced at a late magmatic stage as a result of mineralizers, and that the ore-minerals replace the silicates. This replacement, however, differs from that caused by destructive pneumatolytic or hydrothermal processes in that quartz and secondary silicates are not formed at the time the ores are deposited. It is conceived that the process of formation of plutonic rocks consists of stages and that rock differentiation and ore formation are the results of an orderly series of events. The stages in the norites and gabbros which contain the magmatic sulphide ores are as follows : (1) The first minerals to form are olivine, pyroxenes and feld- spars. (2) Magmatic alteration of the silicates, as the change of pyrox- ene to hornblende, often takes place prior to the formation of the ore-minerals. (3) Later magmatic products include interstitial pegmatite material, interstitial quartz, tourmaline, garnet, analcite, epidote and calcite. (4) The introduction of the ores by mineralizers is later, in general, than the minerals of group (3) and is unaccompanied by any secondary silicates. —~ (5) Pegmatite dikes often come later than the magmatic deposits of the basic rock itself. f (6) Hydrothermal alteration, subsequent to magmatic ore deposition and which in general is insignificant, includes the development of chlorite, tremolite, anthophyllite, sericite and serpentine. (7) Lastly downward enrichment and oxidation may take place. The second part of the paper is taken up with a discussion of various deposits of magmatic ores and the third part contains the summary of the characteristics of magmatic ores and various theoretical conclusions. WE 8. Origin of Massive Serpentine and Chrysotile-Asbestos, Black Lake-Vhetford Area, Quebec; R. P. D. Granam, Kcon. Geol. xii, 154-202, 1917.—The author presents a summary of the geological structure of this important asbestos field and examines in detail various theories concerning its history. Some of his more important conclusions follow. The-serpentine area has been derived from the alteration of original peridotites. A cross-section Geology. 157 from unaltered peridotite to the chrysotile may be divided roughly into four zones as follows: (1) unaltered peridotite ; (2) partial serpentinization of the olivine, with pyroxene more or less unaf- fected ; (3) olivine completely, and pyroxene partially, serpentin- ized—this zone constitutes the massive serpentine “bands” ; (4) complete serpentinization of both olivine and pyroxene—this zone forms the chrysotile ‘‘ veins.” ‘The serpentinization has proceeded outward from joint planes, etc., through the agency of siliceous solutions which, permeating the walls on either side of such fissures, have first acted’ on the olivine content of the rock and from the resulting concentrated aqueous solutions of olivine, serpentine has been deposited in the form of microscopic fibers. The expansion attending this reaction tended to render the rock more porous and permit the solutions to reach fresh zones of the rock and carry the change further on. The result has been that in the zone immediately bordering the fissures the serpentinization of the rock has been complete, while beyond this there is a more or less gradual transition until the unaffected peridotite is reached. ‘The structure of the rock has changed progressively with the serpen- tinization. Even in the massive serpentine the structure of the original rock can at times be seen, and where, further, more or less unaltered pryoxene crystals yet remain and the magnetite is dis- seminated much as it occurs in the original peridotite. Itis only im the chrysotile vein that the original structure 1s entirely lost, even to the extent that most of the iron ore has here collected along certain definite zones. The chrysotile of the veins is believed to be the result of the parallel position and transverse attitude assumed at the very outset by the minute fibers, and of their sub- sequent growth in only one direction. The ultimate length of the fibers is limited only by the width attained by the completely serpentinized zone. W. EL F. 4. Contributions to the knowledge of Richthofenia in the Permian of West Texas; by Emit Bose. Bull. University of Texas, 1916, No. 55, 52 pp., 3 pls.,1 text fig.— A long and detailed description of the brachiopod genus Kichthofenia and the two Texas species ft. permiana and hk. uddeni. The author, after a review of the geographic and geologic distribution of the species of the genus, concludes that they are characteristic of the Permian and were especially common in the early middle Permian. C8: 5. Contributions to Geology, reprinted from the Johns Hop- kins University Circular, March, 1917, 129 pp., figs.—The first paper in this set is by Professor W. B. Clark, who gives a short history of the rise of geological surveys. E. B. Mathews treats of ‘‘ The use of average analyses in defining igneous rocks,” and J. 'T. Singewald of “ The réle of mineralizers in ore segregations in basic igneous rocks.” F. Reeves writes on “ The origin of the natural brines of the oil fields.” The other seven papers, by Berry, Gardner, Woodring, Thom, Wade, and Dorsey, deal with Mesozoic and Cenozoic str atioraphy and faunas. The pamphlet is replete with new and important information. Cc. S. 158 Scientific Intelligence. 6. Geological Survey of Alabama. Euvcrner A. Smirn, State Geologist.—The following bulletins have been issued: No. 17. Second report on the Water Powers of Alabama; by BensaMiIn M. Hatt and Maxcy R. Hatt, Consulting Engineers. Pp. 448; with map, plates and text fioures. No. 18. Preliminary Report on the Crystalline and other Marbles of Alabama; by Wiriiam F. Proury. Pp. 212; 40 pls., 20 figs. No. 19. Statistics of the Mineral Production for 1915, com- oe from the Mineral Resources of the United States by EvGEne POMEL fee Ts 1d a Bibliography of the Geology and Mining Interests of the Black Hills Region ; by Cinopnas C. O’HarRra. Bulletin No. 11, South Dakota School of Mines ; Department of Geology. Pp. 216, with map. Rapid City, South Dakota, May, 1917.— The obvious value of this bibliography is increased by the fact that in most cases brief digests are given of the papers noted. There is also a map of the Black Hills Region. 8. Story of the Grand Canyon of Arizona: a popular illus- trated account of its Rocks and Origin; by N. H. Darvon. Pp. 81. Kansas City, Mo. (published by Fred Harvey).—An inter- esting account of the Grand Canyon, made instructive by the clear exposition and sections of the writer; it is very attractive in the large number of well-chosen illustrations. 9. Bulletin of the University of Texas, 1916, No. 66. J. A.. Uppen, Director of the Bureau of Economic Geology and Technology. Pp. v, 93; 7 pls. (including map), 7 figs. Austin, Texas.—This bulletin is devoted to the Thrall Oil Field and in- cludes a chapter on this general subject by J. A. Udden and H. P. Bybee; another on the ozocerite by E. P. Schoch; and a third on the chemical composition of the Thrall petroleums by E. P. Schoch and W. T. Read. Il. MisceLttAnrous Screntiric INTELLIGENCE. 1. Kood Poisoning; by Epwin Oaxes Jorpan. Pp. 115. Chicago, 1917 (The University of Chicago Press).—This is a very readable summary of the evidence and probabilities relating to the responsibility of certain articles of food for those physio- logical disturbances which are frequently designated as ‘‘ ptomain poisoning.” It also deals with sensitization to protein foods, poisonous plants and animals, mineral or organic poisons added to food, food-borne pathogenic bacteria, animal parasites, poisonous products formed in food by bacteria and other micro-organisms, and poisons of obscure or unknown nature, including the so-called deficiency diseases. The judgment of the author appears to be sane and well balanced on many topics which are still within the range of debate. The little volume is not too technical to pre- all Miscellaneous Intelligence. 159 vent anyone having a moderate acquaintance with the biological sciences from reading it with profit. LBM. 2. Principles of Agricultural Chemistry ; by G. S. Fraps. Easton, 1917 (The Chemical Publishing Co.). 2d edition. Pp. 501.—A volume dealing with the principles and practices of scientific agriculture from the standpoint of the chemist. It includes such topics as the essentials of plant life, soils, ferti- lizers, the composition of feeds, and the feeding of farm animals. Most of the subjects are dealt with in exceedingly summary fashion, so that no one could profit adequately by the perusal of the book without considerable preliminary training in the sci- ences of chemistry and biology. Historical matter is introduced into some of the chapters. The treatise serves as a guide rather than an exhaustive presentation of a very large group of modern agricultural themes. It is unfortunate that the term “ proteid,” now generally abandoned in favor of “ protein,” has been retained in the new edition. L. B. M. 3. The Secretion of the Urine; by Artruur R. Cusuny. London, 1917 (Longmans, Green and Co.). Pp. ix+241.—No one familiar with the author’s contributions to physiological literature need be told that a volume by Professor Cushny on kidney functions is almost certain to present something of more than conventional interest. The present is one of a new series of monographs on physiology intended, as the editor, Professor Starling, expresses it, not to give an exhaustive account of previ- ous writings, but rather to afford “an appreciation of what is worth retaining in past work, so far as this is suggestive of the paths along which future research may be fruitful of results.” Accordingly we find Cushny departing from the traditional con- troversy on the theory of renal secretion and advocating what he terms ‘‘the modern view,” in which considerations of physical chemistry and physical physiology are employed to combat so- called vitalistic hypotheses. Something of the viewpoint may be inferred from this quotation: “ One part of the kidney filters off the plasma colloids, another part absorbs a fluid of unchanging composition. The kidney exercises no discrimination, but con- tinues these activities through life, just as a muscle exercises no discrimination. The kidney loses somewhat in dignity and romance when it is thus represented as merely a hard-working organ, which is admirably fitted to remove the waste products of the blood, but which is so devoid of judgment that in some con- ditions it acts to the prejudice of the organism by removing the diluent instead of the poison” (p. 56). The chapters cover the following topics: anatomy and histology of the kidney, the chief constituents of the urine and their concentration, the work, gaseous metabolism and blood supply of the kidney, theories of renal secretion, direct evidence on the functions of the tubules and glomerulus, blood supply and kidney secretion, the reaction of the urine, the action of diuretics and other drugs, glycosuria, perfusion of the kidney, albuminuria, notes on nephritis and other 160 Scientific Intelligence. renal disorders. An elaborate bibliography completes a stimu- lating monograph. L. B. M. 4, Kield Museum of Natural History; Freprrick J. V. SxirF, Director. A few of our recent circulars in the various i» departments: 3 : Geology: J-3. Genetic Collection of Rocks and Rock- c forming Minerals. J-148. Price List of Rocks. a Mineralogy: J-109. Blowpipe Collections. J-74. Meteor- e ites, J-150. Collections. J-160. Fine specimens. % Paleontology: J-184. Complete Trilobites. J-115. Collec- a tions. J-140. Restorations of Extinct Arthropods. f Entomology: J-380. Supplies. J-125. Life Histories. sy J-128. Live Pupae. i Zoology: J-116. Material for Dissection. J-26. Compara- 4 tive Osteology. J-94. Casts of Reptiles, etc. Be Microscope Slides: J-185. Bacteria Slides. a Tacidermy: J-138. Bird Skins. J-139. Mammal Skins. Human Anatomy: J-16. Skeletons and Models, . General: J-155. List of Catalogues and Circulars. # | ' Ward’s Natural Science Establishment 4 84-102 College Ave., Rochester, N. Y., U.S. A. 2 7 @ Publishers: WILLIAMS & NORGATE, 14 Henrietta Street, Covent Garden, London, W. C. G6 99 3 SCIENTIA INTERNATIONAL REVIEW OF SCIENTIFIC SYNTHESIS. Jsswed monthly (each ie number consisting of 100 to 120 pages). Editor: EUGENIO RIGNANO. ‘ “‘SCIENTIA’’ continues to realise its program of synthesis. It publishes articles : which relate to the various branches of theoretic research, and are all of general in- a terest; it thus enables its readers to keep themselves informed of the general course . of the contemporary scientific movement. ‘“SCIENTIA”’ appeals to the cooperation of the most eminent scientific men of all countries. It has- published articles by Messrs. Abbot (Washington) —-Arrhenius (Stockholm)- Ashley (Birmingham) -Bechterew (Petrograd) - Bohlin (Stockholm)- Bonnesen (/openhagen) - Borel (Paris) - Bottazzi (Napoli) - Bragg (Leeds) -Bril- louin (Paris) - Bruni (Padova) -Castelnuovo (Roma) -Caullery (Paris) - Chamberlin (Chicago)-Ciamician (Bologna) - Clark (New York) - Costantin (Paris) -Crommelin (Greenwich) - Daly (Cambridge, U. S. A.) - Darwin (Cambridge) - Delage (Paris) - De Martonne (Paris) - De Vries (Amsterdam) - Durkheim (Paris) - Eddington (Greenwich) - Edgeworth (Ozford) - Emery (Bologna) - Enriques (Bologna) - Fabry (Marseille)- Fisher (New-Haven, U. 8. A.)- Foa (Torino)- Fowler (London) - Fredericq (Liége)- Galeotti (Vapolt)- Golgi (Pavia) - Gregory (Glasgow)- Guignebert (Paris) - Janet (Paris) - Jespersen (Gentofte) - Kapteyn (Groningen) - Kidd (Oxford)-Langevin (Paris)-Lebedew (Moscou)-Lodge (Birmingham)-Loisy (Paris)- Lorentz (Haarlem) - Loria (Torino)-Lowell (Flagstaff, U. S. A.)-Maunder (Green- wich) - Meillet (Paris) - Pareto (Lausanne) - Peano (Torino) - Picard (Paris) - Poincare (Paris) - Puiseux (Paris) - Rabaud (Paris) - Righi (Bologna) - Rignano (Milano)-Russell (Cambridge)- Rutherford (Manchester)- Sayce (Oxford) -Schiapa= relli (Milano) - Seligman (New York) - Sherrington (Liverpool) - Soddy (Glasgow)- Svedberg (Upsala) - Tannery (Paris) - Turner (Ox/ford) - Vinogradoff (Moscou)- Vol= terra (Roma)- Von Zeipel (Upsala) -Westermarck (Helsingfors)-Willey (Montreal, Canada)-Zeeman (Amsterdam) - Zeuthen (Kopenhagen), and more than a hundred others. ‘*SCIENTIA’’ publishes, at present, in the section dedicated to sociological articles, a series of studies on the present questions of an international character raised by the war. ‘“SCIENTIA”’ publishes its articles in the language of its authors, and joins to the principal text a supplement containing the French translations of all the articles that are not in French. (Write fora specimen number.) Annual Subscription: 24 sh. post free. Office: Via Aurelio Saffi, 11- MILAN (Italy). CONTENTS. Art. 1X.—Physiographic Development of the Tarumai Dome in Japan; by Hipezé Simoromar (TANAKADATE) - X.—Lavas of Morro Hill and Vicinity, Southern California; by.G. A. :Wapine and ©.:A.- Warne. =e ee XT.—On Tri-Jodide and Tri-Bromide Equilibria, especially in Cadmium Solutions; by R. G. Van Name and W.G. BROWN 222200 2 288 Fe YESS Cae ter eee ee XII.—The Environment of the Amphibian Fauna at Linton, Ohio; “by 69C SGasi 20 3 a ae ee en ee XIlI.—Some Fossil Beetles from the Sangamon -Peat ; by HL Fe Wierwam 2) ee XIV.— Granite in Kansas; by 8. Powers _...-- .---=- eens XV.—A New Method for the Determination of Uy drogen Peroxide; by G. S. JAMIESONIG. soos ee eee SCIENTIFIC INTELLIGENCE. Page & Si 98 Geology—The Coral Reef Problem and Isostasy, G. A. F. MoLENGRAAF, 1538.—A Study of the Magmatic Sulphide Ores, C. F. Toitman, Jr., and A. F, Rocrers: Origin of Massive Serpentine and Chrysotile-Asbestos, Black Lake-Thetford Area, Quebec, R. P. D. Granam, 156.—Contribu- tions to the Knowledge of Richthofenia in the Permian of West Texas, EH. Bose: Contributions to Geology, 107.—Geological Survey of Alabama, E. A. Smita: Bibliography of the Geology and Mining Interests of the Black Hills Region, C. C. O’HarRa: Story of the Grand Canyon of Ari- zona, N. H. Darton: Bulletin of the University of Texas, 1916, No. 66, J. A, UDDEN, 108. Miscellaneous Scientific Intelligence—Food Poisoning, E. O. Jornpan, 158.— Principles of Agricultural Chemistry, G. S. Fraps: The Secretion oi the Urine, A. R. Cusuny, 159—Field Museum of. Natural History, Annual Report of the Director, F. J. V. Sxirr: Chemical and Biological Survey of the Waters of Illinois, E. Bartow: British Museum Publications, 160. Obituary—T. McK. Hughes: H. T. Kennedy, 160. fF VOL. XLIv. SEPTEMBER, 1917. . a Established by BENJAMIN SILLIMAN in- a a woe. fe Ayn son! an 7 VAS AMERICAN || JOURNAL OF SCIENCE. . || _ Eprton: EDWARD §. DANA. 4 , +: Nie ASSOCIATE EDITORS 3 Proressors GEORGE L. GOODALE, JOHN TROWBRIDGE, : _ W. G. FARLOW anv WM. M. DAVIS, or Camprincez, Proressors ADDISON E. VERRILL, HORACE L. WELLS, ei LOUIS V. PIRSSON, HERBERT E. GREGORY. ee a : anp HORACE 8S. UHLER, or New Haven, Pipe cox HENRY S. WILLIAMS, or Iruaca, Proressor JOSEPH:S. AMES, or Battimorg, Mr. J. S. DILLER, or Wasuineron. es Ee, GS ee: FOURTH SERIES gq i VOL. XLIV—[WHOLE NUMBER, CXCIV]. 2 No. 26i1—SEPTEMBER, 1917. = WITH PLATE I. 4 i. e NEW HAVEN, CONNECTICUT. a 1917. 4 THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 128 TEMPLE STREET. te " Published monthly. Six dollars per year, in advance. $6.40 to countries in the i Bocca Union ; $6.25 to Canada. Single numbers 50 cents. sf ees as Sens class matter at the Post Office at New Haven, Conn., under the Act List of Choice Specimens and Minerals from New Finds and New Discoveries. Hibbenite, Salmo, British Columbia. Rare and new mineral, a zinc phosphate described in the September, 1916, issue of the American Journal of Science by Prof. Phillips of Princeton University. $4 to $15. .Kaemmererite, near Murphy’s, California. Sharply developed crystals in matrix, new find. $4 to $12. Caleite twins with phantoms (sharp and distinct), Stagg, Cali- fornia. 24 to 384" in diameter. $1 to $3. Dumortierite, Oreana, Nevada, new find. Crystallized, showing dis- tinct crystals, very rare, from 3” to 5". $5 to $12. Wiluite with Achtaragdite, Wilui River, Siberia. Crystals are coated with the rare mineral achtaragdite. One specimen 34 x 24” with five crystals embedded 1” to14" long. $20. Another speci- " men 34 x 4" with three crystals embedded from 2” to 13” long. $18. Very choice specimens. Anglesite, Gemini Mine, Tintic Distinct, Utah. Museum specimen 33" x 81"; five large crystals embedded from}to1" long. Very fine. $18. Emeralds, Muzo Mine, near Bogota, Colombia, South America. I was fortunate in securing some very fine specimens with crystals of good color and fine termination : Specimen 1}” x {"; one crystal embedded 3” in diameter and projecting gs - $20. Specimen 12 x 14"; one large crystal laying on the matrix 4" long and ~;' in diameter, doubly terminated and with some eae $25. Specimen 1%” x 14"; two erystals embedded, one crystal 3," in diameter, the other 2" in diame: both projecting about 2’. $35. eben 2 x 2"; one ieee crystal embedded, ;%" in diameter, project- ing +4"; crystal is of parallel growth of good color and fine termina- tion. $40. I also have a few other fine specimens up to $200. Crocoites, Dundas, Tasmania. I just received a fine lot of matrix speci- mens priced at from $6 to $10. ALBERT H. PETEREIT 81-83 Fulton St., | _ New York City 172) it Am. Jour. Sci., Vol. XLIV, 1917 Halemaumau, the lava pit of Kilauea. THE AMERICAN JOURNALOFSCIENCE [FOURTH SERIES.] +o Art. XVI.— Voleanologic Investigations at Kilauea, with Plate I (frontispiece)*; by T. A. Jacear, Jr. CONTENTS INTRODUCTION ; MAGMATIC GASES. MECHANISM OF HEATING. Is lava lake hotter below surface ? Atmospheric oxidation of magmatic gas. EVIDENCES OF CONVECTION ; THE DUPLEX Lava CoLuMN. Variable surface radiation. EVIDENCES OF SHALLOWNESS OF Liguip Lava LAKE; THE BENCH MAGMA. Lava islands. Shoals, sinkholes and conduits. Summary, duplex lava column. Consistency of bench magma. EVIDENCES OF HEAT FROM GAS OXIDATION, Types of flames. Gases responsible for flames. Distribution crusting and fountaining. Problems of fountain mechanism. Mechanism of different types of fountains. Construction upon lake bottom. Sinkhole cascades. Condensation by de-vesiculation. Multiple fountaining of 1910 to 1912. Disappearance of liquid lava during low levels. Heating mechanism in bench magma. EXPERIMENTS TO DETERMINE DIFFERENTIAL TEMPERATURES, Queries concerning temperature. Reconnaissance and method. Temperature of lava lake. Temperatures of grottoes and flames. Summary of temperatures. Refusion. Furnace effect. Relative coolness of lake. EXPERIMENTS TO DETERMINE DEPTH AND CONSISTENCY. Queries concerning depth and consistency. Measurement of depth. Confirmation of soundings by subsequent subsidence. Summary of depth and consistency. CoNcLUSION. * For explanation of Plate I see bottom of p. 162. Am. Jour. Sct.—Fourrn Serres, Vou, XLIV, No. 261.—SrepremBer, 1917. 12 162 Jagyar— Volcanologic Investigations at Kilauea. InrTROopDUCTION : MacmatTic GASEs. Tue immediate work of the Hawaiian Volcano Observatory since it was established in 1912 has been concerned with record- ing rather than theorizing. Confirmation or refutation of existing theories is greatly dependent on a knowledge of habits and of the sequence of changes which happen in the course of time and the comparison of charts showing these sequences, with other charts that show such changes as those of rainfall or numbers of earthquakes, or the tides or movements of the sun and moon. It is by comparison of time changes that such great sciences as astronomy and meteorology have been built up. Volcano science has had no such records and our observa- tory is trying to supply the defect. Fortunately, however, five colleagues who have worked at Kilauea have written important articles suggesting theoretical possibilities concerning the mechanism and chemistry of the gases and the lava. Three of these papers deal with the ancient problem of water vapor as a cause of volcanic activity. Dr. Albert Brun* of Geneva believes that water is unessential, Professor Dalyt of Harvard and Mr. F. A. Perrett of Naples believe that gases rising from heated magma in the depths are the main heating and liquefying agents in liquid lava, without being prejudiced as to the amount or origin of the hydrogen and oxygen which undoubtedly exist among these gases. Drs. Day and Shepherd,§ of the Carnegie Institution, after chemical work at Kilauea for two seasons, published results showing that steam certainly exists in the gases blown out from flaming cones on the Halemaumau floor, but the proportion by volume of water among the gases collected was only about four per cent, while the dominant ingredients were sulphurous acid, * L’Exhalaison volcanique, Paris, 1911. + Proc. Amer. Acad. Arts and Sci., xlvii, No. 3, 1911. ¢ This Journal, vols. xxxv—xxxvi, 1913. § Bull. Geol. Soc. Amer., vol. xxiv, pp. 578-606, 1918. EXPLANATION OF PLATE I (fig. 1) frontispiece.—General views, interior of Halemaumau, the lava pit of Kilauea voleano. (a, upper view) Aug. 26, 1915, 9 A.M. Upper surface duplex lava column, Halemaumau pit, from S8.W. rim. Diameters 780 by 730 ft. (238 by 222 m.). Length liquid lake 700 ft. (218 m.), maximum width of same 220 ft. (67 m.), and depression of lake 410 ft. (125 m.). Bench clinging to wall left 18 ft. (0 m.) and general floor level 6 ft. (1°8m.) above lake. The clinging bench marks lava level of 1914, mossy appearance due to mat of Pele’s hair. Western conduit pond on left, sinkhole niches middle and right. Streaming left to right ; minimum incan- descence, fountaining and fuming left; maximum right. Rising activity, crags of bench magma crust tilted and fissured on the right by weight of overflows shown. Panchromatic photo by Jaggar, camera inclined forward. (b, lower view) Jan. 31, 1917, 5 p. m. Depression 45 ft. (18 m.). Diam- eter 1200 ft. (866 m.). Halematmau from §.E., showing islands and benches on the left, overflow floors on the right. Swift current around South Island in foreground. Strong overflow in progress, culmination of winter rise. Compare map, fig. 2, p. 168. Photo, Morihiro. JSaggar— Voleanologic Investigations at Kilauea. 162 carbon dioxide and nitrogen, with a residue of about seven per cent rather equally divided among the inflammable gases sul- phur, carbon monoxide and hydrogen, the latter predominant among these. There was in addition less than 0:1 per cent in all of fluorine, chlorine and ammonia. All of these workers agree that rising gas achieves the work known at the surface of the earth as voleanic activity. I think that they all agree that the gases rise from deep sources, and so far as the water problem is concerned, they all agree, in opposition to the older geological text books, that gases other than water in large measure heat and operate the volcanic engine. In this they verify a conclusion reached many years ago by Dr. Wm. T. Brigham,* Director of the Bishop Museum. and Mr. Wm. Lowthian Green.t Their differences of opinion concern the extent to which the combinations with oxygen above listed are original gases or products of union with air, and more especially the extent to which surface heating, by chemical combination among these unstable mixtures, is respon- sible for the liquid lava pools and flows. Brun believes that original carbon in the form of hydro-car- bons exists in lava, and nitrogen combined with hydrogen in the form of ammonia. He insists that there is no original water from deep-seated sources emitted by lavas. Perret believes that the oxides and hydrates which come forth as gases are the result of union with superficial air and water but that the unadulterated gas from the deep region is more ele- mental and is frequently quite breathable in great volcanic explosions, whereas the oxidation products are disagreeable or poisonous. In some eases, such as Vesuvius and Stromboli, he has breathed the rush of gas from great explosions and per- ceived no chlorine, sulphur or poisonous carbon compounds. This agrees with some recent observations of the writer in Hawaii, when a few feet from and immediately to leeward, of the Halemaumau lava lake on its shore (fig. 12), of a pahoehoe overflow on its border and of an aa flow on Mauna Loa, he found no difficulty whatever in breathing the intensely hot products of small bubblings all over these glowing surfaces and perceived almost no sulphur odors; whereas at a greater dis- tance to leeward of a positively flaming grotto or cone, the bluish fume which condenses is full of intolerable compounds of sulphur with oxygen. The Selby commission determined that one ten-thousandth part SO, in air is intolerable to human beings. It hardly seems probable, therefore, that as much as 50 per cent of the magmatic gas rising directly from fresh lava can be SO,+ (see below). * Kilauea and Mauna Loa, Mem, Bish. Mus., 1909. + Vestiges of the Molten Globe, Honolulu, 1887. + Bull. Haw’n. Vol. Obs’y., Sept. 1914, p. 121. 164 JSaggar— Volcanologic Investigations at Kilauea. MrcHANISM OF HBATING. Is Lava Lake Hotter Below Surface ? Daly points out that a voleano may be a true furnace in that heat-producing chemical reactions necessarily take place where free hydrogen is present. Nevertheless he considers that the heat, generated partly from reaction between the gases under- ground, is continuously distributed by ‘ two-phase” convection with a cooler liquid phase of the melt sinking while a hotter and lighter gas-bubble phase is rising. The rising gas by expanding would have a strong cooling effect, so that if there were no compensatory heat reaction, Daly calculates that a bubble rising at the surface at a temperature of 1200° C. would at depth of 37 meters (120 feet) have a temperature over three times as great or 8700° C. He finds that the loss of heat at the surface of the lava lake is vastly greater than the heat lost by conduction into the wall rock enclosing the volcanic pipe. Accordingly the heavy surface lava, losing its dilating gas and growing denser for that reason and by cooling, sinks at the grottoes and fountains, and pours in subsurface currents down- ward. This hypothesis necessarily makes the lava lake hotter and less dense below the surface for several hundred feet of depth, if the lake is over the conduits. Perret also clearly expresses belief that the lava lake is hotter below for he conceives the islands as floating and extending “to a considerable distance below the surface where the tem- perature and the chemical activity of the lava are much greater.” With reference to chemical activity Perret writes: ‘These gases which issue from the liquid lava of a volcano are not to be considered as juvenile gas in its primal state, but that which, expanded into and worked over with the lava in the voleanic edifice, is subjected to the action of air, water and oxidizing and transforming processes of the most complicated kind resulting in the formation of those oxidized und hydrated compounds of sulphur, carbon, chlorine, ete., which constitute the gaseous emanation of ordinary voleanic activity.* Day and Shepherd, on the other hand, reached the conclu- sion that such oxidized gases as water vapor and sulphurous acid are primal, and that chemical action is still going on among the gases, that at these temperatures free sulphur could not remain inactive in presence of carbon dioxide nor free hydrogen in presence of both of those and sulphur dioxide in addition ; ‘the heat generated . . . . may well be much more than sufficient to counteract the cooling effect of the expansion within the rising lava column, which may thus become hotter and not cooler as it approaches the surface.”+ The absence of * Loe. cit., this Journal, xxxv, p. 146, 1913. + Loc. cit., p. 600. The present writer has measured the thermal gradient of the lava recently (Jour. Wash. Acad., July, 1917). There is hot surface reaction, a sub-surface cool zone, and rise of temperature next below that, until the bottom lava is reached. Jaggar— Volcanologic Investigations at Kilauea. 165 equilibrium among the gases at the surface of the lake is shown by great variety in proportions of the individual gases in dif- ferent tubes collected at the same time, and by increase of maximum temperature in the fountains at those times when the quantity of gas is greater. Discussing the water which condensed abundantly in the tubes, these authors believe it could not have come from reaction of hydrogen with air because such a quantity of hydrogen would produce explosion. But this was all on the stated assumption that the flaming gases had met no atmospheric air below the surface cracks where they were collected. Atmospheric Oxidation of Magmatic Gas. Without entering here upon an exhaustive discussion in criticism of these several writers, I would point out that, except for the statement quoted from Perret, these investiga- tors seem to me to take no sufficient account of the certain and obvious reaction of the volcanic gases with oxygen of the air, nor of the extent to which sulphur, hydrogen and carbon may by less obvious mechanism be brought into contact with air within the edifice of highly porous rock that encloses the lava conduit for some thousands of feet above sea-level. Day and Shepherd, however, point out that in the gas reactions there is an enormous store of volcanic energy “which reaches its maximum temperature at the surface itself.”* The abundant flamest through cracks in crusts over the lake, over the grotto fountains and central fountains, through the border cones, and the blowing cones which form above cracks in the floors, and the myriad flaming orifices even at the fronts of some of the flows (block lava of Mauna Loa 1916),t are the obvious evi- dences of reaction between gas and air. A second method of indraught of air downward into the lava is produced by downflow at the grottoes, and in times of subsidence at the sinkholes, when violent cascades rush down border pots from the lake, the cataracts tumbling vertically 30 or 40 feet (9 or 12™) into a boiling, flaming and fuming cauldron (fig. 90), car- rying down the surface crusts, and obviously engulfing air by downsuction as in a waterfall. A third mechanism which carries air downward into the liquid lavas of lakes and flows = Op cite np, GOO: + Dr. Wm. T. Brigham was the first to point out these flames, which were never positively seen by Professor Dana, though they were eventually accepted by him. (Characteristics of Volcanoes, 1891.) The insistence by Brigham on flames, and by Green on absorption of air, against geological opinion which would have it that steam must be the active agent, illustrates the great advantage in science of continuous observation over closet theoriz- ing. Brigham, Green and Coan lived on the field and knew their voleano ; the foreign geologists came for short visits, intent on seeing, with precon- ceived opinion for a guide. ae {Lava flow from Mauna Loa, 1916, by T. A. Jaggar, this Journal, xliii, pp. 200-288, April, 1917. 166 = Jaggar— Voleanologic Investigations at Kilauea. is furnished by the cracking, streaming and foundering pro- cess (fig. 12d) whereby great quantities of porous hardened lava crust carry air down into the gas-charged melt and probably discharge it rather slowly, owing to the sealing of the pores by chilled glass as the slabs first sink in the liquid just below the surface. This hquid is at a temperature so low (see experi- ments hereafter) as to be solidified by such contact, and has no power to melt up the deeply chilled crusts which are several inches thick. A fourth mechanism which brings oxygen and volcanic gases into contact beneath the surface of the lava col- umn is furnished by all those cavings-in and crackings of the older rock whereby avalanches (fig. 4a) or single blocks are precipitated into the liquid lava, er where the liquid lava by percolation through newly-opened crevasses, or by “ stoping,” gains access to broken surfaces of fragments filled with air. ‘A fifth process, not at all obvious but possibly very effective, is the indraug ht of air from the wall through the pores of the lava column, creating a blast furnace and incessantly com- pensating a tendency to vacuum created by chemical reactions,* or by convectional gas pumping, within that column. It is interesting to note that Wm. Lowthian Green in his “ Vestiges of the Molten Globe” held that the fountaining of the lava at places of descent was due to air being carried down with the lava and expanded. He lays great stress on the quan- tities of air carried down in the lava and even insists that vesi- culation is due to air. Expansion of air would not account for the flames observed, and the temperature is not high enough for dissociation of water vapor, which Green insisted on. Neither will quiet foundering of crusts, however, nor a lower- ing of the general level by escape of gas, as suggested by Per- ret and Daly, account for the violent downsucking at the fountains and grottoes. This action is frequently sudden, or graduated from slow to fast in a very short distance. Continu- ous or spasmodic disturbance of equilibrium of the lava by oxidation of gases with devesiculation would account for the phenomena observed. The chief gases collected at Kilauea in 1912 from a flaming cone on the floor by Day and Shepherd had the following approximate average compositiont by volume for 1000 liters of gas which was pumped : A CW es E Mea teh Dae ME MLS a heap ec yes Sul phintdioxitle> 27. e: sis 50) Carbonsdioxide +. 7s) = 25 *Dr. Shepherd informs me that in the reaction 2H, + O. = 2H,0, the vol- ume shrinkage is one-third even at high temperatures. + Kindly estimated for the writer by Dr. A. L. Day. The SO: is doubtful, much of it being probably free S. Assuming derivation from air, this partly accounts for the excess of combined O for the N present. Saygar— Volcanologic Investigations at Kilauea. 167 Carbon’ monoxide: +2. 2.225. 2 1 DSCC Nerd2) 019 es © ey ane ee a 3 SUI pli ste i=... 8. eee es 2 INNCTOOENES 2 so). ee ge 14 These are in two groups of three each, namely carbon mon- oxide, hydrogen, and sulphur, all inflammable, and the three produets of the combustion or oxidizing of these, namely car- bon dioxide, water vapor, and sulphur dioxide. The only other abundant ingredient is nitrogen (14 per cent), which is a constituent of air. The lava was splashing within the glowing spatter cupola whence the gas was drawn. ‘This lava was below the level of a porous “shattered floor” made of lava shells. A channel “just below the surface crust” connected this pot with the lake directly. The roof of this channel collapsed to a chasm a few days after the gases were collected. “As the larger bub- bles rose and burst from the liquid lava within the dome, the jar could be felt on the floor where the collectors stood and a splash could be plainly seen through the cracks.”* In other words, very hot gas only 7 per cent of which was combustible and 79 per cent of which was a combustion product, was rush- ing out through the cracks of the dome above the level of a tunnel leading to the open, cooler air above the lake and all the surrounding floor was porous shattered rock shells full of air above the liquid level and doubtless communicating through scores of openings with the tunnel and the dome which throbbed and jarred. And in addition, the lake, consisting of the same gas-charged lava as that in the dome, was incessantly engulfing air-filled crusts and skins and doing so with especial vehemence, by streaming in the direction of this grotto channel, and thereby also pumping air through the tunnel. Under these circumstances it is impossible for the writer to conceive of the 79 per cent of oxidized gas as having had ao contact with air before emerging from the blowing cone. There is no question but that this was the most perfect experi- mental collection of voleanic gas ever made, but I do not believe that any apparatus at the surface of the voleano can collect uncontaminated juvenile gas. The writer does not wish to be understood to imply that a/Z the oxygen compounds in the gas are atmospheric combustion products, but that a larger propor- tion of them are so than these authors have admitted. The facts demonstrated, however, by the Day and Shepherd collec- tion, of the presence of the three combustibles and the three combustion products, and the argument which they present of heat reactions that reach a maximum at the surface, make a contribution to voleanoiogy which bids fair to revolutionize the science. * Loc. cit., p. 588. 168 Jaggar— Volcanologie Investigations at Kilauea. Fig. 2 “ig mm . —— 2 OVERFLOW BENCH z HALEMAUMAU JAN.(2 97 4 HORIZONTAL ANO VERTICAL SCALE SS et FEET = ETERS CONTOUR INTERVAL IC FEET LAKE DEPRESSION 87 FEET CONDUIT CRAG MASS 5S. SLANO SINAHOLE = LAKE epee 12 17 = == om \ “BENCH MAGMA _ TAs \ LAKE JUNE 23,1918 { Jaggar— Volcanologic Investigations at Kilauea. 169 EVIDENCES OF CONVECTION ; THE Dupiex Lava CoLtumy. Daly terms “ two-phase convection ” the mechanism whereby deep gases rise as bubbles with increased rapidity toward the surface of a lava column and the liquid, released by surface collapse of the bubbles and loss of the gas, tends to sink because it is heavier than the rising froth. Both he and Perret made diagram sketches* showing the lava basin in the Kilauea pit to be shallow and saucer-shaped with one conduit beneath smaller than the visible pool. If conduit remains small and saucer remains shallow, the pool being bordered by a bench of overflow within the pit, how is it that the bottom of the saucer is built up while the lava surface rises 600 feet in six months? (See section, fig. 2.) As will be shown below, there is every evidence that just this relation, but with several tubular conduits, existed from June to December, 1916, and that the lava lake remained con- tinuously shallow. The bottom of the lake builds up by accretion of the relatively cooled and denser sinking surface layers of the convection, step by step with the building up of the shore bench or “floor” by the spasmodic overflow which takes place from time to time owing to the inflation of the liquid part by the rising gas-bubble currents of the convection. The obvious surface evidence of such convection Jies in the persistence of deep wells at certain fixed sites from which vesiculated lava springs up and streams outward across the surface of the lake, while on or near the opposite shores the solidified crusts and skins founder in the grottoes and the fountains. There have been such source wells or spring holes at the west and north sides of Halemaumau for many years and the dominant convectional streaming has been away from them. ‘That they are definite wells in the fresh bench lava has been repeatedly demonstrated at times of faster lowering of the level of the lake, when they were revealed as local small pits. The inflow pits frequently persist for long periods as separate ponds of lava (fig. 6a). : * Loc. cit., Daly, p. 77; Perret, p. 340. Fic. 2. Map and diagrammatic section of Halemaumau, Jan. 12, US are Lava lake in black, crusted conduit ponds shaded, overflow benches diag- onal lines, raised crags contoured. Coarse dotted outline lava lake of Feb. 18, 1912. Fine dotted outline June 23,1916. Rectangle (5) site of lava spring of June 5, 1916. Rectangle (6) west corner of pool June 6, 1916. Note that N.W. corner has been conduit source on all these dates. Slight slope lake surface from conduits W. to overflow bench E. Bench magma elevated on conduit side W.S.W., subsided on sinkhole side E.N.E. Sec- tion without vertical exaggeration, lower profile shows simple rising pool of June 23, 1916. Shoal shown in lake bottom, upper profile, was revealed by subsidence February, 1917. Depths from soundings and subsidence records. Note progressive shoaling from W. to E. Diagrammatic sinkhole E. shows ridge of accretion on lake bottom margin which produces cascade ledge when subsidence takes place. Surveys with transit by T. A. Jaggar. Bench marks (B.M.) U. S. Geological Survey, trig. stations Hawaiian Volcano Observatory. Meridian approximately 155° 17’ 8" W., lat. 19° 24’ 33" N.; 10-foot contours above lake as datum plane. 170 = Saggar— Volcanologic Investigations at Kilauea. Variable surface Radiation. It has been said that the maintenance of the lava lake in a fluid condition is product of an incessant struggle with cold.* The two-phase convection hypothesis imagines a lagged system like a hot-water heating plant with circulation evenly main- tained by rising of hot fluid to the radiator (crater) and un- obstructed sinking of cooler fluid to the furnace. Daly has not discussed the possibility, which seems to me a certainty, that uneven radiation of the system at the surface is what builds the annular bench around the lake and the semi-solid hot lava column within the pit and under the lake, with per- forations through it at the wells of uprising foam. The meven radiation is due to shifting and changing radiators, the craters, which vary continuously in size, shape and debris accumulation as the lava rises, overflows, falls or shifts its vent; and which themselves occasionally by gas reaction become localized heaters. In other words, the downflow material of the convection becomes very viscous near the surface and actually hardens a¢ the surface in the form of overflow benches and islands. When the net effect for a given period is a rising In a crater which widens upward, the downflow column, plastic but stiff within, may build by accretion under and around the hot froth column, and either encroach on the latter or the reverse accord- ing as the heat supply and the cooling are balanced or not (Plate I and fig. 2). It is evident that if this view is correct, then a profound subsidence of the entire lava column, induced at Kilauea by al extraneous cause, like rock tide stress or the relief by over- flow of connecting tubes at Mauna Loa, should carry down the semi-solid lava body as well as the more liquid lava lake. This is just what happened at the time of the great sinking in Halemauman which took place June 5, 1916, simultaneously with the conclusion of the lava flows on Mauna Loa. The lake for two days resolved itself into a shallow streaming puddle, and remained so for 400 feet (122 meters) of sub- sidence. The supporting floor of the liquid lake and the islands which protruded through the lake, sank steadily and undermined the peripheral bench, which tended to cling to the outer funnel walls of the rock pit, so that the bench and lower walls crashed inward in great avalanches. The bench rock was incandescent and semi-solid; the old walls of the encompassing pit were not. For two days of subsidence, the valine slopes from such tumbling were always supported by the sinking lake bottom. When avalanches fell into the lake its shallowness was instantly revealed by the wave which earried all of the liquid up the far slope and by distribution congealed it. The * Daly, loc. cit., pp. 68; 71 and 92. Jaggar— Volcanologic Investigations at Kilauea. 171 islands, mere crusted summits of irregularities on the main lava column, were eventually buried under the talus. The foaming gaseous member of the column appeared towards the end as a lava spring, trickling down the western talus like a mountain torrent (fig. 2, loc. numbered 5), and within a few days this and other springs cascading into the lake restored to it its central position in the pit (fig. 2, loc. numbered 6 west corner lake June 6) surrounded by debris slopes (fig. 4a). There is reason to believe that during the progress of a com- plete eruptive cycle of Kilauea between two repose periods, like the interval 1907-1918, the adjustment of rising gaseous lava to sinking viscous lava becomes most perfect at the cul- mination of the eruption, when effervescence is general and uniform, surface heat is distributed and at a maximum, and border benches are absent. Such a culmination was reached in December—January, 1911-12, when the lava lake extended from wall to wall of the pit, and the subsidence thereafter for eighteen months involved a very rapid dwindling in the size of the liquid phase (fig. 2, heavy dotted outline) and a final retirement of the entire column to the depths amid the wreckage of the solidified portion. EVIDENCES OF SHALLOWNEsSS OF Liquip Lava LakE; THE Brencn Macma. Lava Islands. The writer has studied the formation of islands in the Hale- maumau lava lake repeatedly during the last five years, and has determined that they are capable of extraordinary shifting of position both vertically and laterally in the course of twenty- four hours. He has never, however, seen any evidence that they floated as independently buoyed objects in the liquid part of the lava column, after the fashion of an iceberg in water. That they could not be so buoyed is shown by the facts that they do not equally rise and fall with the risings and fallings of the lake, that they tend to acquire and retain fixed positions, by survey, while the convection currents stream about them, and that movements of tilt in the island escarpments have been repeatedly diagnosed, the tilt according in direction with a place of overloading of the border bench with lava flows. They are thus integral with the shore and bottom of the lake. Their horizontal shiftings are rare, when they move pivotally as crust blocks of the bench magma. Also these islands pass by gradations into craggy scarps in or partly in the border bench where all stages of their forma- tion have been photographed in sequence. Here the relation to weighting by overflow has been repeatedly demonstrated. The effect is strikingly like the supposed relation of sedimenta- tion to isostatic uplift. On this small scale, moreover, the 172 Jaggar— Voleanologie Investigations at Kilauea. mechanism is indeed isostatic, for the uptiltings and upheavals of bench blocks and lake bottom in Halemauman are nothing more than adjustments of the crust of the semi-solid lava column, in which the lake is a saucer, to the mobile matter beneath which tends to flow and rearrange the surface features whenever overweighting takes place locally. Thus heavy flows repeated on the bench northwest for weeks, August to October, 1916, heaved up out of the shallow bottom of the lake a large crag mass, which eventually stood as a hill 70 feet (21 meters) above the lake, but began as a small flat islet (Pl. 1d, figs. 8, 4,5). Later asimilar islet suddenly appeared in the southeast corner of the lake immediately opposite a shore of continuous overflow, and this rapidly rose until it was a pulpit rock 40 feet (12 meters) high (Pl. Id and fig. 7). The border bench itself on the east, fissured off from its southeast extension, rose gradually for five months after August, 1916, its surface tilting to the northeast where there was ceaseless overflow, and finally lifted a pinnacle corner clear above the edge of the pit while its back slope was inclined at forty-five degrees (figs. 8@ and 10). Most remarkable of all was the adjustment of February 18, 1917, when after a fortnight of subsidence of other features, this crag suddenly parted from its supports against the old wall and in the course of twelve hours or less subsided 380 feet (9 meters) upon its viscous foundation, and at the same time opposite to it in the lake a low flat island of the previous day rose 40 feet (12 meters) above the lake to become a towering, flat-topped, steep-sided mass (fig. 8b, see also S.W. islet fig. 5). In appearance this mass was like the lava dome of Tarumai in Japan, which rose in 1909.* The conclusion was inevitable that deep flow from beneath the crag became inflow beneath the island, the distor- tion affecting, not the lake, but the lake bottom. Other rising features corroborated this by survey. The cause of this reverse movement was the unloading of the lake bottom by the faster sinking away of the liquid. The equilibrium, which had balanced lake bottom versus border bench while the basin was full, was disturbed when the lake sank differentially to its saucer. Shoals, Sinkholes and Conduits. Other evidences that the lake of liquid lava is at all times shallow, whatever its depression within the pit, were furnished by the afore-described subsidence of June 5, 1916, and by rela- tively sudden subsidences at other times. On June 5, 1916, a shoal appeared of glistening black lava flats, not infallen debris, after sudden subsidence of 60 feet (18 meters). Tn early Feb- ruary of 1917, rapid subsidence from depression 45 feet (14 * See article by Simotomai in August number. Jaggar— Volcanologic Investigations at Kilauea. 173 meters) to depression 94 feet (29 meters) revealed shoals over the whole southern arm of the lake (fig. 9). In the midst of these shoals was a circular sinkhole 60 feet (18 meters) in diameter, which became the scene of a spectacular downpour- ing of the liquid part of the melt, through a river-like channel from the remainder of the lake, across the shoals. This revela- tion of sinkholes, often with vortical whirling (fig. 3), is a common feature of sudden subsidence. Hight minor pits were revealed by the subsidence of February, 1917, four of them Fie. 3. Fie. 3. Feb. 2, 1915, 8.30 p.m. Hast pool of Halemaumau, dimensions 500 by 200 ft. (153 x 61 m.), from the S.E., looking down. Devression 440 ft. (1384m.). Lake converted into whirlpool of subsidence over ‘‘ Old Faithful” sinkhole, the supply torrent fellowing a circuitous channel froma northern conduit, and entering the sinking pool tangentially as a cascade from the E. Blocks of crust on surface of torrent, ‘‘ Old Faithful” burst- ing in center of whirl. Photo Jaggar. known orifices of inflow for nine months previous, three at known grottoes of downrush, and one a sinkhole under the bottom of the lake. The downflow holes were in general at the opposite end of the pit from the inlet tubes. At times of rapid rise, however, downflow is not persistent in the same holes, and the circulation becomes stagnant or reversed. More- over, during rapid subsidence an inlet tube may become a sink- hole (fig. 2). There appears to be no escape from the conclusion, fully borne out by the accounts of the older writers,* that the lava *Dana Lake, New Lake, 1880 to 1890, Brigham, loc. cit., Hitchcock, ‘* Hawaii and its Volcanoes,” 1909. [Text continued on p. 183. | 174 JSaggar— Volcanologic Investigations at Kilauea. MiGe:. a Fie. 4. 1916-17. Stages in development of great crag mass, Photos Jaggar. (a) North margin lava lake, June 24, 1916. First stage bench terraced by uplift. Talus slopes from collapse of June 5 emerging through bench magma at borders. Jepression 592 ft. (181 m.). (b) Same scene on Aug. 16, 1916, terraced bench raised and fissured. Islets in foreground from raised shoals beginning of crag-mass. Depression 390 ft. (119 m.). 4 JSaggar— Volcanologic Investigations at Kilauea. 175 Gears Fie. 4. (c) Telephoto of northern islet Sept. 138, 1916, looking down from N. Depression 326 ft. (99m.). Sameas tiny islet seen as mere dot inb. Note erust wrinkling, and cracking and foundering blocks in lower left hand corner. Fountain against islet. (d) Same viewpoint as a and b, Feb. 24, 1917, after great crag-mass had passed its maximum, Note main tilt of upraised surface to W.S.W. Depression lake 106 ft. (82 m.) Summit knob identical with island in 0. 176 =JSaggar— Volcanologic Investigations at Kilauea. Fie. 8. a Fie. 5. Growth of the crag-mass and uplift of southwest islet, September, 1916. (a) Sept. 14, 1916, 10.30 a. m., from S.E. Depression of lake 326 ft. (99 m.), crag-mass 52 ft. ({6m.) above lake. North islet on right attached to crag-mass, lifted suddenly since day hefore (fig. 3, c). The S.W. islet, shown onthe left, was formed on Sept. 5 from a collapsed promontory; it was submerged during a rise Sept. 8; it reappeared again, along with rise of neighboring bench, as shown, on Sept. 9. (o) Sept. 26,1916, 3 p.m., from S. Depression of lake 312 ft.(95m.). N.W. pond with two fountains in background, outward streaming from crusted channel in foreground. Southwest islet now 20 ft. (6 m.) high andrising. Its high walls, now wholly without shore-line markings, due to upthrust as semi-solid mass through lake bottom shell, here shallow. Photos Jaggar. Fic. 6. 1916. Relations of conduit pond N.W. to lake streaming and growth of crag-mass. Photos Jaggar. (a) Aug. 23, 1916. Interior of Halemaumau from the north. N.W. pond overflowing on the right, rising crag-mass in middle, outward streaming of lake from W. arm on the left. Note Y form of the lake and tilting of crag-mass toward the region of overflooded bench. Depression 345 feet (105 m.). Left photo with deep yellow filter, showing bright lines. (b) Dec. 15, 1916. Mature crag-mass looking W. The summit pinnacle is identical with larger islet in fig. 3b. Lake depression 138 ft. (42 m.) Am. Jour. Sct.—FourtTH SERIES, Vou. XLIV, No. 261.—SEETEMBER, 1917. 13 178 = Saggar— Volcanologic Investigations at Kilauea. HGae Wie. 7, 1916-17. Development stages of S. island in lava lake. Photos Jagegar. (a) Nov. 7, 1916. South island and beginning of E. point, from 8S. margin Halemaumau. Depression 193 ft. (59 m.). (b) Jan. 5, 1917, from E. rampart at lake level looking S., showing E. point left, S. island right, and fresh overflow in foreground. Depression 101 ft. (81 m.). Saggar— Volcanologic Investigations at Kilauea. 179 Fie. 7. (c) Feb. 8, 1917. S. island above rim of pit, looking eastward. Depression 45 ft. (14 m.). Culmination of rise. (d) Feb. 28, 1917. §. island and S.E. cove after subsidence, looking S.W. through gulch in tumbled benches. Depression 115 ft. (85 m.). 180) Sagyar— Volcanologic Investigations at Kilauea. HnGeese a Fic. 8. Jan.—Feb., 1917. Showing conditions before and after sudden uplift of E. island. Photos Jaggar. (a) Jan. 27,1917. Upraised E. crag from rim Halemaumau looking N.N.W., summit 1 ft. higher than rim of pit, Lake depression 50 ft. (15 m.). (b) Feb. 22,1917. On right subsided E. crag 30 ft. (9 m.) below rim of pit and on left upraised island of bench magma 40 ft. (12 m.) high which rose 40 ft. from a shoal in the lake in a single night, Feb. 17-18, 1917. Lake depression 106 ft. (82 m.). Fic. 9. Feb. 1917. Bottom of lava lake abandoned by subsidence, showing sink- hole cascade. Photos Jaggar. (a) Feb. 28, 1917. Bottom of lava lake of previous month looking S.W. from S.E. margin Halemaumau. Depression 115 ft. (35 m.). (b) Feb. 14,1917. From S. looking down at sinkhole pit 60 ft. (18 m.) in diameter. Channel from lake in background, cascade deflected by obstruction and falling 40 ft. (12 m.); condensation from rising sinkhole gases on left. Depression 94 ft. (29 m.). Great banners of flame rising above sinkhole visible only at night. Saggar— Volcanologic Investigations at Kilauea. A “SUIRITIIM “ff OJOU “FSI, suruoaoe Ul syoouuny 104;vds Jo JUOULpALqUOY JUaTOLA puv ‘MoPAVAO Suor}s Sutanp yurod ‘gp saoys osye oyv[d yey} : AMoLA LOMO] ‘QT “[q PUNOAGOLOF UL UMOYS SOY} 0} ALILUUIS OTT MOLS pInG v UE pesaoutat sva odid oyy, “guiod qy Jo yy Sut olf} OF PUB puosdg UMOYS syoouUNY 10}}uds SAV] JO opis 4Jo[ UO OPLUE SVAL LOISAOULUL PUODAS OIL], “WOISLEUITUL 4say Suryeur yurod “] Wo edid Suoy YA ytoM yu Aztvd SurpuNos SMOYS “punoasyovq UT La}v1o VEL] JO WLI Ioyno puv voy vunvyp, ‘(UtgT) 47 09 uorsseldep oye “gid Jo ULL “W'S ULOLJ “AN SULYOOF “480} SUTPUNOS SULINP nLUINVUMETVA, “Wd CLS. LIGL “SS UCL “OL ‘DIA El TES JSayggar — Voleanologic Investigutions at Kilauea. 183 lakes of Kilauea are truly shallow pools, with feeding conduits beneath of small size. Where these lakes are confined in the cylindrical crater of Halemaumau and within it are surrounded by or include benches and crags of solid lava developed by overflow and accretion, these latter are merely subaérial exten- sions of the semi-solid lava column, which forms the bottom of the lake, and through which feeding conduit tubes are main- tained open. This lava column is incandescent within, graded in viscosity from high to low in transition to the liquid lake through its bottom, from high to solid in transition upward to the crags, and probably from high to lower viscosity in depth where the convectional circulation must become adjusted to a uniform and relatively slight radiation outward into the retain- ing walls. It might be asked, how do the inner bench overflows, as a part of and contributory to the upward growth of this non- liquid lava column, differ from any other surface lavas? The answer is that they are surface layers of a continuously mobile body; that by their weight they subside and displace the mobile incandescent matter into which they grade downward; away from them this semi-solid paste actually pushes upward locally so as to maintain a balance among islands, crags, benches and lake bottoms; there is hence a slow circulation, itself essen- tially convectional, in the matter which supports the liquid lake, in addition to the much more rapid convection continu- ously stirring the lake itself. Where such overflows become buried for hundreds of feet of such compensated subsidence, under other flows of the same kind, remaining integral parts of the lava column and reheated within it, in marked contrast to the older rock walls adjacent, they must be considered a part of the live Java column and of the mechanism of its circulation, and quite distinct from surface lavas which become frozen and dead. (See description of collapse of June 5, 1916, above.) Summary, Duplex Lava Columia I have attempted above to develop proof that there are two phases of the mobile lava column generally visible within Halemaumau, the one apparently liquid, the other apparently solid. Investigation has shown that the liquid lake is largely a gas-charged froth, and that the rocky benches are only a surface hardening or crust upon a stiff but mobile and incan- descent column of lava. The liquid phase maintains a shallow saucer, fed by a few spring holes beneath, in the top of the semi-solid phase. Both phases are in circulation. Heavy crusts founder with much effervescence in the liquid lake while quiet upwelling of lighter lava goes on on the opposite side, and sur- face streaming pours from the locus of upwelling to the locus of foundering. In the case of the semi-solid bench lava, over- 184 Jag ggar— Volcanologic Investigations at Kilauea. flooded occasionally by the liquid, there is tendency to inces- sant but very gradual subsidence of the flooded plains, and elsewhere to uplift, especially near the inlet vents (fig. 2, sec- tion). The places of flooding shift, however, and in the course of a long term of rising lava in Halemauman pit, the net effect is a subsidence of the bench region compensated by an inward flow under the lake bottom sancer, keeping the latter shallow, and even lifting islands above the lake surface. [ shall speak of the two phases as the lake magma and the bench magma. Consistency of Bench Magma. When the lava column subsided 400 feet (122 meters) sud- denly as above described, from depression 300 feet (91 meters) within the pit to depression 700 feet (213 meters), after months of building up by repeated risings and overflowings, there was revealed incandescence thr oughout the bench and island magma. This substance fell inward by opening narrow vertical fissures back from the edge of the bench, until the slabs so loosened tottered and descended crumbling ‘to the slopes below. There was always revealed a bright red luminous fracture surface and the larger slabs, many tons in weight, would flex outward slightly and then fall, disintegrating to a glowing talus whence arose great billows of chocolate-brown dust. These avalanches were noisy but less so than might have been expected, owing to the peculiar consistency revealed by the arching out and crumbling, for which the writer can think of no better simile than that of hard cheese breaking. The entire absence of heavy quaking at the upper rim of the pit, even when a whole quadrant of the bench fell immediately below the observer, showed that the attachment to the side walls was slight. I have elsewhere seen a very small remnant of an older rock bench, long firmly attached, produce a strong earthquake on falling. It was evident in this collapse of June 5 that the piled up overflow strata of months previous were a uniformly incan- descent and mobile stiff magma throughout, and that burial preserved the inner heat of the flows and reheated the crusts. These crust layers of flows are of thicknesses varying from a few inches to several feet (fig. 17u, foreground), like other p2ahoehoe flows and are full of ar wm vesicles. A phenomenon, occasionally seen during such subsidence at orifices in this bench magma, is an outflow which also is like crumbled cheese, and incandescent, but starting out like a liquid and falling like gravel or sand. It trickles out from fresh vertical breaks in falling benches. This substance appears to be identical with or very like aa lava on cooling, but I have never had access to it for verifying the resemblance.* Some of this fell from the glowing wall shown in background of fig. 15e. * See ‘‘ Live aa lava at Kilauea,” by T. A. Jaggar, Jr., Jour. Wash. Acad. Sci., vol. vii, No. 9, May 1917, pp. 241-248. JSaggar— Volcanologic Investigations at Kilauea. 185 EvipENCES oF Herat From Gas OxIDATION. Considering what a poor conductor porous basalt is, as illus- trated by the outer walls of the pit, and granting greater radia- tion at the lake surface than laterally, and particularly if the lake is at its hottest near the surface, as suggested by Day and Shepherd, this reheating of flow crusts and buried crags demands special consideration. Furthermore the whole ques- tion of a special mechanism of oxidation heating would seem to be worthy of examination, in view of the new facts concern- ing bench magma and lake magma. The ditticulty of intro- ducing water vapor from meteoric sources into a lava column has been found insuperable. Does the same difficulty apply to air, an uncombined mixture of fixed gases, and among them oxygen ? in the early part of this paper the writer has indicated five methods by which air may be brought into contact with sul- phur vapor, with hydrogen or with carbon gas in the voleano, and the first two of these are not hypothesis but fact, for sur- face flames are abundant, and suction at grottoes and sinkholes is unquestionable. Types of Flames. It is singular that geologists who saw Kilauea active could have disputed the existence of flames. They are generally invisible by day and they vary greatly in visible abundance by night. Their color by night is of two distinct kinds, blue- green and yellow. The common banners of flame over the grottoes, seen against a dark background, appear from bluish- green to violet. The sharp flame spears which burst out and play only a few seconds, when crusts on the lake are rent apart over accumulated gas, are bright blue, sometimes appearing in series like jagged saw teeth. These bluish or blue-green flames are very common above spatter cones built over cracks in the benches, and at wall chimneys where the lava froth has per- colated into high talus or cliff cracks and developed a flaming aperture, sometimes a hundred feet (80 meters) or more above the lake. It is a question of great interest whether such climbing vents are only the differential mounting of the froth by gas expan- sion, favored by small size of the crevice selected, or whether the heat of oxidation itself, by a sort of blowpiping, fuses and lines its way with a melt re-fused from the rock penetrated. If such refusion by oxidizing gases is possible amid talus open- ings, for example, it is possible on a considerable scale amid the interstices of the crusts of the still hot bench flows, buried by subsidence and overflow, and full of air cavities large and small. [Text continued on p. 192.] a b Fie, 11. Sept. 14 and 18,1916. Telephotographs of S. grotto at night, looking down from §.E. rim, bench on left, lake on right (see fig. 5a extreme left). Depres- sion 526 ft. (99 m.). Arrows on bench pvint to flames (blue-green variety) which were rising through recesses among SPACE lumps. Dallmeyer lens of 11°4¢™ diam- eter, 43 focal lensth, working at F 3°8 (Series B). Wratten ‘‘ M” plates. (a) Deep blue filter Wratten C. 24, aperture F 4, two seconds, 8 P. m. Sept. 14. Shows stalaztite drip. Grotto flaming, blue rays. (6) Bright green filter Wratten B 3, aperture F 5, one second, , sept. 18. Note incandescent ‘splashes on bench, and jagged edge of crust on ees from which there was inrush to grotto. Grotto flaming, green rays. Visually the continuous spec- trum of the melt shows much green, but almost no blue. Hence the rendering of the lines and splashes in contrast to (a). (c) No filter, aperture F 3°8, 1/50 sec., Sept. 18,8 Pp. m. Crusts downsucking with ee fountaining migratory to the left toward grotto. (d) No filter, aperture F 5:8, 1/50 sec., Sept. 18, 9 Pp. m. Hexagon of crust being engulfed at grotto. Note in cand d graded incandescence to centers of gas oxida- tion. Photos Jaggar. 7 hs a? ae * ia. 2 re ‘ + 2? % Fie. 12. Jan. 26,1917, 6 p.m. Series of photographs, Wratten Pancbro- matic plate, aperture F 6°38, exposure 1/35 sec.: lava fountains 100 ft. distant in N. cove of Halemaumau, observer standing on E. point (see fig. 2). Pro- longed migratory type of fountaining. Depression 50 ft. (10 m.). (a) Shows sub-crustal explosions breaking surface ; note high viscosity and depression in crust. (0) Second stage, flinging and spraying, area elongated with migration to the right, actual fountain 40 ft. (12 m.) long. (c) Blowing and flaming, with increased elongation, diminished viscosity, and down-suction. (d) Termination of fountaining against wall in background, with expulsion of gas, flames, and much production of filamentous glass. In taking these pictures the writer was at the edge of the lake to leeward of its whole cen- tral region, but was not at all inconvenienced by the gas from its smaller bubblings, Photos Jaggar. Fie. 13, Fie. 18. Oct. 5, 1916, 12 noon. Telephotographs of travelling fountains in N. cove, from N. rim of Halemaumau, looking down. Depression 283 ft. (86 m.). Fountains migrating from left to right, conflicting currents meeting and crusts foundering along conflict line. Large Dallmeyer lens, aperture F 5, exposure 1/20 sec., red filter Wratten F. Photos Jaggar. (a) Skins downfolding, line of fountains, crusts drawn to explosion center. (b) Second stage, fountains have moved to right and enlarged. Crusts forming behind them. . (c) Last stage, fountains expending themselves against bank, chiefly in one great explosion center, ca) Fie. 14. Aug. 23, 1916, 8P.m. From the north. Grotto, dome, and openwork fountains. Telephotographs with large Dallmeyer lens, aperture ¥3°8, 1/50 second exposure. Lake depression 345 ft.(105 m.). Photo Jaggar. (a) Night view of distant shore shown in fig. 6a, magnified. Central and border dome fountains, and in background interior of very large half-dome grotto overhung by spatter rampart, with curtain of stalactites on each side. The hazy interior above isflame. The central fountain slowly migrated to the grotto. (6) Openwork of nearer large fountain, bursting in north cove. Shows multiple expanding bubbles of glass, not detectable by eye. Fountain approximately 380 ft. (9m.) high, seen from above at a high angle. Note decrease of blurring motion and of luminosity from center of explosion below to upward limit of fling. Solidification in air is taking place. Such bubbles do not explode to fragments but expand within a ropy network. Fie. 15. Intense multiple fountaining of Halemaumau July 17-20, 1912, looking W. From maximum depression of 329 ft. (100 m.) June 22, there had been strong rising to minimum depression 192 ft. (58 m.) July 12, fol- lowed by subsidence to depression 273 ft. (84 m.) July 20. Thereafter subsi- dence continued. (a) Daylight July 17, 1912, 9 a. M., shows inner pit of violently effervesc- ing lake magma, surrounded by floor of bench magma which was fuming and collapsing. Photo Jaggar. (b) Shows same scene at night, July 20, 1912, 10 Pp. m., exposure 1/5 sec., streaming from right to left; note increase in size, height and numbers of fountains from right to left. Length of straight shore on left 510 ft. (155 m.), width of lake 525 ft. (99 m.). Larger fountains 20 ft. (7:62 m.) in diameter. (c) July 20, 1912, 10 P. M., same as (6) but longer exposure 2 sec., shows rapidity of streaming towards left, down-suction at line of shore grottoes and at zone of travelling fountains across middle ; crusts parting at right shore; spatter, stalactite curtain, flame, and bombarded incandescent wall; fume above. The same violent turbulence was present in (b) but short expo- sure arrests motion. Photos J. T. Warren, Eastman film, aperture F 4°5. JSaggar— Volcanologic Investigations at Kilauea. 191 Hie. 16: (ah 5 Fie. 16. Jan. 16,1917. Details of lake surface and rampart grotto at time of experimental work. Depression 78 ft. (24 m.). Photos Jaggar, taken from lake shore. (a) Heavy crusts of lake surface folding downward along line of small fountains 50 tt. (15 m.) away. (6) Rampart grotto with splashing fountain 30 ft. (9 m.) from observer. Iron pipe for experiments shown in lower left corner both pictures. 192 JSaggar— Volcanologic Investigations at Kilauea. The second type of flames is comparatively rare and is dis- tinetly yellow in color like an ordinary coal-gas flame. I saw this first in 1912, flaring out from an aperture in the tumbled high east benches, far above the lake level, and lasting the greater part of a minute. Twice in 1916 during the long ris- ing spell of the summer, I saw a yellow flame suddenly burst through the crusts of the lake, and then flitter a short distance as a flaming foam or spume, dividing into several smaller flam- ing masses before going out. The yellow flames, luminous possibly with hydrocarbons or other impurities, are so rare in the present epoch that they must be considered curiosities. On the other hand, the blue flames are very common and on certain nights when streaming is slow and crusts are heavy, they play in gigantic banners or blankets above blowing grot- toes, or out horizontally from overhanging spatter margins of the lake, under which the gas from beneath crusts is escaping with a rush against the bank, and so is deflected outward. I have seen such blankets of flame for a length of 20 feet (6 meters) of shore jetting over the lake surface out from the shore for a distance of from 10 to 15 feet (8 to 4 meters), as though from a great flat-mouthed Bunsen burner. Frequently such flaming orifices jet lava spray to great distances, and the evidence of gas pressure is shown further by puffing noises and ceaseless variation in the length of the flame banners. Ordinary flames shoot up three to four feet (one meter) above the foun- taining border grottoes (figs. 14a, 12d, 11a, 6). I am unable to agree with Mr. Perret that visible flames always occur above bursting dome-shaped fountains, which break through the crusts in the middle region of the lake sur- face (figs. 14a, 150, 180). I have stood on the spatter rampart at the lake margin a few feet from such fountains, and have watched and photogr aphed fountains for years, and made special efforts to photograph the flames with color screens (figs. lla, 6, 13). My experience is that shooting blue flames break out just as crusts part at the beginning of fountaining, but that when the doming and spattering phases of large or small fountains occur, a flare of flame sometimes is visible, but quite often is not so. It does not seem probable that this observa- tion is due merely to differences of seeing, for the same fact is in less measure true of the border fountains which sometimes are without visible flames. The central fountains most addicted to flaming are the continuous or “ perpetual” kind (fig. 12). I am inclined to attribute the absence of flames above some fountains to the fact that the gases inflating them are more completely oxidized than in the flaming cases. This would of course mean that the combustible gases are variously OXx1- dized in depth. With reference to the composition of the gas burning in the blue flames, as contrasted with that of those “smaller bubblings JSaggar— Voleanologic Investigations at Kilauea. 193 of the liquid lava which emit gas without flame, there is in daylight a distinct yellow-brown fume at the places where the blue flames occur (fig. 96, over sinkhole), and higher there is a condensation of pale blue fume which expands above into a larger cloud faintly blue. This blue smoke is very hot and may be nearly invisible, but it is the most irrespirable of the Kilauea fumes, and the odor is that of sulphur dioxide. It probably contains also the trioxide. The gas from the smaller bubblings makes no perceptible fume and has very little odor, but is sometimes oppressive as though with carbon dioxide. The white smoke that rises from cracks in the benches and crags is weakly sulphurous but quite respirable, and appears to be in the main a mixture of moist air and unburned sulphur. In working to leeward of Halemaumau one learns to dread the intolerable blue fume from the sustained continuous fountains ; at night these are seen to be surmounted by flames, so that the conclusion appears warranted that the surface oxidation which they represent is mainly that of sulphur vapor. Gases Responsible for Flames. It thus appears that of the three combustible gases hydro- gen, carbon monoxide and sulphur, the last is most in evidence as surface flames, the second, along with impurities, may be represented by rare flames but mostly achieves its combustion . below the surface, while the first, namely hydrogen, flashes to water vapor in depth, and is not (unless by spectroscopic means) to be diagnosed by itself as flame at the surface. A residue of both gases is mixed with the sulphur. By its stronger affinity for oxygen, hydrogen would certainly be the first of the combustible gases in the mixture to achieve oxidation below the surface if oxygen were available ; we find it in the Day and Shepherd analysis of even the surface gases in larger amount than the carbon monoxide. In view of the evidence that some oxygen reaches the lava column below the visible lake surface from downward suction and from engulfed talus slabs and crusts, and that in the gas-collecting apparatus water vapor condensed abundantly, it seems hardly admissible that none of the water should be produced by combination of hydrogen with atmospheric oxygen. If any of it is produced by sub-surface combustion of hydrogen, then we have in such combustion a formidable heating agency to be reckoned with for whatever depth air enters the lava. In less degree but in like fashion air would at these high temperatures (850° to 1150° ©.) react powerfully with carbon monoxide and sulphur vapor to produce heat.* * Dr. Shepherd suggests (oral) that silicon hydride (SiO. + 4H2=SiH, + 2H20) is a combustible unstable gas formed at high temperature, possibly present at Halemaumau, and that he has detected the odor of carbon oxy-sulphide (COS) there. The latter isa lower temperature product, also inflammable, and might be expected in the bench magma. Am, Jour. Sct.—Fourts Srrizs, Vou. XLIV, No. 261—Srpremeer, 1917. 14 194 = JSaggar— Volcanologic Investigations at Kilauea. Distribution Crusting and Fountaining. On the lake surface over or near the inlet conduits, and hence at the points of departure of the surface streaming, a very marked cooling effect is generally observed in the form of dark crust which remains stationary over the uprising gas- charged melt (figs. 50, 6a, 15c, on the right, 3 on the right, 18 on the leit). Effervescence and incandescence are both at a minimum at the locus of fresh rising lava. Radiation cooling and expansion cooling both achieve their maximum immed1- ately over the inlet vents. This was profoundly puzzling to the writer during several years of continuous record and close observation, when he supposed that rising juvenile gas was the main source of heat, and hence the inlet wells should yield high incandescence and much bubbling. Such multiple bubbling of bright luminosity oceurs in fact (1913) over the inlet region a when the lava is subsiding, and convectional outletting i is rreatly dominant over inletting—when a shallow rapid cireula- ‘on mixes air with the voleanic gases. During rising periods the whole lake surface is apt to be quiet and crusted. (Con- trast fig. 10 rising, with fig. 15 sinking.) These facts imply that the rising magma and gas from deep sources are only moderately hot, the gas is evenly distributed in small bubbles, and such lava is quickly solidified and crusted on exposure to atmospheric temperatures. Let us now watch what happens as the crusts form, thicken and stream across the lava lake to founder with sudden tearing, downsucking, flaming and violent effervescence at a border grotto, or to be engulfed at a central fountain. Such a foun- tain forms, first, by conflicting currents tearing a crust asunder ; then a spurt of flame appears, followed by a little bubbling ; then a general downsucking of skins takes place toward a point, frequently tearing a symmetrical hexagonal opening in the crust, fifty feet ( (15™) across or more (fig. 12, a and 6). The crusts are heavy, tough, membraneous bodies like a doormat (fig. 16@) and when they rift, the stiff vesicular slag below wells up in the wake of the block which is foundering and draws out variously cooling shreds of its own substance from the sohd jagged edge of the mother crust, a glassy, veined membrane of | ereat toughness forming instantly over the newly exposed lava through shades of cherry red to purple and black (fig. 11, 6, ¢ and @). Cool air is incessantly circulating in contact with the crusts to replace the hot wprush from the whole lake. The crusts are three to four inches (eight to ten centimeters) thick (fig. 17a), vesicular above, frequently folded or festooned and so involv- ing large cavernous space, and their under sides, when they are seen to turn up edgeways and sink, are covered with . Bie, £7: a) Fic. 17. Jan. 1917. Experimental work on E. rampart of lava lake in Halemaumau pit. (a) Jan. 4, 1917. Looking northwestward across fresh overfiows in foreground at ram- part. Part of crag-mass and N.W. outer wall in background. Dipping up liquid lava from lake; 20-foot (6 m.) pipe with iron pot attached. Fountain splashing over rampart on the right. Note thickness of broken crusts in foreground; these are like the crusts which cover lake. Lake depression 101 ft. (31 m.). Photo Stotts. (ob) Jan. 11, 1917. First measurement of temperature with Seger cones. Shows bend in pipe where heated most at lake surface ; straight below and above. Cylinder with cones on end of pipe. Lake depression 87 ft. (26m.). Photo Jaggar. (c) Jan. 16, 1917. Third test with Seger cones. Pipe immersed in lava on left, anchored with rope right. Lake depression 78 ft. (24 m.). Photo Jaggar. 196 = JSaggar— Volcanoiogic Investigations at Kilauea. a porcellanous glaze which is stalactitic and impervious. It is gas-tight, for the crusts are frequently ballooned upward by gas. “Doubtless the blocks become completely encased in such a glaze when they sink. There can be no question, to the writer’s thinking, but that such up- -ending blocks, spongy in their vesicularity, exposed to sweeping winds, are ‘filled with air. Green was of the same opinion. The mechanism. of their solidification implies a coolmg interstitially by air which replaced the hot gases of the unsoliditied state. The persist- ence of those hot gases in tension in the vesicles is unthinkable, particularly as the crusts, after reaching a certain thickness, always acquire the impervious glaze beneath, and are openly porous above. Simultaneous with the foundering of the crust area, a dome- shaped mass of bubbles, or “fountain,” bursts having quite the appearance of an ebullition center in boiling milk; this dome may end the explosion, or it may be continued with a series of high flings of drawn slaggy openwork (fig. 140), or prolonged gushes of spray and flame, the noise being like surf accompanied with puffing. Lava fountaiming is not the expan- sion of a single great bubble, but that of a swarm of bubbles, a true effervescence, and as iit progresses the fluid heats and loses viscosity markedly, flinging small droplets of glass such as the normal lava never makes. While the fountain lasts, engulfment and inward sucking to its center continues, and the surface currents set towards it, increasing in speed at the center, from some distance away. Foundering at the spot does not always precede fountaining, and fountaining may occur abortively without dow nsucking, raising convulsively the surface without breaking it, or merely throwing back a flap of crust without exploding - through the under layer. Great balloons of erust are sometimes blown by quiet gases, the bellying mass finally tearing at the end, when the gas escapes and burns. For an instant a glowing cavern is seen within with a bubbling liquid floor, and then the skin col- lapses. The true fountain, however, doming and spurting, is always followed by indraft of crusts. A continuous fountain is always a place of rapid downsucking and engulfment. I have said that a fountain forms where currents conflict. Such conflict may be of three kinds, surface meeting of two or more streams, deflection of a current against the shore, or sub-surface meeting of currents following down opposed bot- tom slopes with the convection. This “last case reaches a maximum when such submerged downstreaming meets from various directions at a sinkhole where the united flowings descend. A fountain over such a sinkhole reappears from time to time in the same place, and if the streaming process down the sinkhole is constant the fountain will be rhythmic Jaggar— Voleanologic Investigations at Kilauea. 197 This is. the case with “Old Faithful,’ a rhythmic fountain which, with occasional lapses, has been frequently seen in Kilauea. The fountaining border grottoes are places wlicre there is conflict of a current or currents with the resistant shore, and commonly also these are over submerged sinkholes. The meaning of a conflict of streaming currents will be apparent when we remember that convection is the motive power. At @ holes a lighter fluid is rising; at y holes belowa y) shallow lake a denser fluid is sinking; atmospheric cooling, 5-9 radiation and gas expansion change the « fluid into the y fluid ; and a more viscous, partially cooled magma descends. Such is normal convection, and some such convection progresses in Halemauman. Solidification increases weight. The lighter finid rises through the conduits west: crusts form rapidly over it: these are drawn toward the several sinkholes east : they are heavy and founder in the lighter magma just beneath them, when two currents, developed by this sinkhole distribution, bring crusts together and rend them or bend them so as to release the gas accumulated beneath and start an engulfment, which progresses rapidly edgewise when once a portion of a sheet of crust is submerged. This engulfment is often a flexible downfolding (fig. 16a). The result is to draw away along a fissured line a wide sheet of crust, and hot magma wells up the fissure and itself quickly crusts over. This crack- ing and foundering process is seen in flexible skins, heavy blankets and hard brittle crusts according to their thickness, the supporting power of the fluid next below, and the length of time that the crusts are allowed for solidification without dis- turbance. The surface streaming is maintained as a part of the convection by this mechanism, with accelerations at the fountains. Maximum speed of surface streaming is attained during general subsidence, when sinkhole mechanism is dominant over conduit mechanism; .when the « holes lose pressure of rising magma and the y holes themselves subside and downflow through them is at a maximum, uncompensated by any tendency of inflow to add y holes to the # group. Problems of Fountain Mechanism. Why should fountains burst at all, and why at points of in- terference of currents? Furthermore why should accelerated streaming rush centripetally with engulfment to the fountains 4 Why should continuous fountains emit banners of flame? And why should the fountains be places of highest incandescence and highest apparent liquidity? (Fig. 18.) This is evidenced by their spatter which is a perfect glass, often drawn into fila- mentous floss. (See glisten of fresh lava, Plate Ia, upper view.) 198 Suggar— Volcanologic Investigations at Kilauea. It has been stated above that some fountains do not burst through the surface, but merely heave it in a sudden abortive effort without breaking the crust or emitting gas. These in- dieate that explosion takes place below the surface with such balance of reactions that the gas product replaces the re-agent gases. There is no evident ballooning over them, the crust is simply heaved suddenly and sinks back and the gas produced is disseminated below. There is no evidence from artificial stirring of the surface that a condition of general tension exists immediately under the surface crusts. Stirring with an iron pipe does not generate Fic. 18. Fic. 18. Mar. 10, 1916, 6 p.m. Night view upper surface lava column Halemaumau from S., rising and overflowing lake, strong activity, depres- sion 450 ft. (187 m.). Dimensions and conditions extraordinarily like those of six months before (Pl. la, upper view), though a marked subsidence of lava column hadintervened. Western conduit pond on left crusted over and dark, middle region moderate luminosity and fountaining, maximum oxida- tion shown on the right by flaming, sinkhole fountaining, breaking up of crusts, and fuming. The ]uminous line on the extreme right is a cascade rivulet pouring down from the lake among tumbled crags. Photo Jaggar. a fountain. A fall of rock tumble into the lake carries down great quantities of air in vesicles, and fountaining immediately results. A log of wood thrown into the lake end on remains below the crust, carbonizes, produces jets of red flame and in- duces fountaining. In both of these last cases reaction between the hot gas of the magma and an oxidizing agent is responsible for fountaining. Mechanism of Different Types of Lountains. With great quantities of air-filled crust ceaselessly founder- ing in a gas-charged melt cooler than their fusing point, the JSaggar— Volcanologic Investigations at Kilauea. 199 gases of the melt including hydrogen, carbon monoxide and sulphur at high temperature, explosive reactions are inevitable. and these will be a local, temporary and incomplete oxidation, or a general, definitely situated, prolonged and flaring blast- furnace effect, according as the supply of oxygen is temporary and inadequate, or continuous and effective. The conflict of streaming convection currents which determines foundering, will thus produce subsurface oxidation of gas, more or less complete, according as the downrush of air and air-filled crusts is rapid or slow, general or sporadic, localized or shifting, con- tinuons or temporary. It must be borne in mind that the only significance of cur- rent conflict is as an observable and measurable evidence of crust breaking; and that this breaking of crusts by convec- tional interference is not limited to the surface, but takes place wherever the glazed and air-charged blocks are subjected to fracturing or erosion, as when they are drawn down sinkholes in the bottom of the lake. On this analysis it will be seen that fountains should burst where violent oxidation is produced by contact at high temperature of magmatic gas and air. Thus central fountains may commonly be attributed to such accumulation of air-filled crust beneath the surface, as to start the oxidation reaction ; this involves violence and a consequent further breaking up of the air-filled block; the reaction may involve condensation and devesiculates the melt, consequently downsucking from the surface ensues, wlich supplies more oxygen ; localized heat is liberated, inducing localized convec- tion and expansion, and vertical escape of combustion products. The two processes are opposed and produce the opposed effects seen at a normal lava fountain, namely violent upward dis- charge of more or less burned gas and convulsive downward indraft of crusted magma. It will be seen that when two surface currents of different velocity come together at an oblique angle and along the line of meeting continuous downfolding of huge blankets of crust ensues, the two currents merging into one and carrying along the submerged foundered skin in the new direction, the con- dition is ideal for generating a line of travelling fountains (fig. 13). And just such travelling fountains are there gen- erated, the explosions migrating with the line of meeting, accelerating the inrush of the skins along the line, sucking down free air with them in the violence of the reaction, and increasing in vehemence of fountaining until raw glowing air- free melt has been exposed on both sides, the fountain usually ending its career at some permanent grotto on the bank. The two currents streaming at different speeds in this case furnish erosion mechanism to break up the crusts under the surface 200 Saggar— Volcanologic Investigations at Kilauea. along the line and so liberate imprisoned oxygen for union with the magmatic gas. Rhythmic fountains like “Old Faithful” have commonly at, Kilauea intervals between explosions of from thirty to ninety seconds and occur immediately over sinkholes in the bottom of the liquid lake some 50 feet (15 meters) below the surface. They are particularly in evidence during a term of prolonged and continuous subsidence when the sinkhole is steadily acting as such, but their regularity may continue dur- ing rising. The crust material which founders at the lake margins of the cove in question sinks to the bottom and on being drawn into the sinkhole is broken up in contact with the gases of the lava. The reaction is continuous in supply of materials. The excess of combustion products and localized heat effect extend to the surface of the lake vertically in spasms, instead of continuously, because the excess of uncom- bined oxygen is insufficient to maintain a continuous stack or furnace through the 50 feet (15 meters) or so of liquid melt above. Such central fountains occasionally become continuous. What factors are dominant in contributing to the rhythm of the surface explosion is as yet unproved, but it seems probable that the sinkhole itself acts rhythmically in its downflow, becoming clogged at intervals. The greatest amount of oxygen, heat, expansion and local convection would be produced during its more rapid downsuckings, and these times would surely upset the surface equilibrium of the pool above and produce a fountain. Continuous fountains, at the lake margin especially, form grottoes of spatter glass in beehive, or half-dome, form, open as glowing caverns on the side toward the lake and frequently perforated with small flaming crevices, the walls of which are incandescent (figs. 14a, 166, 10 right “center). Usually such grottoes surmount border sinkholes of some permanency as revealed by sudden subsidences, and there is reason to suppose that the mechanism of a prolonged building up of the bench lava, with persistence of such a grotto, constructs such a sink- hole by upward growth of the cavern in the bench and of the floor of the lake by accretion (fig. 2, section) under the cavern’s mouth. Towards such a grotto there is continuous suction, within there is continuous but variable fountaining and spray- ing trom outrush of gas, through the apertures and above the entrance cavern there are continuous but variable spears and banners of flame. Still higher appears the bluish fume which retains its heat for hundreds of feet above and to leeward and is painfully acrid with SO, and SO, in its effect on the human mucous membrane. The temperatures in these grottoes are very high as shown by their luminosity, the magma within them is at minimum viscosity as shown by its mobility, and the glassiness and freedom from vesicles of the spatter product are JSaggar— Volcanologic Investigations at Kilauea. 201 remarkable. Moreover the linings and stalactites of such erottoes indicate refusion, and oxidation of the iron to the ferric condition. The variability in fountain action at grottoes may be from zero, when all is quiet and crusted, to tremendous spraying and blowing with towering banners of flame (fig. 20) and an inrush which amounts to a cascade from the lake to a cauldron within the cavernous dome. Times of extreme quiet are characteristic of rising, times of extreme activity are char- acteristic of sinking. This last antithesis is true also of all other fountains. The grotto fountain appears to be a true case of a stack or furnace built at the margin of the lava lake remote from inlet conduits. The furnace is fed by relatively continuous downflow of lava vesiculated with combustible gases, and supplied with oxygen by continuous downsucking with this maema of large quantities of air. The reactions are quite as in the other cases cited, but the recess constructed by spatter around the margins permits a concentration of the blast at one point. It is quite probable also that chemical activity of gas with lava, reheating the substance of the grotto walls, adds to ‘the heat supply in these specialized edifices. The opening of a confined glowing cavern toward the air over the lake, with continuous indraught of that air induced by the downflow, brings about very complete combustion of the less active gases such as sulphur vapor at the orifice of the stack, and prevents the condensation of free sulphur which elsewhere appears as a white smoke or kind of gaseous emulsion when the vapor rises gradually cooled through border cracks in the bench lava. Construction upon Lake Bottom. The great quantities of air-filled crust sucked under the lake margins at the border grottoes probably descend to the bottom with much of their oxygen unconsumed and with their relatively cool and heavy substance incessantly adding, by a backward or eddying subsurface circulation, to layers of more or less com- minuted material on the bottom of the lake increasingly viscous downward (fig. 2). This subsurface crustal material is what builds up shoals in those portions of the shallow lake least eroded by streaming and least heated by oxidation. _Com- minution, by destroying the buoying effect of vesicles, accelerates sinking. We should thus expect shoals to be built out at the sides of inflowing fresh lava from the conduits, where fountains are absent, and where quiet eddies would permit construction upward from the bottom to join with the thickening crusts at the surface. There bas been repeatedly demonstrated a ten- dency for the lake to grow from the western conduits eastward 202. Saggar— Volcanoslogic Investigations at Kilauea. during rising, in the form of a Y with arms curved outward, the stem being the current from the source. This is what happened in the summer of 1916 when a simple oval lake developed powerful streaming outward from the west side, then shoals appeared and peninsulas were built out (fig. 5) on both sides of the current, until the Y or T form was produced (fig. 6a). Sinkhole Cascades. The extreme case of the grotto fountain is the sinkhole eas- cade such as was described on pp. 172, 173 above. Such cascades from the lake into holes at its border, or in its bottom, have repeatedly appeared when general subsidence began. Com- monly the place of cascading is a border pot which may have been a place of rising lava pouring out into the lake during a previous rising spell. When sinking begins the magma in the smaller tube sinks lower than the lava in the lake with the result that a torrential fall is precipitated from the lake into a glowing void, intense effervescence or fountaining is In progress in the depths of the pot which receives the fall, great sheets of. crust from the lake are drawn over the rim, and break up in the abyss, and a column of flame and fume rises above with great heat and much sulphnrous acid gas (fig. 96). All the phenomena of such a cascade are those of a border grotto exaggerated (fig. 11d). For a long time in watching such cascades the writer was in doubt whether the lake lava actually poured over a ledge or whether the appearance of such an edge were not merely the more rapid fall of surface layers over the slower liquid beneath where the magma in the border fissure, open on the lake basin side, sank as a froth by loss of gas faster than the magma of the lar ger Jake body. I am now convinced that commonly there is an actual lip of bench magma which the lava of the lake pours over, and this lip is usually the margin of the lake bottom, revealed by reason of the sinking of the liquid lake within its basin in the bench magma. There has recently (February 21-28, 1917) been a remark- able case of such cascade action after 46 feet (14 meters) of subsidence of the lava lake—just sufficient to uncover much of the bottom and leave the lake a very shallow body coursing like ariver. At the northeast margin under the cliff of bench lava there was revealed a cavern in the wall into which the lake lava cascaded for seven days and longer. ‘The fall was shaped in plan like an obtuse V with the point toward the lake; the cascade poured over both arms of the V, while in one place an outcropping of ledge rose through the fall and divided it. This clearly indicated that the liquid was cascad- JSaggar— Volcanologic Investigations at Kilauea. 203 ing over a solid or semi-solid obstruction. And this was just at the level where the irregularities of the margin of the lake bottom of the previous higher stand of the lava might be expected to appear. Incidentally the cavern was at a location of large and continuous border grotto building for many previous months. During the progress of such cascading the lake may main- tain its level, or even rise, showing that the cascade is not an outlet. It merely reveals during excessive shallowness, con vectional mechanism commonly concealed by greater depth in the lake. That the violent downrush during cascading, how- ever, is swifter than at other times, is shown by the fact that when such a cascade is in action, it is apt to satisfy completely. or nearly so, the fountaining requirements of the lake mechan- ism, for almost all other fountains cease playing while the cas- cade is falling. This might be expected, as the rest of the lake at such times is very stagnant and crust foundering is largely limited to the one place where erusts and air are being sucked down in prodigious quantities and the heat and flaming are excessive. Probably excess of downward convection over inflow or some equivalent mechanisin in the relation of bench magma construction to lake magma effervescence, brings about the cascade phase of downflow. If, however, it is maintained by an extreme phase of oxidation of gases, which keeps the level of lava in the sinkhole 20 to 50 feet (6 to 15 meters) lower than the level of the lake, then we have to make careful inquiry as to how this could be brought about. Condensation by de-vesiculation. In the above discussion of oxidation fountaining, it was pointed out that two opposed processes would operate in oppo- site senses, when foundering crusts along with convection downflow mixed air with explosive gases, and so induced oxida- tion, excessive heat, consequent gas expansion, and consequent lozalized upward gas convection. Mention has been made in addition of condensation resulting from the explosion of un- stable mixtures of oxygen with hydrogen, carbon monoxide and sulphur vapor. Whether there would be such condensa- tion at the high temperatures prevailing in molten lava is a question for the physical chemist. We know that two atoms of hydrogen uniting with one of oxygen produces a molecule of water which in its liquid state occupies greatly diminished space in contrast to its gaseous progenitors. In the case otf water vapor at 850° C., however, condensation would be less, and the same is true of the transformation of sulphur with oxygen to SO, and of CO with oxygen to CO,. If in the heated condition there is less change in the vapor tension resulting when these reactions take place, it is evident 204. Saggar— Volcanologic Investigations at Kilauea. that the violent downsucking at the grottoes and fountains must be accounted for by some process other than chemical condensation. The only conceivable process other than gas condensation, in view of the obviously excessive discharge of gas and libera- tion of heat which occur at these places, to acconnt for the loss of volume which produces downward suction, is a partial de-vesiculation of the lava, a loss of gas which decreases its bulk, and to that extent increases its density. It is plain that if the furnace of a grotto stack could reheat the vesiculated lava for any depth, we would have our convection reversed and so would become involved in paradox. It is evident that the reheating does not prevent convectional downflow of the liquid, and this would clearly not be the case if we had to deal with mere thermal convection, rather than two-phase convection. but in the union of the oxygen of crust vesicles, with the combustible gases of viscous lava vesicles, to induce violent reactions which expel great quan- tities of burned and unburned gas vertically, the hard crusts must be comminuted, and the melt must be largely robbed of its vesicles. The burned gases es cape in large bubbles, while the local heating and lowering of viscosity favor expan- sion and escape of the small bubbles. Therefore loss of gas plays an important part in shrinking the heated lava of the grottoes and sinkholes. The expansion of the combustion products, moreover, is a cooling effect. All of this may account for the lowering of the level in the sinkholes , when extraordinary oxidation 1s maintained by the cascading pr OCeSs. Furthermore, the faster the shrinkage: within a sinkhole, the heavier the cataract resulting and the more voluminous will be the direct suction of air downward. This in turn increases oxidation and so the process ot acceleration is self-propagating to the limit of combustible gases available. Multiple Fountaining of 1910 to 1912. Before leaving the question of fountaining, mention must be made of the very remarkable distributive fountaining, gener- ally accompanying subsidence, in Halemaumau lake in 1910 aud 1912. This may be considered a time when the lava column had reached its highest level and was entering upon a general decline that reached its lowest in 1918. There was a sudden spurt to nearly the 1910 level in December—January, 1911-12, and there was another rise to a lower level in July, 1912. The characteristics of this extreme activity were abund- ance of bubble fountaining and numerous large fountains, rushing and streaming surface currents in changing directions, and frequent lines of tray eling fountains when two or more such currents would meet and ¢ oppose each other. In July of JSaggar— Volcanologic Investigations at Kilauea, 205 1912 this process reached its maximum, when in a small oval lake about 600 feet (180 meters) long, depressed 200 feet (60 meters) below the rim of the pit, there were several hundred large fountains and many more small ones (fig. 15). All of these were playing at once with a roaring noise, intense incandescence, tumultuous rush of surface currents, and bom- bardment on the bank, first on one shore and then on the other. - To account for this phase of activity we must suppose that convectional downflow was dominant, a shallow, air-filled and superficial convectional circulation was very rapid, and that accordingly oxidation was at a maximum. This was the sum- mer when the Day and Shepherd gas collection was made and excess hydrogen was found among the combustible gases. Temperature measurement with Holborn-Kurlbaum _pyro- meter gave 1185° C. This was the time when Day and Shepherd found greatest emission of gas (maximum fountain- ing) to comeide with highest measured temperatures. There had been repeated inward tumbles of great quantities of air- filled bench rock in 1910, 1911 and 1912, the lava each time recovering so as to rise and submerge this talus debris. It seems likely that the excessive fountaining during the season of smking of 1912 was due to the breaking up by subsidence of this submerged air-filled debris, and this, coupled with and contributing to extreme rapidity of two-phase convectional circulation, as outlined above, brought about an admixture of much oxygen with combustible gases. Vast quantities of air were drawn down directly with downflow currents, under the banks and at the traveling fountains, owing to the speed of the torrents, which increased to cascades in August. It is possible that at certain seasons the supply of some unstable gases such as hydrogen is larger than at other times, and the heating effect greater. Routine photospectroscopic work on flame spectra might prove this. Disappearance of Liquid Lava during Low Levels. These terms of violent effervescence coincided with culmina- tions of rising and the beginnings of sinking. With the low levels of September, 1912, and of the entire year following May of 1918, gas pressure dwindled to a minimum and the pit at some 700 feet (213 meters) of depth was floored with frozen lava surrounded by talus slopes. There was much sulphur fume and water vapor, but even the hissing of confined gas at times ceased completely. This condition, merely because the lava column had retired topographically to a depth one-sixth of the height of the mountain above sea-level, could hardly be attributed to depression alone. Why, in other words, should depression induce inactivity? Why should not the same violent fountaining, and vast discharge of gas and heat, go on 206 Saggar— Volcanologic Investigations at Kilauea. within the mountain at 700 feet of depression, if the phenomena of streaming and fountaining in the liquid lake were functions independent of the form of the higher lava pit, and of contact with the atmosphere ¢, The lava was sluggishly present in the depths of the pit for it would occasionally begin puffing and glowing, and its sulphur smoke was always rising. The answer to these questions would seem to imply that lake magma convection at these low levels becomes very slug- gish or stationary, owing to lack of stimulation with oxygen, and that at the higher levels with increasing sinkhole suction of free air, larger area of surface crusts, and assimilation of air-filled talus blocks, the lava gradually acquires an increased fluidity in proj portion as its upper edifice becomes increasingly a furnace. Meanwhile the bench magma, matrix for the furnace of the liquid lake, remains always a sluggish, heavy and stiff substance such as lay dormant in the depths during the intervals of low level. Heating Mechanism in Bench Magma. The heating mechanism by oxidation of combustible gases which has here been outlined, was first mentioned in relation to the heating of the bench magma. The discussion so far has been directed chiefly to the rapid engulfment of air in the liquid lake. We have seen that an excess of air engulfed in glazed crusts becomes built into the viscous bottom of the lake and that heavy crusts and talus blocks, full of air in vesicles, become buried and ineandescent, by a gradual subsidence under overilows, in the mass of the bench magma column. The maintenance of this inner heat in the bench magma column is accomplished by the slow circulation of its lava, by con- duction from the conduits and sinkholes which perforate it, by the burial of incandescent lava flows within it, by the burial of hot bottom layers of the lake upon it, and lastly by actual percola- tion through it of the voleanic gases under pressure. It is probably this last process which discovers buried air cells, and the gases, uniting with the oxygen so encountered, set up a distributed liberation of heat. This reduces viscosity and so aids the slow circulation of the mass, which as before men- tioned, proceeds by peripheral depression under weighting, and central upflow under the lake. It is difficult to prove this, for the interior of the bench magma is rarely under observation. This much, however, is certain, that the rock of the islands and bench crags, when revealed by collapse only a few weeks after it has solidified, is intensely oxidized. The ferrous iron in the basalt and the magnetite crystals have gone over to masses of earthy brown limonite and red hematite. The lava when it solidified was on the surface black, glistening and glassy, while within it was gray-black and lithoidal. Such rapid decomposition through JSaggar— Volcanologic Investigations at Kilauea. 207 and through for scores of feet one month from the time when live lava was overflowing, as in the collapsed west pit of February, 1917, points to profound and rapid oxidation with strong evolution of heat. Moreover, on March 14, 1917, in this same region, after subsidence of the iake of 84 feet (26™) in six weeks, the weathered pahoehoe flow surfaces, dead since January, began in places to glow with red heat, though 70 feet (21™) above the lake and quite remote from any live lava. This reheating appeared to be gas oxidation through small crevices. The writer as yet has no clear vision of the gradation down- ward of temperature and viscosity in the bench magma nor of its transition stages downward to where bench and lake magma are one. A two-phase convection in the deep region com- plicated by such opposed processes as adiabatic expansion and gas fluxing, or the assumptions on the one hand of increased temperature in depth from a heated substratum and, on the other, of decreased temperature owing to incomplete gas reac- tions, are too difficult for the student of observable facts. Apparently the bench magma and the lake magma in Hale- maumau, with their respective circulations, are definite facts, and for the present, speculation as to what is beneath may be postponed. ‘That the lava column is a stiff body with its gases in solution, and at no great depth, appears extremely probable. EXPERIMENTS TO DETERMINE DIFFERENTIAL TEMPERATURES. Queries Concerning Temperature. There have been hitherto presented in this paper evidences from observation at Kilauea volcano of a duplex lava column, of convection, of shallowness of the liquid lava lake, of heat supply and of oxidation of gases. The evidence from five years of recording, summarized above, necessarily leads to queries, answerable in part, at least, by some very simple ex- periments, if opportunity offer for direct contact with the lava lake, grottoes and blowing cones through the agency of simple instruments thrust into them. Such an opportunity of easy access to the fire pit was first presented to the writer in January, 1917, and for three weeks, while the lava remained high, a series of rough experiments was made with a view to answering leading questions. Such questions are concerned primarily with temperature, viscosity and depth. Would experiment show the lava lake to be hotter or cooler below the surface crusts, than the recorded temperature of about 1000° OC. measured with thermo- element by Perret and Shepherd in 1911 in the “ Old Faith- ful” fountain region? Would the grottoes give temperatures higher or lower than the maximum 1185° C. recorded with optical pyrometer by Day and Shepherd at the time of intense multiple fountaining in July, 1912% Is the temperature below 208 JSaggar— Volcanologic Investigations at Kilauea. the surface crusts high enough to melt those crusts when foundered? How are the temperatures araded from lake to grotto, or to blowing cone on the bench? How hot are the burning gases 4 Are they hot enough for refusion of the basalt 7 Reconnaissance and Method. The floor of Halemaumau around the liguid lake was recon- noitered from the east by a steep trail, and from the west with rope ladders, on January 3, 4, 5 and 7, 1917. The north- eastern rampart, a place of repeated overflow and largely built by spatter over grottoes, was found to be a favorable location for work at the actual edge of the liquid lake. We were here standing on the bench magma, and could thrust iron pipes into the lake magma. On January 4 an iron pot of lava was dipped up from the lake (fig. 17a), with three immersions, in the hope that such a pot might imprison the volcanic gas. sae pot was sent to the Geophysical Laboratory. On January 7, the high western crag mass (fig. 60) of the bench magma was climbed to its summit, and photographs made in all directions from this central position, It became evident that Seger cones enclosed in series in the ends of common “ gal- vanized* iron” (rolled steel) pipes, would be serviceable for determining approximate temperatures, and as the tempera- ture in the lake magma was expected to be at least 1000° C; the first Seger cones used were of fusibilities near that hgure (fig. 19). The conditions of this sort of experimentation are arduous, and the thrusting of a one-inch (2°54) steel pipe into the lava and withdrawing it, are by no means simple operations. The floor or bench was covered with pahoehoe lava flows freshly overflowed nearly every day, crusted with shells three to five inches (eight to thirteen centimeters) thick, and red hot beneath (fig. 17a). Several times I crossed flows which had poured out within four or five hours. The rampart was three to eight feet (one to two meters) high, built of glossy fling from the fountains, which incessantly spattered over it, with changing locations of activity. Consequently the surface of the rampart was hot enough to burn shoe leather, but one could stand on it with hob-nailed boots, by keeping in motion. The high lake when near to over flowing (Pl. Id, lower view) was most favorable for access and on account of its quietness, brimming close to the rampart edge and above the level of the floor behind. Work at such times, however, had to be rapid, as at any moment sudden rising might start overflow, or the rampart might collapse, or fountaining might begin at the * Zinc oxide would act as flux and make readings too high. Black steel was used later. JSaggar— Volcanologic Investigations at Kilauea. 209 place selected (fig. 20). The heat radiating from the lake sur- face at the rampart was very great and could be endured for a few minutes, but not for an unlimited time. On the surface of the lake, drifting with the current slowly along the bank, were heavy crusts like elephants’ hide (tig. 16a), which wrinkled and crackled and were hard stiff bodies several inches thick. No pipe would penetrate them and it was always neces- Fie. 19. Wiha g Fic. 19. Seger cones and partially fused one-inch steel pipe used in tem- perature experiments. Pipe as described for Jan. 26, 1917, fused through at bend, which was at flaming entrance to gas cupola. Fusion progressively less to end of pipe, which was farthest within the cupola. Cap over end destroyed. The upper row of Seger cones was fused in this pipe. The six rows- of Seger cones numbered from above downward are those resulting from tests; the short cones of the lower four rows are mechanically broken- but not fused; the fusibilities in degrees centigrade from left to right in each row are as follows: Le 1st row (uppermost), Jan. 26, 1917, 620, 710, 800, 870, 970, 1090, 1180. 2d row, Jan. 22, 1917, 590, 680, 770, 870, 970, 1070. a 3d row, Jan. 18, 1917, 2d test, 590, 680, 770, 870, 970, 1070. 4th row, Jan. 18, 1917, 1st test, 590, 680, 770, 870, 970, 1070. 5th row, Jan. 15, 1917, 770, 800, $70, 950, 970, 990. 6th row, Jan. 11, 1917, 990, 1030, 1070, 1110, 1180, 1150. Am. Jour. Sct.—FourtH SERIES, VoL. XLIV, No. 261.—SEPTEMBER, 1917. 15 ; 210 Jaggar— Volcanologie Investigations at Kilauea. sary to wait for a rending of the crusts, which revealed the viscous glowing slag beneath, or to find a place, otherwise con- venient, where one “of the ragged glow lines of upwelling was maintained (tig. 11d and 20). It was necessary, moreover, for Fic. 20. Fic. 20. Jan. 31,1917, 6p. m. Depression 48 ft. (15 m.). Looking N.W. Violent fountaining, blowing, flaming and overspatter in grotto under 8.E. rampart. Shortly after time of Pl. 1b; when quiet overflow had ceased and depression had begun. Note trajectory of flying spray. in the midst of which were great banners of flame. Wide glow lines under 5. island in con- trast to Pl. 1b: indicates incandescent overhang at lake margin due to temporary subsidence. This glow rim appears all around lake. Generai subsidence and fountaining follow intervals of quiet inflation. Photo Morihiro. such a liquid place to persist for some minutes, in order to leave the pipe submerged long enough to insure thorough heat- ing of the Seger cones within, and then it was problematical as to whether the pipe could be withdrawn, as heavy crusts were apt to pile against it and to freeze around it. It commonly took the united strength of four or five men to withdraw a pipe (fig. 17¢), and on two-occasions the pipe was lost. For- tunately the surface current proved shallow and while the heated pipe always bent during immersion, it was commonly held by border crust at the lake surface, so as to prevent its being swept along. JSaggar— Volcanologic Investigations at Kilauea. 211 Temperature of Lava Lake. The first three tests with Seger cones were directed to deter- mining the temperature of the liquid lava just beneath the surface crusts, with immersion of about three feet (one meter) of pipe. Six Seger cones were used in each test, of fusibilities differing progressively by about 40°C. The following were the results of these three tests (fig. 19): Jan. 11—Six Seger cones of fusibilities 990° C. to 1150° C., im- mersed 6 minutes, 3 feet (1™) below crust of lake; cones blackened but no trace of fusion (fig. 174). Highest glow and sharpest bending of steel pipe were at point of immer- sion, where gas burned in air, showing zinc flame coloration from galvanizing on pipe. Jan. 15—Six Seger cones of fusibilities 770° C. to 990° C., im- mersed 11 minutes, 3 feet below crust of lake; cones blackened but no trace of fusion. Jan. 16—Six Seger cones of fusibilities 770° C. to 990° C., im- mersed 30 minutes, 3 feet below crust of lake ; (fig. 16a) apparatus frozen into crusts and incandescent pipe rent asunder in the effort to recover apparatus by twisting. Cylinder containing cones lost (fig. 17¢). | In all these tests the cones were held by their bases in a special riveted iron container so that the tips were free within a three-inch (eight centimeters) steel cylinder eleven inches (28) long, covered with a screw cap at the lower end and serewed on a one-inch galvanized steel pipe at the upper end. This pipe, 20 feet (6 meters) long, was open at its upper end and was used as a handle for thrusting the cylinder beneath the lava. A rope was attached to the upper end of the pipe so that several men could pull in recovering it. These tests were surprising, especially that of January 15, when a cone, of fusibility 770° C., wholly failed to fuse though immersed three feet beneath the surface in the quiet streaming lava of the lake for eleven minutes. On the same day, January 15, two temperature readings were made with Bristol portable pyrometer, a commercial thermo-element. The element, without covering, was thrust directly into the liquid lava at the edge of the lake and left there until] the recording needle of the galvanometer came to rest. The first trial gave maximum temperature 860° C., the second trial 910°C. The upper end, or “cold junction,” of the apparatus was necessarily hot exposed in air to the radiant heat from the surface of the lake and the rampart. The makers furnish no correction chart for this source of error, and hence the actual temperatures, if the instrument is accurate, were somewhat higher than the figures recorded. But the range exhibited by the thermo-element, though higher 212 = Saggar— Volcanologic Investigations at Kilauea. than that of the Seger cones, is still much lower than was ex- pected. And probably the Seger cones afford the more reliable data. The writer recalls assisting Dr. Daly in 1909 in some temperature readings on the fountazns of Halemauman, made with Fery pyrometer, when a. temperature of 940° C. was registered, and this was believed at the time to be much too low. In general these first experiments of 1917 show that the lake magma is much cooler than has been supposed. — Temperatures of Grottoes and Flames. The next temperature tests with Seger cones January 18, 22 and 26, 1917, were devised with a view to determining the relative heat of the lake magma twenty feet (6™) from a grotto, ot the lava inside a erotto, and of the burning gas from a blowing cone. Supposing that possibly the three-inch (8) cylinder of the first experiments made too large an air space around the Seger cones, and also because such a cylinder was awkward for recovery when it became clogged with crusts, a simpler device was used for holding the cones. The six Seger cones were placed end to end in a spiral of steel wire nineteen inches (48%) long and this was inserted directly in the end of the one-inch pipe which was closed with a screw cap. The upper end of the pipe was open. From the results of this second series of experiments, which for the lake lava showed the same tendency to low temperatures, but elsewhere fused the Seger cones abundantly, the writer is convinced that the air space in the larger pipes somewhat invalidated their results. The following are the records of the second series (fig. 19) : Jan. 18—First Test. Six Seger cones of fusibilities 590° C. to 1070° C.,. heated close to lake surface 30 minutes, then - dipped under lavaone minute. Dipping was not complete, so that three upper Seger cones of low fusibilities were less blackened than the three lower ones. The three lower cones which were blackened. were of fusibilities 870° to 970° C. Only. the cone of 870° fusibility showed any fusion effect and this was very slight on the angular edges of the cone. As the pipe became involved in crust and the zinc galvanizing of its outer surface burned with a bright greenish yellow flame, it is possible that this fusion of cone 870° was induced by an artificial reaction. Jan. 18—Second Test. Repetition of first test, six Seger cones of fusibilities 590° to 1070 C. heated 30 minutes close to lake surface, then dipped completely for one minute so that all the cones were immersed in the lake. Again the pipe was entangled in crusts, bent and recovered with great difficulty. All of the cones were blackened, but none of them was fused, not even that of. fusibility 590°. Jan. 22—Six Seger cones of fusibilities 590° C. to 1070° C. in the end of a pipe 40 feet (12™) long were placed in the Jaggar— Volcanologic Investigations at Kilauea. 218 interior of the northeast grotto (like fig. 16) 20 feet from the locality of immersion of Jan. 18. The terminal was just over the boiling lava for 10 minutes; then 5 feet (1°5™) of the pipe was submerged in the lava for 5 minutes. Jan. ) The pipe was drawn down violently and recovered, much bent, with great difficulty. The Seger cones were all fused in some measure, that of fusibility 1070° least so. The temperature of the grotto was thus at least 1100° C. 6—Seven Seger cones of fusibilities 670° to 1130° C. were placed in the glowing intericr of a blowing dome (fig. 21) which had been built over the northwest inlet well or pond above a roof of arched lava flows. About 3 feet (1™) of the end of the pipe was thrust for 9 minutes in a flaming orifice 10 inches (25°) in diameter in the side of the dome. The flame was under pressure and of the blue-green variety. ‘The pipe was of ordinary commercial galvanized steel covered with a screw cap of annealed cast iron, and the Seger cones within were in a spiral of “spring steel ” wire. Lava was splashing in a cavernous space 12 feet (4™) below. The galvanizing first burned off the pipe, then the surface of pipe was seen to be continuously dripping with molten incandescent drops, especially at the window, and after 9 minutes the pipe was eaten through, apparently Jused, for a length of about 9 inches, just at the window orifice where the visible gas flame began (fig. 19). The incandescence was highest opposite the window, a strong yellow heat; that inside the dome was orange. The pipe was at once removed. ‘The molten drops appeared to have been molten iron. The cast iron cap was almost com- pletely gone. The pipe showed gradations of fusion from a maximum at the flaming window to a minimum at the inner end. The spring steel spiral inside showed no sign of fusion. The Seger cones were all fused, that of fusibil- ity 1130° C. least so, indicating a probable temperature over 1200° C. It was at first thought that the corrosion of the pipe was oxidation, but in view of the dripping fusion and the differential effects on three classes of iron, and the fact that after the cap was destroyed, the inner spiral had every opportunity to oxidize, but did not do so, it seems that actual fusion took place. Annealed gray cast iron (the cap) fuses at 1220° to 1275° C., rolled steel (the pipe) fuses at 1325° to 1375° C., and hardened steel (the spiral) fuses at 1425° to 1500°C. Wherefore we may conclude that the temperature of the burning gas at the flaming window was close to 1350° and that inside the cupola near 1250°. Even if the metallic reaction was largely oxidation, these figures would not be materially changed, as the Seger cone 1130° was greatly fused within the dome, and the incandescence of the pipe at the window was much higher. The dripping melt in such case was an iron-sulphur oxide flux of fusibility probably near 1250° C. 214 JSaggar— Volcanologic Investigations at Kilauea. Summary of Temperatures. By summarizing the results of these two groups of tests it appears that the lake magma immediately below the surface glow lines has a temperature range from 750° to 850° C.; that the lava in the fountaining grottoes has a temperature range ron INOOs yo. 00° (Che and that the oxidizing gases in the blowing cones have temperatures ranging from 1250° C. within the cones to 1350° C. or higher at the flaming orifices. Reverting now to the queries asked at the beginning of this section, it would appear that the temperature range measured by Perret, Day and Shepherd, namely 1000° to 1185° C., is correct for the fountaining lava, but that the temperature of ordinary lava beneath the crusts is not the same as that of the fountains, but much lower. As the melting point of the basalt BiG Fie. 21. Jan. 7, 1917. Looking W, at blowing cone built above arched lava flows roofing N.W. conduit pond. A similar roof was seen to fall in suddenly a few weeks before. The middle figure is looking into flaming ori- fice, 5 ft. (1m.) across, in summit of cone and 12 ft. (8 m.) below could be seen the fountaining lava. This cone, built higher by spatter, with lateral window instead of summit orifice, was the scene of temperature measure- ment which fused steel pipe Jan. 26, 1917. Photo Jaggar. is near 1100° C., the lake melt would rank as superfused, or fluent and unsoliditied below its fusing point, if it could be treated asa pure melt. Furthermore it becomes plain, on the other hand, that the hot gases rising through this same lava become vastly hotter, when confined in a spatter cupola above the lava, and hotter still when liberated for complete combus- tion in air. The conclusion seems justified that the heating effect is due to union with oxygen, and that this union begins below the fountains, increases within the blowing cones, and culminates in the visible flames. JSaggar— Volcanologic Investigations at Kilauea. 215 Refusion, Two deductions may be made at once from the temperature distribution in a blowing cone: First, that the more slowly oxidizing gases at 1250° CO. inside the cone are well above the melting point of the lava walls of the cone, and are also able to oxidize the ferrous iron of the interior surface. The results are seen in the secondar y glazes, often brown or silvery with ferric iron, found in the interiors of caverns; they are seen also in the “peculiar cavern stalac- tites; these glazings mark a refusion of cavern linings so as to destroy the visible vesiculation of a rough surface and to coat the whole with a smooth mantle of porcellanous aspect. These secondary linings are found everywhere on Kilauea, where the pahoehoe lava has withdrawn from its crusted portions and left cavernous spaces. The linings and _ stalactitie forms are in great variety of color and texture, and it has been for years apparent to the writer that nothing would account for them satisfactorily except a mechanism of refusion, because of their lithologie distinction from the molten rock. When refusion in a blowing cone goes far enough it results in the glowing “filagree ” cone, fused through from within, which may col: lapse and reform, during rising of the lava, again and ae The rising lava is cooler and construets : the gas chamber of a subsidence is hotter and destroys. Furnace effect. Second, that a blowing cone is an Sroclient furnace stack for creating a draught of air into the interior from below, the hottest portions being at the orifices in the summit. This would create a strong updraught, and as all blowing cones contain gas chambers over Java connected in some way with a larger pool, and commonly at the pool level, there would always be an air channel. to supply oxygen to the base of the stack, during the temporary subsidences of the liquid column, which from hour to hour alternate with risings. There- fore rising lava in the stack would yield the purest magmatic gas ; sinking lava should produce more combustion within the editice. The putting which has given these cones their name may be a combustion phenomenon ; it is sometimes strikingly like the noise made by a locomotive. Relutive Coolness of Lake. As to the normal lava lake temperatures, it is clear that foundering crusts in a melt of 800°, 900° or even 1000° C. would have no chance for refusion ‘(melting point 1100° to 1200° C.). On the contrary, in a superfused pure melt their tendency, if not opposed by reheating mechanisms, would be 216 JSaggar— Volcanologic Investigations at Kilauea. to start solidification. They would pile up on the bottom, were they not disintegrated by explosion and local reheating, through the union of magmatic gases with the oxygen in their vesicles. As has been explained above, in view of the fact that great quantities of crust have been seen foundering without immediate fountaining effect, there is probably such piling up along with storage of the oxygen through gas-proof glazing, which coats the exterior of foundered blocks. This would be especially expectable if the lake is shallow, and more viscous near the bottem, in places removed from the sinkholes. EXPERIMENTS TO DETERMINE DerrprTtH AND CONSISTENCY. Queries concerning Depth and Consistency. In order to settle this point with reference to the liquid lake, shallowness being a seemingly indispensable condition to account for the facts observed, an experiment was devised having in view the bold project of actually sounding the lava pool of Halemaumau. The preliminary work with iron pipes and Seger cones had shown that the crusts are heavy and stiff, the lava beneath sufficiently plastic to be dipped up in an iron pot, but by no means as mobile as molten lead, while the lava of the grottoes seemed much more liquid when stirred with an iron pipe. The grotto lava which solidified around the pipe terminal exhibited dull oxidized surfaces, smaller vesicles, finer texture, and a dense selvage next the iron, while the accumula- tions of normal lake lava were lustrous, lighter, glassy and coarsely vesicular. The streaming currents in the lake are neither strong nor deep in contact with the submerged pipes, but in the grottoes they are very powerful. The queries suggested concerning viscosity and depth, possi- bly answerable by experiment, are as follows: (1) when the lake is at a high level, at what depth would a sounding rod strike bottom? (2) Is the lake magma stiffer or more liquid in depth 4 Measurement of Depth. On January 23, 1917, with the surface of the lava lake de- pressed only 50 feet (18 meters) below the rim of Halemaumau the writer tried sounding with the aid of half-inch (1:27) “black iron” (steel) pipe. .A total of 200 feet (60 meters) of piping was screwed together and this was laid out across the northeast floor adjacent to the eastern rampart of the lake margin. In the first test Sever cones were inserted in the end of the pipe nearest the lake and confined with a screw plug, but as this pipe was lost, the temperature in depth was not determined by this means. The extremity of the pipe remote from the lake was attached by a rope toa block and JSaggar— Volcanologic Investigations at Kilauea. 9217 tackle and nine men volunteered to assist the writer in carrying the pipe end on, so as to immerse it in the lake at a high angle, try for the bottom, and withdraw the apparatus with the aid of the rope. | In the first test the pipe was thrust out over the bank at the east point in a southwest direction toward the center of the lake, the bank here being 15 feet (4 meters) high (fig.10). The terminal was thrust downward striking the lake surface some 20 feet (6 meters) horizontally out from the shore, at a glowing liquid zone free from crust, and for from 30 to 40 feet (9 to 12 meters) of depth the pipe descended freely at an angle of about 45°, but sagging to the vertical as the length increased. Be- yond this depth an increasing resistance was gradually encoun- tered, and finally caused the pipe to arch upward and fail to penetrate farther. Sixty feet (18 meters) of piping was sub- merged, corresponding to a vertical depth of approximately 50 feet (15 meters). At this place, then, the lake was 50 feet deep some 30 feet (9 meters) from the shoreline, (the pipe be- neath the surface extending beyond the point of immersion), Subsidence during the following month indicated that this shoreline beneath the liquid was precipitous. Therefore there was no error due to shelving shore. On withdrawing the pipe some 30 feet (9 meters) of the sub- merged portion was recovered quite uninjured by any sign of oxidation or fusion, but the remaining three 20-foot lengths became imprisoned against the bank, owing to the pressing in and piling up of crusts around the pipe, so that it was quite impossible to recover the terminal length which contained Seger cones. In order to prove that the shallowness of the lake and vis- cosity of the bottom layers, seemingly indicated by this first test, is general and not local, a second trial was made on the same day at a point 40 yards (87 meters) farther north, and about where the temperature tests had been made. ‘This place was only a few feet trom the northeast grotto domes, towards which there was a strong current. The pipe was thrust over the bank more rapidly than in the first test, and catching the current tended to arch downward steeply and to bow somewhat toward the north. The results were quite the same as in the first test. In both cases the writer stood at the lake margin manipulating the pipe as it descended and the increasing resistance was very marked during the last 10 feet (3 meters) of its descent. In this second test 61 feet (18 meters) of piping was submerged before the impenetrable pudding was reached. The pipe was withdrawn completely this time but only through the most strenuous hauling by nine men in line. Twice it stuck owing to accumulations of lava which caught against the bank. It must be remembered that in such with- 218 JSaggar— Volcanologic [Investigations at Kilauea. drawal the pipe cannot be pulled up hand over hand, because it emerges from the lava red-hot. The only way to pull it out is by walking away from the rampart and leaving the hot por- tion to trail over the bank. The end of the steel pipe emerged clean and showed no fusion or especial oxidation, nothing to indicate that the depths of the lake were comparable in tem- perature with the flaming cones. This second sounding appara- tus was not equipped with Seger cones. Confirmation of Soundings by subsequent Subsidence. The second sounding test, like the first one, but at a different place, indicated a depth of liquid lava, in the lake opposite the northeast rampart, of 50 feet (15 meter s), with the lower few feet more viscous than the fluid above, and consequently of presum- ably lower temperature. This place, like the east point which was the scene of the first test, became during the following month, February 1917, a vertical cliff over the subsided lake. The sounding was made January 23, and on January 19 thelake had been 68. feet (21 meters) below the rim and was rising approximately two feet (61™) per day, so that on January 23 the depression was 60 feet (18 meters). On February 22 ae depression of the lake was 106 feet (82 meters) or 46 feet (14 meters) lower, and the day before, ne y 21, a cascade devel- oped from the lake over the submerged ledge of the lake bottom into a cavernous recess northeast, nearly under the locality where the second sounding test was made. In other words, the revelation of a month of subsidence was the actual outcropping through the lake at its margin of a por- tion of its bottom 46 feet below the lake level of the previous month. And the ee in two places at this margin in the previous month revealed about 50 feet (15 meters) of depth. Here we have close accordance between the results of experi- ment and the evidence afforded by changes in time. Incidentally this development of a cascade, at a marginal sink- hole after about 50 feet of subsidence, is nothing new, but has happened repeatedly before as one ‘of the characteristics of subsidence. The evidence of the existence of a ledge over which such a cataract one fall (fig. 2, see section of sinkhole) has in the past been puzzling, but this is fully explained when we realize that the bottom builds up with the rising pool, that ie latter is habitually shallow and that when subsidence sets , the lake magma becomes even shallower, by sinking more rapidly than the semi-solid bench magma. The lake at such a time assumes the appearance of a rapidly streaming river. These preliminary experiments provide a method and open the way to interesting further work relative to gas composition, viscosity and temperature in depth, direction of the subsurface JSaggar — Volcanologic Investigations at Kilauea. 219 currents, depth and temperature of the sinkholes themselves, differences between conduits and sinkholes, and variations in depth in the lake under different conditions and in different places. It is hoped that work along these lines will be con- tinued as opportunity offers. Summary of Depth and Consistency. Apparently our two queries are tentatively answered, and the answer accords with the conception of lake magma and bench magma outlined in this paper. The liquid lake in January, 1917, had a depth of about 50 feet (15 meters), with sinkholes as downflow shafts, and conduits as inlets, distributed under it and around its margins to the number of at least eight. Secondly the lake magma, judging by its resistance to a sound- ing rod, was of uniform viscous consistency beneath the crust layer for approximately 40 feet (12 meters) of depth and increas- ingly stiff to semi-solid for the remaining 10 feet (3 meters). ConcLUSION. The main contributions of this paper to volcanologic science as elucidated at Kilauea volcano are as follows: (1.) The Java column in Halemaumau pit at times of charac- teristic activity is duplex. It consists, first, of a main semi- solid incandescent body (bench magma), filling the whole true crater from side to side for an unknown depth, perforated ver- tically from below by several small shafts leading to a saucer in its summit. Saucer and shafts are filled with the minor liquid lava body (lake magma), which exhibits a more rapid convec- tional circulation than the main body. The lake magma contrib- utes substance to the bench magma by overflow and accretion during a rise of the lava column. A slow circulation, which resembles isostatic adjustment in its mechanism, is discernible in the bench magma. During subsidence the lake magma sinks more rapidly than the bench magina, uncovering portions of the saucer. | (2.) Magmatic gases, circulating by convection in the lava column, are in tbe upper crater brought into contact with atmospheric oxygen mechanically by the circulation, which acquires a superticia! acceleration presumably shallow, and the heat effect is sufficient to produce most, if not all, of the dis- tinctive phenomena of the lake magma. Two antithetical conditions are realized in the crater by these generalizations. The bench magma is a product seemingly ot ‘solidification above normal viscosity, and the lake magma a product of liquefaction below normal viscosity. The normal lava column would then seem to be of a sub- stance not commonly revealed at the surface on Kilauea 220 JSaggar— Volcanologic Investigations at Kilauea. voleano.* An exceedingly fascinating field of speculation is here brought into view, containing the possibility that the great aa, or block-lava, flows of Mauna Loa, emergent under high gas pressure and sudden release, are of the normal magma. The intense and distributive oxidation of the gases through such a flow, when released to atmospheric attack, and the ill adjusted colidification and expansion-cooling effects from within outward, added to the fact of acquisition of pahoehoe, or fluent lava, characters near the vent, when furnace conditions are estab- lished there, lend color to this view, which, however, cannot be advanced as anything more than a suggestion in the present stage of inquiry.t The application of the principles developed by this study of Kilauea to the mechanism of other volcanoes, and the termin- ology of less local limitation to be adopted when stiff bench magma and reheated fluent magma are recognized elsewhere, form topics too far-reaching to be discussed in this paper. The writer has in preparation a contribution on these subjects. The question of viscosity vs. gas reaction has been the key- note of voleanology since the eruption of Pelée in 1902, and the classification of volcanoes in genetic series is dependent on discovering, in this relation, the meaning of voleano distribu- tion. The crags of bench magma described in this paper are much like the steep peaks rising ‘from the floor of many lunar craters, and the selenologists can surely help to solve the volcano prob- lem. Jam convinced from what I have seen of Kilauea that the problem will never be solved by expeditions or closet theoriz- ings. The record of voleano process, as I have tried to show all too briefly, involves the registration of change and measure- ment of dimension in relation to passage of time. The only sound mode of attack is through permanent laboratories in the actual voleano field. Such laboratories, adequately equipped, have not yet been established. * Since this was written an extraordinary confirmation of the hypothesis here stated was afforded by the east island (fig. 8b), the base of which proved to be typical block-lava of the Mauna Loa variety. This prior to Feb. 17, 1917, had been the Kilauea lake-bottom. See Jour. Wash. Acad., 1917. + This Journal, xliii, pp. 255-288, April, 1917. Hawaiian Volcano Observatory, March 7, 1917. Browning and Porter— Gallium. 221 Arr. XVII.—On the Qualitative Separation and Detection of Galloum ; by Puirip KE. Browntne and Lyman E. Porrer. [Contribution from the Kent Chemical Laboratory of Yale Univ.—cexci. | GALLIuM, discovered in 1875 by Lecoq de Boisbaudran,* is found in nature most closely associated with the elements aluminium, iron, manganese, zinc, lead, and indiuin.+ Analytically it falls into the aluminium group. It may be separated from the bases giving sulphides in acid solution by hydrogen sulphide; from nickel, cobalt, zinc, manganese, the alkali earths, and the alkalies by ammonium hydroxide in the presence of ammonium chloride; and from iron, titanium, thallic thallium, uranium, indium, and the rare earths by sodium hydroxide in excess, in which reagent the hydroxide of gallium is soluble. In the ordinary course of analysis it appears In the group containing aluminium, beryllium, chro- mium, and vanadium. From the last mentioned two elements it may be separated by ammonium hydroxide after their oxida- tion to the acidie condition. | 2 . This narrows the problem of separation down to the separa- tion from aluminium and beryllium; and the practical absence of beryllium from products containing gallium leaves the most important separation, that from alumininm. Lecoq de Boisbaudran, in a series of articles published soon after the announcement of his discovery,t suggested many methods of separation from the other elements, and recom- mended especially for the separation from aluminium the use as a precipitants of potassinm ferrocyanide in the presence of strong hydrochloric acid to about one-third of the volume of the solution. | In previous papers| one of us has shown that silver, lead, zine, copper, and indium have been successfully separated from gallium by various applications and modifications of known methods. The object of this paper is to give the results of some work upon the application of potassium ferrocyanide to the separation of gallium from aluminium and beryllium, and to describe the outcome of experiments upon the delicacy of the test for gallium by the ferrocyanide method, upon the * Compt. rend. (Paris), Ixxxi, 493. ; | + Boulanger and Bardet, Compt. rend. (Paris), clvii, 718 ; Hartley & Ram- age, Jour. London Chem. Soc., 1897. 533, 547; Hillebrand and Schnerrer, Ind. Eng. Chem., viii, 225. t{Comp. rend. (Paris), xciv, 1154, 1228, 1439, 1628; xev, 157, 410, 503, 1192, 1882: xevi, 152, 1696, 1838; xcvii, 142, 295, 522, 623, 730, 1463. S Comp. rend. (Paris), xcix, 526. | Browning and Uhler, this Journal, xli, 351, Apr. 1916; Uhler and Brown- ing, ibid., xlii, 389, Noy. 1916. 222 =Browning and Porter— Qualitative Separation decomposition of the ferrocyanide when formed and upon the application of strong hydrochloric acid to the separation of gallium and aluminium. It was found that when solutions containing about 0-1 grm. of aluminium or beryllium were strongly acidified with hydro- chloric acid and treated with potassium ferrocyanide no precip- itation took place, while 0001 grm. of gallium in the presence of 0-1 grm. of aluminium was easily precipitated and detected at once. Amounts of gallium as small as 0:0001 grm. could be detected after the solution had been allowed to stand an hour or so. These tests were generally made in a volume of liquid from about 5° to 10°, of which from one-quarter to one- third was strong hydrochloric acid. With traces of zine present, the use of potassium ferrocyan- ide as the precipitant may lead to erroneous conclusions, because zinc ferrocyanide is almost as readily precipitated as the gal- lium. The presence of zinc may be avoided by the careful application of the ammonium chloride and ammonium hydrox- ide process. Should, however, traces of zinc remain, we have found that they may be satisfactorily detected and removed by treating a sodium hydroxide solution with hydrogen sulphide, which removes the zine without precipitating the gallium. The filtrate, which must still be alkaline, is acidified, and free hydrogen sulphide removed by boiling. The sulphur is oxi- dized by hydrogen dioxide in sodium hydroxide solution, and the boiling is continued to remove the excess of hydrogen diox- ide. This solution is then acidified with hydrochloric acid and the usual ferrocyanide test may be made for gallium. Solnu- tions were prepared containing gallium and zine and were analyzed by the experimenter without knowledge of the con- tent. The results follow: Issued Found (1) 0°001 grm. Ga Zn absent, Ga present (2) 0°001 grm. Zn + 0°001 germ. Ga ~—— Zn _ present, Ga present (3) 0°001 grm. Zn Zn present, Ga absent (4) Distilled water Zn absent, Ga absent (5) Gu. ~.m. Zn Zn present, Ga absent (6) 0:0%9 erm, Zn Zn present, Ga very faint indication (7) 0°050 grm. Zn + 0°001 grm. Ga Zn present, Ga present (S) 00528 crm. Zn + 0:0002 grm.Ga Zn present, Ga present The faint indication ot tne presence of Ga in experiment (6) seemed to indicate a trace ot that element in the zine. It is of interest to note that this ind‘cation was not obtained until the solution had stood twenty minutes, while in experiment (8) the and Detection of Gallium. 923 test for the gallium was unmistakable and practically imme- diate. A number of reactions were investigated leading to the decomposition of the gallium ferrocyanide and the recovery of the gallium as the hydroxide, such as treatment with: bromine and with nitric acid, and fusion with sodium peroxide and with ammonium nitrate. The most satisfactory proved to be fusion with ammonium nitrate, which destroyed the ferrocyanide rad- ical, and subsequent treatment with sodinm hydroxide, which precipitated the ferric hydroxide and left the gallium in solu- tion, from which the hydroxide could be readily precipitated by adding ammonium chloride in excess and boiling. Gooch and Havens* have shown that iron may be separated from aluminium by saturating solutions containing these ele- ments with hydrochloric acid gas, adding ether and again saturating. The chloride of aluminium is completely precip- itated by this method, and the iron remains in solution. This process was applied successfully to the separation of aluminium from gallium. The presence or absence of gallium may be determined by evaporating the filtrate to dryness on a steam bath and dissolving the residue in dilute hydrochloric acid. This solution, which is free from aluminium, may be tested for gallium by means of potassium ferrocyanide. The following series of unknown solutions was tested by this method, the aluminium chloride used having been purified by the hydro- chloric acid precipitation : Issued Found (1) 0°0005 grm. Ga Al absent, Ga present (2) 071 grm. Al Al present, Ga absent (3) 0-1 grm. Al + 0°0005 grm. Ga Al present, Ga present (4) Distilled water Al absent, Ga absent (5) 0°02 grm. Al + 0°001 grm. Ga Al present, Ga present (6) 00001 grm. Ga Al absent, Ga present During the first trial of this method aluminium nitrate was used in hydrochloric acid solution, and it was found that no vrecipitation took place with potassium ferrocyanide. After treatment with the hydrogen chloride gas and evaporation, however, an indication of gallium was found. This led to an investigation which showed that there was some gallium pres- ent in the aluminium nitrate, but that the nitric acid formed by dissolving it in hydrochloric acid was suflicient to prevent the precipitation of the gallium as the ferrocyanide. In the evaporation process the nitric acid is destroyed and the test becomes very delicate. * This Journal, ii, 416, 1896. 224 Browning and Porter—Gallium. - An investigation was then made of the effect of nitrates in general on ferrocyanides. It was found that when one drop of potassium ferrocyanide is treated with 0-4 germ. of ammo- nium nitrate in the presence of 6° of 1:2 hydrochloric acid it is oxidized completely to the ferricyanide within two min- utes, as may be shown by the use of a ferric salt and a ferrous salt. If 0°2 germ. of ammonium nitrate is used under the same conditions, the ferr ocyanide is broken up in less than an hour, while if only 0°-l grm. of ammonium nitrate is used a longer time is reqnired, but complete oxidation finally takes place. Other experiments showed that 0°0001 grm. of gallium cannot be precipitated as the ferrocyanide in 5°™* of dilute hydrochloric acid in the presence of 3 drops of dilute nitric acid, whereas it is readily precipitated in the absence of it. It was further found that if 0°000L grm. of gallium is precipitated and 1°™ of dilute nitric acid is added, the precipitate is decomposed and dissolved within forty-five ‘minutes. It is thus seen that in detecting gallium by the ferrocyanide method care must be taken to have no nitrates or nitric acid present, and that these may be successfully removed by evaporation with hydrochloric acid. June, 1917. O. C. Lester —Emanation Electroscopes. 995 Art. X VIII.—On the Calibration and the Constants of Emana- tion Hlectroscopes ; by O. C. Lester. In the summer of 1914 the author began a fairly exhaustive investigation of the radioactivity of the numerous mineral springs found .chiefly in the mountainous region of Colorado. Most of the work was done in the field with instruments of the usual types constructed in our laboratory. However, owing to certain peculiar conditions encountered in Colorado, it was found necessary to make a careful study of the behavior of the apparatus employed and to determine the significance of cer- tain calibration constants. Further emphasis was given to this study when it was found often practically impossible to com- pare the work of different observers not only in different coun- tries but even in the same country. This is particularly true of the work of observers who express their results in terms of the mache unit. Such a state of affairs ought not to exist in any branch of science and an attempt is here made to point out the reasons for the discrepancies, the remedy for which is suggested by a study of the calibration constants and the cor- rections which should be applied in the use of ae elec- troscopes. The emanation electroscopes used consisted of somone loniza- tion chambers of different sizes, to each of which could be attached the same electroscope head, including microscope, leaf, and electrode. In. general the appat atus is similar to that described by S. C. Lind* except that there is only one insula- tion plug instead of two. Its essential features are shown in fig. 1. Lis an air-tight cylindrical brass ionization chamber having stop cocks V near the top and bottom. Altogether four such chambers were used, all of them taking the same electroscope head but each having its own electrode E. The inside dimensions of the chambers and the outside dimensions of their electrodes are given in the following table. The elec- trodes are made of light brass tubing capped at each end. TABLE I. Electrodes I Length, cm. Diam., cm. Toh aoe |. Diana No. 2 24°8 cay | LO? 1°6 Noi. 3 25°1 15°6 TSG 1°6 Nos. 6 & 21°2 13°6 Lost 1°6 *U. S. Bureau of Mines, Tech. Paper 88, Mineral Tech. 6, p. 17, 1910. Am. Jour. Sct.—Fourts Series, Vou. XLIV, No. 261.—Srpremser, 1917. _O. C. Lester—Emanation Electroscopes. 226 Fig. 1. O. CO. Lester—Emanation Hlectroscopes. 927 Most of the work was done with vessels No. 2 and No. 3, the latter being the most sensitive of the four. Nos. 6 and 7 were constructed after the experience of the first summer, and are, on the whole, the most satisfactory. They were designed to have approximately a volume of 3 liters and a distance of 6°™ between the electrode and the outer wall. Each electroscope consisting of the common head and an ionization chamber with its appropriate electrode was carefully ‘standardized a number of times by means of known quantities of radium emanation obtained from pitchblende as suggested by Boltwood.* For this purpose some finely-ground pitch- blende giving 2°10X10~" curie of radium emanation per mg. on direct solution was kindly furnished by Richard B. Moore of the United States Bureau of Mines, Denver. The emana- tion from several milligrams of pitchblende was introduced into the electroscope and the activity observed every few min- utes until it reached its maximum. In each case a complete and typical activity-time curve was platted. Then from the known amount of emanation present and the observed maxi- mum activity, that fraction of a curie which will produce a movement of the leaf of 1 div. per min. at maximum activity ean. be caleulated. This fraction of a curie is called the con- stant of the electroscope. Numerous trials show that readings must always be taken between the same points on the scale or symmetrically about the middle point of the portion used in calibration. Thus if the electroscope is standardized for the portion of the scale lying between 70 and 30, the same maximum activity will be found and hence the same constant will hold for readings taken between 60 and 40 but not, for example, if they are taken between 70 and 40 or between 60 and 30. The shorter distance is sometimes convenient when dealing with weak activity. The constants of each ionization chamber as determined at Boulder at a pressure of 62°5° and at a temperature of about 22° are 2-34 <10-" curie for No. 2, 1-89 10-* curie for No. 3, and 2°0710-"° curie for Nos. 6 and 7. These values are the means of six or more concordant determinations for each chamber. Strictly speaking these constants hold only for a given pressure and temperature in the case of chambers whose volume or air density is not large. As the springs examined are at elevations varying approximately from 5,000 ft. to 10,000 ft. which causes changes in pressure from about 64% to 53™ the constants given above were of little value in the field work. This made necessary an investigation of the way in which the “ constants” varied with the pressure. Previous investigations on the variation of ionization with pressure such as those of * This Journal (4), vol. xviii, p. 378, 1904. 228 -O. C. Lester —Emanation Electroscopes. Rutherford* and Owenst do not fit the conditions of the pres- ent work as they used radiations from layers of solid substances in vessels of wholly different shape. The investigations of W. Wilsont C. T. R. Wilson§, McLennan and Burton n| and Patter- son* deal with the general question, but again under different conditions. Furthermore they are not all in agreement. In order therefore to find how the activity at its maximum varied with the pressure when emanation was mixed with air. in cylindrical vessels and incidentally also to see how nearly the maxima were proportional to the amount of emanation present, a series of tests were run in each chamber. The pro- cedure followed was similar to that described by Madam Curie.** However, what was here sought was a relation which would give the “ constant ’ corresponding to any barometric pressure and thus permit the reduction of the results of observation immediately to curies, rather than a correction term to be applied to the observed ionization current as in Madame Curie’s procedure. After the pressure in the chamber had been reduced to a few centimeters a known amount of emanation was introduced. During this operation the pressure increased to 10 or 20™, After the electroscope had stood charged for a little more than three hours the activity was measured at various pressures determined by a mercury monometer. The relations between pressure and activity in vessel No. 2 may serve as a typical example. These relations for varying amounts of emanation are shown in Table II and by curves in fig. 2. The figures in the body of the table are maximum activities in divisions per minute taken from the curves. In fig. 2 the actual experi- mental data are represented by the continuous lines. TABLE IT. Mes ano | “500 | 600° | 700!" 800“) See 5:06 3-14 3-82 | 4-40 Ae860 17 G5 -oHl 10°35? || 642 | 7-73 | 878 | 9-42 | 9-71 15°35 961 11-49 | 13-01 | 1402 | 14°58 20°57 1310 1600 | 17:90 19:18 | 19°95 25°20 | 15°84 19°31 | 2219 23°93 24-93 * Phil. Mag., vol. xlviii, p 109, 1899. + Ibid., p. 360. t Ibid., (6), vol. xvii, p. 216, 1909. § Proc. Roy. Soc., vol. lxix, p. 277, 1901. | Phys. Rev., vol. xvi, p. 184, 1903. “| Phil. Mag.. (6), p. 281, 1903. *% Traité de Radioactivité, vol. i, p. 286. Fic. 2. 10U0 | A san NES ER 2 1200 800 : evaneac Eo Sent O. C. Lester—Emanation Hlectroscopes. co ie Aqratqoyv 16 | Sones SCENE SS cS 600 Ns ceaie Sy) save is SaNesenane eglels|_ NOSE EERT CT ¢ A ie Bele ey ‘950 0°971 0°990 0 0°934 0°950 Pressure in nim. O°Si7 1 0°880 ~ 600 500 0°732 0776 0-766 TABLE III. 400 0°637 0°628 ~ 0-608 Pressure in mm... . Mean 230 O. C. Lester— Emanation Electroscopes. Table III shows that the ratio of the activity to the amount of emanation is approximately constant at a given pressure, at least for the range of activity here examined. There is good reason to suspect an error in the weight for the second sample. The maximum activity multipled by the “constant” of the electroscope and divided by the volume of water or gas taken gives the number of curies per unit volume which is a fixed quantity. However, since the maximum varies with the pres- sure the ‘‘ constant’ must vary also, but we should always have activity X constant = curies or mk = © ee which is the familiar equilateral hyperbola or Boyle’s Law equation. Now the constants of each chamber are known accurately for a temperature of 22°O anda pressure of 62°5°. From the curves of fig. 2 we find the corresponding mean maximnm activities per milligram of pitchblende to be 0°881 divisions per minute for chamber No. 2 and 1-077 division per minute for chamber No. 3. Hence the constant K, for any pressure p is found from Key and KM, = 2°053 & 107" for chamber No. 2. D036" DP. : +H. Mache and St. Meyer, Rad. in Biol. und Heilk., vol. i, p. 350, 1912. Gockel, Radioactivitat von Boden und Quellen, p. 82. O. C. Lester—Emanation Electroscopes. 233 the air in them, and the difference of potential between the electrodes and walls that there was neither opportunity for the production of all the ions possible nor a sufficiently strong elec- tric field to remove all those that were produced. This makes no difference for measurements in curies* if the chambers have been properly calibrated. However, it is possible from the work of Duanet and of Duane and Laborde? to caleulate for such chambers as were used the relation between the maximum ionization current actually observed and the number of curies which would produce it if the radiations had been completely absorbed and saturation had obtained.$ In this way it was found that the loss, due to Jack of range and saturation, in the ionization current upon which the mache unit is based, amounted to 48 per cent in chamber No. 2, to 46°7 per cent in chamber No. 3,and to 44 per cent in chambers No. 6 and 7. Hence in all the chambers used except No. 2 it happens that the loss in activity due to the above-mentioned causes is almost exactly counterbalanced by that added by the active deposit. The ionization current in electrostatic units is given by j= gem PesndOe:t (3) where g is the drop in potential in volts per scale division, m is the number of scale divisions passed over by the leaf in 7 seconds, and ¢ is the electrical capacity of the instrument. In the case of chamber No. 2, g = 1:22, C = 8-4, and if m is the number of divisions per minute passed over by the leat at maximum activity due to emanation from v liters of water or gas (3) becomes eee: (1°22){8°4) m 2 as me ieaeas 10 (4) (80U)(60} * v Oi diter When corrected by Duane’s factor for the absorption due to the walls of the vessel and for the activity due to the decay products according to the percentages given above (4+) becomes 17 bored tO (5) Vv where I denotes the total ionization current which could be produced by the emanation alone if all its radiation was absorbed in air. Equation (5) holds for a barometric pressure of 62°5™ only and the observed values of m must be corrected by a factor 6 * W.R. Barss, this Journal, vol. xxxiii, p. 546, 1912. + Loe. cit. Loc. cit. : $G. Berndt, Ann. der Phys. (4), vol. xxxviii, p. 958, 1912. See also Gockel, loc. cit., Chap. VIII, and H. Mache and St. Meyer Phys. Zeitschr., vol, xiii, p. 320, 1912. 234 O. C. Lester— Emanation Electroscopes. which varies with the pressure in exactly the same way as the “constant”? expressed in curies. The values of 6 for various pressures may be taken from a curve easily derived from the activity-pressure curves (fig. 2) or better from the constant- pressure curve (fig. 3). The latter curve and the d-curve have exactly the same form since the value of 6 is directly propor- tional to the value of the ‘‘ constant” at a given pressure. For chamber No. 2 the values of 6 at pressures of 40, 50, 60, 70 and 80 ecm., are respectively 1°41, 1:166, 1:025, 0°949, and 0-905. On putting in the pressure factor equation (5) becomes I = 0635 x 10-75 = (6) This again must be multiplied by 1000 in calculating mache units. Hence from the ionization current Mache units = 0°635 6 = (7) But from the generally accepted relation between the mache unit and the curie and from the calibration of the chamber directly in curies we have Mache units: == (2°7>x< 107), S@(curies) 1257 )Gal0. a (8) v Where & is the constant of the electroscope defined by equa- tion (2). 3 In Table IV are given a few results of the calculations in mache units by both methods for chamber No. 2. The data given are based upon observations taken in the field. TABLE IV. Mache units Bar.Px., em! = Kex dm 2 b From eq. (7) | From eq. (8) 64 7°965 2a 0°989 5°0 4°97 60 i AO 60 2°39 1°025 26°40 26°20 c¢ 5°58 : (74 ‘x4 3°63 3°60 ne 2s) DAG? 1°0575 3°64 3°60 DAL ao 51°60 2 1:099 36°0 35°61 53°6 2G 20) DeSione hs 188°7 186°7 It will be noted that the values computed from equation (7) run slightly higher than those from equation (8). If we equate the two expressions for mache units writing w in place of 2°7 xX 10° we get 0°65 O. C. Lester—Emanation Electroscopes. 235 If we now substitute corresponding values of 4 and J from the above table and compute the several values of « we find that they agree closely and give as a mean value = PIG SC" 10" (10) which is the relation between the mache unit and the curie necessary for exact agreement between the results by the two methods if we assume that the saturation ionization current is accurately determined by the constarts found and the correc- tions applied. This value is about half way between the theo- retical value 2°75 & 10° sometimes used and the value 2°7 « 10° usually taken. To be exact, 2°75 < 10° gives the theoretical relation between the mache unit and the curie on the Rutherford-Boltwood standard. According to Rutherford* it is 2°89 x 10° on the International Standard and Mache and St. Meyert give it as 2°67 & 10° on the Vienna Standard. Of course this relation is nothing but the saturation ionization current due to one curie of emanation without disintegration products, multiplied by a thousand. Among investigations in the radioactivity of mineral springs, and in particular among those on European mineral springs, there can be found often the results of several observers on the same water or gas. It is rarely that these results agree closely and those of one observer may range anywhere from many times to a fraction of those given by another. With precautions field work can be made practically as accurate as that done in the laboratory. Hence discrepancies in the work of equally careful observers have often been attributed to varia- tions in the activity of the source. On the other hand, there are springs which have shown no appreciable variation in activity when examined systematically at different times of the year by the same observer using the sameapparatus. Undoubt- edly-some springs do vary in activity but the question of their variability and even the amount of their activity can scarcely be determined from the work of different observers so long as there is no uniformity in standards, in methods, and in “the nature and the number of the corrections to be applied to the observations. This is particularly true of results expressed in mache units based upon ionization currents. In many cases mache units are apparently calculated from the observed ioni- zation current and not from the saturation ionization current when all radiation is absorbed in the air of the chamber. In the first case the mache unit is dependent upon the dimensions of the particular apparatus used and upon the potential applied * Phil. Mag. (6), vol. xxviii, p. 820, 1914. + Phys. Zeitschr, loc. cit. 236 O. C. Lester—Emanation Hlectroscopes. to the insulated system which is clearly not intended by its definition. For the reasons just mentioned the .work of European observers in general presents an almost hopeless confusion when accurate comparisons are attempted. It is trne that much work had been done before suitable units and methods were devised, and we find therefore many results expressed in terms of the fall of the leaf in volts per unit time or in units even more arbitrary. Such results can not be compared with other work. Still other units used are the Milligram-Second, Milligram-Minute, Gram-Second, etc., meaning the amount of emanation produced by a given amount of radio-active sub- stance in the specified time. The substance is usually the ele- ment radium or a radium salt, and when this is specified, as well as its degree of purity, measurements based upon such units can be reduced to curies. Most European observers, outside of France and England, express their results in terms of the mache unit. Generally the corrections which have been made are clearly stated, but not always. Furthermore the correction for absorption by the walls of the chamber (Duane’s factor) has usua!ly been omitted in work where most of the other corrections have been applied. This has been pointed out by Berndt* in an elaborate series of calculations undertaken with the aim of making possible the comparison of the results of different observers. He shows that, depending upon the size of the ionization chamber, the correction for absorption alone may amount to from 10 per cent to 155 per cent. A given instrument can be calibrated simply and accurately in terms of a known quantity of radium emanation. If the mache unit is to be retained it would seem easier and more accurate to reduce results measured in curies to this unit by ineans of the theoretical relation between them, than to calcu- late mache units from the ionization current which involves the determination of several more constants and the applica- tion of troublesome corrections. As has been shown above, the two methods, when all corrections are applied, give identical results within the limits of experimental error. For the drawings which accompany this article and for effi- cient aid in securing the data upon which it is based the author wishes to express his indebtedness to Mr. J. H. V. Finney, instructor in Physics in the University. Hale Physical Laboratory. University of Colorado, Boulder, Colo. * Loe. cit. Querke and Finkelstein— Radioactivity of Meteorites. 237 Art. X1X.—Measurements of the Radioactivity of Meico- rites ; by Terence T. Qurrxe and Leo FinxersreErn. AxrnoucH the determinations of the radioactivity of various rocks are now numerous, the radioactivity of meteorites is known for only two varieties of meteorites and for only three meteorites. The meteorites examined were tested by Strutt* His results are :— Locality Material Variety Quantity Ra per gm. Taken material Dhurmsala Stony Met. Intermed. 50gm_ 1:12X10-" gm Chondrite Thunda Iron Met. Medium 60 0 : Octahedrite Staunton Bee as 30 0 Augusta Co. Va. Santa Catarina, Iron 50 0 Brazil Disco Island, Native Iron 200 0°424 x 107-" om Of these materials Santa Catarina is generally accepted to be terrestrial iron and not a meteorite. Our information in regard to the radioactivity of meteorites could be summarized as follows; one intermediate chondrite, a stony meteorite, has a radium content of 1:12x10-" gm. Ra. per gm. meteorite, and two medium octahedrites, iron meteorites, are free of radio- activity. Through the generosity of Dr. Oliver C. Farrington of the Field Museum of Natural History, the authors had the follow- ing varieties of meteorites placed at their disposal :—fifteen aérolites, or stony meteorites, one chladnite, two eukrites, two white chondrites, one carbonaceous chondrite, one black chon- drite, three spherulitic chondrites, three spherulitic crystalline chondrites, one crystalline chondrite, two siderolites or stony- iron meteorites, five iron meteorites, one finest octahedrite, two medium octahedrites, one coarse octahedrite, and one nor- mal hexahedrite. Chemical Preparation of the Materials The iron meteorites were dissolved in hot cone. HCl and HNO, in a platinum dish. All the stony and _ stony-iron meteorites were treated consistently. Stony meteorites differ from igneous rocks in that most of them contain a considerable amount of metallic iron and nickel. Each sample was ground * Strutt, Proc. Roy. Soc, A, 1xxvii, p. 480, Mar. 1916. 238 = =Quirke and Finkelstein—Measurements of the to sizes between 100 and 200 mesh, and boiled with cone. HCl and HNO,. The acid solution was filtered, and set aside, and later added to the residue. The residue was fused with 20 gm K,CO,, 20 gm Na,CO,, 2 gm Na,B,O,, and 0°5 gin Ba(OH),.* The fused mass was broken up, ‘and dissolved in 200° water. The solution was filtered, and the residue added to the acid solution in which the metal portion of the meteorite had been dissolved, and boiled untilsolution was complete. Sulphuric acid was added to the boiling solution, and the barium sulphate filtered off. The barium was separated from the aqueous soln- tion in the same way. The barium sulphate precipitates were combined, and fused with 10 gm KHSO, in a hard glass test tube. The tube was provided with a rubber stopper fitted with two glass tubes. One of the tubes reached nearly to the bottom of the test tube, and the other just penetrated the stopper. Each tube was bent at right angles to the fusion tube, and the exposed ends which were drawn out to fragile points were sealed. The granite and diorite samples were treated in the same way, except that the preliminary treatment with acids was left out. In the treatment of the granite it was found necessary to use hydrofluoric acid to get rid of the silica in the residue. Purification of Reagents One of the principal sources of error apt to be introduced is the use of reagents which might contain appreciable amounts of radium. This is especially true of the barium salts. Blank tests were made on the reagents used to determine their radium content. Only the barium salts were found to contain enough radium to make the radioactivity of the charge appreciable. It was therefore essential that the barium salts be completely freed from radium. This was done by fractional crystallization as hydroxides according to the method of McCoyt. The barium chloride was dissolved in a special flask, and a 50 per cent sol. KOH which had been previously freed from carbonates by addition of barium was added. The solution was cooled in ice and allowed to crystallize. The mother liquor was filtered off, and discarded. After six crystallizations the barium hydroxide was found to be completely freed from radium. *In some preliminary experiments, each of the stony meteorites received from Ward’s Natural Science Establishment, and the rock samples, labora- tory numbers 105, 108, 109, and 110, were fused with 0°5 gm of BaCOs in- stead of 0'5 gm Ba(OH).. It was found that this brought the radioactivity of the charge up to 0°34x10~” grams of radium element. However, with the use of radium-free Ba (OH), the radioactivity of the charge was found to be imperceptible. + U. S. Patent No. 1,103,600 (1914). Radioactivity of Meteorites. 239 Estimation of Radium The estimation of radium is usually carried out in one of two ways: 1. By the gamma ray method. 2. By the emanation method. The gamma ray method is entirely unsuited for measuring minute quantities of radium. The emanation method, how- ever, can be used to measure extremely small amounts of radi- um with a fair degree of accuracy. When radium decomposes, the first disintegration product is radium emanation. If the radium-containing substance is kept in a sealed tube, the ema- nation reaches its equilibrium amount in about thirty days. The amount of emanation N present at any time ¢ is: INS Nett) where N’ is the maximum amount of emanation, and 2 a con- stant. The emanation can be separated completely from the solution or fused material by bubbling air through the liquid. The air may be introduced into a suitable electroscope, and the ioniza- tion due to the included emanation measured.* Standardization of the Hlectroscope The electroscope has been in use for a number of years, and has a very slow natural leak. It consists of a brass cylinder 13°" high, and 9°" in diameter. The gold leaf is supported by an | shaped brass strip which is insulated by an amber pillar. The leaf is charged by a battery of small dry cells giving 400 to 500 volts. The movement of the gold leaf is observed with a low power microscope, with a scale in the eyepiece. The electroscope was standardized by means of standard radium solutions furnished by Dr. H. N. McCoy. Three dif- ferent solutions were used : No. 1—3°78 & 10-" gm Ra element No. 2—3°78 & 107" “ * ‘ N0g 3223:78) xX 10D att < The standards were made by diluting a standard radium solution, and adding a few ee. of dil. HCl. The electroscope was first evacuated, and dry air from out- doors admitted. The natural leak was then determined by timing the rate of discharge over five divisions of the scale. The electroscope was then evacuated, and the standard which had been sealed for over a period of thirty days was inserted in the air line. The ends of the sealed tubes which had been * Cf, Herman Schlundt, 26th meeting, Am. Electrochem. Soc.,1914. H.M. Plum, Jour. Am. Chem. Soc., vol. xxxvii, p. 1811, 1915. 240 Quirke and Finkelstein—Radioactivity of Meteorites. drawn out to fragile tips were broken, and air allowed to bub- ble through the solution, and pass into the electroscope. The activity rapidly increases and reaches a maximum after three hours. This is due to the accumulation of RaA, RaB,and RaC. The rate of discharge was measured when the maximum had been reached, that is after an interval of three hours. During this time the leaf was kept charged. The leak was measured in exactly the same way as in the case of the natural leak. Estimation of Radium in the Sample The natural leak was determined before each measurement. The tube containing the fused material which had been sealed for a period of thirty days was inserted in the air line. The tips at the ends of the tube were broken off, and the tube heated. When the material had liquefied, air was allowed to bubble through into the electroscope. After the electroscope was filled, the time of discharge after au interval of three hours was measured in the same manner as the standard. The amount of radium in the sample was evaluated from: Amt. of Ra = ere) x S. t, (A-t,) where: ¢, = time of discharge of standard. (Equilib. amt. Ra) t — ce (<4 ¢ ce unknown. ( (<4 74 ee ) A, = natural leak of standard. Bos ee ‘eS cunikuiown: S = amount of radium in standard. The results show that the radioactivity of stony meteorites varies considerably. The meteorite, Juvinas, contains 217X | 10-" gm of radium element to a gram of meteorite, and the meteorite, Farmington, contains only 7:34 10—" om of radium element to a gram of meteorite. Including the determination for the radioactivity of Dhurmsala made by Strutt, seventeen stony meteorites have an average radioactivity of 7°61 x 10—" am of radium to a gram of meteorite. Excluding Dhurmsala, six- teen meteorites have an average radioactivity of 7:39 X10-" om of radium to a gram of meteorite. Two iron-stone meteorites, Estherville and Llana del Inca, have an average radioactivity equivalent to 6°88 X10-"* gm of radium to a gram of meteoritic material. Five of the seven iron meteorites examined are non- radioactive, so far as can be determined, i.e., they do not ex- ceed 10-“ grams of radium per gram of meteorite. The other two iron meteorites, Toluca, and Coahuila, seem to be radio- active ; Toluca decidedly so, and Coahuila so feebly radioactive ab. , Noe wy 220 Bishopville, Sumter Co., 8. C. 205 Juvinas, Ardiche, France. 208 Stannern, Moravia, Austria. 204 Mauerkirchen, Upper Austria. 209 Mocs, 206 Mocs, t¢ 203 Pultusk, Poland, Russia. 214 Farmington, 217 S. Russia. 201 Hessle, Upsala, Sweden. 202 Forest City, 207 Tabory, Ochansk, Perm. Russia. 216 Beaver Creek, 219 Holbrook, Navajo Co., Arizona. 200 Saline, Sheridan Co., Kans. 215 Long Island, Phillips Co., Kans. 218 Estherville, Emmet Co., Iowa. 108 Llana del Inca, Atacama, Chile. 218 Mukerop, Gibeon, 210 Tonganoxie, 212 Toluca, Xiquepelco, Toluca, Mexico. 107 Canon Diablo, Coconino Co., Ariz. 211 Coahuila, State of Coahuila, Mexico. 110 Whiskey Lake, Ontario, Canada. 105 Espanola, ° Name and Locality Transylvania, Austria, Transylvania, Austria. Washington Co., Kans. Mighei, Elizabethgrad, Winnebago Co., Iowa. W. Kootenai Dist. B. C. Gt. Namaqualand, S. A. Leavenworth, Co., Kans. Ontario, Canada. Date of Discovery fell Mar, 15, 1848. fell June how L821: fell May 22, 1808. fell Nov. 20, 1768. fell Feb. 3, 1882. fell Feb. 3, 1882. fell Jan. 30, 1888. fell June 25, 1890. fell June 9, 1889. fell Jan. 1, 1869. fell May 2, 1890. fell Aug. 30, 1887. fell May 26, 1893. fell July 19, 1912. fell Nov. 15, 1898. found 1894. fell May HO. 1879: found 1888. found 1899. found 1886. found 1890. found 1891. found 18387. Class Stone Stone Stone Stone Stone Stone Stone Stone Stone Stone Stone Stone Stone Stone Stone Stone lron- Stone Iron- Stone Tron Tron Iron Tron Tron Variety Chladnite Eukrite Kukrite White Chondrite White Chondrite White Chondrite Gray Chondrite Black Chondrite Carbona- RADIUM CONTENT OF SPECIMENS. Donor K * ry re ee ey ceous Chondrite Spherulitic Chondrite Spherulitic Chondrite Sphernlitic Chondrite Crystalline Spherulitic Crystalline Spherulitic Crystalline Spherulitic Crystalline Chondrite Mesosiderite Mesosiderite Finest Octahedrite Octahedrite Medium Octahedrite Coarse Octahedrite Normal Hexahedrite Granite of Pre- cambrian age. Diorite of Keweenawan age. * F stands for Field Museum of Natural History. W stands for Ward’s Natural Science Establishment. + The two determinations of the radioactivity of Mics do notagree. ever, the two samples were obtained from different specimens, and were not ground or mixed together. SS ast ee eee ar a Ss ry W Wet. of sample 8°66 {hs 12°32 8-77 10°62 11°56 9°98 ol y oO Gms. of Ra el, per gm sample 5°01 x 10-13 2°17 x 10-9 2°41 x 10713 1°45 x a°12 x 0°67 x 7°34 x 1°06 x 8°36 x 1°8 x 5°12) x 1°33 x 3°20 x 3°97 x 9°80 x ole x Ae 7-69 x 10-” 3°20 x 10-2 fg yt Ce How- The difference in radioactivity may be due to a lack of uniformity in the distribution of radium throughout the meteorite. Am. Jour Sci.—Fourtu SERIES, Vou. XLIV, No. 261.—SEpremBer, 1917. 242 Quirke and Finkelstein—Radvoactivity of Meteorites. that its activity may be due to local siliceous inclusions. How- ever, Strutt found much greater activity in the native iron of Greenland, and it may be that radioactivity is not foreign to many of the so-called iron meteorites. It is clear, nevertheless, that most of the iron meteorites are free of radioactivity, and the siliceous meteorites are all radioactive, so far as known. Summary 1. A method for the estimation of radium in rocks and meteorites has been described. 2. The radium content of twenty-two meteorites not here- tofore analyzed has been determined. | 3. From the data it appears that the average stony meteor- ite is considerably less radioactive than the average igneous rock, probably less than one-fourth as radioactive as an aver- age granite, and that the metallic meteorites are almost free of radioactivity. The authors are indebted to Professor Herbert N. McCoy, under whose direction this work was carried out, and to Dr. Farrington and the Field Museum authorities for the use of authentic meteoritic material. 3 Kent Chemical Laboratory, University of Chicago. Miller and Knight— Eumenite. 243 ART. X X.— Occurrence of Humenite in South Sherbrooke Township, Ontario; by Witier G. Mriier and Oyrri W. KNIGHT. In June, 1917, the writers paid a visit to a feldspar quarry, in which euxenite occurs, three miles from the village of Maberley, Ontario. The quarry is near the center of lot 13, concession V, of the township of South Sherbrooke, Lanark county. The mineral was noted some months ago, during the working of the quarry, by Mr. J. A. Morrow, on whose land it occurs. Mr. Morrow sent samples to the Ontario Bureau of Mines for identification. Through the kindness of Dr. Dunstan, Director of the Imperial Institute, London, England, the min- eral was analyzed there and identified as euxenite. The euxenite occurs in a granite pegmatite dike which has a width of about 75 feet, and cuts banded gneiss of pre-Cam- brian age. The dike contains mainly feldspar, which has been worked as a source of potash and for pottery, and quartz; in addition to these minerals there are found considerable black tourmaline, three varieties of mica, black, white and green, and pyrite. The euxenite occurs in masses from the size of a pea to about two inches in diameter, and is embedded in pink feld- spar and in black scaly mica. Pyrite is closely associated with the mineral. Owing to their brittleness, crystals of the min- eral were not obtained, but occur occasionally embedded in the pink feldspar. The mineral occurs here and there in the dike, but was found in largest quantity near the center of the dike in a zone about two feet wide. The zone consists of pink feldspar, more or less stained by decomposition of pyrite, in which are nearly parallel seams of black scaly mica from an elghth to a half an inch thick. The results of the analysis of the South Sherbrooke min- eral are given in column I. The analysis in column II is quoted from Dana.* South Sherbrooke, Ontario Alve, Norway I II Per cent Per cent Ta,O, ronal EAE 5 SARTO 13°89 ey Nb,O, nL MEN he eine 12°73 35°09 (Nb,O,) TiO, SHOE NSLS OE ag 27°70 21°16 AiO, Dynan oahu eae 1°34 Ce,O, : , ican t cae. 0-62 3°17 (Ce,0,) * Descriptive Mineralogy, 6th ed., p. 744. 244 Miller and Knight—Euxenite. South Sherbrooke, Ontario Alve, Norway I Per cent YO ete 2s a ee at PeOnt ante egene 2°63 HeO!t aa vee eelaee ae 0°51 Mn Oy ed sees trace PbO wishes oe ee ae, 020 UE © Pave eae ey Sr gee 10°50 CaQt. cynic Spe, tae 0°09 MOO Oe ee een 0°12 DIOR) Sie ie se ee 0°74 EL Oletess. pens 3°00 Meg Specific gravity .---- 4:99 Provincial Geologist’s Office, Toronto, June, 1917. II Per cent 27-48 (Y,0,) 3-40 (Er,O,) 1°38 4-78 (UO,) 2-63 (HO) 99°09 5°00 Hord—Remarkable Crystal of Apatite. 245 Art. XXI.—A Remarkable Crystal of Apatite from Mt. Apatite, Auburn, Maine ; by W. E. Forp. Recentiy the writer, through the courtesy of Mr. A. H. Petereit of New York City, has been privileged to examine an Hie. i extraordinary crystal of apatite from the well-known locality near Auburn, Me. It was such an unusual crystal in respect to its size, color and crystal development that it seemed worthy of the following brief descriptive note. 5 246 _ford—Remarkable Crystal of Apatite. The crystal measures 3°8°™ by 4:3 in the horizontal diree- tions and 8° in the vertical direction and weighs slightly over 100 g. Its color is the wonderful deep amethyst characteristic of apatite from this locality. The crystal contains cracks and flaws and is cloudy in portions, but also in many small areas it is perfectly clear and of a gem quality. The erystal forms observed include the following : c(0001), m(1010), a (1120), h, (8120), 7(1016) %, x (1011), y (2021), 2 (3031), s (1121), u (2131) and m,(3121). The accompanying figure represents ‘he crystal as nearly as possible in its true proportions and about 1$ times its natural size. On one side of the crystal there are several oscillations between the faces with conse- quent parallel growth between different portions of the crystal. Three of the edges between the base and the pyramid & are replaced by narrow faces of a very low pyramid. ‘These faces are curved and do not yield a definite reflection on the goni- ometer. The only known form that they might be, however, is the pyramid 7 (1016). The faces of the pyramid z, as is ustal on crystals from this locality, show marked horizontal stria- tions and the prisms @ and 4, commonly show faint vertical striations. All of the other faces are plane with no distinctive markings. Both the right and left forms of the third order pyramid, w and w,, are present, with no apparent distinction to be made between them in regard to their luster, ete. While the figure shows the majority of the faces occurring upon the crystal, several very small and narrow faces of the second and third order pyramids had to be omitted from the drawing. A small erystal showing the same forms is attached at one side to the lower part of the large crystal but is not represented in the figure. A small amount of cookeite is attached to the crystal. Mineralogical Laboratory of the Sheffield Scientific School of Yale University, New Haven, Conn., June 15, 1917. William Bullock Clark. bo — -~J WILLIAM BULLOCK CLARK. Dr. Wittiam Buttock Crarx, professor of geology in the Johns Hopkins University and State Geologist of Maryland, died suddenly of heart failure on July 27 at his summer home at North Haven, Maine. He was born at Brattleboro, Vermont, on December 15, 1860. Entering Amherst College in 1880 he received his A.B. degree in 1884 and immediately went abroad, spending the next three years in von Zittel’s laboratory at Munich where he received his doctorate in 1887. After spending several months in London and Berlin he returned to America as instructor in the newly founded Department of geology at the Johns Hopkins University. With the death of George H. Williams in 1894 Clark became a full professor and head of the department. He was connected with the U.S. Geological Survey in various capaci- ties from 1888 until his death. Although coming from an old New England stock—the Bul- locks haying settled in Salem in 1643 and the Clarks at Plymouth in 1623—Professor Clark was for thirty years a citizen of Balti- more, and it is doubtful if there has been anyone who has per- formed a greater service than he to the commonwealth of Maryland. It needs but an enumeration of his many positions of responsibility to appreciate this unique service. He organized a State Weather Service in-1892 and was its director for 25 years. He organized the State Geological Survey in 1896 and was its director for 21 years. He organized the State Bureau of Forestry in 1906 and was its executive officer for 11 years. In 1898 as State Geologist he was instrumental in starting the good roads movement in Maryland and successfully steered through the shoals of possible political waste in the expenditure of about $2,000,000 in the making of state highways. In 1910 a State Roads Commission was organized to take over the rapidly expand- ing work of the Highway Division of the Geological Survey, and for four years more he was a very active member of this commis- sion. He represented the state in the resurvey of the Mason and Dixon line, was a member of the State Conservation Commission, was instrumental in forming the state exhibits at the Buffalo, Charleston, St. Louis, Jamestown and San Francisco expositions, and in arranging the state mineral exhibit in the State House at Annapolis. He took an official part in the White House confer- ence on conservation in 1908. In civic affairs he served as a member of the emergency com- mittee appointed by the mayor at the time of the great Baltimore fire in 1904, and aided in the rehabilitation and improvement of the burnt district. In 1905 he was appointed by the mayor a member of the committee to devise a sewerage system for the city. In 1909 a like appointment resulted in the plans for the develop- ment of a civic center for Baltimore. For 16 years he was presi- dent of the Henry Watson Children’s Aid Society, He was also 248 ; William Bullock Clark. a member of the executive committee of the State Tuberculosis Association and an officer of the Federated Charities. Professor Clark was a member of the National Academy of Sciences and chairman of its Geological Section, a fellow of the American Academy of Arts and Sciences, a member of the Phila- delphia Academy of Natural Sciences, Washington Academy of Science, American Philosophical Society, Deutsche Geologische Gesellschaft, Paleontologische Gesellschaft, and American Asso- ciation for the Advancement of Science. He was a councillor and treasurer of the Geological Society of America, of which he was a charter member, and a foreign correspondent of the Geological Society of London. He was, for several years, president of the Association of American State Geologists. Amherst conferred its LL.D. on him in 1908. In 1897 Professor Clark was an official delegate to the International Geological Congress and spent sey- eral months in an extended trip through Russia. He was abroad several summers and attended the centenary of the Geological Society of London. Hespent the summer of 1906 in Alaska, and traveled extensively in Mexico and throughout the United States. He was married October 12, 1892, to Ellen Clark Strong of Bos- ton, and had four children, all of whom survive him. Professor Clark was eminently social and had the gift of inspir- ing affection in men of all walks of life. His influence on the progress of geology was unique. Starting as a paleontologist he soon became an authority on the Echinoidea. He was early diverted to more strictly stratigraphical work and prepared a correlation paper on the Eocene for the U. 8. Geological Survey on the occasion of the Washington meeting of the International Geological Congress in 1891. After studying the Upper Creta- ceous of New Jersey for the U. 8S. Geological Survey he attacked the Coastal plain formations of Maryland and Virginia with char- acteristic energy, and the results of this work were eventually embodied in the systematic reports on the Lower Cretaceous, the Upper Cretaceous, the Eocene, the Miocene and the Pleistocene, published by the Maryland Geological Survey. With the multi- plication of administrative duties as head of the Geological Department and member of the Academie Council of the Uni- versity, as well as the increasing widening of the work of the Maryland Geological Survey, the Weather Bureau, the Highway Commission and the Forestry Bureau, most of his time was engrossed in organization rather than in research, and undoubtedly his greatest monuments are the reports of the Survey he organized and the contributions to Science by a host of younger men who came under his influence—draw ing material aid as well as inspira- tion from his example and ideals. With the outbreak of the war Professor Clark became actively interested in problems of defence and economic preparedness. He was appointed a member of the National Research Council, was chairman of the subcommittee on road materials, and a mem- ber of the committee on camp sites and water supplies. He was also chairman of the committee on highways and natural resources of the Maryland Council of Defense. E. W. BERRY. Warps Narurat Science Estas.isument A Supply-House for Scientific Material. Founded 1862. Incorporated 1890. A few of our recent circulars in the various departments: Geology: J-3. Genetic Collection of Rocks and Rock- forming Minerals. J-148. Price List of Rocks. Mineralogy: J-109. Blowpipe Collections. J-74. Meteor- ites, J-150. Collections. J-160. Fine specimens. Paleontology: J-134. Complete Trilobites. J-115. Collec- tions. J-140. Restorations of Extinct Arthropods. Entomology: J-30. Supplies. J-125. Life Histories. J-128. Live Pupae. Zoology: J-116. Material for Dissection. J-26. Compara- tive Osteology. J-94. Casts of Reptiles, etc. Microscope Slides: J-135. Bacteria Slides. Taxidermy: J-188. Bird Skins. J-139. Mammal Skins. Human Anatomy: J-16. Skeletons and Models. General: -J-155. List of Catalogues and Circulars. - Ward’s Natural Science Establishment 84-102 College Ave., Rochester, N. Y., U.S. A. . ty Publishers: WILLIAMS & NORGATE, 14 Henrietta Street, Covent Garden, London, W. C. “SCIENTIA” INTERNATIONAL REVIEW OF SCIENTIFIC SYNTHESIS: Issued monthly (each number consisting of 100 to 120 pages). Editor: EUGENIO RIGNANO. ‘“*SCIENTIA”’ continues to realise its program of synthesis. It publishes articles which relate to the various branches of theoretic research, and are all of general] in- terest; it thus enables its readers to keep themselves informed of the general course of the contemporary scientific movement. **SCIENTIA”’ appeals to the codperation of the most eminent scientific men of al? countries. It has published articles by Messrs. Abbot ( Washington) - Arrhenius. (Stockholm) -Ashley (Birmingham) -Bechterew (Petrograd) - Bohlin (Stockholm)-— Bonnesen (Kopenhagen) - Borel (Paris) - Bottazzi (Vapoli) - Bragg (Leeds) -Bril- louin (Paris) -Brani (Padova) - Castelnuovo (Roma) -Caullery (Paris) -Chamberlin (Chicago)-Ciamician (Bologna) - Clark (New York) - Costantin (Paris) -Crommelin (Greenwich) - Daly (Cambridge, U. S. A.) - Darwin (Cambridge) - Delage (Paris) - *De Martonne (Paris) - De Vries (Amsterdam) - Durkheim (Paris) - Eddington (Greenwich) - Edgeworth (Ozford) - Emery (Bologna) - Enriques (Bologna) —- Fabry (Warseille)—Fisher (Vew-Haven, U. 8. A.)-Foa (Torino)-Fowler (London) - Fredericq (Ziége)- Galeotti (Napoli)-Golgi (Pavia) - Gregory (Glasgow)- Guignebert (Paris) - Janet (Paris) — Jespersen (Gentojfte) - Kapteyn (Groningen) - Kidd (Ozxford)-Langevin (Paris)-Lebedew (Moscow)-Lodge (Birmingham)-Loisy (Parvis)- Lorentz (Haarlem) -Loria (Torino)-Lowell (Flagstajf, U. S. A.)-Maunder (Green- wich) — Meillet (Paris) - Pareto (Lausanne) - Peano (Torino) - Picard (Paris) - Poincare (Paris) - Puiseux (Paris) -Rabaud (Paris) - Righi (Bologna) - Rignano (Milano)-—Russell (Cambridge)- Rutherford (Manchester)-Sayce (Oxford) -Schiapa= relli (Milano) - Seligman (New York) - Sherrington (Liverpool) - Soddy (Glasgow)- Svedberg (Upsala) - Tannery (Pavis)- Turner (Oz/ord)- Vinogradoff (Moscou)- Vol= terra (Roma)-Von Zeipel (Upsala) -Westermarck (Helsingfors)—-Willey (Afontreal, Canada)-Zeeman (Amsterdam) - Zeuthen (Kopenhagen), and more than a hundred others. ‘“*SCIENTIA’’ publishes, at present, in the section dedicated to sociological articles, a series of studies on the present questions of an international character raised by the war. : _ **SCIENTIA’’ publishes its articles in the language of its authors, and joins to the principal text a supplement containing the French translations of all the articles that are not in French. (Write fora specimen number.) Annual Subscription: 24 sh. post free. Office: Via Aurelio Saffi, 11- MILAN (italy). CON TEIN ess Art. XVI.—Volcanologic Investigations at Kilauea, with - Plate I (frontispiece) ; by T.-A. Jacear, Jr. ...-.-.-- 161 XVII.—On the Qualitative Separation and Detection of Gallium ; by P. E. Brownine and L: E. Porrsr._--.- 221 X VIII.—On the Calibration and the Constants of Emanation Electroscopes; by O.-C.. Lesrur..-_22 222 2 ee 225 XIX:—Measurements of the Radioactivity of Meteorites ; by ‘TT Quirke and LEQ FInKnLSTRIN =) <2 2 eee 237 XX.—Occurrence of Euxenite in South Sherbrooke Town- ship, Ontario ; by W. G. Mitier and C. W. Knicut_.. 243 XXI.—A Remarkable Crystal of Apatite from Mt. Apatite, Auburn, Maine ; by W. E. Forp..-- - Sy Spe ie te nea 245 Witiiam Buurock CLARK 220 32 eae ee ee OCTOBER, 1917. Established by BENJAMIN SILLIMAN in 1818. THE AMHBRICAN JOURNAL OF SCIENGE. Epitorn: EDWARD S. DANA. ASSOCIATE EDITORS Proressors GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW ann WM. M. DAVIS, or Camsrivaz, Proressors ADDISON E. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT E. GREGORY AND. HORACE 8S. UHLER, or New Haven, Proressor HENRY S. WILLIAMS, or Irnaca, Proressor JOSEPH 8S. AMES, or Battimore, Mr. J. S. DILLER, or Wasnuineron. FOURTH SERIES VOL. XLIV—[ WHOLE NUMBER, CXCIYV]. No. 262—OCTOBER, 191% NEW HAVEN, CONNEC GS Ie at THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. - Published monthly. Six dollars per year, in advance. $6.40 to countries in the Postal Union ; $6.25 to Canada. Single numbers 50 cents. . i Entered as second-class matter at the Post Office at New Haven, Conn., under the Act 4 of March 3, 1879. LIST OF CHOICE SPECIMENS . Remarkable calcite stalactites from the famous Copper Queen Mine, at Bisbee, Arizona. Coral-like formation and rich in colors; some spec- imens have three and four colors in different shades of er een and blue. A description can give only a faint idea of their beauty. Only a very small amount was found. They will make a wonderful centerpiece in any cabinet. Specimen, 9” lone and 5” wide with small calcite xls. attached; white, blue and green.- $22.50. Another 74 x 44’, 4" high; beautiful shades of light and dark green, $15. Another 5 x 44"; 33" high: shades from white to dark green. $8. Another 5 x 33"; 24’ high: shades from white to dark green. $4. Opal, Barcoo River, Queensland, Australia. Beautiful specimen 6" long, 3” wide and 1” thick. The whole face is covered with fire opal full of color. This is the most beautiful specimen that has been received from Australia in a number of years. A number of large and valuable gems could be cut out of it. It is really worth $150. $100. Topaz, Thomas Range, Utah, very choice specimens. Specimen, 2 x 23"; 12” high; four large crystals of golden color embed- ded. §$8. Specimen, 3 x 2"; 12" high; three crystals of yellow color embedded. $8. Corundu:n var. ruby, Mysore, India. Doubly terminated, opaque, pyramidal crystal in matrix, +8" long. Specimen measures 14” x 13” x12" high. $18. Emerald, Muzo Mine, Bogota, Colombia, = A. Matrix 13x 4’, }’ high; crystal is 4" in diameter and projects ;3,"; fine termination and of good color and ¢ gem quality. $18. Emerald, Muzo Mine, Bogota, Colombia, S. A. Matrix 2x 12", 13” high; crystal is 4" in diameter and projects 2’; transparent, fine termination and good color, gem quality. , $50. Emerald, Muzo Mine, Bogota, Colombia, 8. A. 1% x 14"; 1" high; four crystals of fine color and termination embedded, associated wlth pyrite. Crystals are from + to}" in length. $130. Emerald, Habachthal, Tyrol. Specimen eek ey 3 high, four crystals of good color and double ter- mination embedded. 5 to #" in diameter. $18. Specimen 3 x 22’, }' high: two crystals embedded. $8. Emerald, Takowaja River, Ural Mts., . Crystal 12” long, $" in diameter; good termination. $7.50. Crystal 13" long g, a” in diameter: good termination. $10. Apatite, Auburn, Maine. A fone fine specimens with beautiful amethys- tine crystals embedded. $6-$10-$12-$20. Tellurium, crystallized, Colorado City, Colorado. 13x 1’, ;%” high, weighs ¢ ounces. It is a wonderful artificial product. The crystals are beautiful. Only a very few specimens were made by a professor. $10. : Nagyagite and gold, Nagyee; Transylvania, 24 x 14"; 1" high, very rare. $15. ALBERT H. PETEREIT 81-83 Fulton St., New York City THE AMERICAN JOURNAL OF SCIENCE pL COFU ES Ee Sok Rt E'S) | —_¢++—___ Art. XXII.— Block Mountains in New Zealand ; by ©. A. Corron, Victoria University College, Wellington, New Zealand.* CONTENTS. Introduction. Part I. Block Mountains and Related Forms. 1. Structure. 2. The Initial Surface. 3. Possible Types of Drainage. True Consequents. Anteconsequent Drainage. Antecedent Drainage. 4. Sequential Forms. Stripping. Salients on Stripped Plateaus. Mature Dissection of the Undermass, Fault Scarps and Fold Scarps. Composite Fault Scarps. Fold Scarps. 0. Trough Filling. Part II. The Block Mountains of Central Otago. 1. Historical Sketch. 2. Structure. 3. Major Tectonic Features. The Central Otago System. The Northern Highland of Otago. 4, Drainage. The Central Otago Chain of Depressions. The Clutha River System. The Waitaki River System. 5. The Surfaces of Uplifted Blocks. Stripped Plateau Surfaces. Details of the Surface on Schist Blocks. Salients on Block Surfaces. Scarps of the Schist Blocks. The Scarps of Greywacke Blocks. 6. The Floors of the Central Otago Depressions. List of Papers to which reference is made. * References are listed in full at the end of the article. Am. Jour. Sct.—FourtsH Series, Vou. XLIV, No. 262.—Ocrtossr, 1917. 18 250 C0. A. Cotton—Block Mountains in New Zealand. INTRODUCTION. Tue geologic structure of New Zealand has been described by the writer in a recent article (1916 a, pp. 319, 320) as “a concourse of earth blocks of varying size and shape, in places compressed ; the highest blocks lying in the northeast and southwest axis of the land masses, so that the whole structure may be termed a geanticline ; the blocks initially consisting of an older mass of generally complex structure much denuded and largely planed, and concealed over the greater part of the area by covering strata which had not been disturbed before the ‘blocking’ took place; the whole since these movements considerably modified by erosion somewhat complicated by the effects of later movements of uplift and subsidence.” In another article (1316 0), the block mountains and asso- ciated features in a small area in northern Nelson have been described. This paper deals with Central Otago, where the “ block” features are unusually well preserved. As an aid in presentation, the description and interpreta- tion of selected types of New Zealand mountains is preceded by a discussion of the physiographic development of block mountains in general based on the established principles of geomorphology. 3 Part lI. Brock MounrAINS AND RELATED Forms. 1. Structure. The structure is postulated as an undermass of rocks with varied structures which before being covered were denuded enormously and reduced to small relief. Though the final planing was accomplished in some parts of New Zealand by the sea, subaérial agencies probably prepared the large planed areas for the final marine planation. Remnants prove the presence of a former widespread over- mass or cover upon the eroded surface of the undermass. It is not necessary to assume that this overmass was laid down as a continuous sheet over the region or that it is entirely of marine origin. It may be postulated, mdeed, that the cover- ing beds are in part fluviatile, resting upon a peneplain,* and that other parts of such a surface may never have been covered. Perhaps the most important factors to be considered in the processes of uplift and denudation are: (1) the nature of the *It is not to be inferred from the above statements concerning covering strata in New Zealand that the writer believes submergence and the initia- tion of sedimentation to have been everywhere simultaneous. Itis probable that overlap in the covering strata resulted from submergence of successive ‘‘blocks” of the undermass. Further, the statement of Speight (1915), based on the results of prolonged study, that in North Canterbury islands rose through the Upper Cretaceous and Tertiary sea, must not be overlooked. C. A. Cotton— Block Mountains in New Zealand. 251 Fie. 1. sere - ; au iri Mtns “Bowns, is : ST eee Mtns 2D SOUTH ISLAND of NEES ON NEW ZEALAND. . ‘ oo ag Se BS: i @ — Dunedin Fic.1. Locality map of the South Island of New Zealand. Seale 1 inch = 100 miles. 252 ©. A. Cotton— Block Mountains in New Zealand. uplift, (2) the relative strength of, or resistance offered to, erosion by the undermass and the cover, and (8) the nature of the structures in the undermass where the overmass is rela- tively weak. The covering strata in New Zealand are relatively weak, consisting in great part of mudstones and incompletely indu- rated sandstones, also of thin limestones, relatively resistant but soluble, mterbedded with weak clastics. A few thick masses of indurated conglomerate which occur in places at the base of the cover offer great resistance to erosion. In some districts thin lava flows occur; but these like the limestones are weakened by interbedding with clastics. In contrast with the overmass, the rocks of the undermass are generally highly resistant. In northern Nelson these are indurated argillite, quartzite, quartz schist, crystalline lime- stone, and intrusive granite. In western Otagd, eneissic and plutonic rocks occur, and in Central Otago, the undermass consists entirely of schist—relatively a very resistant rock com- pared with the unconsolidated sands and clays largely devel- oped in the overmass of that district. 2. The Initial Surface. The form of the initial surface depends upon the nature of the uplift. Two types of uplift may be distinguished: (@) The blocks are differentially elevated, depressed, or tilted—the dis- placement being solely by faulting or by faulting replaced toa minor extent by monoclinal flexures. The initial surface must be a mosaic of plane areas at various attitudes, some perhaps horizontal and many inclined, separated by initial fault scarps facing in different directions. (6) Strong warping—perhaps better termed folding—attains considerable development ; faults though present pass into or replace the limbs of folds, the uplifted blocks being in part anticlinal and the fault angles and trough depressions being in part synclinal. The surfaces of the structural units, which here as well as in (@) may be termed “blocks,” will be in part warped or flexed, though there may still be notable plane areas, and the fault-scarp boundaries between adjacent blocks will be replaced in part by monoclinal slopes. This is the type of deformation concerned in producing many of the New Zealand block mountains, and, according to Gilbert (1874), a somewhat similar type is not unknown in the North American Great Basin. The types of initial and sequential forms for regions of uplifted and tilted blocks without cover have been deduced by Dayis (1908, 1905, 1912) and matched with examples from the Basin Ranges and ‘elsewhere. Louderback (1904)‘has described faulted blocks with a cover of resistant lava. In the present C. A. Cotton— Block Mountains in New Zealand. 23 paper, attention will be directed principally to forms devel- oped in a region in which the predeformational surface is a plain of deposition, being the surface of a weak overmass lying on a planed undermass. 3. Possible Types of Drainage. The early stages of the drainage of a region in which the initial forms have been produced by the faulting and folding of a weak overmass on a planed undermass must be entirely consequent if the overmass is continuous and if its surface is a plain of marine deposition. Parts, however, of the Otago dis- trict were probably emergent prior to the deformation, and some of the streams may have persisted during and after the deformation as antecedents. Such a preéxisting drainage sys- tem would be present also in any part of the area where the upper layers of the covering strata were of subaérial origin, or im any part where regional elevation antedating the deforma- tion had brought the upper layers of marine strata above sea level. Few streams are powerful enough to persist through strong deformation, and though a few large streams may be antecedent, the greater part of the drainage, including most if not all of the minor and tributary streams, will be consequent. As movements of deformation occupy some time, erosion can accomplish much while uplift is in progress. A drainage sys- tem will be established as soon as any portion of the region is exposed as land, and this early consequent drainage tends to be perpetuated during the continuance of the movements. As movements probably will not go on continuously in all parts of the region, if the consequent drainage resulting from the deformation be alone considered, the early drainage pattern is not necessarily the same as that which would have come into existence had the final structure been instantaneously assumed. In a deformed region may be found, therefore, true antecedent streams, true consequent streams, and streams that were conse- quent upon the form of the surface assumed as the result of early movements but are antecedent to later movements of the same series. Such streams might perhaps be appropriately termed antecedent consequents or antecconsequents. Spill-over courses resulting from the overtopping of divides by alluvial accumulations are also -possible. After cutting down through the covering strata, streams of any or all of these types will be superposed in places upon the structures of the undermass. Besides, all may be expected to develop insequent tributaries, and as the cycle progresses sub- sequent streams will form upon weaker structures in the under- mass and such portions of the covering strata as have escaped complete destruction. 254 ©. A. Cotton— Block Mountans in New Zealand. It is necessary to consider criteria for distinguishing the types of drainage. Subsequent streams guided by weaker strata in an inclined series are readily recognized, but when they are guided by shear-zones, ancient fault-planes, or master joints they are generally included with the insequents. Leay- ing these aside, we may consider the important types of drain- age in the early stages of the cycle of erosion introduced by the deformation, namely, true consequents, true antecedents, and anteconsequents, all possibly superposed. - True Consequents.—On the highly improbable assumption that deformation is simultaneous and instantaneous and without contemporary erosion, it follows that consequent streams will make their way down the tilted and warped surfaces and that some consequent lakes will be formed. Under normal condi- tions of humidity, these lakes will spill over at the lowest gaps along consequent courses, which will be superposed later on the structure. of the older mass, gorges will be cut through the higher blocks, and systems of consequent streams will be estab- lished. Exactly similar drainage patterns are to be expected if the deforming movements are simultaneous though not imstan- taneous. If the movements are sufficiently slow no lakes may result, as early-formed consequent courses across the lowest sags in the crest lines of the rising blocks will be continuously deep- ened by corrasion. Consequent lakes considerably above base- level are short-lived, so they are not likely to leave permanent records of their existence. Anteconsequent Drainage. There is as little justification for the postulate that movements have been simultaneous throughout a period of deformation as there is for that of in- stantaneous deformation. Movement may be well-advanced in some parts of a region while other parts are as yet unaftected. There may even be a rhythmic passage of waves across the land surface. (In regional movements such oscillation is well attested; but the present discussion is concerned with strongly differential as distinguished from regional movements.) In the case of a low-lying block surrounded by differentially rising blocks, the movement of which is not necessarily simul- taneous, the lowest gap in the basin rim (erosion being left out of account) may not always be in the same position; and the consequent outlet of a basin established during an early stage of deformation and persisting by rapid corrasion may not at a later stage be situated at the lowest sag in the surrounding blocks. Antecedent Drainage. Antecedent drainage channels on a surface of the kind postulated must be inherited from a simple centrifugal system of subparallel streams radiating from the old land or that portion of the undermass which escaped burial C. A. Cotton—Block Mountains in New Zealand. 255 during the period preceding the deformation. The approx- imate position of such a nonburied area of the undermass or the source of the material forming the covering strata furnish evidence for the direction of possible antecedent streams. Without such knowledge only streams which cannot be placed with certainty in any other category may be classed as possibly antecedent. The relation of valley directions to the trend of elongated blocks may give information. If, as in Otago, the deformation has produced longitudinal tectonic features—par- allel elongated arches and troughs or long tilted blocks sepa- rated by fault angles—both true consequent and anteconsequent drainage will follow generally longitudinal courses, though per- haps breaking across here and there from one linear series of depressed areas to another. Antecedent drainage, on the other hand, may cross the longitudinal features diagonally or trans- versely. 4, Sequential Forms. Davis has recognized two main elements in the form of a simply tilted block mountain, the back slope and the front or scarp. These are the two main elements in the whole initial landscape of a region of tilted blocks. In a region that has been affected by diverse movements, the elements of the initial landscape are horizontal areas (high or low-lying), back slopes (areas of surface with a more or less uniform and gentle slope), fold surfaces (areas of steeper slope, not necessarily uniform), and fault scarps. There will be transitions from fold surfaces through what may be termed fold scarps to fault scarps. The present problem is to trace the development of topographic features from an initial surface comprising these various elements, upon which asystem of con- sequent drainage becomes established. After postulating an initial relief caused by instantaneous deformation, it is possible to consider the effects of erosion and a continuation of the deformation going on simultaneously. In the stage of extreme youth, the block fronts (fault and fold scarps) will experience the most rapid changes of form and, as a result of slumping and the formation of consequent gullies, these will supply the largest quantity of waste. Very early, also, the perhaps closely spaced consequent streams of considerable length, which will have come into existence all over the relatively large areas of inclined block surfaces, will be actively engaged in grading their courses (see fig. 2). From the rapidly deepened consequent valleys, insequent tributaries will be developed, and probably also subsequents. Since the spacing of the consequents alone may be close and the texture of dissection becomes finer when insequents and subsequents have also been developed, and since entrenchment of the whole system beneath the sloping surface of the weak 256 C. A. Cotton—Block Mountains in New Zealand. covering strata must rapidly take place, maturity of dissection will be rapidly attained, first in the middle parts of the steepest slopes and later over the whole area of sloping upland. On level upland surfaces, where streams may be widely spaced, on gentle slopes, where stream grade may be attained at no great depth below the initial surface, and also on low-lying blocks, where the surface is at no great height above base-level and where, consequently, deep dissection is impossible, maturity of dissection may be relatively long delayed. Fie. 2. Fic. 2. Diagram of the development of the back slope of a gently tilted block with a relatively resistant undermass and weak overmass from the initial form A through sequential forms B, C, and D to a stripped floor (earlier denudation plain) with shallow dissection, E. During this stage of the cycle, the troughs will generally be aggraded on account of an enormous quantity of waste from the upland surfaces. During the whole period of deformation, the troughs will be filled as they sink (see fig. 2, B, C, D). The waste may be laid down in part on the floors of lakes, wholly or in part conformable to the deposits of the prede- formational period, or it may be deposited wholly subaérially as fans growing outward from the margins of the surrounding blocks, coalescing and forming an agegraded plain, the deposits of which will, in general, accumulate upon a maturely dissected surface developed as a sequel to the deforming movements. Stripping.— When the consequent and other streams of the sloping uplands cut through the cover and become superposed C. A. Cotton— Block Mountains in New Zealand. 257 on the resistant oldermass, the rate of further downward cut- ting will become comparatively slow. Before this stage the measure of the relief has been increasing progressively with downward cutting and may still increase slightly if maturity of dissection of the surface is still to be attained. After the attainment of maturity, reduction in height of the interfluvial areas of weak covering strata may go on more rapidly than vertical stream corrasion on the resistant underlying rocks. Even if the streams had attained grade without cutting through the cover in the early stages of dissection, after the cover has 1né, By \ “ Sarr YG ND 8D 0 —— Giorcottonaois, 1 Ny a - Fig. 3. A northwestward-sloping stripped plateau surface forming the back slope of the Rough Ridge block in Central Otago, and descending beneath covering strata planed by the Ida Burn. been largely removed from the higher part of the block they will be forced to cut deeper and will eventually become super- posed. With the complete removal of covering strata, all the streams will be incised to some extent. Owing to the resistant nature of the undermass the ravines will long remain narrow, while inclined flat areas on the interfluves will survive. This stage will be attained earliest at the middle parts of the slopes of block surfaces. On lower slopes, undissected interfluvial areas are likely to be larger, and on slightly inclined higher slopes where there is little concentrated wash, remnants of the cover may be expected to survive for some time longer (see fig. 2, E). Eee ocine plains of denudation, almost entirely stripped of their cover, and crossed by many steep-sided, generally con- sequent ravines, which increase rapidly in depth as followed upstream, with here and there remnants of the covering strata relieving the otherwise flat interfluvial surfaces, though com- mon in New Zealand, have received scant attention. They have been noted by McKay, Bell (1907), the writer (1916 6) in the Aorere district of northern Nelson, and by Thomson (1914) in South Canterbury. The more level portions of similar sur- Bf \ NN : 258 0. A. Cotton—Block Mountains in New Zealand. faces in Otago have been termed by Park (1906, 1910) the Barewood Plateau or Central Otago Peneplain; and Speight has recognized a similar surface in the Kaimanawa Mountains in the North Island (1908). The most perfect examples known to the writer are those forming the back slopes of some of the block mountains of Central Otago, for example, the western slope of Rough Ridge (see fig. 3); a long northeasterly slope from the broken pla- teaus of Central Otago to the fault-angle depression followed by the Shag River; a Y similar slope southeastward to the Taieri Plain ; a similar slope northeastward from the Kakanui Moun- tains towards the Oamaru district (see fig. 4); the westward Fig. 4. Fic. 4. Part of stripped plateau surface descending northeastward to Oamaru district, Otago, with adeeply incised gorge, that of the Waianakarua River, on the right, and a large residual mesa of the overmass just to the right of center. slope of the Hunter’s Hills, noted by Thomson, with which are associated similar slopes surrounding the Waihao basin ; the surface of the Gouland Downs, Northwest Nelson ; and the northwestward slope towards the fault-angle depression followed by the Aorere River, northern Nelson. yelients on Stripped Plateaus.—The generally flat surface of a stripped plateau will be broken by a pattern of reéntrant ravines. Salient features may or may not be present. Such may have originated as monadnocks on the eroded surface of the undermass which has been lately reexposed, or as small isolated fault blocks in which the surface of the oldermass has been uplifted above its level in surrounding blocks; or they may be remnants of cover not yet removed, thus closely resem- bling monadnocks in form. Certain remnants of cover may owe their preservation to local induration or local thickening of a relatively resistant stratum and to the presence of a lava flow of small extent. Mesas or buttes of covering strata may thus be scattered sporadically over a surface. Other salients may owe their preservation to slight inequalities of uplift which have caused unusually wide interfluvial spaces between conse- quent streams. Salient features developed from monadnocks and from small uplifted blocks may be distinguished with great difficulty even when first laid bare, for monadnocks may C. A. Cotton— Block Mountains in New Zealand. 259 have been cliffed by wave action (Noble, 1914, p. 62), and fault-line scarps with a very similar form may be present on one or more sides of a small block. Moreover, salients of both kinds may be dissected by insequent ravines, and thus soon lose their initial form. In New Zealand remnants of cover are common as salients projecting above stripped plateaus. In northern Nelson these are generally limestone mesas; in eastern Central Otago vol- canic rocks cap many small buttes and protect large areas. On the more level plateaus are small salients of indurated quartz conglomerate. The writer has not recognized with certainty any monadnocks. ; Mature Dissection of the Undermass.—Stripped plateau surfaces traversed by ravines of moderate depth will persist for a long period if the surface slope is rightly adjusted to the volumes and grade of the streams. In a region of small rain- fall an initial slope of a block surface as high as 10° may have small consequent streams and a large number of subequal, graded, shallow ravines occupied by intermittent streams. Though these break up the stripped surface to some extent, unless the stream spacing is very close, the plateau remnants will be relatively stable. Grading of the ravine sides by soil creep will cause the sharp shoulders bounding the plateau remnants to disappear and the interfluvial areas to be reduced. But still their summits will _ be accordant with one another and suggest the reconstruction of the tangent surface of the undermass. Such conditions exist in the stripped plateaus of South Canterbury and Otago (see figs. 3 and 4). Later, a surface tends to waste away very slowly unless destroyed by erosion working back from initially steep portions of tbe same block. Under other conditions plateau remnants are relatively short- lived. If, owing to abundant rainfall, to steepness of initial slope, or to initial irregularities of surface which have resulted in concentration of consequent drainage along a few channels, the graded profile for such streams lies far below the stripped plateau surface, the plateau remnants will be cut up by conse- quent, insequent, and perhaps subsequent ravines. An early and complete dissection of the surface over the whole block will result, and the block will become an asymmetrical moun- tain ridge with strong relief throughout (see fig. 5). Block surfaces in this stage of dissection are common in northern Nelson, examples being the surfaces sloping easterly from the Pikikiruna Mountains to Tasman Bay. Of the same kind are the northwestern slopes of the Kaikoura and Seaward Kaikoura Mountains, in Marlborough. A special case of large stream volume leading to deep dis- section of a plateau surface is that in which the descent from 260 ©. A. Cotton—Block Mountains in New Zealand. the highest portion of a block to a neighboring trough takes place by a fault or flexure followed by a sloping surface (see fig. 6). Streams of large volame from the higher block will destroy the lower, gently sloping surface with its deep ravines. Such is the dissection of the northwestwardly sloping surface which descends to the Aorere fault-angle depression in north- west Nelson (Cotton, 1916, pp. 66-68). Hig. 5. EN LEB SS 3 Y LCE XS oN ae CNS ENS : A eR Vi) file NEN ANN Lox ee = SS aN _— SSS y i 7 => SSN = WW YEN Hilly, Fic. 5. Dissection of the steeply inclined back slope of a faulted block by consequent and insequent ravines. The initial form is shown on the right. fault Scarps and Fold Scarps.—Except for the possible case of antecedent drainage crossing an uplifted block in a direction opposite to that of the general slope of the tilted sur- face and emerging from gorges on the steeper side of the biock, the dissection of the steeper sides of asymmetrical blocks with the structure postulated will be effected entirely by consequent streams. When they have steep fronts the blocks are bounded C. A. Cotton— Block Mountains in New Zealand. 261 by faults or steep monoclinal flexures; when the crest line-is some distance back, the descent is formed either by a fold sur- face alone or by a fold surface broken by one or more faults. To the former class belong the fronts of simply tilted block mountains. They are dissected by consequent ravines which in the stage of early youth divide a scarp into sections and later reduce it to a linear series of triangular facets. Fic. 6. ani Fic.6. Dissection of a gently sloping block surface (with a cover over- lying a planed undermass) by extended consequent streams from a higher block behind it. The counterpart of this surface is formed in the Aorere fault-angle depression in northern Nelson. Davis (1912; 1913) has distinguished fault-line scarps of generally similar form to dissected fault scarps, but originating as a result of the removal of weak rocks from one side of a fault which has brought’ weak and resistant rocks in contact. Composite Fault Scurps.—In New Zealand many more or less dissected scarps occur, which agree in general form with either fault scarps or fault-line scarps, which may be in their 962 C0. A. Cotton—Block Mountains in New Zealand. lower parts fault-line scarps, though where the displacement on the faults is considerable, they are true fault scarps in their upper parts. When faulting occurs in a region where a weak cover over- lies a resistant undermass, and the base of the resulting fault scarp is at a considerable height above base-level, removal of the covering strata at the base of the scarp may be going on while inovement is in progress. If such removal is prevented by an abundant supply of waste spread out on the covering strata in Pre (. Fie. 7. Diagram of the development of a composite fault scarp, in its upper part a fault scarp and in its lower parta fault-line scarp. the form of fans or a piedmont alluvial plain, after movement ceases degradation will soon take place. Thus a fault-line scarp will be exposed below and continuous with the already dissected fault scarp, the two constituting a single morphologi- cal feature which may be called a composite fault scarp. During the early stages of stripping of the fault surface, the ravines by which the upper portion of the composite fault scarp is already dissected may have cut nearly to or perhaps below the surface of the undermass on the down-throw side of the fault. If so, these ravines will divide the fault-line por- tion ot the scarp before it is exposed into sections that will be’ downward prolongations of the facets of the upper scarp. In some cases these lower fault-line portions of facets may be recognized by their steepness (see fig. 7 C); but the facets will soon dwindle and remnants of the fault-line portion only be left (see fig. 7 D). Where there is a continuous covering of vegetation, soil creep plays an important part in producing convexly rounded C. A. Cotton— Block Mountains in New Zealand. 263 surfaces, as has been pointed out by Davis (1892; 1912) and by Gilbert (1909). In small facets of fault scarps blunting of the edges destroys the flatness of the whole facet, though the dis- section of the scarp by ravines may be incomplete. This rounding of facets is particularly well illustrated by alow scarp at Waimate, South Canterbury (see fig. 8), which was tirst diag- nosed as a fault or fault-line scarp by Thomson (1914). Sharp- edged facets seem to occur only in the case of scarps resulting from movements that have been renewed in very recent times. Fie. -S: Fie. 8. An eastward-facing composite fault scarp near the southern end of the Hunter’s Hills block in South Canterbury. If the crest line—the divide between consequent drainage of the back slope and of the block front—is some distance back from the base-line, the initial descent from an arched ‘crest to the top of a fault scarp may be gentle. Consequent streams will arise on the upper slope, and the ravines will cut deeply and rapidly into the upper slope and after removing the cover dissect the undermass. Thus, in a comparatively short time, the mountain front will be maturely dissected by steep-graded ravines which will continue to work headward and push the crest line divide down the back slope of the block. On the steep mountain-face spurs will descend between these ravines to end in a line of rapidly dwindling facets at the fault trace, and the stage of maturity will rapidly be attained. The front of the Kakanui-Horse Range facing the Shag Valley in eastern Otago appears to be a scarp of this kind (see fig. 9), also the southeastern front of the Kaikoura Mountains (Cotton, 1913). 264 (. A. Cotton—Bl--k Mountains in New Zealand. fold Scarps.—Simple fold surfaces and fold surfaces broken by a succession of small faults will give rise to forms very similar to those just described except in their earliest stages. Removal of a weak cover from a fold surface will be rapid and result in the exposure of some portions of the sloping floor beneath; but the graded profile of the streams may be so steep that deep cutting is not favored. Further increase in the depth of ravines will then take place only as a sequel to head- JEG, €.@..Cotten. 1915. Fie. 9. Maturely dissected scarp of the Kakanui-Horse Range descending southwestward to the Shag Valley fault anglein north Otago. The stripped plateau descending in the opposite direction to meet the fault scarp is seen in the foreground, and, to the right of the center of the foreground, the gorge of the Shag River superposed on the undermass. ward erosion, which may be so slow that flat areas will persist for a comparatively long period, though they may be steep compared with the stripped plateaus of the back slopes. With steeper initial fold surfaces, consequent graded streams may be so deeply incised as to maturely dissect the surface of the oldermass, except, perhaps, for a few facet-shaped remnants. The resulting form is indistinguishable from a maturely dis- sected fault scarp. Vigorous streams may be expected to push back the crest line divide, which will recede in the stage of maturity down the back slope of the block. A good example of a submaturely dissected fold scarp is the eastern face of the northern end of the Blackstone Hill block (see fig. 10). Of the same nature are the side slopes of a broad saddle of cate- nary form separating the stripped plateau of the Gouland Downs from the depression of the lower Aorere Valley in northern Nelson (Cotton, 1916 6) and probably many other scarps in the same region. C. A. Cotton— Block Mountains in New Zealand. 265 5. Trough-filling. An enormous amount of waste results from the stripping of back slopes and dissection of faulted and folded fronts. Ex- ceptionally, such waste may be all removed as it is supplied but in most places deep aggradation of troughs will take place progressively with deformation and with the degradation and dissection of the higher blocks. Fie. 10. Fie. 10. Submaturely dissected fold surface towards the northern end of the Raggedy-Blackstone block in Central Otago. View looking northward. Where initial depressions are open towards the sea and por- tions of low-lying blocks are submerged, or where an area un- affected by movement borders a region of uplifted blocks, the new deposits will overlie the sediments of the predeforma- tional period without stratigraphical break but with abrupt change in the nature of the detritus. Conglomerate may over- lie mudstone and pass upward into fanglomerate, as in Marl- borough (Cotton, 1914). The passage of fine-grained sediments upward to conglomerate in the upper Tertiary rocks of many parts of the South Island of New Zealand is another example. In marginal areas of shallow water, however, the predeforma- tional cover will be eroded by wave action, while unsubmerged areas will be eroded by subaérial agencies, so that. when fans and deltas of waste from neighboring high blocks are built forward these will generally rest unconformably upon a denuded surface. When the phase of maximum aggradation is reached, inter- mont basins will be occupied by alluvial plains through which the eroded summits of small, isolated blocks may project. At this stage, portions of the divides between basins may be buried, the waste spilling over from one basin to another in the man- ner described by Davis in his discussion of the arid cycle Am, Jour. Sct.—Fourtu SERIES, Vou. XLIV, No. 262—Ocrtossr, 1917. 19 266 C. A. Cotton—Block Mountains in New Zealand. (1905 6) and well illustrated in some aggraded depressions in California described by Lawson (1906, p. 455). When such spilling over occurs, the main stream draining a basin may take a spill-over course, become fixed in it during a period of de- gradation succeeding one of aggradation, and abandon its former outlet. Upon the surface of the alluvium filling an intermont basin the aggrading streams will flow in braided, ever-changing channels. In the case of a trough opening to the sea the waste will form a delta or a piedmont alluvial plain. Similar forms will also result along initial fault coasts facing the open ocean. During degradation the filling from some troughs may be removed and the underlying rock exposed. One of the largest intermont basins in New Zealand, the Upper Taieri and Maniototo plain in Central Otago, has reached the stage of dissection at which little alluvial filling remains, and there is little evidence to show whether it has ever experienced exten- sive aggradation. The planed surface of the covering strata of the low-lying blocks forming the floor of the basin is covered generally by a layer of flood-plain gravel. The planed surfaces form several terraces whose slope indicates that they were formed during intermittent regional uplifts by the same small streams which have since dissected them. In places, portions of the undermass project above the lowland surface. A similar stage of reduction has been reached by most of the Otago intermont basins. In North Canterbury the large Waiau-Hurunui basin, which Speight describes as having “an origin in deformational move- ments either of folding or faulting” (1915, p. 348), is floored almost entirely by alluvium. A few “islands” of eroded covering strata project above the basin plain. The alluvial filling has been trenched to some extent as a result of late movements of uplift, but this affords no evidence that the period of maximum waste supply from the mountains to the northwest has been passed. The transverse course of the Waiau and Hurunui rivers across this basin and the adjacent uplifted block indicates an antecedent course (Cotton, 1913). These streams are possibly anteconsequents, for the Hurunui outlet is situated at a sag in the crest of the range. The basin of the Hanmer Plain in North Canterbury, which is enclosed by high blocks, is completely floored with- alluvium, and the covering strata that probably exist are completely buried. With long-continued stillstand, piedmont alluvial plains, like basin plains, must be subject to dissection after the upland blocks have been reduced by erosion. The writer can not point with certainty to any example of it in New Zealand. The dis- section of the Canterbury Plains by the Waimakariri, Rakaia, ©. A. Cotton—Block Mountuins in New Zealand. 267 and other rivers is perhaps an example (Speight, 1908).* It is equally probable that this dissection has been the result of a disturbance of the nicely balanced proportion of water to waste brought about by some climatic change. Dissection may be caused at a much earlier stage of the cycle by coastal retrogradation or by regional uplift. A good exam- ple of deep dissection of piedmont plains occurs in an aggraded tectonic depression which extends inland in a southwesterly direction along the base of the Seaward Kaikoura Range and is followed by the coach road from Kaikoura to Waiau. It has been described by Park (1911) as the Waiau “ Glacial ” Valley. The streams which supplied the alluvium forming the floor of the depression descend from the mountain range, some of them uniting and reaching the sea as the Kahautara River, which leaves the depression at its northeastern end, and others uniting and flowing inland to join the Conway River, Owing to recent uplift the streams are now all deeply entrenched below the aggraded surface which is being rapidly destroyed by the headward erosion of insequent ravines. Part II. Tuer Brock Mountains or CENTRAL OTAGO. 1. Historical Sketch. The origin of the relief of Central Otago has been studied by various writers, and various theories of origin have resulted. Hector (1862, 1869, 1881, 1890) considered the structural fea- tures as the result of normal erosion affected by later regional earth movements. Oneof Hector’s publications (1870) includes an indefinite statement assigning a tectonic origin to the major topographic features. Beal (1871) ascribed the smooth ridges and hillsides to the work of ice and the depressions to stream erosion. Hutton (1875) also believed that the rock-bound depressions of Central Otago,.‘‘old lake basins,” had been excavated by ice. McKay (1884, 1884 5) recognized the oro- genic movements to which the present relief is due and formu- lated a hypothesis which agrees in some essentials with the views presented in this paper. In 1897 McKay recognized a number of great faults in this region, and called attention to the importance of these movements in determining relief. He apparently did not recognize the evidence of tectonic origin in the forms of the mountains and described the depressions as “Jake basins” and the higher fluviatile gravels as “lake ter- races.” Park in 1890 and Gordon in 1893 noted the high inclination of some of the beds of the covering strata and appear to have * In a recent publication, Speight (1911) ascribes the building of the Can- terbury Plains to a postglacial pluvial period and the deep incision of the rivers below the plains surface to a later period of dry climate. 268 C. A. Cotton— Block Mountains in New Zealand. been convinced that it was evidence of deformation. Gordon mentioned an interstratified leaf bed at St. Bathans “ at about the same inclination as the face of the schist rock against which the quartz drift is lying” (p. 119), and appears, however, to have regarded the covering strata as local fluviatile deposits, and the predeformational relief as strong. Park in 1906, 1908, and 1910 described the ranges as block mountains and recognized that the initial forms roughed out by the deforming movements still determine the general forms of many of the mountains of Otago. He pointed out that the upland surfaces in Central and eastern Otago are portions of a dislocated plain of erosion which he termed a peneplain. This plateau had been recognized by Andrews (1905, p. 192), and mentioned also by Marshall (#1905, p. 1038), both of whom regarded it as a peneplain ; but neither of these authors appears to have understood the manner in which the surface had been dislocated, both regarding the relief of Central Otago as the work of erosion. With regard to the origin and filling of the so-called lake basins, Park favored an explanation involving con- temporaneous deformation along lines of fault. He regarded the whole of the cover that had been affected by the deforma- tion as lacustrine and, therefore, younger than the initiation of deformation. | Though Park (1910) has described the South Island as coy- ered by an ice sheet in the Pleistocene glacial period. he has not specifically appealed to glacial sculpture to account for the erosional features of the Otago block mountains, merely credit- ing the excavation of the Dunstan gorge to the work of ice (1906), and describing some morainic accumulations in the adjacent portion of the Manuherikia depression. The Dun- stan gorge was also described as guided by a fault line (1908). The writer’s hypothesis involves planation, sedimentation, and deformation followed by a period of erosion, during which large areas of the planed undermass have been reéxposed by stripping of the cover. 2. Structure. For a great part of its length, the Otago Central Railway fol- lows a chain of broad tectonic depressions and to these the road systein of Central Otago is also mainly confined. It is in connection with this chain of depressions that the salient “block” features occur that present the closest analogy with the block mountains described in other parts of the world (see fies. 11 and 13). Throughout a great part of the area the undermass consists of metamorphic rocks irregularly but not generally closely folded, and moderately though not highly resistant. To the C. A. Cotton— Block Mountains in New Zealand. 269 northeast, folded, unaltered or but little altered,.and more resistant sedimentary rocks form the undermass. The transi- tion from typical greywacke to typical schist is in some places quite sharp. Such junctions in Otago are, perhaps without exception, fault junctions. Along the boundary between the main schist area and the main greywacke area the two types of undermass may be distinguished at a distance of several miles by the details of the relief forms developed on them, and the Hie. 11, a Wins - N ara Ly aS ee Sp 3 enon aniototo Plain rr a —$—$—$—$—$— $$$ — Upper Taicrif i }—Plain—oy\— J aan, Si Fie. 11. Geologic sketch map of the block mountains associated with Central Otago chain of depressions. (Boundaries after McKay, with slight modifications.) The areas in which schist undermass rocks reach the surface are marked by waved north-south lines, the areas of greywacke (unaltered or little altered) by straight north-south lines, and those in which the overmass forms the surface, or is thinly covered by alluvium, by straight east-west lines. Volcanic rocks of the overmass are shown in black. presence is thus indicated of a mosaic of blocks separated by a system or systems of relatively ancient faults to which the present relief is largely or wholly indifferent. Some of the blocks of the mosaic are too small to be shown on the small- scale maps hitherto published. The relief due to movements on these ancient fault lines (together with that resulting from earlier folding) probably 270 C. A. Cotton— Block Mountains in New Zealand. had been almost or wholly destroyed prior to the deposition of the overmass or cover. Though the later faulting to which the existing relief is largely due appears to have followed the lines of the older faults in some places, the displacement has generally been reversed. Upon a planed surface of the undermass rests an overmass of covering strata, which are preserved in the tectonic depressions, but are almost entirely removed from the uplands and high- lands. Over a great part of the area the overmass consists of beds of fine quartz, sand, clay, brown coal, and greywacke gravel. These are generally weak and incoherent with the exception of an indurated Jayer—in places possibly more than one layer—several feet in thickness of quartz grit with a sili- ceous cement. Basalt, relatively a very resistant rock, occurs towards the east interbedded with the covering strata over a considerable area. The overmass of Central Otago could not have been continu- ous unless an emergent land-mass ontside this area furnished the detritus. To the northeast, east, south, and west of Cen- tral Otago, however, are remnants of a sheet of marine-cover- ing strata, and to the north the undermass is composed almost entirely of greywacke. ‘The lower beds of the cover through- out Otago are accumulations of detritus resulting from denuda- tion of schist similar to that on which they rest. It is quite probable, therefore, that some portion of the eroded surface of the undermass has never been covered and so is a true pene- plain (possibly with monadnocks). But the overmass was much more extensive than now. It is significant that this con- clusion, at which the writer arrived independently from a study of the geomorphology, had been- reached much earlier from a study of the beds themselves by McKay, and clearly stated in his later writings. McKay writes of the so-called quartz drifts (a term applied in New Zealand to superficial and bedded auriferous deposits of fine conglomerates or grits—not glacial deposits) as follows: ‘‘These accumulations are so disposed that they are in a large measure—what remains of them—protected from being destroyed by ordinary denuding agents, being either overlain by younger deposits or involved between older and younger strata, so that the same result is effected. That their area in past times was much greater than at present there is abundant evidence in the disjointed scattered patches that are preserved and in the great abundance of cement stones |“ sarsen stones” of cemented quartz grit| over surfaces considerably distant from any deposit of loose quartz drift, and the quantities also of this particular kind of rock in the newer drifts and recent gravel deposits in interior Otago.” (1897, pp. 88-89.) C. A. Cotton—Block Mountains in New Zealand. 271 To the writer’s mind conclusive evidence as to the former wide extension of an overmass is afforded by the occasional preservation of small outliers of the cover on the upland pla teaus; by the wide distribution of the “sarsen stones” (fig. 12) derived from the cemented quartz grit; and by the manner in which planed surfaces of the older rocks emerge from under beveled clay and brown coal strata, as well as quartz grits, with the same inclination. lee, 1, Fic. 12. ‘‘Sarsen stones” on the back slopes of the Rough Ridge block. 3. Major Tectonic Features. As in most parts of New Zealand, the larger features in Otago with the exception of the volcanic massif of the Dune- din district are undoubtedly of tectonic origin. The positive forms fall under the head of block mountains of simple or moderately complex types, and in the Central and northeasten district more than oue of the types of block structure reccg- nized by Gilbert (1874) in the ranges of the Great Basin are distinguishable. These initial forms have been also sculptured in detail by erosion. The area is characterized by uplifts of relatively simple types—uplifted masses either tilted or bounded on both sides by faults forming a set of elongated blocks with a definite south- west and northeast orientation and enclosing elongated depres- 272 CO. A. Cotton— Block Mountains in New Zealand. sions. Blocks bounded on all sides by faults have not been recognized; most of the boundary faults diminish in displace- ment and are evidently replaced by flexures (fig. 13). Towards the southwest, owing to diminishing displacement on the faults and flexures bounding them, the depressions are no longer distinguishable. With the mountain blocks they merge into a high, broken plateau, in which detailed investiga- HG elt. Fic. 18. Generalized diagram of the Central Otago chain of depressions and the associated block mountains. D, Dunstan block; M, Maniototo depression ; R, Raggedy-Blackstone block ; 1V, Ida Valley depression; RR, Rough Ridge block ; Mo, Maniototo depression ; HP, Rock and Pillar block ; T, Strath Taieri depression and Taieri River; P, Barewood Plateau ; B, St. Bathans block; H, Hawkdun block; I, Mount Ida; C, Clark’s Diggings fault angle; K, Kakanui block; S, Shag Valley fault angle; Cl, Clutha River. tion doubtless will reveal the presence of a number of distinct blocks. Towards the northeast, on the other hand, the crests of the mountain blocks slope down so as to merge more or less com- pletely in the lowlands and form a chain of nearly continuous depressions, which is a fault angle between the northeasterly slope of the southern block-complex and a great scarp with a general northwesterly trend, or a series of scarps broken by several offsets, which forms the boundary of a northern com- plex of high blocks. In the latter fault-block complex, which may be termed the northern highland of Otago, the general trend of the disloca- C.. A. Cotton—Block Mountains in New Zealand. 273 tions is northwesterly; but still farther to the northeast the first-mentioned system again makes its appearance with perhaps a more northerly trend. The northeasterly trending system of late uplifts due to dislocations and folds is widespread in New Zealand, while the northwesterly system seems practically confined to Otago, making its appearance again to the south- west of the schist area of Central Otago.* The dislocations of the two systems do not, as a rule, occur together so as to define a regular rectilinear mosaic, but the blocks within the area traversed by the dislocations of each system are elongated like anticlinal and synclinal folds. It is noticeable that while the block boundaries along the great northwest and southeast scarp that forms the boundary between the depressions and the northern highland coincide more or less exactly with the lines of junction of greywacke and schist areas of the undermass, the metamorphics, which presumably were originally the more deep-seated rocks, occupy the downthrown side. That is to say, the dislocations have followed the lines of more ancient faults of even greater throw, but have displaced the crustal blocks in the reverse direction. The Central Otago System.—Oonfining our attention to the blocks somewhat directly connected with the Central Otago chain of depressions, we find that the most westerly of these— the Manuherikia depression—is bounded on the northwest by a highland block, the Dunstan Mountains, with an average height on the crest of 5,000 feet, or 4,000 feet above the floor of the Manuherikia depression. The even crest of the block indicates that its upper surface is flat, and any cover that for- merly lay on the planed surface of the undermass has been stripped from it. The southeastward slope from the crest to the depression is a fault scarp, probably not of the simplest type. The floor of the Manuherikia depression (fig. 14) has an average height of about 1,000 feet above sea-level. It is about 40 miles long and 8 miles wide. It contains a great thickness of the overmass, the beds of which are cut into terraces and dissected into residual hills, a relief that reaches a height of about 600 feet. The overmass is much obscured by a veneer of postdeformational alluvium. The depression is bounded on the southeastern side by a narrow, elongated upland block 38 or 4 miles wide, which a transverse stream divides into two portions. The southwest- ern portion is called the Raggedy Range and the northeastern, Blackstone Hill. The whole may be termed the Raggedy- Blackstone block. Its somewhat undulating crest line is gener- ally about 2,000 feet above sea-level, but rises to 3,200 feet in *In a recent paper, Speight (1916) ascribes the courses of some of the rivers of Canterbury to dislocations with a northwesterly trend. 274 ©. A. Cotton—Block Mountains in New Zealand. Blackstone Hill. The back slope has an inclination of about 10° towards the northwest, where the stripped surface of the under- mass passes under the covering strata of the Manuherikia de- pression. The southeastern face, or front, of the block isa fault scarp (fig. 15) against which are upturned the covering strata in the next depression. This depression, known as Ida Valley, is traversed by the Ida Burn and Pool Burn. The mean height of its floor above Fic. 14. Fie. 14. View looking westward across the Manuherikia depression from the back slope of the Raggedy-Blackstone block. The southern end of the Dunstan block is seen in the distance and the valley of the Manuherikia River in the foreground. the sea is about 1,500 feet. It is about 25 miles long and 3 or 4 miles wide and has the same northeast and southwest trend as the associated blocks. It is floored largely by postdeforma- tional alluvium, but the covering strata appear at a few points along the margins. At the southwestern end the stripped floor appears and rises to merge with the upland plateau. To the east another upland block forms Rough Ridge. _ Its crest is very even for many miles with a height of about 3,200 feet above the sea. On its northwestern and northeastern sides this block is similar to the Raggedy-Blackstone block. C. A. Cotton— Block Mountains in New Zealand. 275 Farther south, however, it is complex and relatively wide, two broad splinters descending towards the northeast and forming offsets on the lowland level between the northeast-trending fault-scarp portions of the boundary line between the Rough Ridge block and the next depression to the east. This depression, about 250 square miles in area, includes the Upper Taieri plain and the Maniototo plain—the whole being termed here the Maniototo depression. The lowest part of its floor is about 1,000 feet above the sea; the mean height of the Upper Taieri plain is about 1,200 feet and of the Mani- Pies to: C.a.Cetran.1 Ss. Fic. 15. The scarp or front face of the Raggedy-Blackstone block (Black- stone Hill portion). ototo plain about 1,500 feet. Considerable portions of the sur- face are covered by a layer of postdeformational alluvium, but at many points the beds of the overmass appear, among which on the eastern side are sheets of basalt. At the northwestern side low islands of the undermass emerge through the cover, the largest being almost continuous with the first splinter from the Rough Ridge block. To the south, as in the neighboring Ida Valley depression, the stripped surface of the undermass emerges, and the line of division between the depression and the southern highland plateau is not simple. A high block southeast of the Maniototo depression forms the Rock and Pillar Range, about 8 miles broad. Its western boundary is in part a fault scarp replaced towards the north by a steeply dipping “ fold-surface,” which passes under a sheet of cover preserved by basalt. The top of the range, presumably a stripped plateau, is gently inclined towards the northwest. The whole drainage of the highland surface is led away in that direction, and profound gorges are cut in the western scarp. In the highest portion of this block, the southeastern edge is about 4,500 feet above the sea, more than 3,000 feet above the floor of the depression on the western side and nearly 4,000 feet on the eastern side. At its north- 276 ©. A. Cotton—Block Mountains in New Zealand. eastern end the block surface passes beneath the cover. A small portion, like the neighboring cover, is here maturely dis- sected. On the southeastern side a very regular and apparently simple fault scarp about 20 miles in length, with an average height of 3,700 feet, descends from the even crest to a long, narrow fault-angle depression named the Strath Taieri, fol- lowed longitudinally by the Taieri River. Its planed floor, beneath which is preserved a narrow strip of the overmass, lin, 20, Fic. 16. View looking west across the Barewood Plateau. The Rock and Pillar Range is on the right, and the Lammermoor and Lammerlaw Ranges are seen in the distance, their fronts being the scarp of a high plateau. descends from about 1,000 feet above the sea at the northern end to about 600 feet at the southern end. On the southeastern side of the Strath Taieri depression the surface of the stripped undermass rises very gently to a plateau (fig. 16) varying in height from about 1,000 feet to 2,000 feet above sea-level. The plateau is traversed by a num- ber of low, southeastward-facing fault scarps (fig. 17). In the fault angles some strips of cover are preserved and there are also lava-capped remnants on the uplands. This area with that to the southwest has been called by Park the Barewood Plateau. Its southwestern continuation is bounded on the western side by the fault-scarp margin of a higher plateau area practically continuous with the top of the Rock and Pillar block, though at a slightly lower level. This is the same broken plateau into which merge all the upland and low- land blocks enumerated. The conspicuous 2,000-feet fault scarps along the boundary between the higher platean and Barewood plateau have caused portions to be known as the Lammermoor and Lammerlaw ranges. C. A. Cotton— Block Mountains in New Zealand. 277 The Manuherikia, Ida Valley, and Maniototo depressions are practically continuous towards the north, being separated only by gently upwarped areas in which the undermass rises sufti- ciently high to be stripped of its cover. The Maniototo depression is separated from the Strath Taieri, however, by a considerable area in which the covering strata, now maturely ao have survived as uplands owing to the presence of basalt. The Northern Highland of Otago.—The northern boundary of the chain of depressions described above is formed by a Iie, 1, Fic. 17. Bird’s-eye view of the slightly dissected stripped surface of a gently tilted, narrow fault block descending northwestward (toward the camera) and meeting alow fault scarp, the crest of which is seen in the fore- ground. Note that the dissection is entirely consequent and insequent. The point of view is a small lava-capped butte of rounded form, a remnant of a former more continuous cover. series of scarps with a predominant northwesterly trend. Beginning at the western side, the northeastern end of the Dunstan block is a narrow, northwest-trending fault-angle depression occupied by Dunstan Creek separating the Dunstan block from the St. Bathans Range. The St. Bathans block merges at its southeastern end into the Manuherikia depres- sion. It is elongated in a northwesterly direction and to the northwest merges into the highland region. It presents a fault scarp to Dunstan Creek and is strongly tilted to the northeast, the back slope descending in that direction to a fault-angle depression forming a northern prolongation of the Manuherikia depression, which is bounded on the opposite side by a long, straight, and conspicuous fault scarp, that of the 278 C. A. Cotton—Block Mountains in New Zealand. Hawkdun Range (fig. 18). This scarp forms the northern boundary of the Ida depression also. The straight front and remarkably even crest of the Hawkdun block have been noted by McKay (1897) and by Park (1906). Professor Park informs the writer that the surface of the sum- mit plateau is strikingly flat and horizontal. The Hawkdun plateau descends southward from an average altitude of 6,000 feet to something over 5,000 feet in a distance of about 12 miles. At its southern end is the dissected dome of Mount Ida, rising 300 or 400 feet above the plateau (fig. 19). It is perhaps a monadnock as suggested by Park (1906, p. 6). The boundary between the depressions and the northern highland area swings around Mount Ida and extends due east Hic. 18. —— Catton 191 Fie. 18. Maturely dissected fault scarp of the Kakanui block. View looking north across the eastern margin of the Maniototo depression. for 10 miles, forming a great reéntrant occupied by a portion of the Maniototo depression, though part of the relatively low-lying area along the base of the scarp is occupied by a number of low-lying, subsidiary blocks, now forming uplands of moderate height in which the undermass is exposed. The surface of the Hawkdun block slopes gently eastward from the Hawkdun scarp and Mount Ida down to a fault angle over 4,000 feet above sea-level, bounded by the fault scarp of the Kurow and Kakanui Mountains, which faces west-south- west and rises about 2,000 feet higher. Remnants of the overmass are preserved in this fault angle on the highland sur- face at the locality known as the Mount Buster or Clark's Dig- gings (McKay, 1884, ce). The Kakanui fault scarp continues south-southeastward (fig. 20) and forms for 15 miles the eastern boundary of the Manio- toto depression, having in this portion an average height of 5,000 feet above the sea. The great reéntrant in the highland rim occupied by a portion of the Maniototo depression is thus completed and the general southeasterly trend of the great scarp is restored. The scarp continues southeasterly to the sea, C. A. Cotton— Block Mountains in New Zealand. 279 a distance of nearly 30 miles. From the crest line of this por. tion of the Kakanui fault scarp, the back slope of a great tilted block descends gently to the northeast as a somewhat undulating, stripped plateau surface and dips beneath marine strata near the coast. A branch of the Maniototo depression extends towards the coast as a fault angle along the base of the Kakanui fault scarp, forming Shag Valley. Its south- western side is a gently sloping and slightly undulating stripped Fie. 19. Fie. 19. Thescarp of the northern highland with Mount Ida in the center, as seen from subdued hills of covering beds in the Maniototo depression, just east of the northern end of Rough Ridge. plain descending from the Barewood Plateau. In the fault angle some remnants of marine covering strata are preserved. 4, Drainage. The Central Otago Chain of Depressions.—The main lines of drainage in the chain of depressions of Central Otago appear to be entirely consequent on the deformation ; but this state- ment does not exclude the possible occurrence of ‘ anteconse- quent” streams. The existence of true antecedent streams is not very probable. The Manuherikia depression is traversed longitudinally by the Manuherikia River, flowing southwestward to join the Clutha (or Molyneux) River. The head of the Manuherikia occupies the fault angle at the base of the Hawkdun fault scarp, and its largest tributary, the Dunstan Creek, coming in 280 C0. A. Cotton—Block Mountains in New Zealand. from the northwest emerges from the fault angle between the Dunstan and St. Bathans blocks. These streams are fed by a large number of small tributaries which with the exception of the Pool Burn are consequent on the slopes of the neighbor- ing block mountains. The Pool Burn, which enters the Manuherikia depression from the southeast, cuts transversely across the narrow Rag- gedy-Blackstone block ina deep gorge. This transverse course Fie. 20. Ze ne ee = SSS ——— —_- ESS TRU wi TOTO Ne c.a. Cotton. IS16. Fie. 20. Maturely dissected fault scarp of the Kakanui block. View look- ing north across the eastern margin of the Maniototo depression. has been described by Park as a capture (1906, p. 13), but none of the features generally associated with recently effected cap- tures are to be seen in the valley system. Neither of the longitudinal streams of the Ida depression—the Pool Burn and the [da Burn—which unite to flow through the gorge, can have been reversed by capture unless the capture took place before the excavation of the covering strata. Park says, “The Ida Valley basin, as shown by the river terraces and surface contours, at one time drained into the Maniototo Basin.” He describes the Ida Burn for a distance of over 12 miles as a reversed stream; but in that distance it has a fall of 500 feet, and it flows the whole distance in a flat- floored valley opened to the full width of the Ida tectonic depression. Such terraces as survive descend in the same direc- tion, giving no indication of reversal. There is, moreover, no stream nor abandoned valley of erosion that can be inter- preted as the beheaded former course of the Pool Burn. Furthermore, it is difficult to explain why a tributary of the Manuherikia should breach the Raggedy block. The climate C. A. Cotton— Block Mountains in New Zealand. 981 is arid; other consequent streams on the back slope of the block are intermittent and feeble; and have not even begun to cut gorges through the range. As the Pool Burn gorge (fig. 21) crosses the Raggedy-Black- stone block at a low sag in the crest line, there can be little doubt that it is consequent on the deformation. Whether the present transverse stream marks the point at which a conse. quent lake in the Ida Valley depression spilled over or whether INTs, ill. Fic. 21. The course of the Pool Burn. the excavation of the gorge kept pace with the deformation cannot now be determined. The centripetal consequent streams of the Ida Burn and Pool Burn systems are of insignificant size and in dry seasons practically disappear. Both the Ida Burn and the Pool Burn are fed by numerous wet-weather consequents which run down the front of the Raggedy-Blackstone block and the back slope of the Rough Ridge block; but the main volume of the Ida Burn comes from the greywacke highland to the north and of the Pool Burn from the schist plateau to the south. Their courses are thoroughly graded on the weak material of the overmass in the depression, broad valley plains have been developed, and the streams are bordering on senility. How- ever, the gradient of the valley floor of the Ida Burn is consid- Am. Jour. Sci.—FourtTH Srriss, Vou. XLIV, No. 262.—OctosBEr, 1917. 20 982 ©. A. Cotton—Block Mountains in New Zealand. erably steeper than that of the Pool Burn. This difference is due to the much greater load of waste which the northern stream has to transport, and which comes from the great fault scarp of the greywacke highland. The little dissected schist plateau from which the Pool Burn flows supplies a relatively small quantity of waste. The numerous parallel, southward-flowing, consequent streams in the northern or Maniototo Plains portion of the Maniototo depression are exactly similar in their gradients and their arrangement to the streams of the northern end of the Ida Valley depression. The Ida Burn and the Wether Burn are separated by a narrow interfluve consisting of the outcropping undermass of a northerly continuation of the Rough Ridge block. The Wether Burn and the other similar streams form the northern members of a centripetal consequent system in the Maniototo depression and are tributaries to the Taieri River. Those streams also consequent on the deformation, which enter the depression from the schist upland plateau to the south, like the Pool Burn, flow with very gentle gradients across the wide valley plain which there forms the floor of the depression. The chief of these, the Upper Taieri, wanders freely, and oxbow lakes abound upon its flood plain. Asa result of the steeper gradient of the northern streams owing to larger supply of waste, the east-and-west axis of the lowland in the Maniototo depression has been pushed far to the south. The drainage of this depression finds an outlet by way of the Taieri River into the Strath Taieri depression, crossing the area of mature dissection around the northern end of the Rock and Pillar block in the fault angle between that block and the Kakanui fault scarp to the north. This portion of the course of the Taieri is probably consequent on the deformation, being determined either by the spilling over of a consequent lake in the Maniototo depression or cut down during slow deformation. The floor of the depression was obviously much higher in an early stage of the cycle before part of the overmass had been removed. It is unnecessary, therefore, to suppose that the present rather young gorge was cut contemporaneously with the deformation. Its cuttmg must, however, have gone on while the floor of the Maniototo depression was being excavated to its present depth. The southwesterly course of the Taieri River through the Strath Taieri depression is obviously consequent on the defor- mation, as also are most if not all the larger tributaries this river receives in its course across the Barewood plateau. The course of the Taieri River itself in a southeasterly direction across the plateau appears to be consequent, as it follows the lowest sag in the surface. It is graded, or almost so, with a gradient of C. A. Cotton— Block Mountains in New Zealand. 283 about 20 feet per mile, but it is still in the narrow-floored, rock- walled Taieri Gorge (fig. 22), which is incised at one place to a depth of about 1,000 feet below the plateau surface. The Shag Valley fault angle is drained by the Shag River, which rises in the maturely dissected uplands overlooking the Maniototo Plain and follows the line of the depression in a northeasterly consequent course to the sea. Most of the tributaries of the streams in the Central Otago depressions flowing over the rocks of the undermass might be Rig. 22. _ Fie. 22. View looking down the Taieri River at the upper end of the great gorge in which it traverses the Barewood Plateau. considered superposed consequents instead of normal conse- quents, if it can be shown that their directions were determined by the slope of a former cover. They are indifferent to both the dip and the strike of the underlying rocks, but as they follow the slope of the surface of the undermass they may be classed as simply consequent, whether they were initially guided by the slope of the planed and tilted surface now visible or by the slope of an overmass for which that surface formed a floor. A distinctly. different kind of superposition is shown by the Manuherikia River and the Ida Burn, flowing parallel with the tilted Raggedy-Blackstone and Rough Ridge blocks, and also 2984 @C. A. Cotton—Block Mountains in New Zealand. by the Shag River. Portions of these streams are now super- posed on the undermass of the sloping plateau as a result of shifting far over to the side of the guiding depression and fail- ing later to slip off the gently inclined surface while cutting vigorously downward in response to rejuvenation. Each such portion is now imprisoned in a narrow rock-walled gorge, while reduction of the general level of the adjacent portions of the depressions is being carried on by minor streams. The case is closely analogous, though not strictly homologous with that of the horseshoe bend of Hoxie Creek, described by Gilbert as an exception to the general rule of monoclinal shifting (Henry Mountains, 1877, pp. 187-188). For some portions of the course of the Manuherikia thus imprisoned in schist gorges Park has given a different expla- nation (1906, p. 12). He says, ‘The river leaves its old course and plunges suddenly into a deep, narrow rift” which is formed by “a number of intersecting faultlike fractures, along which the river runs, passing from one to the other.” Similar superposition has been described in the Clarence Valley fault-angle depression in Marlborough (Cotton, 1913, pp. 2383-234). The Aorere Valley fault-angle depression in northern Nelson apparently affords another example, for which, however, Bell (1907, p. 27) has proposed a theory of capture similar to that advanced by Park for the Manuherikia. The Clutha River System.—Beyond the area with which we are immediately concerned are the largest rivers of the region —the Waitaki to the north and the Clutha (Molyneux) to the south. The latter crosses the Manuherikia depression at its southern end, where the Clutha is joined by the Manuherikia River. The upper course of the Clutha is guided by a smaller similar chain of depressions which may be termed the Upper Clutha chain. A large tributary, the Kawaran, also receives the drainage from a number of narrow fault angles. Inter- mediate portions of its course may be in great part consequent on the deformation, being guided by the lower sags or fault angles in the highland plateau, considerable areas of which, as Park has shown (1908), still survive in the various flat topped ranges. The Kawarau is now the outlet for Lake Wakatipu, which formerly overflowed at its southern end by way of the air gap leading to the upper Mataura. The upper Kawarau, now deepened into a gorge throughout its length, is thus prob- ably of recent origin and perhaps subsequent. The combined waters of the Kawarau and the Clutha form a river of great volume and enormous energy. The portion of this stream which cut the deep Dunstan Gorge (fig. 23), lead- ing from the tectonic depression of the Upper Clutha to the Manuherikia, seems to have been guided by a sag in the high- C. A. Cotton— Block Mountains in New Zealand. 285 land plateau between the Dunstan block and the Old Man Range to the southwest. It may, therefore, be classed provi- sionally as consequent. The same origin may be assigned to the gorge in which the Clutha leaves the Manuherikia depres- sion in a southerly direction. Farther downstream the river traverses a succession of narrow tectonic depressions. The Clutha river system may, therefore, be regarded as almost entirely consequent. The Mataura and Oreti also flow generally in broad tectonic depressions and must be classed as generally consequent. Hires 23: Fig. 23. View looking southeastward down the Dunstan Gorge, by way of which the Clutha River breaks into the Central Otago chain of depressions after leaving the Upper Clutha chain. The Wartaki River System.—The great Waitaki River occupies in its upper course a broad tectonic depression, the details though not the general form of which have been modi- fied by glacial action.* In its middle and lower course the Waitaki is guided by a linear tectonic depression irregularly bounded by the fault scarps and back slopes of a complex of blocks in the bottom of which some low-lying remnants of covering strata are preserved (Marshall, 1915). The Waitaki River is thus probably wholly consequent on the deformation. It follows a complex graben along the north- * Kitson and Thiele (1910) describe the Waitaki Basin as ‘‘due to pre- glacial erosion, faulting, with probably some warping, modified by glacial action” (p. 551). 286 C. A. Cotton—Block Mountains in New Zealand. ern boundary of the block-complex with predominating north- westerly trend which forms the northern highland of Otago. In the Middle Waitaki graben there are a number of small blocks in various attitudes. On the northern or Canterbury side of the graben the dislocations seem to belong to the north- easterly system, though trending somewhat irregularly, while farther north the highest blocks are elongated generally in a north-and-south direction. One great depression, occupied by the consequent course of the Hakataramea River, a tributary of the Waitaki, is of great length and has a width on the floor of about five miles. It is bounded on the west by an imposing fault scarp of the block forming the Kirkliston Range, 6,000 feet high; but its chief peculiarity is a bar of the nndermass about half a mile wide and rising about 400 feet above the level of the river, which forms a kind of sill across the outlet, cutting off the excavated low- land within from the Waitaki depression and river. The sill is traversed by a narrow, rock-walled gorge cut by the outflow- ing Hakataramea River. A layer of bedded fluviatile gravel on the top of the sill proves that the great lowland of the Hakataramea Valley upstream from the sill has been eroded by the river during the time occupied in cutting the outlet gorge through the sill. A number of terraces in the depression mark stages in the excavation of the lowland. In the earliest postdeformational cycle, the undermass of the sill of the Haka- taramea depression had probably not been revealed by erosion. 5. The Surfaces of Uplifted Blocks. Stripped Plateau Surfaces.—W ith the exception of the lava- protected areas, the cover has been stripped from the uplifted block surfaces. The only indication of its former existence is the occurrence of ‘“‘sarsen stones.” The planed surface of the undermass is thus revealed, and in a few places subma- turely or maturely dissected. The stripped plateau surface survives over large areas, even on slopesexceeding 10°. The precipitation is small and the siopes are drained by systems of numerous, intermittent, parallel con- sequent streams and a few unimportant insequents; but no distinctly subsequent streams or positive relief forms on the surface of the undermass have been observed. The edges of the interfluves are almost invariably rounded by soil creep. The type of stripped sloping-platean surface with very shal- low dissection is of general occurrence in Central Otago and the neighboring area to the northeast. (See figs. 2, 3, 17.) This type of surface appears to have a close analogue in that found on the flanks of the resequent or stripped anticlinal C. A. Cotton— Block Mountains in New Zealand. 28% ridges of Table Mountain sandstone in southeastern Cape Colony (Davis, 1906, fig. 6). Details of the Surface on Schist Blocks.—The same general type of relief is found on the schist and on the greywacke blocks. The surfaces of tlie latter are smooth and soil-covered (fig. 24); on the former are many residual tors—great castle- like, joint-bounded masses of bare schist (fig. 25). Tors are characteristic of a Central Otago landscape, and they have sug: gested the names given to many of the schist mountains—as, Fie. 24. Fie. 24. Youthfully dissected stripped surface of a small greywacke block in the larger fault angle between the northern end of the Raggedy- Blackstone block and the scarp of the northern highland. Rough Ridge and Raggedy Range. In the humid area towards the coast, tors are not found ; they are largest and most abun- dant in the most arid districts, a gradual transition being trace- able from the humid to the arid area. The development of the schist tors is, therefore, controlled by climate. They are found on both horizontal and sloping block surfaces, and their form is most characteristic and regular where the schist lies nearly horizontally. The uneven resistance of the schist to erosion is probably not due to differences in composition, as suggested by Rickard (1893, p. 419). Finlayson (1908, p. 73) and Park (1908, p. 12) have noted that the tors are bounded by joint planes. It 288 ©. A. Cotton—Block Mountains in New Zealand. would appear, therefore, that the schist is more susceptible to weathering in the interior areas on account of some peculiarity in the jointing, and it may be that the surface was deeply and irregularly weathered prior to the deposition of the covering strata. The occurrence of scattered logs of timber on the block surfaces (Park, 1908, p. 23; Speight, 1911) where the climate now borders on aridity, proves the existence of forest established in an earlier, moister period. If the forest cover- HaGoecos Fig. 25. Schist tors on Rough Ridge. ing was continuous, deep and irregular weathering may have taken place even in the current cycle. Whether differential weathering is or is not still going on, weathered material is being rapidly removed from the surface of poorly protected soil, and it is thus that the tor pattern has been etched out. As greywacke does not occur in the most arid area, the con- trast between greywacke and schist surfaces may be seen only in the vicinity of the scarp of the northern highland where the climate is semiarid. It may be that tors would form on a greywacke surface in the arid district. But as the greywacke is always thoroughly jointed, possibly no portions of the rock are sufficiently free from joints to survive as tors. It is to be noted that the presence of tors on the plateaus and sloping uplands indicates a general lowering of the sur- face. Tors 20 and 30 feet high are very common, and on the higher plateaus some reach a height of 70 to 80 "feet (Park, C. A. Cotton— Block Mountains in New Zealand. 289 1908, p. 12). This reduetion of the block surfaces is not the result of channeling. The convexity of the edges of the inter- fluvial areas indicates that soil creep plays a part; but on the more nearly level interfluves, erosion is more probably of the sheet-flood kind. Salients on Block Surfaces.— Preéxisting salients on a plain that has been strongly and in places irregularly deformed can not always be recognized with certainty even in an early stage of the postdeformational cycle (p. 257), and some may have been overlooked by the writer. On such of the Central Otago blocks as have been evenly uplifted, however, salients are absent. Mount Ida, on the edge of the northern highland, is perhaps a monadnock. Park suggests that Mount St. Bathans also is a monadnock rising above the level surface of the Dunstan Range (Park, 1906, pp. 6 and 8), but this is a distinct block (p. 278). on the northern end of the Rock and Pillar block and on the gently tilted blocks of the eastern or Barewood Plateau are small prominent salients which are probably without exception lava-capped remnants of the overmass. Scarps of the Schist Blocks.—The majority of the scarps following lines of dislocation in Otago are at the present stage of denudation in part fault-line scarps revealed by the removal of the covering strata. Since the fault scarps on the same lines had not been destroyed by erosion prior to the develop- ment of fauit-line scarps, there is no danger of misinterpreting the geologic history in considering them entirely as simple fault scarps. Fault scarps form the boundaries of a number of blocks in Central Otago, and sufficient evidence of faulting is furnished by the attitude of the covering strata at the bases of the scarps. Along the front of the Dunstan block these beds are shown by Park (1890 c) and McKay (1897, figs. 20, 21, 26, 27) to be steeply inclined, vertical, and even overturned. The dip of the schist-foliation has been shown also by Park (1906) to become steep in the vicinity of the fault line, show- ing that the scarp is at least at its southwestern end a composite fold and reverse-fault scarp similar to that forming the south- eastern slope of the Kaikoura Mountains in Marlborough (Cot- ton, 1913). This block boundary is the line of the Manuheri- kia fault of McKay (1897). The evidence of compression found here is in accord with what is known of the dislocations farther west. Along the front of the Raggedy-Blackstone block the strata are upturned, as shown by Park (1890 a, p. 21), along a fault termed by McKay (1897) the Blackstone Hill fault. McKay writes : “ This line runs nearly parallel to that along the south- east base of the Dunstan Mountains, and both lines have on the opposite side of the valley an outcrop of quartz drifts dip- ping at a lower angle in a northwest direction, or away from 290 ©. A. Cotton— Block Mountains in New Zealand. the range that bounds the valley on the southeast side” (p. 112). There is no definite evidence that the fault is reversed. In other cases the beds of the cover lie practically horizon- tally close to the bases of the scarps, as in the eastern scarp of the Rock and Pillar block, and in the searp of the Hawkdun block (McKay, 1897, fig. 16), indicating probably normal fault- ing Fold scarps of small extent and not yet maturely dissected occur near the northeastern ends of the tapering Raggedy- Fie. 26. Fie. 26. The eastern portion of the Maniototo Plains and the scarp of the Kakanui Range. Blackstone, Rough Ridge, and Rock and Pillar blocks on their eastward-facing fronts (fig. 10), and on the northwestward- facing portion of the last. The fault scarps of the schist blocks, especially where the schist is weak, have crumbled to very gentle slopes. The great scarp, nearly 4,000 feet high, forming the long, straight southeastern face of the Rock and Pillar block, has an average slope of 20°. The Scarps of Greywacke Blocks.—The scarps of the grey- wacke blocks along the margin of the northern highland of Otago are little-dissected fault scarps forming the fronts of tilted blocks, as are also the scarps of the block complex in and beyond the Waitaki Valley depression. Among the well-preserved scarps which form the fronts of simple tilted blocks there is a decided family resemblance, and they contrast strongly with the scarps of schist blocks. The eS C. A. Cotton— Block Mountains in New Zealand. 291 average slope of greywacke fault scarps is nearly twice as steep as that of schist scarps. Like the schist fault scarps, though generally not deeply dissected, these may be described as sub- mature or mature with graded slopes. The difference in steepness between the schist and greywacke scarps is to be ascribed, therefore, to the much steeper gradient which in the greywacke blocks represents a condition of equilibrium after the first rush of scarp destruction by crumbling and slumping initiated by the deformation. In low scarps (up to about 1,000 feet) the average slope of the salients (scarcely to be dignified by the term spurs) is about 40°, and that of the intervening reéntrants somewhat less. Definite sharp-edged facets are not found, the salients being generally of even slope and broadly convex as a result of grad- ing of the slopes by soil creep. Instead of ravines the reén- trants are usually funnel-shaped shoots that merely scallop the edge of the upland surface. They are occupied by streams not of water but of angular fragments of rock. Such a scarp is found in the front of the southern end of the Hunter’s Hills block in Southern Canterbury (fig. 8), and a much higher one forms the western slope of the St. Bathans block, the shallow- ness of the dissection of the front of which is due to the strong backward tilt of the block. In high scarps which do not conform to this extreme type, mature dissection has been effected by steep-grade permanent streams whose heads penetrate only a short distance back into the highland plateau. Of this nature are the Hawkdun and Kakanui fault scarps (figs. 18, 20, and 26). 6. The Floors of the Central Otago Depressions. In general the depressions contain no great accumulations of alluvium as a result of postdeformational aggradation. The vast quantity of waste resulting from the stripping of the overmass from the uplands and its erosion in the lowlands, as well as that derived from the erosion of portions of the under- mass, has been removed in the course of ages by streams, many of which seem puny and almost powerless. In those large» areas in which the covering strata extend below local base- levels and have thus escaped complete removal, planation of their surface is far advanced. This is particularly true of the southern parts of the Ida Valley and Maniototo depressions, floored by the level valley plains of the Pool Burn and the Upper Taieri River. The northern portions of the Ida Valley and Maniototo depressions and of the Manuherikia depression, which are prac- tically continuons, may be described as a local peneplain, the erosion of which has been lately revived several times. The upper courses of the streams vf the northern part of the Manio- 292 O. A. Cotton— Block Mountains in New Zealand. toto depression are at present confined to shallow valleys, the flood-plain floors of which, though of no great width, open out downstream and become continuous. They forma plain which slopes towards the confluences of the streams with the Taieri and is accordant with the valley plain of the Upper Taieri. The recent revival of erosion along the base of the northern fault scarp is perhaps due to climatic changes affecting the sup- ply of waste to the headwaters of the streams. The valley plains and the terraces are covered by a layer of greywacke ravel. i In the southern part of the Manuherikia depression the relief is considerable. There are broad areas of high terrace owing to the recent deepening of the valley of the Clutha, of which the Manuherikia is a tributary. The effect of this may extend into the upper Manuherikia; but it has not yet appeared to affect the streams of the Ida Valley depression ; while those of the Maniototo depression do not belong to the Clutha system. There has been a small rejuvenation in the Taieri River where it flows through the Strath Taieri depression ; but the floor of this narrow fault angle is for the most part a valley plain. LIST OF PAPERS TO WHICH REFERENCE IS MADE. Andrews, E. C.: Some interesting facts concerning the glaciation of south- western New Zealand, Australasian Assoc. Adv. Sci. Rept., 10th (Dunedin) Meeting, pp. 189-205, 19095. Beal, L. O.: On the deposition of the alluvial deposits of the Otago gold- fields, New Zealand Inst. Trans., vol. iii, pp. 270-278, 1871. Bell, J. M.: The geology of the Parapara Subdivision, New Zealand Geol. Survey, Bull. 3, 1907. Cotton, C. A.: The physiography of the Middle Clarence Valley, New Zealand, Geogr. Jour., vol. xlii, pp. 225-246, 1913. — On the relations of the great Marlborough conglomerate to the under- lying formations in the Middle Clarence Valley, New Zealand, Jour. Geology, vol. xxii, pp. 846-565, 1914. — The structure and later geological history of New Zealand, Geol. Mag., vol. iii, pp. 248-249 and 314-320, 1916 a. — Block Mountains and a ‘‘ fossil” denudation plain in Northern Nel- son: New Zealand Inst. Trans., vol. xlviii, pp. 59-75, 1916 6. Davis, W. M.: The convex profile of bad-land divides, Science, vol. xx, p. 245, 1892. — The Mountain ranges of the Great Basin: Harvard Coll. Mus. Comp. Zool. Bull., vol. xlii, pp. 142-177, 1903. The Wasatch, Canyon, and House ranges, Utah, Harvard Coll. Mus. Comp. Zool. Bull., vol. xlix, pp. 15-67, 1905 a. The geographical cycle in an arid climate, Jour. Geol., vol. xiii, pp. 381-407, 1905 b. Observations in South Africa, Geol. Soc. America Bull., vol. xvii, pp. 377-460, 1906. Die erklirende Beschreibung der Landformen, Leipzig, Teubner, 1912. —- Nomenclature of surface forms on faulted structures, Geol. Soc. America Bull., vol. xxiv, pp. 187-216, 1913. Finlayson, A. M.: Some observations on the schists of Central Otago, New Zealand Inst. Trans., vol. xl, pp. 72-79, 1908. Gilbert, G. K.: Progress report for 1872, U. 8. Geog. Surveys W. 100th Mer., p. 50, 1874. C. A. Cotton— Block Mountains in New Zealand. 293 Gilbert, G. K.: Report on the geology of the Henry Mountains: U. S. Geog. and Geol. Survey Rocky Mtn. Region, 1877. The convexity of hilltops: Jour. Geol., vol. xvii, pp. 344-350, 1909. Gordon, H. A.: The goldfields of New Zealand, New Zealand Parl. Paper C.—3, Wellington. 1893. Hector, J.: Notes relative to the geology of the Manuherikia Valley, Otago Prov. Gov. Gazette, Sept. 3, 1862. Progress report, New Zealand Colonial Mus. and Geol. Survey Rept. Geol. Exploration 1868-1869, p. vi, 1869. On mining in New Zealand, New Zealand Inst. Trans., vol. ii, pp. 361-384, 1870. On the distribution of the auriferous cements in New Zealand, New Zealand Inst. Trans., vol. xiii, p. 429, 1881. — — On the deep-sinking at Naseby, New Zealand Colonial Mus. and Geol. Survey Rept. Geol. Exploration 1888-1889, p. lvi, 1890. Hutton, F. W.: Geology of Otago, Dunedin, 1875. Kitson and Thiele: The Geography of the Upper Waitaki Basin, New Zealand, Geog. Jour., vol. xxxvi, pp. 587-558, 1910. Lawson, A. C.: The geomorphogeny of the Tehachapi Valley system, Cali- fornia Univ. Dept. Geology, vol. iv, No. 19, pp. 481-462, 1906. Louderback, G. D.: Basin Range structure in the Humboldt region: Geol. Soe. America Bull., vol. xv, pp. 289-346, 1904. Marshall, P.: The geography of New Zealand, Christchurch, 1900 (?). Cainozoic fossils from Oamaru, New Zealand Inst. Trans., vol. xlvii, pp. 377-387, 1915. McKay, A.: On the northeastern district of Otago, New Zealand Colonial Mus. and Geol. Survey Rept. Geol. Exploration 1883-1884, pp. 45-81, 1884 a. On the origin of the old lake basins of Central Otago, ibid., pp. 76-81, 1884 6. On the auriferous quartz drifts at Clark’s Diggings, Maniototo County, ibid., pp. 91-95, 1884 c¢. Report on the older auriferous drifts of Central Otago, 2d edition, Wellington Govt. Printer (1st ed. Parl. Paper 1896). 1897. Noble, L. F.: The Shinumo Quadrangle, Grand Canyon District, Arizona, U.S. Geol. Survey Bull. 549, 1914. Park, J.: On the Ophir district, Otago, New Zealand Colonial Mus. and Geol. Survey Rept. Geol. Exploration 1888-1889, pp. 17-22, 1890 a. On German Hills alluvial gold-diggings, ibid., pp. 24-26, 1890 b. On Tinker’s alluvial gold-diggings, ibid., pp. 27-30, 1890 c. The geology of the area covered by the Alexandra sheet, New Zealand Geol. Survey Bull. ii, 1906. The geology of the Cromwell Subdivision, New Zealand Geol. Survey Bull., v, 1908. The Geology of New Zealand, Christchurch, 1910. Some notes on the Marlborough coastal moraines and the Waiau glacial valley: New Zealand Inst. Trans., vol. xliii, pp. 020-524, 1911. Rickard, T. A.: The goldfields of Otago; Am. Inst. Min. Eng. Trans., vol. xxi, pp. 411-442, 1893. Speight, R.: ‘‘ Geological History ” in L. Cockayne’s ‘‘ Report of a Botanical Survey of the Tongariro National Park,’ New Zealand Parl. Paper, C.—11, p. 7, 1908 a. Some aspects of the terrace-development in the valleys of the Canter- bury Rivers: New Zealand Inst. Trans., vol. xl, pp. 16-438, 1908 0. eee climate of Canterbury, ibid., vol. xliii, pp. 408-420, Ue The intermontane basins of Canterbury, ibid., vol. xlvii, pp. 886-353, 1915. The orientation of the river-valleys of Canterbury, ibid., vol. xlviii, pp. 187-144, 1916. Thomson, J. A.: Coal prospects of the Waimate district, South Canterbury, New Zealand Geol. Survey Eighth Ann. Rept., p. 160, 1914. 294. Shuler—Dinosaur Tracks in the Glen Rose Limestone. Art. XXIII.— Dinosaur Tracks in the Glen Rose Limestone near Glen Lose, Texas; by Ex.tis W. Souter. Tue finding of dinosaur tracks is sufficiently rare that it is worth while to record all new localities, and especially forma- tions carrying such tracks; the unique occurrence of these tracks in limestone rock seems to merit a somewhat detailed description. Thanks for calling the attention of the writer to the locality and for helpful discussion is due to Prof. J. D. Boon of the Texas Woman’s College of Fort Worth, Texas. The dinosaur tracks figured in illustrations 1 and 2 are exposed in the flat bottom of a ravine near Glen Rose, Texas, where they are locally known as the “ bird tracks.” The tracks are found in a hard white limestone which forms the bed of the stream. The layers immediately over the limestone are less resistant to stream erosion and the stream has stripped back to steep banks a flat bottom more than fifty feet wide, one side of which is covered by stream gravel and sand. Eight tracks were found, five in the uncovered portion and three covered by stream gravel. The tracks show a uniform step of 4 feet and 2 inches. The dinosaur was moving N.E. and the tracks are directly in line, each step following consecu- tively and not alternately from side to side. The tracks show three toes with a measurement of 16 inches from the anterior end of the middle toe to the heel and a width of ten inches. Imprints of claws occur at the end of each toe. ‘There is no evidence of a fourth toe or dew-claw. All tracks are surpris- ingly alike, there being small evidence of right and left foot. The beds in which the tracks are found are placed in the middle third of the Glen Rose Formation, which belongs to the Trinity division of the Comanchian. According to R. T. Hill* the Glen Rose Formation has in this locality a thickness of 315 feet. The middle third of the formation is composed of thick and massive indurated limestone. Figure 3 shows in detail the section immediately above and below the layer carrying the tracks. The measured section is as follows: 3 Section of Glen Rose Limestone where dinosaur tracks are found near Glen Rose, Texas. a) Shell breccia, indurated. 89% calcium carbonate._. 13 inches b) Shale, calcareous; greenish yellow. Limy atthe top 8 inches - c) Shale, greenish yellow. Weathers into a fine- grained clay. 33% calcium carbonate ._--.--- 5 inches *2ist Ann, Rept: U: SG os., pt. Vip. Los. Shuler— Dinosaur Tracks in the Glen Rose Limestone. 295 d) Limestone, white, indurated. Contains dinosaur tracks. 75% calcium carbonate ________- apa 24 inches Eyimestone, shaly, hard .:_........-. eee A inehes ie pumestome, oray; impure. _..........__.:24.:.4... 1 inch 7); uimiestone, shaly, yellowish gray_........_-.-_..- 5 inches Fie. 1. Fie. 1. Dinosaur tracks in the Glen Rose Limestone, near Glen Rose, Texas, As given in the table above a chemical analysis of the lime- stone layer (@) shows that it contains 75 per cent of calcium car- bonate, while the shaly layer immediately above contains 33 per cent of calcium carbonate. The layer below will probably show more than 50 per cent calcium carbonate. 296 Shuler—Dinosaur Tracks in the Glen Rose Limestone. A microscopic examination shows that both the indurated layer with the tracks and the shale layer above it contain finely comminuted angular fragments of quartz, which in the shale measure up to 1/10 mm., while in the limestone only up to 1/50 mimes cling section the limestone layer shows an occasional minute bivalve shell in a groundmass of fine granu- lar calcite with fragments of quartz and clay material. The common association of dinosaur tracks has been with sandstones and shales, rocks which bear visible evidence of Fie. 2 Fic. 2. Near view of Dinosaur track in limestone. littoral conditions. But the dinosaur tracks near Glen Rose, Texas, are found in a hard white limestone, which at the time of the passing of the dinosaur must have been in the form of a stiff, tenacious lime mud forming a surface layer of about three inches into which the foot pressed to depths of two to two and a half inches. Had no dinosaur tracks been found the section would probably have been interpreted as having been laid down in deep marine water and at a distance from shore, since no part of the section shows a calcium carbonate content less than 33 per cent. What then were the conditions under which the nels were made? First of all, the dinosaur was walking on a mud surface, = 2 Shuler—Dinosaur Tracks in the Glen Rose Limestone. 297 lime mud to be sure, but mud. The foot was pressed down into the lime mud sliding forward somewhat, as is seen in the imprint of the heel, figure 2, until the ends of the toes were buried beneath the surface of the mud. The mud was suffi- ciently plastic to hold the form when the foot was withdrawn. The tracks were probably made while the mud was under- neath the water. There are neither sun-cracks as evidence of Fie. 3. 7 sepia (aso? 4, 3C Cs Jz |Shetl Breceia | TTITL Outline of track ae “3 ° ‘ ° ; ea . mane FO did . e ° ass oh 2 Fete : Lye neh ee . = OF : S si ae ‘d M35 mors Caleareous Yellow Shale Weathers | 25 Hard White 4S| | TTT TT | Fie. 3. Cross section of beds in which the dinosaur tracks are found. Outline of track one-twelfth natural size. desiccation or emergence of the muds; nor ripple marks or other evidence of current action. The limy clays immediately above the limestone, which were deposited after the passing of the dinosaur, show little stratification and were evidently laid down by very gentle currents. That the dinosaur was a land animal seems highly probable from the character of the foot. That it was wading also seems probable. A range of depth of water five to ten feet does not seem unreasonable but a further important question is the dis- tance away from shore. The Glen Rose formation was laid down in a transgressing sea, which if the surrounding area were near to base level, would sweep over wide areas with Am. Jour Sci.—Fourts SERiEs, Vou. XLIV, No. 262.—Octossr, 1917. 21 298 Shuler—Dinosaur Tracks in the Glen Rose Limestone. shallow depth in a brief period of time. If the waves were cutting against limestone rocks, lime muds would be deposited near the shore. On the other hand it is true that the muds in question might have been chemically precipitated. There is at present no critical test to distinguish detrital lime muds from those chemically precipitated unless triturated quartz grains be valuable for that purpose. If the dinosaur was wading near shore then, since the dip of the beds is very gentle it would be necessary to think of the seas of Glen Rose time as very shallow, extending over many square miles but only a few feet deep, perhaps not more than five to ten feet deep. Current action except under stress of storm was probably slight; yet the brecciated zone at the top of the section does show current action. Shel] fragments up to an inch in cross section were transported. There is still a further problem to be solved and that is the position of the Trinity sands during the deposition of the Glen Rose limestone. RK. T. Hill’s position that the Glen Rose lime- stone passes laterally into Trinity sands seems to be substan- tiated by many facts. It is noteworthy that in the vertical transition from Trinity sands into Glen Rose limestone there are abrupt alternations of limestone and sand without interme- diate shale beds. It seems quite probable that the lateral tran- sition was equally abrupt and that while Trinity Sands were being laid down along shore Glen Rose limestones were form- ing near shore, so near in fact that only the transition beach of Trinity sands and marlsintervened. If such were the conditions then it would have been easily possible for a land animal to have waded out far enough in a shallow sea to have left tracks in Jime muds. Eubrontes (?) titanopelopatidus sp. nov. For this lower Comanchian dinosaur track the following name is offered: ubrontes (?) titanopelopatidus sp. noy., which being interpreted is “the lime mud strider.” Southern Methodist University, Dallas, Texas. G. Stefanini— Geological History of Venetia. 299 Art. XXIV.—Outline of the Geological History of Venetia during the Neogene ;* by GiusEPPE STEFANINI. Ar this time, when a fierce struggle is in progress and most of the European nations are fighting to return to the geographic and natural limits of their frontiers, it is perhaps not without interest to cast a glance at the geological history of one of the most important battle-fields, one, where, perhaps, the struggle is fought most fiercely from the point of view of natural diffi- culties, namely the region of Venetia. The ideas which I am going to set forth rapidly are a résumé, or rather the conclusions of a somewhat exhaustive memoir, which was published more than a year ago, and which is the fruit of studies which have taken many years of work.+ If, then, from the following pages, based entirely upon objective studies, made without any foreign prejudice regard- ing the purely geological subject which I treated, the reader will be led to draw the conclasion that the Trentino and Val- sugana, from the geological point of view (as well as from the ethnological point of view), are inseparably attached to -the Venetian region, having been a part, since the most remote time, of the same maritime basin, one cannot accuse me of having been guilty of unscientific prejudices or of having wished to bring forward proofs in support of a national issue, which, moreover, has no need of such arguments for its support. Nature and Extent of the Neogene Deposits in Venetia. The Neogene series is represented in the Venetian region by its two series—the Miocene and the Plocene—in very different proportions. ‘The Miocene sediments occupy, all told, an extent of approximately 360 square kilometers; the Plio- cene, on the contrary, is represented by only a very small rem- nant of an almost inappreciable extent in the province of Treviso, to which can be added two others, likewise very restricted, in the province of Brescia. The maximum thick- ness of Miocene beds in central Venetia—where the complete series is represented—totals perhaps 3000 to 3500 meters, while the Pliocene is only a hundred meters in thickness. All of these deposits, from the lithologic point of view, pre- sent a very monotonous facies; they are clastic deposits, marly or molassic,t originating evidently from the denudation of a eke tein heals by Prof. Edward W. Berry; translated by George E. Orsey. io: Memolie dell’ Instituto Geologico della R. Universita di Padova, vol. i Thies tern molasse is applied to marine, brackish and freshwater littoral, either calcareous or argillaceous rocks, easily worked, with interbedded con- glomeratic lenses, characteristic of the Alpine region during periods of oscilla- tion of the strand. 300 G. Stefanini—Outline of the Geological History of continent which was rising. In the Lower Miocene are found sands and “ caleaires grossiers” with glauconite, passing into “ molassic ”’ deposits, or into marly limestones which are some- times employed in the manufacture of hydraulic lime. A fauna rather rich in pectens and echinoids, with some corals and barnacles characterises the stages of this period, fairly easily recognizable,—the Aquitanian and the Langhian (Burdigalian). The Aquitanian rests on the npper Oligocene in Vicenza, Tre- viso, and Belluno; to the west of this central region (Verona, a entino) the Aqnitanian is transgressive on terrains of differ- ent ages, more or less ancient; to the east (Friuli) it is absent, and it is the Langhian that is the immediately succeeding stage resting directly on the Eocene Flysch or on the Cretaceous limestones. And yet in western Venetia there are spots where the Langhian i is transgressive, for example, in the immediate vicinity of Verona. The middle Miocene presents a facies entirely marly at the base (Helvetian) ; a marly sand at the top (Tortonian) ; glaucon- ite appears to be absent even in the clastic condition. In the Tortonian the “molasses” alternate with layers of pebbles, more and more thick and frequent. The fauna is especially rich in molluses of a tropical or subtropical type. The middle Miocene is largely developed in eastern and central Venetia; to the west deposits of this age are known only in Valsugana and in the neighborhood of Bassano; from all the remainder of Vicenza, Verona, and the region around the Lago di Garda it has been removed by erosion. The upper Miocene is represented by an alternation of con- olomerates, marls, and ‘“ molasses,’ many times repeated, rich in impressions of leaves and in terrestrial and fresh water shells. Occasionally teeth of Dinotherium and of mastodons are found in lignite beds, which form a very characteristic horizon at the base of the series. The upper Miocene forma- tions cover large areas in Friuli, Treviso, and in the Trentino ; they have not yet been recognized farther to the West. Finally, the Pliocene, reduced by erosion to extremely restricted remnants, has its usually marly fossils, and presents a curious mixture of echinoids and stenobolic marine, almost bathyal molluses, together with leaves of continental trees. Because of the often displaced position of the beds, and of the nature of the rocks,—all more or less easily eroded,—the Neogene deposits as well as the Paleogene generally do not form outcrops of great extent in the Venetian region. On the contrary they are separated into numerous remnants, spread in part at the edge of the prealpine chain, facing the plain (extra- alpine remnants), and in part in the interior of the mountain region (intra-alpine remnants). Venetia during the Neogene. 301 Concerning the structure, one can say, in general terms, that the Miocene by its mere presence marks synclinal zones; for, indeed, it is in the synclines protected by the older and more resistant rocks, that the Miocene beds, always more or less marly or ‘‘molassic,” have been preserved from complete destruction. Four synclines can be recognized affecting the Miocene terrains of the intra-alpine remnants,—from Friuli (or the province of Udine, or eastern Venetia) through Belluno (province of Belluno) and Valsugana, to the Lago di Garda. A band of Neogene deposits of varying width, in general elevated or overturned, borders almost without interruption the prealpine chain, forming chains of hills, elevated some hun- dreds of meters, “en échelon” in many lines, from Udine to Bassano and beyond; further to the west this band divides into many small fragments, composed of horizontal or nearly horizontal beds, surmounting the Paleogene formation of Vi- cenza, of Verona, and of the province of Brescia. The extra- alpine remnants can in general be considered as the southern wing of the prealpine anticline, although the latter may some- times be complicated by secondary folds. Lhe Preadriatie Gulf. Contrary to accepted opinions, I have been led to think that he intra-alpine remnants, enclosed in general in the bottoms of the valleys, are not evidence of little gulfs, or of narrow arms of the sea, which the surrounding mountains must have formed before the deposition of the Miocene beds. The Miocene beds of the intra-alpine remnants have been ruptured, elevated, overturned, at the same time as the Paleogene or Mesozoic beds on which they rest, and which in turn compose the sur- rounding mountains; they are only the remains, protected. from the forces of erosion, of a formation which must have been continuous or almost continuous, at least from the Taglia- mento to the Chiese, and extending in the interior of the coun- try to Belluno, Borgo Valsugana, and Riva di Trento; that is to say (fig. 1) over the whole extent of the Venetian region in the broad sense of the word. The fragment at C. Caulana, elevated to a height of 1065 meters, is a most striking proof of this. This group of fragments of greater or less extent, is more- over well characterized compared to other Neogene basins. To the east are found only marine deposits in the Balkan Peninsula, in Servia; to the north, the nearest marine Mio- cene formation is found in the Vienna Basin. Everything causes one to think that the Venetian Neogene basin had an individuality all its own, and that it was bounded by a coast on 302 G. Stefanini— Outline of the Geological History of the east, as well as one on the north and the west, while it opened broadly to the south. The characteristics of the facies of the different remnants, compared with their distance from the supposed edge of this gulf, confirms this hypothesis; for the beds of the same horizon, which a minute study has often permitted me to recognize in the different areas, show in general a facies less littoral in those remnants farthest removed from the shores, and vice-versa. In conclusion, the Preadriatic Gulf was only an arm of the sea, which during the Neogene occupied the valley of the Po, and in which were deposited the now classic beds of the Col- line de Turin, and of Langhe, Sarravalle, Tortona, Plaisance, and Asti; beds with which the formations of Venetia are very closely allied paleontologically and geologically. HIsToRY OF THE PREADRIATIC GULF. Form and extent of the Gulf. What has been the history of this gulf during the Neogene period? And, first, how did it originate ? Presumably during the lower Oligocene the Venetian region was almost entirely submerged by the sea, which deposited the beds of Montecchio Maggiore. Perhaps, connected with the basic eruptions of Vicenza, a regression occurred at the end of the lower Otigocene since the deposits with lignite and impressions of leaves, so frequent among the beds of this period in Vicenza, indicate close proximity to a continent. Islands or peninsulas then occupied the neighborhood of Verona and the Tessini hills; eastern Venetia had also emerged. The upper Oligocene, or Chattian, marks an incursion of the sea, and is characterized by beds with Nullipores, and Lepido- eyclina, accompanied by small nummulites. The shores of this Chattian gulf, occupying a part of Vicenza, Treviso, and of Belluno, are relatively easy to identify, for they correspond to the boundary between the region where the lower Miocene is conformable on the upper Oligocene, and the region where the Miocene is transgressive. The Preadriatic Miocene gulf was due to a transgressive movement of the sea and it reproduced in the closest manner, the form of the preceding Oligocene gulf. This Miocene transgression, if my deductions are correct, does not appear to have been produced suddenly. The western borders of the sulf were already more or less completely invaded by the sea at the beginning of the Agquitanian while the eastern border remained dry until the Langhian. Indeed, in the beds in the Province of Verona (as distinguished from those of Verena proper, where itis the Langhian which rests in ravines in the Fig. 1. The Preadriatic Gulf.* Region in which the Aquitanian is conformable on upper Oligocene. Region in which Aquitanian is transgressive. Region in which Langhian is transgressive on Hocene or earlier rocks. Approximate limits of Chattian Gulf (upper Oligocene). Approximate limits of the Preadriatic or lower Miocene Gulf. * It should be noted that the portrayed outline of the Gulf is truncated by the folding of the beds, which is parallel for the most part with the struc- tural axis of the Carnic Alps (W.S.W.—E.N.E.) and has produced a great shortening in the North and South direction. BOOP 304 G. Stefanini— Outline of the Geological Mistory of upper Eocene) the Aquitanian rests on the lower Oligocene beds, and although the transgression may not always be marked - by an angular unconformity, it is in general evident because of the nature of the basal beds—conglomerate or “ breccioles” with littoral fossils, and often formed of débris of reworked fossils from the underlying formations. In Friuli, on the contrary, it is the Langhian whose beds, having at the base a transgressive conglomerate, rest at times unconformably, and at others apparently conformably, on middle Eocene or Cretaceous formations. In western Venetia also the sea invaded the Jand during the Langhian and the islands around Verona were submerg ed. It was thus during the two sub- vanes which form the lower Miocene that the transgression became complete and the sea probably attained its maximum extent. This probably coin- cides with the time of greatest depth—namely the upper Lan- ghian. The Fauna. The fauna which invaded the Venetian area aly the Mic- cene trangression is a warm sea fauna. This is evident in the Aquitanian and in the lower Langhian, but especially in the Tortonian,—that is to say in the stages which are represented by the most littoral facies. The mollusean fauna of the Tor- tonian is a truly tropical fauna, extremely rich in Terebras, Cones, Mitras, Ficules, Cerithiums, Melanias, Turritellas, Cardi- tas, Arcas, Aviculas, ete., which both by their development and by the extraordinary thickness of their shells, indicate a warm climate, and waters rich in carbonate of lime. And, likewise, the remains of plants and terrestrial vertebrates, which the rivers carried into the sea, though rather rare, are an indica- tion of a similar climate; they show that the coasts were then covered with forests of pines and palm-trees,* inhabited by Rhinoceros, Dinotherium, Mastodons, Tragulids, ete. But to return to the marine fauna: in addition to the ideas which they give us of the conditions of the climate, they prove also that broad and easy communications then existed between the Preadriatic gulf and the rest of the Mediterranean Basin, which in turn had broad connections with the oceans. Indeed, it is in these latter, and particularly along the shore of Sene- gal and of the Gulf of Mexico, that one finds recent faunas which show the closest affinities to those which lived in the Miocene Mediterranean, compared to which the fauna of the *E. g. Sabal, confined to America in the existing flura and reaching its northern limit of range in North Carolina, occurred in Europe in the lower Miocene a distance of at least ten degrees farther north and thrived through- out the whole Alpine area. Venetia during the Neogene. 305 Venetian gulf shows insignificant differences due to iocal races or varieties. The terrestrial faunas are most striking, at least from a paleo- geographic point of view. The relative frequency of species identical to or closely allied with those of the Rhone Valley, of Switzerland and of the Vienna Basin, associated with others of still wider range, proves that the Alpine region at that time formed probably .a single biological province. These conti- nental faunas of the Pontian, as well as those strictly littoral ones of the upper Tortonian, by their similarities indicate very close relationships with the faunas living today in the warm countries of the “ancien continent ” (Melania, Terebralia), although one could cite as well in Trivia a bond of union with the faunas of North America, due probably to some collateral phylum of slow evolution. Succession of facies. Some very interesting conclusions can be drawn froma study of the succession of facies in the Neogene series (see table fig. 2), and from their geographical distribution during each stage of this period. The Miocene beds of Venetia represent a complete cycle of sedimentation. The series commences in the lower Miocene with “ molasses” of shallow water origin in the center of the gulf, by littoral conglomerates or sands along the shores. A difference of facies is evident from a comparative study of the rocks, it is confirmed by certain details of the fauna. I shall cite only one illustration—the near shore barnacles, which with the Nullipores abound in the beds of the Aquitanian and of the lower Langhian in the intra-alpine remnants, but are always replaced in the extra-alpine deposits by species of the same genus, but living on sponges or on shells, and consequently able to live in an environment farther removed from the shore. But on this rocky foundation, with beating waters, rich in organisms, peopled with sharks, cetaceans, and sirenians, the depth rapidly creased; in the center of the basin the psam- mophyllic (sandy) faunas with Clypeasters and Scutellas are replaced before the end of the Aquitanian by Spatangids, Pholadomyas, and Pleurotomids. At the end of the Langhian the depth reached its maximum, and in the marly or marly-caleareous beds of this stage abound the simple corals, Pinodonta, and in the extra-alpine region, Cephalopods (Aturia, Nautilus) and bathyal Cirripedia. In the intra-alpine region, on the contrary, in spite of its consider- able depth, the shore is still reasonably near: here and there (Belluno) the large and deep estuary of a stream of some import- 306 G. Stefanini—Outline of the Geological History of ance is recognizable by the nature of the deposits. These are sandy gravels, enclosing organic remains of continental origin, —plants and animals, which attracted a large voracious horde, for sharks teeth and whale bones are found in the sands of Belluno associated with teeth of Rhinoceros, shells of Trionyx and of Emys, pine cones, leaves aud trunks of palms (Sabal, Palmacites). But, in general, elsewhere in the extra-alpine deposits the upper Langhian corresponds, as regards depth, to the bathmetric zone with brachiopods and corals. From this time, the evolution of the facies undergoes a sequence the exact opposite of that of the lower Miocene, the beds of the middle and upper Miocene indicating a slow and continual fill- ing of the basin. The Helvetian is represented by marly beds, deposited in a tranquil sea, in the zone of Laminaria or sea- grass, inhabited by bivalves. The lower Tortonian corresponds to the highest part of the same zone, and the deepest part of the littoral zone in the broad sense; in Friuli one can recognize in this sub-stage two facies only very slightly different—a facies more strictly litto- ral, richer in pelecypods, and a somewhat deeper facies, whose fauna is almost exclusively of gastropods. ‘The first corresponds to a more northern zone, nearer shore; the second is located farther to the south and farther offshore. And this shore was a coast with an extremely gentle slope with very fine sediments such as that of the present western Adriatic. In the upper Tortonian the sediments suddenly become very coarse, the sandy or sandy-caleareous beds become of greater and greater thickness and frequency until,—after a horizon with a littoral facies in the strict sense (intertidal zone) charac- terized by marine, brackish water, or continental faunas inter- mingled, with Auriculides, gigantic oysters, ete..—the advance of the deltas continued, and at the end of the Tortonian the whole region had passed into continental conditions. The delta, first sub-marine, had become sub-aerial in the Pontian, the calcareous conglomerates of which it is formed alternating with marls and variegated clays with fossil leaves, with ‘ molas- ses” with terrestrial fossils and fresh water beds with layers of lignite. | At the close of the Miocene a continental phase returns analogous to that at the beginning of the period; but the lower Miocene possesses all the characteristics of a deposit of a rocky coast, while the Tortonian and Pontian, on the contrary, cor- respond to delta deposits respectively sub-marine and sub-aerial. Thus the entire Miocene Preadriatic gulf was filled with sediments, brought to the sea at first by rivers more or less mature, carrying sand and clay, then (after the Helvetian) by young torrential streams bringing pebbles torn from regions Venetia during the Neogene. 307 recently emerged. Among the various talus slopes which com- posed the delta, there appeared here and there, in the midst of a tropical vegetation, lakes or ponds, in which lived Melanias, Planorbis, Pisidium ; into which fell with the leaves and branches of Cinnamomum and other trees destined to become buried and to form beds of lignite, the shells of Helix and Clausilia ; and where herds of Hyomoschus, Dinotherium, and Mastodons came to water. The early drainage system. It is in these early rivers, accordingly, that the beginning of the recent drainage system can probably be fonnd, whose origin has not been overlooked by the geologists of the region, par- ticularly Mms, Taramelli, Rossi, Dal Piaz, ete. The lithologic nature ‘of the pebbles forming the Pontian conglomerates. in the extra-alpine region of Venetia varies greatly from place to place,-according to the nature of the rocks outcropping in the corresponding river basins. Thus there are eonglomerates formed almost entirely of dolomite, and of limestones with flints, oolites, and Hippurites in eastern Vene- tia. In eastern Treviso they become polygenetic, and mixed with these limestones are Triassic or Permian sands, red lime- stones and even nummulitic limestones. They enclose mica schists, gneiss, granites and quartz porphyries in western Tre- viso, evidently connected with streams rising in the Valsugana and Cima d’Asta, where many of these crystalline rocks are - only exposed to-day. Definite traces of this early drainage sys- tem can elsewhere be recognized in certain patches of a con- _ glomerate with rounded pebbles which one meets here and there in the interior of the region, in the mountain region of Friuli, at an elevation of 1500-1600 meters above sea-level, every ‘where on the divide between the valley of upper Tag- lhamento, and the head of the valleys of the Meduna and the Arzino. The study of the rocks which are found as pebbles in the Pontian conglomerates permits us to make still other deductions. The Mesozoic limestones (particularly the Turon- ian and Senonian limestones with Rudistes) form, it will be remembered, the periphery of the Alpine chain in this region,— that is to say, the base of the intra alpine zone,—but they never appear farther in the interior, where the Upper Cretaceous (locally called Seaglia) is transgressive on Jurassic or Kocreta- ceous limestones. The great * frequency of pebbles of this Rudistes limestone in the Pontian conglomerate of Friuliis a further proof in support of the hypothesis just enunciated, of the partial emersion of the intra-alpine zone before the end of the Tortonian ; for it is only in the Tortonian that these peb- bles begin to ‘appear. On the other hand, it must be noted that the marine Tortonian beds, lying conformably on the 308 G. Stefanini— Outline of the Geological History of earlier Miocene, are found forming a part of the intra-alpine deposits, in central and western Venetia. Thus, it is toward the end of the middie Miocene that the peripheral region of the gulf underwent an elevation which carried it well above sea level, while the central part of the gulf was only filled up in the upper Miocene with sediments which the young streams brought into it in large part from the recently emerged region. Age and nature of the Miocene Displacements. As to the duration and the nature of the movements which affected the region at the time of the emergence, one must be warned lest he believe that the succession of facies, everywhere not so thick as are observed during the middle and upper Mio- cene, is an indication of a slow and general elevation of the bottom. On the contrary, the Tortonian alone measures 700 meters in Friuli. Now these lower beds exhibit already a strictly littoral character and could never have been deposited in a depth of several hundred meters. Upon as good grounds this same reasoning could be continued for the complete series of middle and tpper Miocene beds, a thickness of 1800 to 2500 meters. : There is no doubt in my mind that the deposition of this enormous quantity of clastic material, corresponding to the emergence and the erosion of the Alpive continent, was accom- panied by a siow but general and continuous sinking of the bottom; a subsidence, however, which was not sutticient to pre- vent the formation of deltas which continued until the com- - plete filling up of this part of the basin. A gradual sinking of the bottom is, according to the studies of Barrell, an indispensable condition for the formation of delta deposits of enormous thickness, such as the sub-Himalayan formations, which suggest, in many respects, resemblances to the subalpine formations of Venetia. In conelusion, during the middle and upper Miocene, the Preadriatic gulf was the seat of two opposite movements. While the peripheral region underwent the emergence of which I have spoken above, other areas farther seaward, and especially those which corresponded to the central part of the gulf, the sinking of the bottom, begun throughout the entire region at the beginning of the period, continued, and the materials which the youthful streams tore from the recently emerged country contributed largely to the filling of the basin which persisted in the extra-alpine zone. The Pliocene transgression. The geocratic* period just mentioned was not of long dura- tion. At the beginning of the Pliocene the sea returned, bear- * Geocratic = uplift, or a negative movement of the strand. Venetia during the Neogene. 309 ing a fauna which was the direct forerunner of the recent faunas of the Adriatic and the Mediterranean province in general, although considerably mixed with Senegalian elements. The argillaceous nature of these deposits, and the phenomena of elevation and resulting intense erosion, which immediately followed, easily explains the small extent "to- -day of the lower Pliocene, now preserved in two extremely limited fragments ; at Cornuda (Province of Treviso), and at S. Bartolommeo de Salo (Brescia). At Cornuda they rest conformably on the Pon- tian conglomerates, dipping at about i0 degrees; at S. Bar. tolommeo they lie in deep gullies in the Senonian “ Seaglia,” and are often found raised “fen bloc” and almost horizontal to 530 meters above the sea. In both eases they are fine marls of a quiet and fairly deep sea, rich in Brissopsis, Schizaster, Amussium, Arca, Natica, etc., bat deposited probably at no oreat distance from the coast, as is indicated by the impressions of leaves of Platanus, Llex, and Rhododendron, which they contain. The conformity of the Pontian and Plaisancian at Cornuda shows that there were no violent movements in this region, nor any considerable interruption in sedimentation between the Miocene and the Pliocene. The return of the sea was accom- plished quietly; the waters simply submerged the deltas of the basin which they had recently built up. This was merely a return of the Preadriatic gulf. Pliocene uplift, and the rejuvenation of the drainage system. It was only a short return, however, for this talassocratic* phase did not last long. At Cornuda as well as at Salo proofs of a phase of intense © displacement are evident. At Salo there was elevation with- out considerable folding, at Cornuda the latter occurred. The combined study of these two fragments permits us to identity, with a precision very rare in our science, the geologi- cal date of these movements. At Cornuda almost horizontal Villafranchian conglomerates (upper Pliocene or lowermost Quaternary) lie in ravines, at times in strongly inclined Plaisan- cian marls, at times on Pontian conglomerates likewise inclined, at times on older formations. The Astian appears to be absent. At Salo the sub-horizontal Plaisancian, lying in valleys in the Cretaceous deposits, has been elevated more than 500 meters, while a short distance from here at Castenedolo, the sub-hori- zontal Astian lies only 120 meters above sea-level. One can thus say that the period of displacement, which is manifested ina simple elevation at Salo, and by folds at Cornuda, took place between the Plaisancian and the Astian. * Talassocratic = hydrocratic = depression, or a positive movement of the strand. 310 G. Stefunini— Outline of the Geological History of It is with this moment in the geological history of the region that we can correlate the rejuvenation of the drainage system, whose old valleys are met at from 600-640 meters above sea- level. The high limestone plains of Friuli and the epigenetic gorges which abound in the whole prealpine zone are the most striking examples. Indeed, throughout the upper Miocene and the Plaisancian the drainage system, which we have seen developing since the upper Tortonian, had progressed toward maturity with the partial peneplanation of the country. The high limestone plains of the Karst, which border the Alpine chain in this region, in my opinion, are only remnants of this peneplain, still only partially completed when the elevation of which I am going to speak was produced, carrying these remains to heights varying between 800 and 1,200 meters above the present sea-level. The erosion of the profound and narrow gorges which open like passage-ways in the Mesozoic limestones at the mouths of most of the valleys in the Venetian plain, date back to the upper Pliocene. Farther in the interior of the region it is not difficult to find similar traces of this old drainage in the very ancient alluvial deposits, lying at 700-800 meters. Post Pliocene Continental Formations. After this phase of active erosion, which in the extra-alpine region is indicated by a well marked angular unconformity and by very evident lithologic differences, another phase of sedimentation ensued at the beginning of the post-Pliocene, and conglomerates were deposited as enormous talus slopes at the mouths of the valleys on the plain. These are the con- glomerates which the older Italian authors called “ preglacial,” but which according to the modern view can be considered as corresponding to a glacial period or a very early interglacial period. Above these formations were deposited the well known morainic ares of Tagliamento, Piave, Garda, etc., while in the interior of the valleys, embedded in these old alluvial deposits, other alluvial deposits were laid down, in turn greatly terraced. But tectonic movements did not completely cease in the Qua- ternary, for the Villefranchian conglomerates, of which I am going to speak, are themselves bent into an anticlinal dome in the hills of Montello and Conegliano. Character of the Tectonic Movements. Summarizing—the Venetian region during the whole of the Neogene was a gulf opening into the sea, which then occupied the basin of the Po. , This gulf was already outlined in the upper Oligocene; it became larger during the lower Miocene, reaching its greatest yFTTGn 311 eu0gZ sutdty Venetia during the Neogene. “Biya JO 7FTTGn y5TTdn *suteTa 4Stex TOUBTH Teroued *uoT}PeuRTdeueg *AyTAn7eW te SninyeA sedey YYTA OTOp-us4sBg FO S}Tsodep suTzeI NVILSV (skeTTea oTzousstde)uoryeusanCay So0eLL9} Teysty°wnNtANTTS WeTYoUsTPeETITA (oTTeqUOW) UOTTBTANTTY ANAOOLSISTd eILE A Nae So eae Oe Moore el bee Eee ¥) SIOPeoy due OTUTBION*sooBTLIEY TAIOT*sreToOeTS JO SUCTSUaz oT “BTITA JO sjUsueoBTAsty SJUSUaAOU DTUO0 JIS, HLYON eoyofsuted YATA SgNT°UT~USOTA JO SuUoTICNTS OTSeR SNZOODINO AIGCIN *sJINy olpPetApyerd ey} JO BUTT INO TEUTITIO gNTOODIIO adden ere ee ee ee *STUOFOePUN}OAINS BTTOINOS —_—_—__— pue STSueTBIAS UCU SNUsoOTIed YI Spups NVINVLINOV (weyoed UTN Spues)etoys syO0L Jo seToOeF TBL077TT OT ry AMTOOTHN 8 *Zodeep ATeAtssIaJoaId sSotoe sr} (StTeer0q BUTTON] pue Tanze wranqy YyTN STzeN) NVIUONVIT BTBZOO DUT SpodoTyoeRra fo seTour ie EEE (FrAnTIy Bory pue TuTpretng snus) ULE STren) BIaeuUTUB DUB BLd}sOZ FO euos G44 FO SyTBoded NVILGATOH ne a : ‘ (Tsneysaoe snuog pue SLozTyOI SuopoLg YITM asseton) (oyet nsues) szFsodep Ter04qqtT UNTOOIN AIGCIN Brsuesula BIL}sQ NOTTeys ATAATSSearZord setoEey SoUSPTSGNsS SUTPSd0% UCTS USUTPES|uoTYeYUSUT pes SUTpeeoxe 9oUepTSsqns NVINOLAO’ pue BUTSSTSSELO BATISO UIT SO}yBTeUOTSu0o pues SOSs BT ON) SYISOCap BLIP IUuTZeUqnS NEOUTST (nt WUNTYLILED YIM SOqQTUSTY *BTUPTSSeD YYTN SOSSETON)SITSOCep TepTyreqUT *STsusutequteys ...2--heportedno Ge. Faint! (10) 0°2 g p y cloudy. No ppt. 1t}=0:000i. orm: Ge. 2.525. i ce foe an ei Reported Ge present. S i8 2 Good test. aac pestis Oe a se 22S Le Reported possible trace of Ge very faint. In all of these tests potassium permanganate was added with the hydrochloric acid before distilling, to provide the chlorine necessary to prevent reduction and distillation of AsCl, in the event of the presence of arsenic. Blanks were made with KMnO, to determine whether the chlorine evolved would cause a precipitate of sulphur in the distillate when hydrogen sulphide was added. ‘The results of these determinations showed in only a few cases a faint cloudiness, probably due to sulphur, but in no case did a precipitate settle as in experi- ments where 0:0001 grm. Ge was present. Experiments (2), (4), (6), (8), (10) and (12) also showed that the use of potassium permanganate does not vitiate the germanium test. The dis- tillates in Experiments (2) and (4) were allowed to stand over night after saturation with hydrogen sulphide without any pre- cipitate separating from the very faintly cloudy solution. Winkler * recommends the following method for the separa- tion of germanium from arsenic. The sulpho salts formed by fusing the mineral containing the elements with sodium carbon- ate and sulphur are dissolved in water, and the solution is care- fully neutralized with sulphuric acid, when the sulphide of arsenic is precipitated leaving the germanium in solution. By treatment of this solution with hydrochloric acid and addition of hydrogen sulphide the germanium sulphide is thrown down. Two modifications of this method were tried by us to bring about this separation. First, a solution of the sulpho salts was saturated with carbon dioxide. This treatment caused the pre- cipitation of from 60 per cent to 75 per cent of the arsenic, while no germanium was precipitated. Second, a solution of the sulpho salts was treated with ammonium acetate, acidified with acetic acid, and then treated with hydrogen sulphide. Under these conditions the arsenic was completely precipitated, and the germanium remained in solution. This process was tried on solutions containing arsenic only and germanium only, as well as upon mixtures, with equal success. Finally, some solutions of content unknown to the experimenter were pre- pared by one of us, and the report by the other showed the accuracy of the method. * Journ, prakt. Chem., xxxiv, 177; xxxvi, 177, June, 1917. 316 _ LH. Ries—A Peculiar Type of Clay. Art. XXVI.—A Peculiar Type of Clay; by H. Rims. Tue clay referred to in this paper was received by the writer from western Texas, and is of such unusual character as to be worthy of record. On superficial examination it appeared to be a gritty clay resembling loess.’ When mixed up with water it developed sufficient plasticity to permit its being molded into bricklets without difficulty, and these when fired at 950° C. baked to a porous body of moderately hard nature, but which on exposure to moisture disintegrated completely, indicating that the clay contained a high percentage of either lime or magnesium carbonates. | It was the microscopic examination of the material that showed its extraordinary character, for it was found to consist almost entirely of small rhombs (fig. 1) of varying size but averaging about ‘008"" in diameter. The quantity of very fine undeterminable clay particles was practically negligible. . “08S Port Molle HWF &¢ Thue Spruwe rise 125° Wi 16 ssh 6 SH : 6-4 \ & et Seaforth: 1/1757 fp EERES) Splines ay Gia, 498) 46, oF 8.0.45 = ‘\ oe Waagris @ oly Sas , Sem" as SE 4 : a ‘ 38\ Burning hy Extensive plain tn = 2 ‘= Low Country ey 2 > £ ARigh Hills s > 2 q = F at the back y = - E NSS Pi REPULSE BU a ad ack, eae 2 17 16 | 14 4 \~ gi NZ Sil, BAY set ae Set es fn Ne ast poe Se eed Nee st as Pt Ro Ga se A BOM Nise BFE é XIDO™ Rise I aus = satisfactorily explained. Vaughan’s view is based on the physiographic investigations of parts of the eastern coast of Australia by Andrews (19038) ; but a careful review of Andrews’ papers and a brief inspection of the region itself in 1914 have convinced me that the explanation just quoted is not the only one that deserves consideration, and that upgrowing coral 342 W. M. Davis—The Great Barrier Reef of Australia. Fie. 3: 1628 1638 {G00 W. M. Davis—The Great Barrier Reef of Australia. 343 reefs may have bordered the Queensland coast not only during its recent moderate submergence, but during a long antecedent period of great downwarping, as a result of which the total thickness of successive reef growths and of the lagoon deposits behind them may be as much as one or several thousand feet. The reasons for this interpretation are here set forth. Andrews was the first to apply modern physiographic methods to the explanation of the coastal highlands of eastern Australia. The region that he more particularly studied is locally known as “ New England,” and lies in the northeastern part of New South Wales, not far south of the southern end of the Great Barrier reef which fronts the Queensland coast farther north. The history of the highlands indicates the history of the off- shore sea bottom or reef foundation also. Andrews’ chief results are that the New England region, of complicated struc- ture and once of mountainous form, as in block A, fig. 4, was reduced to an extensive peneplain, FI’, in Cretaceous time, and that as a result of successive broad uplifts of a north- south belt, GG’, MJ’, RK’, after shorter and shorter intervals of time, three coastal peneplains of Jess and Jess extent, LL’, QQ, UU’, have been eroded on the eastern slope of the uplifted belt at lower and lower levels, so that they now form a series of benches, each one a number of miles in breadth, separated by irregularly dissected escarpments, K, O, S, independent of ‘rock structure, and from several hundred to a thousand feet in height. The lowest and youngest of the peneplains, UU’, now diminished in breadth, VV’, by subrecent submergence, constitutes the present coastal lowland; like the earlier pene- plains it is often surmounted by monadnocks, large and small. Andrews gives only a brief discussion of the off-shore changes attendant upon the successive uplifts of the highland. His most significant statements on this point are as follows: “ Dur- ing late Phocene and Pleistocene times, a differential subsi- dence took place for the coastal area ... The formation of the present Great Barrier reef probably does not antedate this last movement of subsidence, although reefs doubtless existed prior to the cycle of subsidence. Unless the movements of subsidence were accentuated in an easterly direction, the depres- sion which determined the Barrier Reef must have been very moderate in amount... The effect of the late subsidence was ... to give birth to the Great Barrier reef.’’* It is to the extension of this discussion, particularly with regard to an eastward accentuation of subsidence, that the present paper is directed. ‘The successive uplifts of the coastal region must have been of diminishing measure eastward: for *H. C. Andrews, An outline of the Tertiary history of New England. Rec. Geol. Surv. N.S. W., vii, 1903, 140-216; see pp. 214-215. 344 W. WM. Davis—The Great Barrier Reef of Australia. if the uplifts had been everywhere of the same amount, the continent would have been broadened thereby, and the coastal lowland today would show successive littoral belts of younger and younger marine sediments, more or less eroded. The absence of such belts indicates that, while the interior high- lands were uplifted, the adjoining sea floor either stood still or subsided. Let it be noted in passing that the absence of littoral belts of marine sediments suffices to disprove the possibility to which Fie. 4. icon Kon kates == SL I is (ih Za UNS = <== ‘AN Ww Rae ji . WES < some credence appears to be given by Australian investigators, that the uplift of the highlands was not actual but only appar- ent, and that the sloping land really stood still while the sea sank to lower and lower levels; for in such case at each sink- ing the marine sediments previously deposited would be laid bare, and although they would be worn down to lowlands in each interval of local peneplanation, they would still be visible today in the littural zone. No such belts are to be seen; hence the supposition that the land gained height because the sea surface sank is excluded ; the continental mass, not the sea. surface, must be the chief seat of movement. Furthermore, had the coastal lowland and the adjoining sea bottom stood still while the inner highlands rose, the present W. M. Davis—The Great Barrier Reef of Australia. 345 shore line would not be embayed, and a broad fluviatile plain, composed of the confluent deltas of many rivers, would border the coast ; but asa matter of fact, the shore line is well embayed and the delta plains are relatively small and discontinuous, as at V, fig. 4. Hence the sea floor but not the sea surface probably subsided each time that the highland rose; in other words, the littoral belt was gently and intermittently tilted eastward between a broad highland anticlinorium on the west and a deep sea-bottom synclinorium on the east: in short, the littoral belt occupies a zone of intermittent flexure. The line of no change of level or axis of flexure cannot have advanced seaward as the successive flexures took place, as in section E, fig. 1, for in that case as well as in the case of uni- form uplifts, belts of marine sediments would now be found in the coastal lowland. The line of no change cannot have remained fixed, for this would entail a non-embayed coastal lowland bordered by fluviatile deposits ; of less breadth, to be sure, than if no down-flexing had taken place off shore, yet much greater breadth than is actually found in the discontinu- ous deltas of today. Hence, as these various possibilities are excluded, the line of change probably shifted westward, as in section D. This interpretation is adopted in fig. 4, in which nine stages in the physiographic development of the region are represented in successive east-west blocks from background to foreground. The disordered structure of the region demands, as above noted, that it was once mountainous, as in block A. The mountainous mass must have remained long quiescent, for it was reduced in Oretaceous time to an extensive peneplain, FF’. Then as a result of a gentle flexing the seaward area was sub- merged and the interior area was elevated as in block GQ’ ; and in this position the seaward slope was reduced to a second but less extensive peneplain, LL’, separated from the uncon- sumed part of the uplifted peneplain, HH’, by a ragged retreating escarpment. This escarpment of differential erosion is analogous to the mountainous escarpment in North and South Carolina known as the Blue Ridge. Similar changes occurred twice again, resulting in the formation of a third peneplain, QQ’, and a fourth peneplain, UU’. But as the present shore line is much embayed, and as many subdued hills rise as islands from the shallow sea bottom, the eastern part of the fourth peneplain, UU’, must have been slightly down- flexed and submerged at a recent date: the interior part of the peneplain, constituting the present coastal lowland, VV’, together with the benched highland, TT’, back of it, were probably upflexed at the same time, for the lowland, where I saw it, is now trenched by narrow valleys. Hence the latest 3846 W. Mf. Davis—The Great Barrier Reef of Australia. movement of the coastal region seems to be of the same nature as, but of smaller amount than the movements inferred to have taken place at much earlier periods. A slight emergence of very recent date need not be considered here ; it may be the result of the latest upflexing, or it. may, following the principle of shore line changes demonstrated by D. W. Johnson, be the result of diminished tidal range in consequence of the enclosure of the lagoon by the present barrier reef. The problem before us is to determine what opportunity for reef formation may have been offered during the successive down-ilexures of the sea floor and during the long intervening” periods of rest as indicated by the successive peneplanations of the coastal belt; but as this inquiry involves the conception of young barrier reefs, mature reef-plains, and abraded reef- plains, two paragraphs will be given to that aspect of the sub- ject, following and extending a view of reef development first clearly enunciated by Hedley in 1907. Imagine a recently uplifted and reef-free coast, or for eon- venience of graphic illustration imagine a recently constructed, reef-free voleanic cone, of which an undissected sector is shown at D: hes d-0aP he shore is attacked by waves and a cliff and platform are developed, sector E, on which no reefs can be formed* until submergence embays the radial valleys, sector F, when a narrow young reef will make its appearance, enclosing a lagoon where aggradation takes place. ‘Tahiti, I believe, exemplifies this sequence of development. Let it be supposed that the reef thus initiated is maintained in a narrow or youth- ful stage with a widening lagoon by continued subsidence, uniil in sector G a long enduring still-stand period sets in. The youthful reef will thereupon widen to more mature form by outgrowth and overwash; the embayed valleys will be filled by deltas, which will advance farther and farther into the lagoon, as in sectors H, J, and K; and the lagoon will be at last converted into a mature reef-plain, as in sector L. The mature reef-plain is not, however, the limit of orderly change if the island suffers no variation of level. The detritus from the central island will be delivered by the streams to the outer face of the mature reef, and the growing corals may thus be smothered. When that eventuality is reached the waves will cut the dead reef farther and farther back, reducing it toa shallow submarine platform, as in sectors M and N. Finally the central island itself will be again attacked, unless before that goal is reached subsidence sets in anew, whereupon a rejuvenated reef will make its appearance on the outer edge of the widening platform, or on the outer edge of the narrowed * See Clift Islands in the Coral Seas, Proceedings Nat. Acad. Sci., ii, 1916, 284-288. W. M. Davis—The Great Barrier Reef of Australia. 3847 plain, or on the island border, according as the submergence is large and rapid or not. It is interesting to note in passing that Darwin foreshadowed the idea here presented: he noted that, should atolls “remain at their present level, subjected only to the action of the sea and to the growing powers of the coral,. ...it cannot... . be doubted that their lagoons and the islets on their reef would present a totally different appear- ance from what they now do. This consideration leads to the Fie. 9. Tea —————— ——— i | hee Te if il suspicion that some agency (namely, subsidence) comes into play at intervals, and renovates their original structure” (’42). In applying this scheme of reef development to the Queens- land coast, it will be convenient to trace the sequence of events backwards from the present into the past. _ But let it first be noted that the Great Barrier reef is already a mile or more wide in part of its length, and that the coastal embayments are already somewhat encroached upon by deltas; hence the present reef, Z, fig. 4, formed after the fourth flexure of the Jittoral belt, has passed the stage of earliest youth and is advancing toward maturity; if the coast should stand still for a long time to come, the broad lagoon would be converted into a mature reef plain. Now as to the origin of the present reef, there can be little doubt that the opportunity for its upgrowth was given by the latest or fourth flexure of the continental border, whereby the formerly wider coastal lowland, UU’, was reduced to the pres- ent narrower lowland, VV’. But the wider coastal lowland must, before flexure and submergence, have been fronted by a 348 W. I. Davis—The Great Barrier heef of Australia. mature reef plain, Y’, formed during the fourth peneplanation of the coastal slope, UU’; and the mature reef plain must have been developed from a young, upgrowing barrier reef, Y, that had been initiated by the flexure which tilted the peneplain, QQ’, into the coastal slope, RR’, just as the present reef, Z, was initiated by the flexure which tilted the peneplain, UU’, into the coastal slope, VV’. Again, the foundation from which reef Y grew up must, with large probability, have been a down-flexed and submerged mature reef-plain, X’, of an earlier cycle; and so on backwards through the series of flexures. Thus, as far as present features can be genetically linked to the chain of antecedent features, the platform, shown in fore- ground section, from which the visible Great Barrier reef, Z, has grown up, appears, in its outer part at least, to be in large measure the product of coral-reef agencies that were directly or indirectly in operation through a considerable period of past time; for it must be remembered that reef-forming agencies contribute to the formation of a mature reef-plain not only directly by the constructive growth of the reef organisms and by supplying overwashed waste, but also indirectly by serving as a breakwater which encloses a lagoon where a large amount of land waste is locally deposited, instead of being swept off- shore as is the case where unimpeded ocean waves wash and attack the land margin. It is believed that this sketch of reef development takes fuller account of the accepted physiographic history of the adjoining coastal highlands than has been taken by other sketches; and that reef-forming agencies are thus shown to have been, with large probability, so long associated with the development of the off-shore structure of this part of the Aus- tralian continent that they cannot be reasonably limited to the brief period needed for the upgrowth of the present barrier reef from a depth of only 30 fathoms; but it is not intended to imply that the above-outlined sketch of the development of the Great Barrier reef embodies unescapable conclusions, for it is easy to suggest alternative possibilities, one of which is men- tioned below. It is furthermore believed that, in as much as the devel- opment of an off-shore continental shelf along the cooler coast of New South Wales is now contemporaneous with the devel- opment of the Great Barrier reef in the warmer waters along the Queensland coast, the contemporaneous development of these two unlike features, one largely by inorganic, the other largely by organic processes, may have been similarly contemporaneous - each in its own latitudes, for a long time in the past. It is true that this view traverses the opinions of Guppy,* Forbes,t __*H. B. Guppy, The Origin of Coral Reefs, Proc. Vict. Inst., xxiii, 1890, pres O. Forbes, The Great Barrier reef of Australia, Geogr. Jour., ii, 1873, 540-546. 7” W. M. Davis—The Great Barrier Reef of Australia. 349 Vaughan* and others, who seem to have thought that only in recent times could these two unlike features be developed on a single coast line, and that in earlier times an inorganic conti- nental shelf fronted the whole coast; but no good reason has been shown why a continental shelf and a barrier reef can not have been formerly developed contemporaneously along the cooler and the warmer parts of any continuous continental coast, just as well as they are now contemporaneously developed along the Australian margin, and just as well as they are now developed on many discontinuous coasts; hence I am inclined to regard the present juxtaposition of shelf and reef as a normal and not as an exceptional relation, and therefore as a characteristic of the past as well as of the present. As to alternative explanations for the Great Barrier reef:—it is easily conceivable that, until the latest flexure of the coast took place, whereby the coastal lowland was narrowed as in the change from UU’ to VV’, fig. 4, no previous shift of the axis of flexure occurred, as in section C, fig. 1; or if such a shift occurred, the axis of no change of level may have advanced not landward as in section D, but seaward as in section E. In this case, reefs may not have been formed, for they do not as a rule occur along coasts that are bordered by recently formed, uncon- solidated sediments. Then, in the absence of off-shore barrier reefs, the unimpeded waves may have formed an inorganic con- tinental shelf along the northern half of the Australian coast, just as they are now and long have been forming a shelf along the southern half of the coast; and finally after the latest flexure took place, the present barrier reef may have grown up on a continental shelf in the formation of which earlier reefs had taken no part. But in view of the various lines of geo- logical evidence which suggest a large diminution of land areas in the Australasian region during Tertiary time it seems not unreasonable to assume that, on the whole, the earlier flexures of the Queensland coast have, like the latest flexure, involved a westward shift of the axis of flexure and caused an encroach- ment of the sea on the land, as in section D; and in view of the occurrence of greatly uplifted and elaborately dissected coral reefs on the neighboring islands of Australasia, which imply the presence of corals in this region for a considerable period of past time, it seems unreasonable to assume that reefs have occupied the Queensland coast only for the brief period demanded for the formation of the superficial part of the Great Barrier. The Great Barrier reef of today does not in all parts of its extraordinary length rise from the outer margin of its platform, *T. W. Vaughan, The Platforms of Barrier Coral Reefs, Bull. Amer. Geogr. Soc., xlvi, 194, 426-429. 350 W. A. Davis—The Great Barrier Reef of Australia. and this has been taken to show that the platform is an inor- ganic continental shelf and not an antecedent reef plain. But no reason has been adduced to show why a reef should grow up at a moderate distance back from the margin of a smooth continental shelf, and why it should not grow up at a similar distance back from the margin of submerged and almost equally smooth mature reef-plain. On the other hand, reasons for the upgrowth of a young barrier reef at a certain distance back from the margin of a submerged reef-plain are suggested in sectors M and N of fig. 5:—a partly abraded reef-plain might, after rapid submergence, be too deep for coral growth at the outer margin of its abraded platform, but not too deep at the margin of the unconsumed part of the plain. Hence the con- trol of the location of the present barrier reef may well have been the result of other conditions and factors than those which formed the platform that serves as its foundation. There is one way, to which attention has not hitherto been called, in which the inorganic processes now at work in devel- oping ‘the continental shelf of New South Wales may be in the future, and may have been in the past, unfavorable to the development of a mature reef-plain along the Queensland coast. In that part of the coast where the Great Barrier reef of Queensland and the great continental shelf of New South Wales adjoin, the ’long-shore currents are at present engaged in forming extensive sand reefs, which appear to be extending northward. The sand reefs have already, in the relatively short interval since the latest flexure by which the coast was embayed, gained lengths of scores of miles. Hence in an earlier and much longer interval, sufficient for the partial pene- planation of an upflexed coastal belt, similar sand reefs might have extended much farther northward along the coast than they now reach; as fast as they advanced they would certainly kill all the reef-building organisms, and thereafter the waves might be able to cut away the reef. It is, therefore, conceivable that, toward the close of the several still-stand periods in which the lowlands, LL’, QQ’, UU’, fig. 4, were developed, the coral reefs (W’,) X’, Y’, of the Queensland coast may have been encroached upon for a few hundred miles by sand reefs from the south; but inasmuch as the present immature reef has its southern end determined chiefly by temperature, however far its mature predecessors were encroached upon by shore sands, it is quite possible that earlier young reefs (W,) X, Y, may also have been able in their youth to extend as far south as their successor of today. But this transcendental aspect of the Pe lem need not be pursued further. A. P. Ooleman— Wave Work as a Measure of Time. 351 Arr. XXIX.— Wave Work as a Measure of Time: A Study of the Ontario Basin ; by A. P. CoLteman. TueE Pleistocene of the Toronto region has been studied as long and as carefully as that of any part of North America, and the work of waves on lake shores, ancient and modern, and of rivers entering the successive lakes in the Ontario basin has been of great importance in determining the geological rela- tionships. Its study may be said to have begun with the visit of Sir Charles Lyell to the raised beaches (Iroquois) north of the town in 1842 ;* and to have been continued by the distin- guished engineer, Sir Sandford Fleming, who in 1850 described the growth of Toronto island by the transport of materials from Scarboro’ Heights,t and in 1861 gave an excellent account of a part of the Iroquois beach to the north of Toronto.t In 1880 Dr. Spencer aud Dr. Gilbert showed that the beaches of the ancient lake were deformed, and the appropriate name of the Iroquois Water was given to the predecessor of Lake Ontario by Dr. Spencer.§ Since then various papers have been pub- lished by the present writer on matters connected with Lake Iroquois and other lakes and rivers which formerly played a part in the history of the region, the latest in 1913.| In this paper an attempt was made to work out the age of Lake Ontario from the rate of recession of the Scarboro’ cliffs and the distance they had receded, additional, though less reliable, evi- dence being deduced from the growth of Toronto island. The results thus obtained have recently been criticised by Dr. Spencer in an article on the “Origin and Age of the Ontario Shore Line.” Dr. Spencer’s pioneer work on Lake Iroquois was so good that his views on the history of the Ontario basin will naturally receive careful consideration. The question of the age of Lake Ontario is of so much interest that a further discussion of the matter is desirable, since the difference between 8,000 years, as determined by myself, and 2,000, as worked out by Dr. Spencer, is far too great to be accidental. In reading his paper carefully it appears that through lack of personal knowledge of the region he has made some incorrect assump- tions which lead to wrong conclusions, and it is proposed to restate the problem and show how the errors occurred. * Travels in North America, vol. ii, pp. 103-8. + Jour. Can. Inst., 1834, pp. 107-223. tIbid., 2d Series, vol. vi. pp. 247-253. SThe Iroquois Beach, Trans. Roy. Soc. Can., 1889, p. 121, etc. ; and other publications. | An Estimate of Post-glacial and Interglacial Time in North America, Congrés Geologique, XIIe Session, Compte-Rendu, pp. 435, etc. 4] This Journal, vol. xliii, No. 257, pp. 351-362, May, 1917. 352 A. P. Coleman— Wave Work as a Measure of Time: The most striking physiographie feature in the vicinity of Toronto is Scarboro’ Heights to the northeast of the city, extending as bold cliffs of stratified clay and sand capped with bowlder clay, for a distance of nine-and-a-half miles in a nearly straight line, and rising for a short distance near the center to 350 feet above Lake Ontario. These cliffs have been studied by the present writer every year since 1890, and the undercut- ting by waves and the land slips occurring in the spring give clear evidence of recession. During the time of observation the shore has receded in a nearly straight line, and the most satisfactory mode of determining the rate of recession would be to compare the results of careful mapping of the shore, repeated after a lapse of years. Unfortunately no such map has been made, the only accurate measurements available hav- ing been carried to the top of the cliff and not to the shore. However, it has been found in more than 20 years of observa- tion that the slope of the cliffs has remained very uniform, so that results obtained from the edge of the heights must corre- spond closely on the whole with the recession of the shore beneath. It has been noted also that, with the clearing of the forest ravines due to rain, erosions have become more frequent and are cut backwards more rapidly than formerly. In some eases such rain gullies have advanced more than fifty feet inland in a single season, and it is evident that only the actual edge of the cliffs, corresponding to the nearly straight shore beneath, can be taken into account in determining the rate of recession. The first survey of Scarboro’ was made in 1792, as shown by plans at the Crown Lands Department of Ontario, but no accurate work was done until 1862 and *63 when Mr. F. F. Passmore carefully re-surveyed the township, planting corner stones to fix the road allowances and the boundaries of proper- ties. In 1912 I was asked by the City Commissioner of Toronto, Mr. R. C. Harris, to report on the rate of recession of the cliffs, smce the location of a reservoir on the heights was under consideration. On my recommendation Messrs. Speight and Van Nostrand repeated Passmore’s measurements toward the end of 1912, fifty years after the original survey. Beginning at the southwest, the recession at the seventeen points where corner posts were available is as follows: 8, 98, 93, 120, 85, 55, 198, 31, 76, 50, 167, 199, 127, 128, 62, 39, and 89 feet. An examination of the edge to which the measurement was made at these different points showed that the three larg- est recessions, 198, 167 and 199, occurred where ravines were being actively cut, and so should not be included in the com- putation. The first one, of 8 feet, was on low ground at the end of the cliff where little wave work was going on and was left BYy3) A Study of the Ontario Basin. SONNE aes a aL Yo 27 suLoyyw pe aa sBurpunos) igs O Oey ee OF SY FY; / Ory SE o€ st vs eavosel NN KFEaG Oy oy hve 7; svbp? TP f Faroe! A Map of Toronto Region. 354 A. P. Coleman— Wave Work as a Measure of Time: out of consideration also. The thirteen measurements that appear to be normal give an average of 81 feet in the fifty years, which works out to 1°62 feet per annum. The rate of recession having been settled, the next problem to determine is the distance through which the cliffs have been cut back. At first sight this might appear to be almost insol- uble, but a study of the history of the former lakes of the Ontario basin and of the mode of action of waves in such lakes gives quite definite information as to the point where the work began. The earliest lake of which there is positive evidence occurred in interglacial times, when the water stood 150 feet or more above the level of the present lake and a great river flowing from the northwest entered at Scarboro’. This river, an interglacial successor of Dr. Spencer’s preglacial Laurentian river, built a delta of clay and sand out into the lake causing an obstruction to the next advance of the ice. This delay in the motion of the glacier was repeated at each successive advance, and the result was the building of a ridge of bowlder clay and interglacial materials, which begins at the interlobate moraine between the Ontario and Georgian Bay basins and extends with a fairly uniform elevation to Secarboro’ Heights, making the highest point on the whole shore of Lake Ontario. The transformation from a river valley into the highest ridge in the region is an impressive one. That such a long and uniform ridge did not end suddenly but once extended farther into the Ontario basin is in itself probable; and to determine just how far it reached beyond the present shore wili give a measure of the length of time dur- ing which wave work has gone on under present conditions. It is a well-known fact that a cape projecting into a body of water is more strongly attacked than the rest of the shore, and that the materials removed by the waves are shifted in the direction of the greatest reach of the prevalent storm winds and built out across the next bay as a spit. If the bay is shal- low the spit may grow across its whole width and form a bar. If the bay is deep the spit advances only so far across as the lower limit of wave action permits and then bends shorewards forming a hook. Successive hooks are built out into the deeper water forming a platform which rises a few feet above the water and encloses shallow depressions or lagoons. As the work continues and the cape is cut back farther and farther, a gently sloping terrace is carved from it, and at the same time the spit or hook is shifted in a direction inland forming a ter- race also, but in this case one that was first built up from the bottom and afterwards cut down a little by wave action. These two varieties of terrace, though formed in different ways, A Study of the Ontario Basin. 355 join and make a continuous subaqueous shelf along the coast where the process is at work. The under-water forms of the two portions of the shelf may be studied by means of sound- ings, and it is found that the part carved from the projecting shore has a gentle slope to the edye where the attack began, and a steeper slope beyond towards the depths of the lake ; while the built terrace terminates outwards in a much more rapid descent into deep water. The slope just mentioned is usually about at the angle of stability for the materials of which it has been built. Such a shelf consisting of two distinct parts formed in differ- ent ways and having different contours may be recognized under water by soundings, and the edge where the more rapid descent into deep water occurs represents the end of the promontory before the carving of the waves began. Fortunately ancient Lake Iroquois did the same kind of work as Lake Ontario and carried its work to about the same degree of completion, so that one may study these subaqueous earth forms conveniently on dry land. An old shore cliff of Lake Iroquois is well seen at Davenport ridge in Toronto, ris- ing from 50 to 75 feet above a terrace carved from rollin bowlder-clay country, and sloping down for nearly 170 feet to the shore of Lake Ontario. This descent is pretty uniform, and the contours are fairly evenly distributed over the three miles of distance between the old shore cliff and the present water front. The carved portion of the terrace, sometimes veneered with more or less stratified sand, joins on the east the terrace built up of cross-bedded sand of great thickness south of the old Humber bar. These relationships were described by Sir Sandford Fleming in 1864 and a comparison was made between them and the relations of Toronto island to Sear- boro’ Heights.* We are now ready to consider the extent of the recession of Searboro’ Heights by the wave action of Lake Ontario. It will be seen from Dr. Spencer’s map, in spite of the fact that it is taken from an old and very imperfect survey, that the shelf carved from the cliffs extends with a gentle slope for 2 or 3 miles to about 100 feet in depth and then falls off rapidly to greater depths. The southward projection of the contours shown on his map near the middle of the water front is due to - anerror. The edge of the shelf is in reality very uniform as may be seen on the map accompanying my former paper, or the one given here, which has been prepared from the last Hydrographic Chart published in 1916. The point where the carved terrace joins the built terrace, with its steeper slope, is well marked and the extent of the pr omontory removed by the * Journ, Can. Inst., 2nd Series, vol. vi, pp. 247-253. Am. Journ. Sct.—Fourtu Series, Vou, XLIV, No. 263.—NovempeEr, 1917, 356 A. P. Coleman— Wave Work as a Measure of Time: waves measures from 24 miles to about 8 miles. In my earlier account it was given as 24 miles or 13,000 feet in round num- bers, which corresponds with the average distance from shore of the 15-fathom line of the new chart. Dr. Spencer casts this measurement aside and replaces it in his own paper by a line drawn at a distance of 4,000 or 4,400 feet, his only assigned reason for this being that “its slope should corr espond to that of the country between the Iroquois beach and the low shore just west of Victoria Park.” As this slope is on the rapid lakeward descent of an Iroquois gravel bar and not on a carved terrace cut in the clay, such as that south of Davenport ridge a few miles to the west, it is evident that he has not distinguished between the two quite different forms produced by wave-action. As shown before, the carved terrace slopes only 170 feet in 23 or 3 miles, while the built terrace descends much more steeply. As we are dealing with a terrace carved from interglacial clay and sand and the suc- cessive bowlder clays above, it is clear that the slope should correspond, and the more rapid descent 24 miles from shore must be considered the lakeward end of the promontory. Thirteen thousand feet divided by 1°62 feet, the rate of annual recession shown in 50 years, gives 8,000 years as the length of time required to cut back the old Searboro’ promon- tory. When the time required to cut back Scarboro’ Heights to its present position had been worked out, it seemed advisable to take up the other end of the problem, i. e., the length of time needed to build Toronto Island out of the sand trans- ported from Scarboro’. For this purpose an estimate was made of the amount of sand deposited northeast of a pier at the eastern channel into Toronto bay during the thirteen years since it had been constructed. The annual increment of sand against the pier was estimated at about 42,000 cubie yards, and the bulk of the island was roughly put at 337,000,000 ceubie yards, which works out to a little over 8,000 years. Owing to the short number of years during which sand had accumulated east of the pier and to other uncertainties no great stress was laid on this computation, though it seemed to corroborate the conclusion drawn from the Scarboro’ recession. For some reason not very clear to me, Dr. Spencer has reversed things and assumed that my time estimate depended on the growth of the island rather than the recession of the cliffs and states that ‘‘ the enormous mass of sand deposited here is of more than local interest, having given rise to fatal chronologi- cal speculations.” He proceeds to demolish my line of argu- ment by stating that only the upper 30 feet of the island belong to the Ontario beach, the materials beneath being delta deposits from the river Don. A Study of the Ontario Basin. 357 Toronto island is one of the most characteristic examples of wave-work on the Great Lakes, its form both above and below water agreeing perfectly with those of a succession of hooks built into deep water. The only evidence Dr. Spencer gives of its delta character is the small amount of clay occurring interbedded with sand; and this is easily explained when one has seen the waters opaque with clay for miles from shore dur- ing an easterly storm. This drifts right past the island at present and must have been deposited in earlier days over the growing shoal in advance of the island. That the Don has deposited delta materials to a depth of 80 feet or more in an old deep channel beneath the present Don flats is proved by wells sunk near its mouth, but its rather insignificant delta ends a mile northeast of, Toronto island where marshy shoals occur in Ashbridge’s bay. The whole of the small delta is northeast of the pier against which sand is accumulating and the main structure of the island is clearly due to the southwesterly drift of sand from Scarboro’. Why should the little river Don build a huge hook-shaped delta when the much larger Humber and Rouge rivers flowing through the same type of country have done nothing of the kind? There is no hook near the mouth of the Rouge or the Humber because there was no great promontory of sand and clay just to the east from which the storms could obtain build- ing material. This is the only difference between them and the Don. ; If Lake Ontario hes lasted at least 8,000 years it is probable that Lake Iroquois, of about the same size and with shores of abont the same maturity, lasted in the neighborhood of 8,000 years, also; though an argument from analogy must of course be received with caution. The lapse of time between the end of Lake Iroquois and the beginning of Lake Ontario is more uncertain. The thawing away of the ice from the St. Lawrence region, after it had sunken so far as to allow the waters to drain past the Adiron- dacks, seems to have gone on rapidly, since no prominent shore lines are found between those of Lake Iroquois and the marine beaches formed when the St. Lawrence valley became free from ice. Again, the length of time during which the Ontario basin was below sea-level, forming a fresh water extension of the Gulf of St. Lawrence, is not certain. The old marine shores are much less mature than those of Lake Iroquois and of Lake Ontario; which would seem to imply a much shorter time in their construction, perhaps not more than half as long, say 4,000 years. How long a time was required to elevate the outlet of the basin and cut off Ontario from the sea, and after- wards to raise the outlet until the water was ponded back as 858 A. P. Coleman— Wave Work as « Measure of Time: far as the Toronto region, so that wave work could begin at Scarboro’, is very hard to estimate. The guess hazarded in m last paper, that the events occurring in the interval between the . two lakes might require about 8,000 years, is objected to by Dr. Spencer, who thinks the time much too short. In this he may be right, in which case the 8,000 years suggested should be looked upon as a minimum; and the whole period since the ice left the Ontario basin may be considered to have been at least 24,000 years, but probably somewhat longer. There are two other points of interest raised in Dr. Spencer’s paper, that of the constancy of level in Lake Ontario in recent times; and that of the limit to be assigned for the end of the ice age. These may be discussed briefly. He suggests that “researches as to the great accession of water to the Niagara river show that the “terrestrial tilting about the northeastern angle of Lake Huron occurred as late as 3,500 years ago. This earth movement extending to the St. Lawrence river was that which gave birth to the great modern river itself. No appreciable deformation has since occurred. Consequently Lake Ontario is found to be some 3,500 years old.” The fact that all the streams flowing into Lake Ontario near the southwestern end have dead water for two or three miles while their courses are rapid above this point, strongly sug- gests that the water has been backed up by the differential elevation of the outlet of the lake; and the further fact that the lagoons behind the meanders are not yet filled up with mud from the spring floods, nor with peat or other vegetable matter, suggests that the change of level has been much more recent than 3,500 years. The other point is raised in the following rather curious Ila TMeT aa. bag ‘the professor has shown elsewhere that Lake Iroquois was a glacial lake, consequently his 8,000 years as the age of the Iroquois beach must be taken away from his post-glacial time, leaving 16,000 years. Such @ priore philoso- phy leaves a suspicion that its author had some speculation to support, but the analyses of the data show that a confusion is thereby thrown into the problem of geological time when he had within his grasp the material for a lasting scientific contribu- tion of great value, and if the confusion be not expunged such must lead to the retardation of scientific research.” The suspicion suggested that I had “some speculation to support” need not be replied to, but the question as to when one geological period ends and another begins is of consider- able interest. When did the Glacial Period end? When the ice began to disappear, when it had half disappeared, or when it had wholly disappeared? If the last assumption is made the A Study of the Ontario Basin. = 859 ice age still continues, for there are still small glaciers iu Labrador, larger ones in the Rocky Mountains, still larger sheets in Alaska and the Arctic islands, and an ice cap in Greenland. Can the question be settled by a reference to climates, the Glacial Period implying Arctic conditions and ending when the climate became cold temperate? If that is the case the Glacial Period ended for the Ontario region at the begin- ning of Lake Iroquois, for large spruce and tamarack trees grew at Hamilton while the lake was still young, the beaver, the bison and the wapiti, as well as extinct mammals, inhabited its shores before the great gravel bar in front of Dundas marsh had been built, and unios, campelomas, plueroceras and spheriums like those in the waters of Lake Ontario occur in an Iroquois gravel bar at Toronto. | The climate was not Arctic but cold temperate on the shores of Lake Iroquois and the shells and other remains found in the old beaches of Lake Agassiz and of Lake Algonquin in the more northern parts of the province of Ontario point to similar climatic conditions, since they are the same as live in the present waters of those regions. In spite of the presence of a stagnant, slowly thawing mass of ice, reaching from the Adirondacks northwards and block- ing the waters of Lakes Iroquois and Algonquin, the climate was not Arctie any more than that.of southern Alaska, with great ice sheets not far off, is Arctic at the present day. Probably the most satisfactory account of the close of the Gla- cial Period is to make it a progressive event beginning in lowa and adjoining states thousands of years before it reached Ontario and taking place on the shores of Lake Iroquois thou- sands of years before it reached James Bay and Hudson’s Bay and the last remnant of the continental ice sheet finally disap- peared. To the stratigrapher striving to establish sharp boundaries between formations this account of the slow and progressive ending of one geological period and beginning of the next may appear unsatisfactory ; but in reality most physical changes on the earth have been gradual, spreading from point to point, from region to region, requiring many thousands of years for completion. It is only when looking back upon them from far off that these events appear sudden and instantaneous for the whole world. A discussion therefore as to the precise year in which the Ice Age ended is not of much value since it really ended at different times in different places. University of Toronto, Toronto, Canada. 360 TZ. D. A. Cockerell— Arthropods in Burmese Amber. os XXX. —Arthropods in Burmese Amber; by T. D. A. CocKERELL. In the August number of this Journal for 1916, I recorded the occurrence of insects in Burmese amber, obtained from beds of Miocene age. Mr. Swinhoe has since sent additional material, and the species found appear on the whole remarkably primitive. They are, indeed, related to living forms; but in practically every case to precisely those forms which we have thought of as ancient, as remnants of a very old fauna. So far, no ants, bees or wasps have been found; the Hymenoptera are Evaniids and Bethylids. The Hemiptera so far examined belong to the archaic family Enicocephalide. It is too early to express a positive opinion, but it is difficult to avoid astrong suspicion that the amber, though found in Miocene elay, is actually very much older, conceivably even Upper Cretaceous. The species described in the present paper all come from a very large block which has been cut into slabs; they therefore all lived at the same exact time and place. PSEUDOSCORPIONES. Hlectrobisium new genus (Obisiide). Cephalothorax remarkably long and narrow ; apparently no eyes; pedipalps very long and slender, with large trochanter, stout femur, very slender tibia, and long slender hand with bulbous base; first pair of legs with fémur very stout, the others ordinary, but all the legs quite long ; abdomen broad and rounded, the scutes entire; body and appendages not hairy. Type the following. Electrobisium acutum new species. Wood brown; abdominal segmentation marked by darker lines. The following measurements are in microns: Length 1040 ; width of abdomen 530; pedipalp with femur 304 long, tibia 256, and hand 3868; first legs about 496 long, second the same, third about 544, fourth extending about 770 beyond margin ot body. Burmese amber, from R. C. J. Swinhoe. In the same slab as type of Hlectrofenus, but on opposite side, and close to edge of thickest part. This is quite unlike the pseudoscor- pions described from Baltic amber, though there is a slight superficial resemblance to Chelifer ehrenbergit. THYSANURA. Lampropholis (?) burmiticus new species (Lepismatide). Male. Length 3™™ (not counting eerci), parallel-sided, the form asin Lampropholis dubia Koch and Berendt (from Baltic ? | } T. D, A. Cockerell—Arthropods in Burmese ‘Amber. 361 amber), but body apparently without scales; stili can only be seen at apex of abdomen, and certainly are absent from the middle segments. In the following descriptions the measure- ments are all in microns: Antennee as usual in the family, with short setee; maxillary palpi 5-jointed, the, firss joint short, the others measuring, (2) 96, (3) 88, (4) 96, (5) 160 ; width of head about 400; width of thorax in middle about 640; width of abdomen about 528; long slender hairs project from sides of thorax as in Silvestri’s figure of Lampropholis, only they are more numerous; coxe and femora stout, the legs essentially as in Lepisma, the only noteworthy feature being the extremely oblique end of first tarsal joint; hind tibize 304 long, their tarsi 820; cerci about 1440 long, with fine bristles, and three very long bristles, more or less broadened apically, placed about 320 apart ; caudal stili about 240 long, not counting the apical bristle, which is about 65; Inner process of caudal subcoxe long and pointed, its length about 160. Burmese amber, from R.C. J. Swinhoe; in the same piece as the type Alectufenus, and 28"™ from it. This may have lost all its scales, but if it was a scaly form, it is surprising that no scales are to be found, even in the amber round about it. Lepidothrix, from Baltic amber, is without scales, but is in other respects very distinct from our insect. Possibly the Burmese species should be placed in a new genus. HEMIPTERA. Dispherocephalus new genus (Enicocephalide). Related to Hnicocephalus, but differing thus: head strongly constricted behind eyes, so that there is a slender neck before the bulbous expansion; thorax long and relatively slender, much longer than wide, with a marked posterior constriction ; wings (the specimen apparently adult) not developed, repre- sented by pads; rostrum long, in specimen seen extending straight out from head; antenne long and very slender; anterior tibize (shown greatly foreshortened in figure) broadened and angularly produced apically, much as in Hnicocephalus, but their tarsi long and narrow, with two slender straight claws, one much longer than the other; hind femora very stout, with an obtuse angle beneath. The head and thorax are delicately hairy; the antennze are not hairy. Type the follow- ing. | Dispherocephalus constrictus new species. Length (to end of extended proboscis) 5™™; the following measurements are in microns: Eyes to end of first antennal joint 640 (it is not possible to see quite clearly where the joint 362 T. D. A. Cockerell— Arthropods in Burmese Amber. begins, but 400 of this is antenna, at least); second antennal joint, 750; third, 576; fourth, 480; width of thorax about 496; eyes to end of rostrum about 960; width of abdomen about 1070; anterior legs with long claw 192, short 128; length of middle tibia, 960; middle tarsus, 336; width of hind femur 208. The claws of middle legs are ordinary, slender and simple. It is assumed that the antennz are four-jointed, but the region at base is obscured, and there may be a small basal joint. Burmese amber, from R. C. J. Swinnoe. In the same slab as the type of Cryphalites, and 22™™ from it. Lat first thought this singular insect must be a Reduviid, but although the wings are not developed, the structure indicates the primitive family Enicocephalide. Owing to the inclusion of air, the details of the dorsal surface cannot be clearly made out. Phthirocoris Enderlein, from the Crozet Is., is apterous, but is in other respects very different. The following key separates the principal genera of this group : Wings not developed; anterior tarsi with two claws..------ a Wings. developed:... a2: sii Geen. eyes od: aa) er 1. Antenne very short and stout; rostrum short and stout; middle and hind legs with 1- jointed larsiss: | + CH,.CO.ONa: CH..O;CO-CH. CH,OH The excess of sodium hydroxide was then titrated with deci- normal hydrochloric acid using phenolphthalein as an indi- eator. From results obtained the ester was shown to be 99°38 per cent pure. The use of anhydrous copper sulphate as a dehydrating agentt in the esterification of certain hydroxy-acids has been previously described in the literature. In the present investi- gation this dehydrating agent is used for the first time in the esterification of polyhydric alcohols. Another point worthy of mention in connection with the preparation of this compound is the low boiling point given by the various investigators. Atkinson gives 182°, De Mole 180°-182°, and Lourengo 180°, as the boiling point of their * Loc. cit. + Bogojawlensky and Narbut, Ber. d. d. Chem. Gesell., xxxviii, 8344, 1895. Clemmenson and Heitman, Amer. Chem. J., xlii, 319, 1909. Dean, this Journal, xxxvii, 832, 1914. Drushel, ibid., xxxix, 114-117, 1915. Substituted Aliphatie Alcohols. 375 respective products. Glycol diacetate boils at 186°-187° and ethylene glycol at 197°. From the molecular constitution of the monacetin of ethylene glycol we would naturally expect its boiling point to lie between those of ethylene glycol and glycol diacetate. The boiling point of the main portion of the product obtained by direct esterification was distinctly higher than that given in the literature, and lies between that of the glycol and of the diacetate. The @-methoxy-ethyl acetate was prepared by treating the 8-methoxy-ethy] alcohol with the theoretical quantity of acetyl chloride. The ester, boiling at 144°-145°, was purified by fractional distillation. This ester has been previously pre- pared* by treating the corresponding alcohol with acetic anhy- dride. The @8-methoxy-ethyl alcohol was obtained for the prepara- tion of the 6-methoxy-ethyl acetate by preparing monosodium glycollate, and treating it with the theoretical quantity of methyl iodide under suitable conditions according to the method of Palomaa.t <04)520"s5 0:1010 U°1000 071018 0°4817* 0:4811 +0:°0006* 3x0°4 2045 The results of the experiments recorded go to show (1) that the use of an alcoholic liquid saturated with the substance to be precipitated is unnecessary to the attainment of good analyti- cal results; (2) that it is practicable to so restrict the volume of the washing liquid (97 per cent alcohol containing 0:1 per cent of perchloric acid) that the solubility of the precipitated perchlorates is insignificant for practical purposes; (3) that a single evaporation with a moderate excess of perchloric acid (0°1 em* for every 071 grm. of salt) is not sufficient to convert considerable masses of alkali chlorides (e. g., 0°3 grm.) com- pletely to perchlorate, and that in such a case the residue of the first evaporation with perchloric acid should be dissolved in the least amount of water, another portion of perchloric acid added, and the evaporation repeated; (4) that in separations of the larger amounts of insoluble perchlorates (0°3 grm.) from sodium perchlorate the residue left after digestion of the nearly dry mass of perchlorates in the washing liquid and decanta- tion should be dissolved in a small amount of water and the process of evaporation and extraction repeated; (5) that, in the case of rubidium at any rate, digestion of the residue for fifteen or twenty minutes with the washing liquid before effecting the transter is advantageous. It is to be noted that the evaporation of large amounts of perchloric acid in glass may result in a considerable action upon the glass and it has been found that perchloric acid which has stood a long time in glass may yield an appreciable residue on evaporation. * Upon dissolving these residues, again precipitating, and weighing, errors found were —0:0004 grm. and +0°0004 grm. respectively. L. D. Burling—Protichnites and Climactichnites. 387 Art. XX XIV.— Protichnites and Climactichnites; A COritr- cal Study of Some Cambrian Trails ;* by Lancaster D. BURLING. Jupcine the nature of the maker of a trail by peculiarities in its composition may be difficult and the solution false—for example, the writer has watched the larvae of the common may-fly crawling along the mud on the tidal flats of the St. Lawrence and leaving a perfectly smooth sinuous trail or groove which would naturally be associated in the mind of almost anyone with the work of a worm, certainly nothing with the legs of a may-fly larva. That the Upper Cambrian sea was peopled by animals of large size is well known, but the trails upon which this inference is based have so far failed to indicate the true nature of their makers. Indeed they have been the subject of frequent and widely variant conjecture. A critical study of some of the trails in the Cambrian has yielded conclusions so substantial or so different from those in the literature that they appear to be worthy of record. PROTICHNITES. The trails to which this name has been applied were referred to the agency of a tortoise by Owen’',f who later’ assigned them to the work of a crustacean tike Zemulus. In this view he was followed by Dawson* and Dana.’ Dawson later’ assigns them indubitably to the work of crustaceans, but lessens the weight of this reference by suggesting that Climactichnites may have been made by the same animal. With the exception of Chapman,’ who suggests that both Protechnites and Clomac- tichnites are of fucoidal origin, succeeding authors, beginning with Billings in 1870,’ have referred them to the work of trilo- bites. Packard*® thinks they could “ perhaps have been made by the extremities of the feet of a small shrimp-like creature.” Later’ he questions the ability of Paradoxides to make the trail, a question first raised by Dawson.” Walcott*' unhesitat- ingly states that they “ were made by trilobites of the genus Dicellocephatus.” Let us look at the trails themselves and see whether or not their critical study may not yield results of tangible value in the identification of their makers. Protichnites (see fig. 1) is characterized by two rows of footprints paralleling a median groove. They have been found on Upper Cambrian sandstones in Ontario and New York. The trails give us several clues as to the animal which made them, and these facts * Published by permission of the Deputy Minister of Mines, + For references, see the literature at end of article. Am. Jour. Sc1.—FourtTH SERIES, VoL. XLIV, No. 263.—NovemseR, 1917. 27 388 L. D. Burling—Protichnites and Climactichnites. and the inferences they support follow, using the same nota- tion in each case. Facts: (a) the side tracks are frequently three-toed ;” (0) these trifid tracks usually toe in ; (¢) the side tracks are usually 2 or 3 inches apart’® though trails up to 5 or 6 inches across have been observed ; (@) the trails are not straight, and a single trail has been observed to reverse its direction entirely so that the animal moved off in a direction parallel to but opposite to that of its previous track, all in a distance of less than three ies 2b Upper Cambrian Trails. Fic. 1. Protichnites loyananus Marsh x1/6. (After Walcott.) Ausable Chasm, Ni > Us os Nat, Mus: ; times the width of its track, the sharpest curve observed havy- ing a radius but little more than half the width of its track; (e) some of the median grooves are double and very sharply incised, others on the same slab betray no doubling, yet the width of the single groove closely approximates the distance between the double tracks; (7) the median groove does not swing to the side when the trail makes a turn, even on the sharpest curves the median groove lies midway between the L. D. Burling—Protichnites and Climactichnites. 389 side leg tracks; (g) in one trail the median groove is only impressed at intervals, but these are regular and occur 26 times in a horizontal distance equal to 25 times the width between the tracks; (A) where the trail crosses a ripple marked surface all traces of ripple mark are obliterated for a distance a little wider than the extreme width of the trail, and the feet tracks are large and coarse with the median groove deeply incised ; (7) on a surface adjacent to that showing g occurs a trail simu- lating Protichnites, but without the median groove and with the feet tracks small and sharply impressed ; (7) the side rows of leg tracks are not arranged along a single straight line, but appear to be more or less double; (#) where the median groove is deep the side tracks are proportionately deep ; (/) where the median groove is only marked at intervals, as in g, the impres- sions of the legs betray a tendency to be arranged in slightly curved lines concave toward the center, with the crests about as far apart as the intervals dividing the impressions of the median groove and more or less opposite to these impressions ; (m) the number of leg impressions was counted in two places on the trail mentioned in g; in one where a group of 9 median groove impressions was available 65 leg impressions occurred on one side, 63 on the other; in another group of 6 the number of leg impressions was respectively 88 and 40; each of. these groups was crossed by another trail, and the number of legs may be greater on this account. Inferences: (a) Some of the appendages used by the animal in walking were three-toed ; (0) the animal toed in, and toeing in is usually characteristic of heavy low-lying bodies whose feet touch the ground well toward if not beyond the sides of the body; (ce) if the inference in 6 is correct, the animal was neither wider nor narrower than the track and individuals ranged in size from 2 to 6 inches in width ; (d@) its body was either extremely flexible or else short and more or less circular in outline; (e) the animal usually (see 2) did not carry the entire weight of its body on its legs, but allowed a median por- tion to rest on the bottom and this portion was apparently forked in some, club-shaped in others—perhaps a sexual differ- ence ; (7) the part of the body which rested on the bottom was not the telson of a Zimulus-like crustacean or trilobite, but was a process situated somewhere between or very close to the legs ; (g) the animal was able to bear almost its whole weight (all, if 2 was made by the same animal) on its legs, but where its median portion did just graze the ground it did so once for every time it moved forward throngh a distance equal to its own width; this corroborates d in indicating the general cor- rectness of making the animal round or oval in outline; (A) the animal was heavy, and its legs were comparatively short 390 L. D. Burling—Protichnites and Climactichnites. and sank deeply into the bottom; (2) the ends of the legs were more or less pointed and could only support the entire animal when walking on a hard bottom (¢ may have been made by a different animal) : (7) the legs were not all of the same length ; (k) the question of whether the median portion touched the bottom or not was apparently one of whether or not the bot- tom was soft enough to allow the legs to sink in, though it must be recorded that the trail described in 7@ is 6 inches or more across and may have been made by a very large and per- haps strong form ; (/) the front and back legs were respectively shorter than those in the centre; (m) in making this trail, g, the animal was apparently skipping along with the body sup- ported in the water, and the impressions of the feet are proba- bly not confused. If this inference is correct, and remembering that the trails are crossed by others, the number of pairs of legs was in all probability 6, though it averages nearly T. CLIMACTICHNITES. Logan, the first describer of these trails, believed them” to be the work of molluscs, a suggestion which received support as late as 1903 when Woodworth” published the first illustra- tion of the peculiar oval bodies which have been found at one end of the trails at Mooers, New York. Walcott’* has recently figured a similar oval body from the Upper Cambrian at New Lisbon, Wisconsin. He refers them to the work of annelids, a reference which was anticipated by Gratacap in 1901.’ Curiously enough several of the earlier writers believed Climac- tichnites and Protichnites to be different expressions of the trail of the same animal, an observation which received experi- mental confirmation at the hands of Sir Wiliam Dawson’ who discovered that when walking on the bottom the horse-shoe crab used its legs and made a trail like Protechnites, but that in shallow water just covering the body it propelled itself by moving its abdominal gill plates and left a trail resembling Climactichnites, (a) “except that the oblique furrows made by the legs between the median and lateral ridges are directed in the reverse direction”; (6) ‘except that in the track of Limulus the lateral and median lines are furrows instead of ridges.”**° Jones” believed them to be the flattened galleries of burrowing crustaceans, and Grabau” in 19138 suggests that the oval bodies of Woodworth may be collapsed burrows. Dana,” Billings,* and Packard” believed they were to be ascribed to trilobites. Todd*® coucludes that the animal was provided with a rigid caudal shield, with bristles or slender spines, and that the ambulatory organs leaving the last impressions were very perfectly flexible and must have been in pairs, each capa- ble of motion independent of. its fellow. Hall’ says that “the L. D. Burling—Protichnites and Climactichnites. 391 markings under consideration do not appear to have been made by an animal provided with free movable limbs, or otherwise with very short limbs, without the acute appendages belonging to Limulus.” Patten* was the first to suggest an Eurypterid origin, “ the abdominal gill plates making the rhythmic ridges in the sand.” Grabau and Shimer” assign the trails to the work of “some unknown terrestrial or semi-terrestrial animal.” Authors are generally agreed that the oval bodies represent the end of the trail; thus Woodworth” has suggested that they EiGue Upper Cambrian Trail. Fie. 2. Climactichnites youngi (Chamberlin), 5/6 nat. size. (After Wal- cott.) New Lisbon, Wis. U.S. Nat. Mus. (The front end of the trail is toward the bottom of the page.) represent the end of the trail, and Eastman” states that “ the animal, if an Kurypterid, moved toward the sedentary impres- sion and not away from it.” Todd,” who apparently did not have the opportunity of observing the oval bodies, records his belief that the apex of the V-shaped impressions points for- ward, and Walcott™ speaks of the oval bodies as terminal and (p. 284) of the forward-curving transverse furrows made by pressing the beach-sand backward in creeping.” The observa- 392 L. D. Burling—Protichnites and Climactichnites. tions of Kishinouye® indicate that the apex of the V-shaped impressions in the somewhat similar track of Limulus points forward. As already mentioned, however, Packard records” the fact that the oblique furrows of Limulus are directed in a direction reverse to that of the ridges of Climactichnites, and Patten,*’ in comparing the tracks with those of Limulus, states that the tracks showed a beginning in a hollow in the sand, and thus corresponded to those of Limulus “ which remains buried on recession of the tide and upon its first return crawls and then swims away.” Grabau*"* mentions Climactichnites-like trails in the Silurian, which may have been produced by eurypterids with bilobed telsons, myriopodus types, or insecta. Climactichnites may be characterized as consisting of a series of more or less transverse subparallel ridges bounded on either side by a lateral ridge. They have been found on Upper Cambrian sandstones in Ontario, New York, and Wisconsin. Let us examine the trails critically to see whether or not they speak for themselves, dividing fact from inference as we did in discussing Protichnites. Facts: (a) the lateral ridges may be almost absent or may be very coarse, in which case they are regularly swollen at intervals equal to the distance between the transverse ridges, and each swollen portion appears to merge at one end into au adjacent cross ridge; (0) the lateral ridges vary from 14 to 43 inches apart in specimens from Wisconsin, but average 4 to 6 inches apart in specimens from New York and Canada; (¢) the transverse ridges are usually arched or V-shaped but they are frequently very irregular, even sinuous or double bow-shaped, and the angle of the V varies within wide limits; the apex of the V is not always symmetrically spaced, betraying a general tendency to swerve to the outside on curves, but being irregu- larly disposed even on tangents ; (d) there is more or less inter- ruption of each ridge at the apex of the V,so much so that the line connecting the apices sometimes forms a slightly marked ridge ; (¢) the transverse ridges are usually equally spaced, but this again varies greatly and the ridges may be small and irregularly spaced ; (7) lying upon the ridged trail in the speci- men from New Lisbon; Wisconsin (see fig. 2), is @ series of very closely spaced almost semicircular raised lines which cross the transverse ridges without interruption or deflection ; (¢) the convexity of the lines mentioned in 7 is directed in the same direction as the apex of the V-shaped arch in the transverse ridges ; (4) Todd® mentions longitudinal lines which are some- times wavy; (2) the trail completely reverses its direction in a distance almost equal to 5 times its width, the sharpest curve observed being one with a radius of little more than one-half L. D. Burling— Protichnites and Climactichnites. 3938 Gao Upper Cambrian Trail, New Lisbon, Wisconsin. Fia. 3. Climactichnites youngi (Chamberlin), 3/4 nat. size. (After Wal- cott.) U.S. Nat. Mus. (As now interpreted the animal moved toward the bottom of the page.) 394 L. D. Burling—Protichnites and Climactichnites. the width of the track; (7) at one end of several of the trails convex oval bodies as wide as, or slightly wider than, the trail and little more than 24 times as long are present (see figs. 3 and 4). Woodworth” figures these as symmetrically rounded at both ends, Walcott’s specimen* shows the outline at only one Fic. 4. Fic. 5. -or ree, a | a WS = Mean) mesg ¢ Upper Cambrian Trails, Mooers, N. Y. Fie. 4. Climactichnites wilsoni (Logan), x1/25. (After Clarke.) State Museum, Albany, N. Y. Fie. 5. From a photograph of a cast of the slab shown in fig. 1, in the Brooklyn Museum. The animal that made the trail is now believed to have moved away from the oval body end. end, but the slab in the Museum at Albany,” of which there is a partial replica in the Brooklyn Museum (see figure 5), shows specimens with one end rounded, the other (the end toward the trails) unsymmetrically arched outward, symmetrically L. D. Burling—Protichnites and Climactichnites. 395 V-shaped outward, abruptly truncated at right angles to the longer diameter, and even arched inward ; (#) the “ oval bodies” are themselves frequently curved, even broadly S-shaped ; (2) the apex of the V-shaped transverse ridges always points toward the “oval bodies”; (m) the V-shaped transverse ridges may often be seen to extend nearly half way beneath the oval-body ; (n) the trails are nearly always faint and disappear at the end opposite to the one bearing the “‘ oval body.” Inferences: (a) the side ridges were apparently made im the same push that made the transverse ridges, and that both are ridges instead of furrows indicates that the apex of the V-shaped ridges points backward with reference to the line of progress, for this is the only direction in which material could be shoved outward and heaped up along the edges of the trail ; (6) in all probability these figures represent the entire width of the animal; (¢, d, and e) the portion of the animal making the transverse ridges was very flexible and capable of making move- ments differing in amplitude, direction, and form ; the inter- ruptions in the center (d@) are to be expected, and do not require a division of the ridge-forming portion, a view which is corroborated by the wide lateral shifting exhibited by this median ridge; (7) these semicircular raised lines must have been made last or they would have been obliterated or marred, and must indicate the conformation of the back end of the ani- mal; their close spacing would indicate slowness of forward movement or creep ; (g) the apex of the V-shaped ridges there- fore points also toward the rear; (4) probably made, as Todd suggests, by bristles or other portions of the under surface as the animal moved along,—I have not observed them ; (z) the animal, or its ambulatory organ, was very flexible, or else short and more or less elliptical in outline; (7) the evidence seems to warrant us in disagreeing with the concensus of previous opinion (see p. 391), and in supposing these oval bodies to repre- sent the initial resting place ot the animal that made the trails, the round ends, as indicated in @ and f, being the rear and the V-shaped end the front. This front end was capable of being moved from Y-shape forward (ts position in repose) to V-shape backward, and this movement carried the animal along. The convexity of these oval bodies may be explained as follows: If an animal with a very flexible under surface or foot were stranded on the retreat of the tide, scour would obliterate the previous tracks and would reduce the general level of the beach wherever it was not protected from erosion by the disk-like foot, and the edges of this organ would natu- rally be depressed in an endeavor to prevent being washed away. ‘The lens of sand thus enclosed would be left upon the departure of the animal at the approach of the next tide, its 3896 L. D. Burling—Protichnites and Climactichnites. preservation, and that of the tracks made in moving away, being due to fortuitous circumstances; (£) corroborates 7@ in proving the animal or its foot to be extremely flexible; (Z) therefore, if we are right in a, 7, and 7, the apex of the V always points in the direction from which the animal has been moving, not forward as nearly everyone has supposed, Patten™ being the only one to suggest a possible difference ; (7) unless the ambulatory organ or organs occupied nearly half of the under surface this fact alone would prove that the animal moved away from the oval-body end. The preservation of these marks in the portion of the trail where the body must have rested seems to the writer to be explained by supposing the edges of the disk-like foot to be sufficiently extended in repose to protect the last marks made by the animal previous to its rest—the oval body is frequently about one fifth wider than the immediately adjacent trail ; (7) the progressive faintness and disappearance of the trails at the end opposite to that bearing the oval bodies is characteristic of nearly all of the trails; all of those, for example, which exhibit both ends in the specimen at Albany. They certainly seem to corroborate a, f,7, and m, in indicating that the animal started from the oval-body end and rose into the water at the other end, as pointed out by Patten,” and that they could swim. In some trails, notably the one running down the center of the slab in the museum at Albany, the disposition of the ridges is such as to suggest that the animal did move toward the oval body end of the trail. The impressions of the curved margin described under J on a previous page were unknown to geolo- gists until the specimen from New Lisbon, Wisconsin, was figured by Walcott in 1912, and while Walcott adheres to the belief that the New Lisbon animal also moved toward the oval- body end we have endeavored to show that this specimen proves the oval body end to be the initial portion of the trail. The V-shaped ridges in both the Albany and New London specimens point toward the oval-body end, and it is somewhat improbable that the animal should have moved toward that end in the one case and away from it in the other. However, there are certain differences in the trails, and these may be due either to causes dependent on the physical conditions at the time the trails were made or to differences in the animals themselves. The New Lisbon specimen certainly started from the oval body end to crawl away; those in the Albany slab may have come to rest in the manner described by the early observers. Naturally, however, the conclusion that the oval- body end was made last has called forth attempts to explain the disappearance of the animal. L. D, Burling—Protichnites and Climactichnites. 397 Conclusions.—The animals that inhabited the sea some thirty million years ago are known to us with a perfection that is a continual source of wonder, and the discoveries of the past few years in these ancient rocks are little short of marvellous, but that the Cambrian seas were peopled by a host of forms of which we know little or nothing is no less certain than that the waters of pre-Cambrian time were full of life. While certain of these unknown forms offer us nothing more substantial than the record of their reptant efforts, the desire to know is responsi- ble for attempts at their deciphering. The facts are daily becoming more numerous and the inferences surer. That P7ro- tichnites was made by a short, low-lying, and more or less heavy set, approximately 12-legeed crab-like animal, and that Climactichnites was made by the snail-like creep of a flexible slug-like animal which was frequently stranded at low tide, but was able to swim in the waters of the full tide, have passed the stage of guess-work and border on the real. LITERATURE. Quart. Journ. Geol. Soc. London, vol. vii, pp. 250-252, 1851. Idem, vol. viii, p. 224, 1852. Canadian Nat. and Geol., vol. vii, pp. 276 and 277, 1862. Manual of Geology, 1863, p. 185. Quart. Journ. Geol. Soe. London, vol. xlvi, p. 599, 1890. Canadian Journ. Sci., n. ser., vol. xv, p. 490, 1877. Quart. Journ. Geol. Soc. London, vol. xxvi, pp. 484-485, 1870. Proc. Amer. Acad. Sci., vol. xxxv, p. 408, 1900. Idem, vol. xxxvi, p. 61, 1900. 10. Canadian Nat. and Geol., vol. vii, p. 276 and 277, 1862. 11. Smithsonian Misc. Coll., vol. lvii, p. 277, 1912. 12. Smithsouian Misc. Coll., vol. lvii, pls. 46 and 47, 1912. 13. Idem, pls. 48 and 49. Many of the following observations are based on these plates. 14. Canadian Nat. and Geol., vol. v, pp. 279-285, 1860. 15. Bull. New York State Mus., No. 69, pp. 956-966, 1903. 16. Smithsonian Misc. Coll., vol. lvii, pp. 259-262, pls. 38-40, 1912. 17. American Geol., vol. xxvii, p. 89, 1901. 18. Canadian Nat. and Geol., vol. vii, pp. 274-277, 1862. 19. Packard, Proc. Amer. Acad. Sci., vol. xxxvi, p. 66, 1900. 20. Eastman, Textbook of Paleontology, 2d ed., 1913, p. 142 (noted by Dawson). 21. The Geologist, London, vol. v, pp. 188-139, 1862. 22. Principles of Stratigraphy, 1918, p. 1091. 23. Manual of Geology, 1863, p. 185. 24, Quart. Jour. Geol. Soc. London, vol. xxvi, p. 485, 1870. 20. Proc. Amer. Acad. Sci., vol. xxxvi, p. 64, 1900. 26. Trans. Wisconsin Acad. Sci., vol. v, pp. 276-281, 1882. 27. 42d Ann. Rept. New York State Mus., 1888, p. 29. 28. Science, n. ser., vol. xxviii, p. 382, 1908. 29. Index Fossils, vol. ii, p. 248, 1910. 30. Bull. New York State Mus., No. 69, p. 964, 1903. dl. Textbook of Paleontology, 2d ed., 1913, p. 142. 32. Trans. Wisconsin Acad. Sci., vol. v, p. 277, 1882. 83. Smithsonian Misc. Coll., vol. lvii, pp. 259 and 260, 1912. SECO Se 398 L. D. Burling—Protichnites and Climactichnites. 34. It should be noted that these ‘‘forward-curving transverse furrows” are more or less sinuous and bend sharply backward (using the same terminology) at their union with the sides of the trail, which they join at a tangent. Consideration of this one trail alone (pl. 89, fig 2 of Walcott) seems sufficient to prove that the animal must have progressed in a direction exactly opposite to that assumed by Walcott. Under the new interpretation the transverse furrows bend sharply forward at the sides and are bent backward in the center, and are closely covered by the series of curved raised lines left by the posterior margin of the animal (see g of text). 30. Cambridge Nat. Hist., vol. iv, Crustacea and Arachnids, fig. 157, Pp. 274, 36. Proc. Amer. Acad. Sci., vol. xxxvi, p. 66, 1900. 37. Science, n. ser., vol. viii, p. 382, 1908. 37a. Bull. Geol. Soc. America, vol. xxiv, 1913, pp. 463-464. 38. Trans. Wisconsin Acad. Sci., vol. v, pp. 278-279, 1882. 39. Bull. New York State Mus., No. 69, fig. 1, p. 961, 1903. 40. Smithsonian Misc. Coll., vol. lvii, pl. 88,1912. — 41. See Bull. New York State Mus., No. 80, 1905, pl. 3. 42. .Science, n. ser., vol. xxviii, p. 382, 1908. 43. Bull. New York State Mus., No. 80, pl. 3, 1905. 44. Science, n. ser., vol. xxviii, p. 382, 1908. SCIENTIFIC -INT EE ELG HN Cake Il. CHEMISTRY AND: PHYSICS. 1. A New Method for the Recovery of Salts of Potassium and Alumimium from Mineral Silicates——Many efforts have been made to devise methods for extracting potassium from orthoclase feld- spar which occurs in such quantity and purity as to make it a possible source of supply for salts of this element, and the desir- ability of such a process has greatly increased since the time that the supply of German potash salts has been cut off. The methods heretofore proposed for this purpose do not appear to have been successful on the large seale. J.C. W. FRAZER, W. W. Honuanp and E. Miner, of Johns Hopkins University, have now proposed for the purpose a method which appears unusually promising, since comparatively low temperatures are required for the operation, and since, besides the potassium, the aluminium of the mineral may be extracted in the process. The finely ground feldspar is mixed with about 0.8 parts of potassium hydroxide (or an equivalent amount of the sodium compound) and heated for about an hour at a temperature of 275 to 300° C. A reaction takes place whereby practically one-third of the silica of the feldspar is converted into potassium silicate, soluble in water, while the residue corresponds in composition to the min- eral leucite: KAISi,O, -+- 2KOH = KAISi,0, + K,SiO, + H,0 Chenvistry and Physics. 399 Upon treating the mass with water the alkali used in the process goes into solution, largely as silicate. It is then causticized with lime, and, after filtering off the resulting calcium silicate, the liquid is evaporated and the alkali is used again for the treat- ment of feldspar, with a total loss of only about 1%, according to the experiments of the authors, working on a small scale. The residual artificial leucite gives up its potassium very readily to dilute acids, so that the point where the potassium is extracted can be detected by an indicator, such as methyl orange. It is possible, therefore, to extract the potassium as chloride, sulphate or nitrate by the use of the corresponding acids without attacking the aluminium in the compound. The residue, now filtered from the potassium salt solution, resembles kaolinite somewhat in composition, but it is readily decomposed by sulphuric acid with the formation of aluminium sulphate and gelatinous silica. After dehydrating the silica by drying, the two things may be separated by treatment with water and filtration.—Jour. Indust. Eng. Chem., 1x, 935. H. L. W. 2. Electrochemical Equivalents; by Cari HERING and FREp- ERICK H. GeTMAN. 12mo, pp. 180. New York, 1917 (D. Van Nostrand Company).—This little book gives an excellent table of electrochemical equivalents for practically all of the known elements. The international atomic weights of 1917 are used as the basis, and the equivalents are calculated for all the possible valences or changes of valence in each case. The values given are in terms of milligrams per coulomb, coulombs per milligram, erams per ampere-hour, ampere-hours per gram, pounds per 1000 ampere-hours, and ampere-hours per pound. The data just mentioned are given in the principal table, but there are several other tables containing useful information. While the book does not profess to be a treatise on electrochemistry, sufficient explanatory text is given to permit the use of the data without need of reference to other works. There are discussions of fundamental laws and data, the methods of calculation are clearly explained and illustrated’ by numerous examples, the principles of electrolysis and the electronic theory are well presented, while in the appendix the subject of valence and chemical caleula- tions are taken up. The book appears to be a very useful one, not only for the purposes of practical electro-chemical calcula- tions, but also as a reliable and concise source of theoretical information in the field where chemical and electrical sciences are connected. ist gaye 3. A Laboratory Manual of General Chemistry; by WiLLIAM J. Haug. 12mo, pp. 474. New York, 1917 (The Macmillan Company).—This book presents an unusually extensive and advanced course of laboratory work. The experiments are not only very numerous, but there are some rather elaborate quanti- LOO, Sas Scientific Intelligence. tative experiments, particularly near the beginning of the course, and there is also a good deal of work in the direction of the qualitative grouping of elements and radicals. The course of work evidently deserves high praise in regard to its fullness, its instructiveness, and its clear presentation of the operations. It appears, however, that the course as a whole is too extensive for the time that is usually available for such courses, and that the somewhat complicated quantitative work is introduced at such an early stage that the average beginner would very probably lack the manipulative ability and the knowledge to carry them out properly and to comprehend them satisfactorily. However, it may be said that it is easier to omit portions of a too exten- sive book than to add material to one that is too short or too elementary. While practically one-half of the pages of this book are left for the student’s notes, these pages are supplied with printed numbers corresponding to numbers placed in the text of the opposite pages, so that the student may know where to record his observations and answers to questions. This arrangement will facilitate the proper taking of notes, and it will be an aid to the instructor who examines them. H. L. W. 4. A Short Manual of Analytical Chemistry; by JoHN Murer. 6th American Edition, Edited by J. THomas. 8vo, pp. 237. Philadelphia, 1917 (P. Blakiston’s Son & Co.).—This book, which is intended for the use of students of pharmacy, deals with qualitative and quantitative, inorganic and organic analysis. In spite of its moderate size and the wide field that it covers, it gives a surprisingly comprehensive amount of infor- mation. It is a well-known work which has passed through 10 English and 6 American editions. The latter are by no means mere copies of the former, since the American editions are made to correspond with the legal requirements for drugs as fixed by the United States Pharmacopeeia. The book contains so much information about analytical methods that it should be useful as a reference book to all kinds of analytical chemists. H. L. w. 5. Allen’s Commercial Organic Analysis. Edited by W. A. Davis. Fourth Edition. Vol. IX. ‘8vo, pp. 836. Philadel- phia, 1917 (P. Blakiston’s Son & Co. Price $5 net)—This volume of the entirely rewritten fourth edition of this monu- mental work has been issued in order to bring up to date the matter of the preceding eight volumes, especially the earlier ones, since the work of revision was begun in 1907. The articles included are very numerous and naturally vary in length and importance. In many cases the original contributors have fur- nished the revisions, but in some cases others have done the work. An important feature of the volume is a complete general index to the whole series of nine volumes. This will greatly facilitate reference to the work. H. L. W. Chemistry and Physics. «£01 6. The Ionizing Potential of Sodium Vapor.—Several experi- menters have shown that, when the vapors of cadmium, mag- nesium, mercury, and zine 7 vacuo are bombarded by electrons from a hot cathode, a single-line spectrum is emitted, provided the kinetic energy of the electrons does not exceed a certain eritical value. The line appeared when the potential difference involved attained the value required by the quantum relation - and the frequency of the radiation; for example, 4.9 volts for the wave-length 2536.7 in the spectrum of mercury. An extended series of experiments on the vapor of sodwwm has been carried out recently by R. W. Woop and 8S. OKANO. Since they used three or four different forms of apparatus and varied the experimental conditions in many ingenious ways, in order to eliminate the hypothetical sources of error as far as possible, it will not be feasible, for lack of space, to do justice to the experi- mental details of the investigation. By making visual observa- tions with a very efficient Schmidt and MHaensch pocket spectroscope, it was found that the red and green lines of the subordinate series faded gradually as the voltage was decreased and finally disappeared at 2.3 volts. The quantum relation leads to the expectation that the D-lines would vanish at about 2.1 volts. Asa matter of fact the yellow radiation could be detected until the potential difference had dropped as low as 0.5 volt. The lack of agreement between the predicted and observed values may be due to the presence in the cathode stream of a relatively small number of electrons having a much higher speed than the mean value corresponding to 2.1 volts. The authors suggest that this possibility may be tested by separating the electrons magnetically into beams each of which is composed of corpuscles moving with sensibly equal speeds.—Phil. Mag., xxxiv, p. 177, September, 1917. Ease OU. 7. Penetrating Power of X-Rays from a Coolidge Tube.—The paper under consideration was written by RuTHERFORD and it contains an account of some experiments made to determine the maximum penetrating power of the X-rays excited by high voltages in a Coolidge tube, lead being used as the absorbing material. The radiation was excited by a large induction-coil, actuated by a mereury motor-break in an atmosphere of coal-gas, and capable of giving 20 inch sparks. The heating current through the tungsten spiral of the tube was adjusted to give a radiation of maximum intensity at the voltage required. This voltage was fixed by an alternative spark-gap between points. The radia- tion was found to be most constant when a fairly rapid stream of sparks passed between the points. The voltage corresponding to the alternative spark-gap was determined by comparison with the sparking potential between two brass spheres 20 em: in diameter. The ionization current was measured by means of 402 : Scientific Intelligence. cubical electroscopes of the type usually employed in gamma-ray work. For determining the initial absorption, the lead front of the electroscope was cut away and replaced by thin aluminum foil. In cases where greater thicknesses of absorber were neces- sary, lead electroscopes having sides respectively 3 mm. and S mm. thick were employed. The absorbing lead screens were of much larger area than the face of the electroscope and, since they were placed close to the front of this instrument, the greater part of the radiation scattered in a forward direction by the absorber entered the electroscope. The experimental data are tabulated in foar columns which ceive respectively, the maximum voltage (79,000 to 196,000 volts), the range of thickness in lead (0.7 mm. to 10.0 mm.), the absorp- tion coefficient » (27 to 8.5 em.), and the mass absorption coeffi- cient p/p (2.37 to 0.75). The following facts are brought out by this table. In the first place, the thickness of lead through which the radiation was measurable increased with the voltage applhed. This was due not only to the inerease in the penetrating power of the radiation but also to the large increase with voltage of the intensity of the radiation. At 196,000 volts the radiation was detected and measured after passing through 10 mm. of lead. The intensity had then fallen to less than the one-millionth part of its initial value. Again, for the end radiations, » does not change very much between 79,000 (u = 26) and 144,000 volts (u = 22), and between 105,000 and 144,000 voits » remains con- stant. Within the latter range of voltages the radiation is absorbed nearly exponentially with a value of » equal to 22 em.*+ Above 144,000 volts the absorption is no longer exponential, but the value of » decreases progressively with increase of thickness of absorbing layer. For example, at 183,000 volts » decreases from 26 to 12 as the thickness of the absorber is increased from 0.7 mm. to 7.0 mm. These apparently peculiar results are readily explained by taking into account the characteristic absorption band of lead. By assuming that the portion of the sraph (2 = log p/p, y= log A; A= wave-length) corresponding to wave-lengths less than those of the absorption band is an approximately straight line parallel to the segment of the locus on the longer wave-length side of the band, the author extrapo- lates to the value » = 5, for the minimum value associated with 196,000 volts. The experimental value was found to be 8.5. Since this last number is known to be too large and as the value 5 was obtained from the data of other observers, the author concludes that his results are not inconsistent with the quantum relation e V=h/Amin This equation has been shown by Hull and Rice to hold up to 100,000 volts and probably as far as 150,000 volts. Rutherford’s investigation therefore confirms their work and extends the range of validity of the quantum Chemistry and Physies. 403 equation. A few data are also given for the absorption by aluminum of the end radiation after passing through iron. The observations on the absorption of X-rays in aluminum and lead throw important light on the difficult question of the prob- able wave-lengths of the penetrating gamma rays from radio- active substances. The line of argument may be suggested by the following brief account. The numerical data for X-rays obtained by Rutherford and by Hull and Rice are collected in the first six horizontal lines of a table. The first and second columns contain respectively the voitages (84,000 to 196,000) andthe corresponding shortest wave-lengths (0.147 A to 0.063 A), conformably to the equation H—=—hv. The values of the mass-absorption coefficient (u/p) for aluminum and lead are entered in the third and fourth columns. The seventh line pertains to the penetrating gamma rays from radium C. In this line the first and second spaces are to be filled in, while the third and fourth contain the values of p/p given by Ishino. A discussion of all the data in the last two columns leads, by extrapolation, to a general estimate of the voltage required to excite the gamma rays, and from this datum together with EH = hy the order of magnitude of the wave-length of these rays is obtained. The following quotation contains in concise form the final conclusions. ‘‘In our present ignorance of the law of variation of »/p with frequency in this region of the spectrum, it is only possible to estimate the actual wave-length of the most penetrating gamma rays. It is clear, however, tha* the waves are at least three times and may be ten times shorter than those which correspond to 200,000 volts, 2. e., they correspond to waves generated by voltages between 600,000 and 2,000,000 volts, and thus lie between 02 and .0O7 A.U. It is thus clear that the gamma rays from radium C consist mainly of waves of about ;$> the wave-length of the soft gamma rays from radium B, and are of considerably shorter wave-length than any so far observed in an X-ray tube, with the highest voltages at our disposal.’’ The last part of the paper deals with the @ rays from radium B and radium C and their probable relation to the associated y rays. ‘‘The results as a whole suggest that the groups of B rays are due to the transformation of the gamma rays in single and not multiple quanta, according to the relation H = hy.’’ ‘‘Tf the single quantum relation should prove to hold generally for the conversion of y rays into B rays, the magnetic spectrum of 8 rays should afford a reliable method of extending the investi- gation of X-ray spectra into the region of very short waves where the erystal method either breaks down or is practically ineffec- tive, and thus places in our hands a new and powerful method of analysing waves of the highest obtainable frequency.’’—Phil. Mag., xxxiv, p. 153, September, 1917. EL, )85-Ue Am. Jour. Sci1.—FourtH Series, Vout. XLIV, No. 263.—Novemserr, 1917. 8 404 Scientifie Intelligence. 8. Problems in General Physics ; by Morton Mastvus. Pp. vi, 90. Philadelphia, 1917 (P. Blakiston’s Son and Co. ).—This little book contains 1000 problems and it covers the entire field usually taught in engineering colleges. The topics that are more impor- tant or lend themselves more readily to treatment by problems have received greater attention than the relatively less important ones ; for example, certain parts of mechanics and electricity have more space devoted to them than optics. The problems under each heading (equilibrium, calorimetry, etc.) are divided into four groups of equal average degree of difficulty. It is, therefore, possible to select four different sets of about 250 prob- lems each, for use in four consecutive years, or with four divisions of the same class in one year. Since the problems in each group increase in difficulty from the first to the last, an easy course can be made out from the first problems in all four sets, and a more advanced course can be based on the last exercises in the groups. Physical constants required for the solution of the problems have been omitted whenever they are given in the tables in A. W. Duff’s “ A Textbook of Physics.” Answers to the problems have not been incorporated in the volume. As far as one can judge, without testing the book with students, the problems seem to have been carefully selected and arranged. The only obvious drawback to the book consists in the unnecessarily small type used in the printing. | Ho S50: IT. Groxoey. 1. A monograph of Japanese Ophiuroidea, arranged accord- ing to a new classification; by Hikosaicutro Matsumoto. Jour. College of Science, Imperial University of Tokyo, xxxvili, Art. 2, 408 pp., 7 pls., 100 text figs., 1917.—In this excellent work are noted or described the known living species of ophiurians, num- bering 232 species, here grouped into 88 genera. They have been studied from all angles, and on pages 352-365 is also given their geographical distribution. Of especial interest to paleontologists = ene with Paleozoic forms are the following definitions : “Subelass I. (Egophiuroida Matsumoto. Ophiuroidea with external ambulacral grooves and without ventral arm plates. Radial shields, genital plates and scales, oral shields, peristomial plates and dorsal arm plates also absent. Ambulacral plates alternate or opposite ; in the latter case, they may often be soldered in pairs to form the vertebre. Adambulacral plates, 1. e., lateral arm plates, subventral in position. Madreporite either dorsal or ventral, often large and similar in shape to that of an asteroid. “This subclass mostly consists of Palzozoic forms, and lacks all the fundamental characters by which the recent ophiurans are clearly distinguished from the asteroids. Indeed, the distinction Geology. 405 of the present subclass from the asteroids depends merely upon the different development of certain common structures. “Subclass II. Myophiuroida Matsumoto. Ophiuroidea with- out external ambulacral grooves, and with ventral arm plates. Radial shields, genital plates and’ scales, oral shields, peristomial plates and dorsal arm plates usually present ; but sometimes, some or all of them may be rudimentary or absent. Ambulacral plates opposite, usually completely soldered in pairs to form the vertebre. Madreporite represented by one, or sometimes all, of the oral shields. “This subclass includes certain Paleozoic forms and all the ophiurans from the Mesozoic downwards. “The Paleozoic Myophiuroida appear to me to represent a dis- tinct order by themselves.” (pp. 5-6.) “Though it is my purpose to discuss the results of a study of Paleozoic ophiurans in a future paper, I will here enumerate some of the more important structures of Palzozoic ERO roida, as bearing on the question before us. “1, Disk covered with delicate scales or by a naked skin, with- out distinct primaries. ‘9. Radial shields absent. «3. Genital plates and scales absent. ‘4, Oral shields absent. “5. Adoral shields not very distinctly specialized from the lateral arm plates. “6. Oral plates and frames long and slender. “7, Distinct creases probably present between the interbr achial ventral surfaces and arm bases. ‘8. Dorsal arm plates entirely absent, or present only in a few basal joints ; the dorsal side of the arms therefore largely unpro- tected. | “9, Lateral arm plates with prominent spine ridges, which extend to the ventral side of the arm ; those of the two sides not meeting above or below, except in the very distal arm joints. “10. Ventral arm plates higher in position than the lower bor- ders of the lateral arm plates, so that the arm is longitudinally grooved ventrally. ‘“‘T believe that the Palzeozoic Myophiuroida are the stock from which the recent ophiurans have been directly derived, because they show no trace of peculiar specialization and are fairly inter- mediate in their organisation as a whole between the Cigophiu- roida and recent ophiurans. If this view be right, then the most archetypal group of recent ophiurans must be looked for among those forms which have the strongest resemblances to the Palzo- zoic Myophiuroida.” (pp. 367-369.) CS 2. Publications of the Umted States Geological Survey; GEORGE Otis Smiru, Director.—Recent publications of the Sur- vey are noted below. See earlier May, 1917, pp. 418, 419; June, 1917, pp. 489, 490. 406 Scientific Intelligence. GroLocic Fontos—No. 205. Detroit, Michigan (Wayne, Detroit, Grosse Pointe, Romulus, and Wyandotte Quadrangles) ; by W. H. SHERZER. Pp. 22, 12 maps, 12 pls., 20 figs. No. 206. Leavenworth-Smithville, Missouri-Kansas; by HENRY Hinps and F.. C. GREENE. Pp. 13, 4 maps, 10 pls., 10 figs. No. 207. Deming, New Mexico; by N. H. Darron. Pp. 15, 3 maps, 1 structure-section sheet, 9 pls., 11 figs. PROFESSIONAL PAPERS.—No. 94. Economie Geology of Gilpin County and adjacent parts of Clear Creek and Boulder Counties, Colorado; by E. 8S. Bastin and J. M. Huu. Pp. 379, 23 pls., 79 figs. No. 97. Geology and Ore Deposits of the Mackay Region, Idaho; by J. B. Umpuesy. Pp. 129, 21 pls., 14 figs. No. 100-A. The Coal Fields of the United States: General Introduction; by M. R. Campseitu. Pp. 36, 1 pl., 3 figs. No. 108. Shorter Contributions to General Geology, 1917. C, D, KLE. MINERAL REsouRCEs of the United States for 1916. Numerous advance chapters. BULLETINS.—No. 625. The Enrichment of Ore Deposits, by W. H. Emmons. Pp. 530, 7 pls., 29 figs. No. 647. The Bull Mountain Coal Field, Musselshell and Yellowstone Counties, Montana; by L. H. Woo.tsry, R. W. RicHarps, and C. T. Lupron; compiled and edited by E. R. LiLOwD ia. 241.8,c50. DISs, 2. es: No. 652. Tungsten Minerals and Deposits, by F. L. Hass. Pp. 85, 25 pls., 4 figs. No. 653. Chemical Relations of the Oil-Field Waters in San Joaquin Valley, California (Preliminary Report); by G. BS. Rogers. Pp. 119, 7 figs. No. 657. The Use of the Panoramic Camera in Topographic Surveying, with notes on the application of Photogrammetry to Aerial Surveys; by J. W. Baaury. Pp. 88, 15 pls., 22 figs. No. 660. Contributions to Economic Geology, 1907, Paria Asa: No. 661. Contributions to Economie Geology, 1917, Part L. A, B, C, E. No. 666. Our Mineral Supplies. Several advance chapters. WatER SuppLty Paprrs.—Nos. 362, 386, 391, 401, 403, 405. Surface Water Supply of the United States; NaTHAN C. GROVER, Chief Hydraulic Engineer. 1913. No. 362. Part XII. North Pacific Drainage Basins. Pp. 775, 2 pls. 1914. No. 386. Part VI. Missouri River Basin. Pp. 261, 3 pls. No. 391. Part XI. Pacific Slope Basins in California. Pp. 370,.2 pls—No. 394. Part XII. North Pacific Drainage Basins. Pp. 180, 2 pls. 1915. No.401. Part I. North Atlantic Slope Drainage Basins. Pp. 186, 2 pls. No. 403. Part ILf. Ohio River Basin: ‘Pp: Miscellaneous Intelligence. 407 209, 2 pls. No. 405. Part V. Hudson Bay and Upper Missis- sippi River Basins. Pp. 245, 4 pls. ; No. 423. Geology and Water Resources of Big Smoky, Clay- ton, and Alkali Spring Valleys, Nevada; by O. EH. MeEtyzer. Pooler, Vt pls: 10 fies, No. 425-A. Ground Water in San Simon Valley, Arizona and New Mexico; by A. T. SCHWENNESEN, with a section on agri- culture, by R. H. Forpes. Pp. 39, 3 pls. 2 figs. Ill. MiscetLangrous ScrientTIFIC INTELLIGENCE. 1. Eleventh Annual Report of the President, Hmnry S. PritcHeTt, and Treasurer, Ropert A. FRANKS, of The Carnegie Foundation for the Advancement of Teaching. Pp. 172. New York City (576 Fifth Avenue).—This report, for the year ending September 30, 1916, shows a total endowment of $14,250,000, an accumulated surplus of $1,327,000, and an annual expenditure of $779,000. Of this $39,000 was spent in administration, $47,000 in educational enquiry, and $687,000 in retiring allow- ances and pensions. During the year 30 retiring allowances and 16 widows’ pensions were granted, the average grant being $1703. The total number of allowances now in force is 331, the total number of widows’ pensions 127, the general average being $1553. The total number of allowances granted since the begin- - ning of the Foundation is 685, the total expenditure for this ' purpose having been $4,910,000. Part II discusses the general subject of insurance and annu- ities, to which the President has made important contributions in earlier volumes. The report includes official replies from 52 of the institutions associated with the Foundation concerning the new contributory plan of insurance and annuities proposed by the Foundation, and presents the fundamental principles of a pension system which have been approved by the trustees of the Foundation and a joint commission representing the American Association of University Professors, the Association of Ameri- ean Universities, the National Association of State Universities, and the Association of American Colleges. Details are given concerning the new Teachers Insurance and Annuity Association which is to be established, together with an estimate of its prospective service to the teaching profession. In addition to its annual reports, the Foundation has published nine bulletins on special subjects. Bulletin No. 10, now issued (pp. 127), is by J. L. Kanpen, and has as its subject ‘‘A Study of Federal Aid for Vocational Edueation.’’ It traces the legisla- tive history of Federal grants for education and reaches the con- clusion that these grants have always been made for political purposes and without any well-considered educational reasons. 408 Scientific Intelligence. 2. Publications of the Carnegie Institution of Washington.— Recent publications of the Carnegie Institution are noted in the following list (continued from April, 1917, vol. xl, pp. 341, 342) : No. 159. The Mosquitoes of North and Central America and the West Indies; by LELAND O. Howarp, Harrison G. Dyar, and FREDERICK KNas. Volume IV. Systematic Description in two parts. Part Il. Pp. 1064. . No. 175 (Vol. 3). Ocean Magnetic Observations 1915-1916 and Reports on Special Researches; by L. A. BAUER, with the collaboration of W. J. PETers and others. Pp. vii, 445; with 25 plates and 35 text figures. No. 208. A Concordance to the Poems of John Keats; by Dane Lewis BALpDwIN and others. 4to. Pp. xxi, 437; with portrait (frontispiece). No. 215. History of Transportation in the United States before 1860; by CaroLINnE E .MacGinu and a staff of collabo- rators, under the direction of B. H. Meyer. Pp. xi, 670. No. 226. Contributions to Embryology. 4to. Vol. VI. Nos. 15-19. Pp. 1-168, 21 pls., 24 text figs. Papers here included are the following: by F. P. Maun on eyclopia in the human embryo; by Mapce DrG. THURLOW on mitochondria in nerve cells; by Mare@aret R. Lewis on chick embryos; by FLORENCE R. SABIN on origin and development of primitive vessels of the chick and pig; by F. P. JoHNson on a human embryo of twenty-four pair of somites. | No. 250. Ulugh Beg’s Catalogue of Stars, Revised from all Persian Manuscripts existing in Great Britain; with a vocabu- lary of Persian and Arabic Words; by Epwarp BALu KNOBEL. 4to. Pp. 109. No. 251. Papers from the Department of Marine Biology ; ALFRED G. Mayer, director. Vol. XI. Pp. v, 360. Fourteen articles are here included, illustrated by numerous plates and text figures. 3. Publications of the British Museum of Natural History.— The following publications have been recently received: Hco- nomic Series, No. 4; Mosquitoes and their relation to Disease; their life-history, habits and control; by F. W. Epwarps. Pp. 20. No. 5, The Bed-Bug—lIts habits and life-history and how to deal with it; by Brucze F. Cummines. Pp. 20. No. 6, Species of Arachnida and Myriopoda (scorpions, spiders, mites, ticks and centipedes) injurious to Man; by STANLEY Hirst. Pp. 60; with 26 text-figures and 3 plates. Instructions for Collectors: No. 14—Mammals. Part II, Skeletons, with special notes on the collection of specimens of Cetacea; by S. F. Harmer. Pp. 8. No. 13.—Alcohol and Aleo- holometers; by S. F. Harmer. Pp. 8. Warn’s Natural Science EstTaBlisHMENT A Supply-House for Scientific Material. Founded 1862. = Incorporated 1890. A few of our recent circulars in the various departments: Geology: J-8. Genetic Collection of Rocks and Rock- ‘ forming Minerals. J-148. Price List of Rocks. Mineralogy: J-109. Blowpipe Collections. J-74. Meteor- ites. J-150. Collections. J-160. Fine specimens. Paleontology: J-134. Complete Trilobites. J-115. Collec- tions. J-140. Restorations of Extinct Arthropods. Entomology: J-30. Supplies. J-125. Life Histories. J-128. Live Pupae. - Zoology: J-116. Material for Dissection. J-26. Compara- tive Osteology. J-94. Casts of Reptiles, etc. © Microscope Slides: J-135. Bacteria Slides. Taxidermy: J-188. Bird Skins. J-189. Mammal Skins. Human Anatomy: J-16. Skeletons and Models. General; J-155. List of Catalogues and Circulars. Ward’s Natural Science Establishment 84-102 College Ave., Rochester, N. Y., U.S. A. The American Journal of Science Contributors are requested to revise their manuscript with special care with reference to the use of the linotype machine; changes involving overrunning are expensive. In general, the publisher cannot undertake to assume the cost of more than minor corrections in the proof. All corrections should preferably be made in ~ink. Foot notes should be numbered consecutively from 1 up. The following is an example of the system of references employed, the volume being “given in heavy faced type: ; : Am. J. Sci., 44, 249, 1917. The year should be given in every case; this renders a series number unessential. Thirty separate copies of each article will be furnished to the author free of cost and without previous notice from him. They will be provided with a plain cover (but with reference to volume and year). If the author orders separate copies, they will be understood to be in addition to the thirty mentioned above, and he will receive a bill for the extra expense involved, as also for that of a printed cover (with title, etc.), when this is specially ordered ; see schedule below. No. Copies. 60 100 | 200° 300 | 500 Renee a eek $2.25 $2.75 $3.50 $4.25 $5.50 hice ee aa ee 3.50 4.25 525 | 6.25 |- 8.00 | fiat Se rere Rees 4.75 5.75 7.00 8.25 | 10.50 (ets eee ad a 1.00 | $1.25 | $1.75 | $2.25 | $3.00 &@>- During the paper famine the above prices may be increased from 6 to 10 per cent. CONTENTS. Page Art. XXVIII.—The Great Barrier Reef of Australia; by W. M. Davis 5.22.53 23) ee 339 XXIX.—Wave Work as a Measure of Time: A Study of the. Ontario Basin; by A. P. CoLEMAN ~__-~22-22 ee 351 XXX.—Arthropods in Burmese Amber; by T. D. A. COCKERBELL (2). 2222). oe ee XX XI.—A Calcium Carbonate Concretionary Growth in Cape Province 3"by C. J. Mauny 2 2-2-2 eee XXXIJ.—On the Preparation and Hydrolysis of Esters Derived from the Substituted Aliphatic Aleohols; by W.. A. Drusue. and G. BR. BANCROFT, 23. 12 See aeeee 371 XXXTII.—The Perchlorate Method for the Determination of the Alkali Metals; by F. A. Goocu and G. R. Biraxe 381 X XXIV.—Protichnites and Climactichnites; A Critical Study of Some Cambrian Trails; by L. D. Burtine ..-____-. 387 SCIENTIFIC INTELLIGENCE. Chemistry and Physics—A New Method for the Recovery of Salts of Potas- sium and Aluminium from Mineral Silicates, J.C. W. Frazer, W. W. Houianp and EK. MiLuErR, 398.—Hlectrochemical Equivalents, C. Herine and F. H. German: A Laboratory Manual of General Chemistry, W. J. Hate, 399.—A Short Manual of Analytical Chemistry, J. Murer: Allen’s Commercial Organic Analysis, W. A. Davis, 400.—The Ionizing Potential of Sodium Vapor, R. W. Woop and S. Oxano: Penetrating Power of ~ X-Rays from a Coolidge Tube, RurHeRFoRD, 401.—Problems in General - Physics, M. Mastus, 404. Geology—A monograph of Japanese Ophiuroidea, arranged according to a new Classification, H. Matsumoto, 404.— Publications of the United States Geological Survey, G. O. Situ, 405. Miscellaneous Scientific Intelligence—Eleventh Annual Report of the Presi- dent, H. S. Prircnett, and Treasurer, R. A. FRanKS, of the Carnegie Foundation for the Advancement of Teaching, 407.—Publications of the Carnegie Institution of Washington: Publications of the British Museum of Natural History, 408. See Library, U.S. Nat: Museum. vor XV, 495°}. DECEMBER, 10917. Established by BENJAMIN SILLIMAN in 1818. e | _ AMERICAN 3 JOURNAL OF SCIENCE. Bprron: EDWARD S. DANA. ASSOCI:sTE EDITORS PROFESSORS GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. ‘FARLOW anD WM. M. DAVIS, or CampringeE, he Propussos ADDISON E. VERRILL, HORACE L. WELLS, ee) aa LOUIS V. PIRSSON, HERBERT E. GREGORY is 3 - anv HORACE S. UHLER, or New Haven, ||... Prorssson HENRY S. WILLIAMS, or Irxaca, — ¢ ce _.. Prorzsson JOSEPH S. AMES, or Baxrimore, ae By en eres “Mr. J. S. DILLER, or Wasuineron. oe ‘FOURTH SERIES Ge ae. VOL XLIV—[WHOLE NUMBER, OXCIY] é ‘No. 264—DECEMBER, 1917, NEW HAVEN, CONNECTICUT. oat * aati THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET. PBpbhished monthly. Six dollars per year, in advance. $6.40 to countries in the ; Postal Union; $6.25 to Canada. Single numbers 50 cents. i _ Entered as ‘second- class matter at the Post Office at New Haven, Conn. , under the Act : ae oe March 3, 1879. JUST: PUBLISHED. A Practice Book in ELEMENTARY METALLURGY By ERNEST EDGAR THUM, E.M. Assistant Professor of Metallurgy University of Cincinnati. This book will appeal to instructors who are confronted with the problem of building up a respectable laboratory course with a modest amount of equipment, funds, time and assistance. By its use officers in charge of well-equipped laboratories will avoid the nuisance of mimeographed instruction sheets. The book treats of the metallic materials of engineering con- struction; how they are gained from mother nature; how they are further refined and worked; and how their chemical compo- sition and past history influence their various physical properties, and their adaptability for the duty expected of them. viii + 313 pages, 6 by 9, 59 figures. Cloth, $2.75 net: FREE EXAMINATION—NO. CASH IN ADVANCE. This book will be sent for free examination, no cash in advance, to members of national technical societies. If you are not a member of any society you can supply a reference, or state your position. At the end of 10 days you are to remit the price of the book, or return it, postpaid. JOHN WILEY & SONS, Inc. Dept. B.—432 Fourth Avenue, New York London: Chapman & Hall, Ltd. Montreal, Can.: Manila, P.I.: Renouf Publishing Co. Philippine Education Co. shes) AMERICAN JOURNAL OF SCIENCE [FOURTH SERIKES.] soe Art. XXXV.— Origin of the Chert in the Burlington Lime- stone; by W. A. Tarr. TABLE OF CONTENTS. Introduction. I. Theories as to Origin. II. The Chert in the Burlington Formation. A. Definition of chert. B. Megascopic description. C. Microscopic description. D. Mode of occurrence of the chert. KH. Age of the chert. Ill. Origin of the Chert. A. Outline of theory. B. Evidence for theory. 1. Source of the silica. (a) Carried to the sea by streams (6b) Derived from the land through chemical denudation. (c) Relationship to periods of peneplanation. (d) Derived from shore work. . Dispersion of silica in sea-water. . Deposition of the silica. (a) Experimental. (6) Cause. (c) Form of precipitated silica. (d) Associated minerals. . Relationship of chert to enclosing rock. . Fossils and the silica. . Absence of siliceous organisms in the Burlington formation. Conclusions as to origin under the theory. Evidence against the replacement theory. (a) Position of the chert in the limestone. (6) No adequate source of the silica. (c) Adverse evidence of the structure of the nodules. (d) Fossils in the chert. (e) Weathering of the chert. (f) Conclusions as to replacement, IV. Application of the Colloidal Precipitation Theory to Other Cherty Formations. V. Summary and Conclusions. Am, Jour. Sct.—Fourts Series, Vor. XLIV, No. 264.—Drcremper, 1917. 29 C9 DIDO 410 Tarr—Origin of the Chert in the Burlington Limestone. INTRODUCTION. Many explanations have been advanced to account for the occurrence and distribution of chert in limestone and dolo- mite. Since chert is found in these rocks in all parts of the world, geologists have had many opportunities to study its occurrence and to accumulate enough facts to give an ade- quate explanation of its origin. This has not been done, how- ever, for careful descriptions of chert, its mode of occurrence, physical characteristics, and composition are lacking in most references to it. There are a few good descriptions but these are mainly old; very few of the articles dealing with chert, even in a general way, are later than 1900, and these, asa rule, do not present new evidence but cite the earlier views as to origin and conclude that they are adequate. In studying the chert of the Burlington limestone in the vicinity of Columbia, Mo., the writer observed many facts which do not accord with the prevalent view that chert is due to the segregation of silica in limestone into the characteristic chert nodules. The following paper presents the evidence which has been collected. aud its interpretation as the writer sees it. I, THEORIES AS TO ORIGIN. There have been many different views advanced to explain the origin of chert, but the majority can be classed under two heads: (1) theories that seek to explain the origin through segregation by organic means; and (2) theories maintaining _ tbat the material of the chert was a direct chemical precipitate. There are, however, other theories, as the following complete table shows : Chert is the result of direct chemical precipitation. Chert is the result of the secretion of silica by organisms. It is the result of the replacement of calcareous material. It is of organic origin, but not through replacement. It is a spring deposit. It is a mechanical sediment. It is regarded as being due to weathering. © SENG at Oe It would be entirely out of place to present the various theories in full in this paper, but a brief statement of the essen- tial point in each one may be made, and the bibliography at — the end of the paper furnishes many excellent summaries and further citations to the literature. The theory that chert originated through the precipitation of silica in the colloidal form upon the sea bottom was advanced by Prestwich in 1888. The silica was thought to have accu- mulated about the siliceous spicules of sponges, whieh acted as nuclei, or in the absence of such material, decaying organic Tarr Ovigin of the Chertin the Burlington Limestone. 411 matter might also have acted in the same capacity. Once started, the aggregate would, according to the theory, continue to attract silica to itself from the surrounding muds as long as material was available. Hull and Hardman, M. A. Renard, C. A. White, E. O. Hovey, C. R. Van Hise and R. D. Irving, and N. H. and H. W. Winchell have all advocated the same view, although the first two men thought that the silica so pre- cipitated had later replaced calcareous material. Hovey based his theory upon a study of chert from the Burlington and the Cambro-Ordovician formations of Missouri, which makes his conclusions especially interesting by way of comparison, since this paper also is based upon astudy of the chert of the Bur- lington and the Cambro-Ordovician formations. He states: “ Regarding the Lower Magnesian and Lower Carboniferous cherts from southern and southwestern Missouri, the present writer’s conclusion is, that they are due to chemical precipitation, probably at the time of the deposition of the strata in which they occur or before their consolidation. ” The advocates of the theory that the silica was first segre- gated by some organism have presented much evidence in its support. Their main evidence is the finding of the remains of sponges and diatoms scattered throughout the chert. This theory was early advanced in England by Bowerman and was ably supported by the researches of Sollas, Hinde, Jukes- Grown and Hill, and Sorby. It is largely accepted by Ameri- can geologists as 1s shown by the statements made in all our latest text-books. Had these early investigations been made on material collected in Missouri, very likely a different theory would have resulted, because Missouri chert shows no remains of siliceous organisms. The advocates of the theory that the chert and flint nodules and beds are the result of the replacement of calcareous mate- rial of the limestone early called attention to the fact that fossils known to be originally calcareous were often found silicified undoubtedly through replacement. A. H. Church was able to replace with silica the calcium carbonate of a frag- ment of coral by allowing a weak solution of colloidal silica to percolate over it.* On account of the finding of fossils on the mterior of the chert and flint nodules which were originally calcareous, the conclusion was drawn that the silica had replaced the fossil and that the nodule had grown by further additions on the outside, each particle of silica replacing an equivalent amount of caleareous material. The presence of siliceous spicules of sponges in chert and flint suggested a pos- sible source for the silica in the nodules, and since the silica in *F. W. Clarke, Bull., U.S. G. S., 616, pp. 515, 543. 412 Tarr—Origin of the Chert in the Burlington Limestone. the spicules is in a form regarded as being soluble in ground water, it was thought to have been taken into solution and deposited about other sponge spicules which acted as nuclei. Other silica-secreting organisms, such as radiolaria, have been suggested as possible sources of the silica. In this and the preceding theory the organisms were thought to have obtained their silica either from the sea water in which it was held in solution or by decomposing silicates occurring in the muds upon the bottom. There seems to be room for doubting that the latter method is an important mode for obtaining the siliea, for the silicates which accumulate in the muds have already withstood the attacks of solutions upon the earth’s surface and are very stable. Most of the men who held the view given above believe that the segregation of silica by organisms was followed by its solution, removal, and deposition elsewhere, so that these two theories are really no more than one. That the chert layers and nodules may have been due to the accumulation of abundant organic remains upon the sea bottom was an idea advanced by Wallich and has been favored by others. The possibility that a part of the silica was a chem1- cal precipitate derived from soluble sponge spicules was also included in this theory, but the method of precipitation was not explained. Lawson has suggested the theory that the chert of the Fran- ciscan series in California is due to the precipitation of colloidal silica brought to the sea water by thermal springs. The chert contains sponge spicules, and Lawson states that these proba- bly fell into the soft colloidal silica on the sea bottom and were preserved. An unusual mode of origin was suggested by Penrose and Buckley, who thought that the silica might be transported to its present position as a fine siliceous mud. Penrose applies his theory to the Boone chert in northern Arkansas and Bueck- ley, to the dark-colored, secondary chert in the Joplin district of southwestern Missouri. It is difficult to explain the source of such a siliceous mud. Another theory that has been suggested to explain the origin of chert is that the aggregation of the silica is due to the ordi- nary processes of weathering. It is thought by Ulrich, who suggested this view, that the silica, being imsoluble, is concen- trated near the surface as the associated calcareous material is. removed by subaerial erosion. This view is held by some others, but the finding of chert in the lower formations in some of the deepest wells that. have been drilled in Missouri is strong evidence against it. Further, a careful study of the occurrence of the chert in the area described by Ulrich shows that it does not have unequal distribution, though Ulrich reported that it Tarr—Origin of the Chert in the Burlington Limestone. 418 did and interpreted the condition as indicating superficial con- centration by weathering. It is evident from the above brief statements that while there are several views as to the origin of chert, the important ones are the first three given, which should really be regarded as two. The theory which seems to be prevalent at the pres- ent time, judging by the recent text books of geology, is that the silica was originallv secreted by organisms and then aggre- gated into the nodular or bedded form through replacement of the limestone by circulating ground waters, either before. or after the consolidation of the limestone. The supporters of the theory that chert is due to chemical precipitation assume that the silica was in the sea water, and that it has been precipitated by some means or other, which they fail to suggest. Il. Tur CHErRT IN THE BURLINGTON FoRMATION. A. DEFINITION OF CHERT. In this paper the term chert will include those erypto- crystalline varieties of quartz whichare white, gray, or blue- gray in color. Dark gray to black varieties will be called flint, while those which owe their color to iron oxides will be referred to as jasper. } Chert is defined in our latest text books and mineralogies as an impure flint. The writer wishes to express the view that this much quoted statement is wrong and that flint is an impure chert. It is the excessive amount of organic matter in the flint which gives it its black color, and this foreign matter is absent in the chert. Since all these materials are supposed to be pure quartz it would seem that chert is the purest because of the absence of this coloring matter. As to composition chert is practically pure silica, as the analy- ses given below show. The other constituents which are usually present are calcium carbonate, alumina, iron oxides, and a small amount of water, the latter being probably con- tained in the opaline silica present in the chert. B. MEGASCOPIC DESCRIPTION. Megascopically the chert is dominantly white although some of it is gray or mottled. The mottled appearance is a common feature of the chert and is due to the aggregation of organic matter into small areas and to smal] aggregates of disseminated pyrite. Fossiliferous and non-fossiliferous chert oceur. The fossils are still composed of calcite, but in some eases they are entirely of silica. While the greater part of the chert is massive and 414. Tarr—Origin of the Chert inthe Burlington Limestone. Analyses of Missouri and Other Similar Cherts fe ee 3 yy 5 6 7 Silica (Si0,) 98°23 98°17 .98°92 98°71 9946 99°23RR9s sam Alumina (AJ,O,) Z 3 . "QF e . . e y) "59 Iron oxides (Fe,O,)} es 85 48 43 29 2 52 Calcium carbonate (CaCO,) Magnesium carbonate (MgCO,) Magnesia (MgO) ‘O01 02 ‘02 Trace Trace Trace Lime (CaO) 05-08-08, -.:04 | Aageuaaaias Alkalies (Ignition) ‘78 "42 50.84 ‘50 “40 10041 99°84 99°87 99°69 100-13" 99:9 aaa a2 8 9 10 11 12 13 Lh 15 Silica (S10,) 99°13 95°81 91°54 96°88 71-29 94 9l= Ga°67eesee Alumina (A1,O,) loi, Sogo ; 5 Ne 2: 243 9: 2 0 eee Tron oxides (Fe,O,) 16 9-90 "84 £8 $3 Bo 2 Calcium car- bonate (CaCO,) 1:43 Asi0OS bee ors “15 32°12. 1-00 Magnesium car- benate (MgCO;) ~ ‘Trace -17 Trace Trace; Trace jem Lime (CaO) Trace Magnesia (MgO) -01 Alkalies "15 (Ignition) ‘20 "28 «1°64 12 99°50 100°19 100-41 100°47 99°96 98°51 99°56 99°64 Fresh chert, Mississippian, Henderson Mine, Barry County, Mo. Fresh chert, East Hollow, Belleville, Jasper County, Mo. (591). Partly altered, same locality. Altered to “ cotton rock,” same locality. Fresh chert, Surprise Mine, Joplin, Mo. (591). Blue chert, fresh, Bonanza shaft, Galena, Kan. (591). Same locality, fresh (591). Altered chert, same locality (591). Fresh chert, Jonesboro, Montgomery County, Mo., Mo. Geol. Surv. 10. Altered chert, same locality. 11. Fresh chert, Lebanon, Laclede County, Mo., Mo. Geol. Surv. 12. Fresh chert with crinoid stems of calcite, Sulphur Springs, Ark. 13. Slghtly altered Miss. chert, Roaring Springs, Newton County, Mo. 14. Fossiliferous, calcareous chert, Miss., Grand Falls, Newton County, Mo. 15. Fresh chert, Burlington formation, Columbia, Mo. Analyses 1, 9-14, E. O. Hovey, Mo. Geol. Surv., vol. vil, pp. 727-739, 1894. Analyses 2-8, Bull. 591, U. 8. G.S., p. 222. Analysis 15, Mrs. W. A. Tarr. wo - reece © 00 I Tarr Origin of the Chert in the Burlington Limestone. 415 without bands of any kind, some nodules show bands, which are usually horizontal and but rarel y concentric. The nodules frequently contain cavities which are almost always lined with quartz crystals. Fossils in the chert, espe- cially brachiopods, blastoids and horn corals, are often lined with such crystals. The spiralia of brachiopods are often pre- served and covered with quartz crystals. The major part of the nodule is made up of the massive white or mottled chert but small masses of crystals may be included in the interior and very rarely there are concentric bands for a few inches around these masses. Some nodules consist of fragments of chert with more or less concentric banding. These have the appearance of having been formed, then broken up and re- cemented with silica. The exterior is often bleached, soft, and porous, due to alteration. This shell, often an inch or more in thickness, is similar to the tripoli of southwestern Missouri. Such leached zones are generally lacking in the chert of the Ordovician and Cambrian formations in Missouri. C. MICROSCOPIC DESCRIPTION. The microscope shows the chert to be composed of opaline or amorphous silica, chalcedony, and quartz. The amount of amorphous silica in the Burlington chert is small. The chert consists mainly of a granular mosaic of chalcedony and quartz (see figs. 1, 2, and 4). Scattered throughout the slide are cir- cular areas, from *013 to °026™ in diameter, of very fine-grained chaleedony and quartz, the grains ranging from *0006 to -0013™™ in diameter. These circular areas may ‘possibly be interpreted as representing the original form of the colloidal silica. Such S) globular forms have been described by many other authors (see especially, Jukes-Brown and Hill, W. Hinde and J. A. Howe), but have usually been regarded as the tests of foramini- fera, although in some instances writers say that they are uncertain that such is the case. It would seem that ‘these globules, which in some instances are amorphous silica, might show the structure of the or iginal precipitate. Spherulites are rare in the chert and those present are not perfect. ‘They appear to occupy the areas between the fossils. The dark cross is very indistinct and the spherulites are small. Both non-fossiliferous and fossiliferous chert were studied. They were examined especially for remains of originally silice- ous fossils, but none were found. In the fossiliferous speci- mens studied, crinoid stems, fragments of brachiopods, and bryozoans were fairly abundant. Some of the fossils were practically replaced by silica, others were only partially re- 416 Tarr—Origin of the Chertin the Burlington Limestone. placed, and still others were unaltered. In all instances of alteration the calcite was replaced by quartz. The most coarsely crystalline areas and the largest grains of quartz are in the silicified fossils. Scattered through the grains of quartz there are many very smail grains of calcite, these grains being the unreplaced residue of the original fossil. These calcite grains are so minute that they do not show any cleavage planes or other crystalline characteristics, except the interference tints. Iie, Ie Fic. 2. Fic. 1 (x 40). Fresh chert, Burlington limestone, Columbia, Mo. Shows fine mosaic of quartz and chalcedony ; the larger areas of the latter show the interference cross. Fie. 2 (x40). Altered chert, Burlington limestone. Shows fine mosaic of chalcedony in the altered outer portion of a chert nodule. When a fossil is only partially replaced the quartz crystals form a border around the ealcite. This border may occupy fully one-fourth of the area of the fossil. The dense white chert consists almost entirely of a mosaic of chalcedony and quartz. The grains which compose this mo- saic are very minute. Probably not more than one per cent of the chert is amorphous silica. The amorphous silica remains dark during a complete rotation of the stage of the microscope between crossed nicols. Since it may show an interference cross at times, it may be more abundant than is stated. By re- flected light it has a pale milky-white luster. Much of the chert in the Burlington has an outer zone Tarr— Origin of the Chert in the Burlington Limestone. 417 which is usually white and which is a weathered portion of the chert. This zone varies in thickness but is usually about an inch thick. Slides were made of the fresh and the altered por- tions of the same nodule in order to determine if possible what changes had taken place in the formation of this outer zone. The only difference observed was in the granularity of the chaleedony. The entire weathered zone was a mass of gran- ular chalcedony (see fig. 2). The various fossils were readily ie, Hie Fie. 3 (x40). Banded chert, Burlington limestone. A mosaic of quartz and chalcedony. Dark bands due to carbonaceous material. Fic. 4 (x 40). Fresh chert, Burlington limestone. Coarsely crystalline quartz and chalcedony. Large black and white areas are quartz, radial areas chalcedony. seen and many grains of calcite were scattered through the zone, showing that calcium carbonate has probably been removed by leaching. The grains are very small, ranging from :006 to ‘00138™" in diameter. Some of the slides show the banding which is a megascopie feature of the chert nodules (see fig. 3). The bands were found to be due to minute black specks of what appeared to be or- ganic matter but which may be pyrite in part. The black specks were also seen in the lighter colored portions of the chert but were not so numerous as in the darker bands. Other materials observed in the chert were grains of pyrite, usually small and never exhibiting crystal faces; a few grains 418 Tarr—Origin of the Chertin the Burlington Limestone. of glauconite; and a small amount of a yellowish, dust-like material, which is probably clay. D. MODE OF OCCURRENCE OF THE CHERT. The Burlington limestone, of lower Mississippian age, is in this region from 100 to 175 feet in thickness. It is a heavy, irregularly bedded, coarsely crystalline limestone, mainly of a hight bluish-gray color. The upper portion in some parts of Missouri is fine- grained and the lower is coarse-grained. The beds are not well defined, due evidently to a lack of silts, etc., in the waters and to rapid variation in the sizes of the calcare- ous materials. The beds are rather thick, but range from a few inches to several feet. The formation consists dominantly of the remains of crinoids, in fact crinoids are so abundant that the formation was formerly known as the “ Encrinal limestone.” The chert occurs throughout the formation but is especially abundant in the lower half near the top. The following sec- tions of the Burlington at different localities in Missouri will show the distribution of the chert: Section of Burlington limestone one half mile south of Columbia, Missouri. Limestone: medium to coarse-grained; gray; con- tains stylolites and a few chert nodules -_--.-._.- 3 ft. 6 in. Limestone: medium to fine-grained; gray. This is a very cherty zone; contains large masses of chert and considerable pyrite in small nodules; the large chert masses occur along planes-- - --- 13 ttoGeine Limestone: coarse, medium, and fine-grained; usu- ally some shade of gray, bluish tints predom- inating; fossiliferous in places; stylolites rather COMMON: = Seo rs ee gee pe eee 13 ft. Limestone: coarse-grained; fossiliferous; blue gray ; contains stylolites, and chert nodules along a planes 4s chee eee idee TSctee = ee Limestone: fine- -orained, oray, stylolitic. MR eo ae 3 Limestone: coarse-grained, grading upwards into fine-grained; blue gray to buff; contains chert in lareenodales andnissstylolitic™ sess 2 {t. Limestone: fine-grained and medium to coarse- grained; blue gray; contains stylolites and chert . Zp Bowral oh gags Sct os 2 ee ae ee pea Chert: many large nodules AMONG? HONS DIANNE 222 425 = Zont: Limestone: medium to fine- -grained, the texture varying rapidly both vertically and laterally; gray to blue gray; fossiliferous; contains styl- olites and scattered nodules of chert._--.----.- 6 fta6m Limestone: coarse-grained; light to very dark gray; Stylolitie/ 2: = 59 ern eee itt. Grim Limestone: medium to fine-grained; gray; has nu- merous stylolites 2.3 52. y 2 eee em late Tarr—Origin of the Chert in the Burlington Limestone. 419 Section of Burlington limestone, Greene Co., Missouri. EK. M. Shepard, Geol. Sur. Mo., vol. xii, pp. 100-123, 1898. Upper Burlington Shaly limestone: thin-bedded, compact, rarely odlitic, interbedded with layers of chert which are usually. fossiliferous. The lime- stone is very fossiliferous, coarse-grained and varies from white to gray in color. Middle beds are coarse-grained, crystalline, soft, gray to white in color, have many nodules of chert, are stylolitic, and pass downward into a shaly, Coansed-orained simestone= 2224558. -.-5 22-2. 200 tt, Lower Burlington Limestone and chert: consists mainly of brownish yellow, blue, or slate-colored beds; fine to medium-grained; contains some fossil- iferous bands; middle portion thin-bedded; top is mainly yellowish-white chert; much nodular chert and many lenticular layers of chert chies lower part. 22246... helt 70 {t. Section of the Burlington limestone, Calhoun Co., Missouri. C. F’. Marbut, Geol. Sur. Mo., vol. xii, p. 161, 1898. Upper Burlington Gray or white, coarse-grained encrinital limestones in rather massive beds, with chert nodules scattered throughout; very fossilifer- GING sal (eel SiS a aac niger Eee gee cee 80 ft. Lower Burlington Limestone: rather fine-grained; drab; some- what earthy; has many crinoid stems and much chert, the latter occurring as masses through- out the limestone and as thin layers alternat- ing with thin layers of limestone .-.-.--.-.--- 30 ft. Section of the Burlington formation in Pike Co., Missourt. R. R. Rowley, Geol. Sur. Mo, vol. viii, p. 36, 1917. Upper Burlington Thin bands of brown and yellow limestone AMG CINE Deep uipe rey on eee Unga t RE bee Boe So OO A Es Lower Burlington liesimmestonen a: sac asus ss ee. Att. White limestone with very little Chet ee errs Ae tena ea AoE, White and brown layers of lime- stone with considerable chert in some nec sweets ner Ce Oe Oe A eh 80: Ft, Wintesromemenae fee eo Ott; 50 ft. 420 Larr—Origin of the Chert in the Burlington Limestone. The chert occurs chiefly in the form of rounded, elliptical, or lenticular nodules (see fig. 5). The nodules are found always with their longer axes parallel to the bedding planes but they are not necessarily along them, although such is a common occurrence. Rarely the lenticular masses mer ge so that the chert becomes essentially a bed. This is not a very common feature in the Burlington in this vicinity but in other localities it becomes a very important mode of occurrence, especially in ibe, 2 Fic. 5. Banded chert nodules in Burtington limestone, Columbia, Mo. The notebook is 54 inches long. Stylolite at (a). the lower Paleozoic formation, where thick beds occur. Along a given plane the nodules are of a remarkably uniform size, but there are marked variations in size along the different planes. The rounded nodules are rarely more than six inches in diameter with a range of from three to six inches. The elliptical forms are from three to fifteen inches in thickness and from five to twenty-five inches in length. In ground plan the nodules are very irregular, but the dominant tendency is towards a lobed circularity. The lenticular nodules may vary from two to fifteen inches in thickness and some reach a length of several feet, as can be seen in fignres 5,6,and 7. Elliptical and lenticular forms are dominant. The layers of chert nodules may occur along such bedding planes as ‘the limestone shows, or they may occur at various 1n- Tarr— Origin of the Chert in the Burlington Limestone. 421 tervals along planes in the beds (see fig. 7). This lack of a definite relationship to the bedding planes should be noted. A layer of nodules may occur in the central part of a thick mas- sive bed of limestone which shows no evidence of a bedding plane along the plane of the chert even after weathering. The vertical interval between the planes of chert varies consider- ably ; it may be from a few inches to many feet ; but, what- ever the distance, the planes of chert are always parallel to each Fie. 6. Fie. 6. Chert nodules, Burlington limestone, Columbia, Mo. other and to the bedding planes. The horizontal interval be- tween the different nodules appears to be more uniform, as there seenis to be a rather definite spacing of the nodules. This spacing of the planes of chert without reference to the bedding planes of the limestone and yet always parallel to them has a very important bearing upon the origin of the chert. The shape of the chert nodules appears to be the result of flattening. The elliptical and the lenticular shapes are sugges- tive of this, as though the elliptical shape might be due to the spreading of the soft, globular masses of vel under their own weight. The bands frequently seen in the chert are further evidence of flattening. The great majority of these bands are horizontal and approximately parallel to the bedding planes. Those which have the same general shape as the exterior of the nodule always have the bands broadened at the ends of the 429 Tarr—Origin of the Chert in the Burlington Limestone. nodules (see fig. 3). This is a feature which would have de- veloped in a band if the material in which it is found had undergone a certain amount of movement laterally. Since these broadened parts of the bands are always found in those portions of the nodules which must have moved laterally if the nodule has been flattened, it is believed that they were caused by such a movement. The widened portions of the bands are usually of a lighter color than the narrower ones, showing that ~ EG anie Fic. 7. Shows chert nodules along a plane in Burlington limestone, Columbia, Mo. in the former the same amount of coloring matter has been distributed over a wider zone. When circular nodules contain bands, the bands do not show this widening. The majority of the nodules show upon ‘their surfaces many cracks (see figs. 8, 9, and 10) which may be three inches in depth and one- half inch in width at the surface of the nodule. They gradually die out in the chert. These cracks are usually filled with limestone which is always continuous with the lime- stone surrounding the nodules (see fig. 11). Many examples of reopened cracks are found. This reopening occurred after the complete consolidation of the beds and the chert. A coarsely crystalline, yellow or white calcite was deposited in these reopened joints. It may occur on one side or on both sides of the limestone in the original crack but is usually on one side only. The limestone which fills cracks may contain Tarr— Origin of the Chert in the Burlington Limestone. 423 fossils just the same as the other limestone does. The cracks run in all directions through the nodules. Usually they are found on the upper surface but they may occur also on the lower. They are most abundant in the thicker portions of the nodules. The interesting structural feature, stylolite, is wonderfully well developed in the Burlington limestone. Very frequently the Hee of the stylolites may follow a plane of chert nodules (see 5, a). When such is the case the outside of the nodules Fie. 8. Fie. 8. Chert nodule containing cracks filled with limestone. About one- third size. shows the striated faces which are produced by the stylolites. Planes of stylolites have been observed passing through the limestone immediately over the cracks in the chert nodules and showing no evidence of deviation from a plane. This in- dicates that the cracks in the nodule were filled before the development of the stylolites. Itis not thought, however, that the development of these cracks was due entirely to the flat- tening process, but that they were partly developed during the changes which the silica underwent in the early stages of the development of the nodule, especially during the change due to the loss of water. 494. Tarr—Origin of the Chert mm the B urlington Limestone. E. AGE OF THE CHERT. The question as to the time of the formation of the chert is a very important one in considering its origin, since the major- ity of the theories, and especially the commonly accepted one, appeal to the action of circulating ground water In accounting for the chert. If it can be shown that the chert is of a certain age, the time necessary for its formation can be estimated. This period of formation would be the time extending from ides, U.S, Gas. Bull’ Nor 267, p. 27 Ball, S. H. and Smith, A. F.: Mo. Bur. Geol. and Min., vol. i, 2d ser., pp. 34, 85, 145-147, 1903. Bowerbank, Dr.: Trans. Geol. Soc., 2d ser., vol. vi. Buckley, HE. R. and Buehler, H. A.: Geology of Granby, Mo., Area, Mo. Bur. Geol. and Min., vol. iv, 2d ser., pp. 31-82. Chamberlin, T. C. and Salisbury, R. D.: Textbook of Geology, vol. i, p. 426. Clarke, F. W.: Data of geochemistry, Bulls. 491 and 616, U. S..G.S. Petz hws: Lowa Acad’ Sei Pree. vol. ai, p. 1777 1895. Geikie, A. : Structural and Field Geology, p. 66. Grabau, A. W.: Principles of Stratigraphy, pp. 718-720, 764. A. G. Seiler Sel Coz. Ne Yi, Griswold, L. S.: Ann. Rep. Geol. Sur. Ark., vol. iii, p. 177, 1890. Hatschek, Emil: Intro. to Physics and Chem. of Colloids. Hinde, G. J.: Phil. Trans. Royal Society, vol. clxxvi, part II, p. 403, 1885. Geol. Mag., vol. iv, p. 485, 1887; ibid., vol. v, p. 241, 1888. Hovey, E. O.: this Journal, vol. xlviii, pp. 401-409, 1894. Lead and zine deposits of Missouri, Mo. Geol. Surv., vol. vii, part II, pp. 727-739, 1894. Hull and Hardman: Sci. Trans. Roy. Dub. Soc., vol. i, pp. 71, 85, 1878. Irving, R. D.: this Journal, vol. xxxii, pp. 255-271, 1886. Jones, T. Rupert: Proc. Geol. Assoe., London, vol. iv, 1876, p. 489. Jukes-Brown, A. J. and Hill, W.: Quart. Jour. Geol. Soc., vol. xlv, p. 403, 1889. Keyes, C. R.: Iowa Acad. Sci. Proc., vol. x, pp. 108-105, 1903. Lawson, A. C.: 15th Ann. Rep. U.S. G. S., pp. 420-426, 1895. Lee, Wallace: Geol. of Rolla quadrangle, Mo. Bur. Geol. and Min., vol. xii, 2d ser., 1918. Leith, C. K. and Mead. W. J.: Metamorphic Geology, pp. 184-135, ete. Leith, C. K. and Van Hise, C. R.: U. S. G. S. Monograph 52, p. 505. Lindgren, Waldemar: Mineral Deposits, pp. 189, 301, ete. Prof. Paper U.S. G. 8. No. 48, p. 68. . Lyell, Sir Charles: Elements of Geology. pp. 321-322. Mantell, G. A.: Wonders of Geology, vol. i, p. 306. Merrill, G. F.: Rocks, Rock-Weathering, and Soils, p. 110. Moore, KE. S.: Jour. Geol., vol. xx, p. 261, 1912. Ostwald, W.: Colloid Chemistry. Penrose, ie ne ie Ann. Rep. of the Geol. Sur, of Ark., vol. i, pp. 129-1388, 1890. Pirrson, L. V. and Schuchert, Charles : Textbook of Geology, pp. 469-497. Péschel, Victor: Intro. to Chem. of Colloids. 452 Tarr—Origin of the Chert in the Burlington Limestone. Prestwich, Joseph: Geol. Chem. Phy. and Strat., vol. ii, pp. 320-324, 1888. Renard, M. A.: Bull. R. Acad. Belgique., vol. xlvi, p. 471, 1878; ibid., vol. xlvii, p. 562, 1879. Richardson, Clifford: Min. Sei. Press, vol. iii, p. 48, 1915. Sollas, W. J.: The Age of the Earth, pp. 1388-165. Annals Mag. Nat. Hist., (5) vol. vi, p. 438, 1860. Sorby, H. C.: Quart. Jour. Geol. Soc., London, vol. xxxvi. p. 91. 1880. Taylor, W. W.: Chemistry of Colloids, New York (Longmans, Green & Co.): i Van Hise, C. R.: Monograph 47, U. S. G. S., pp. 817, 847-853, etc. Van Hise, C. R. and Irving, R. D.: 10th Ann. Rep. U.S. G.S., Pt. 1, p. 397, 1890. Vaughn, T. W.: Carnegie Inst. of Wash., vol. iv, No. 138, p. 99. Wallich, G. C.: Quart. Jour. Geol. Soc., vol. xxxvi, pp. 68-92, 1880. Way, J. T.: Jour. Chem. Soc., vol. vi, pp. 102-106, 1854. Wkite, C. A. : Geol. Sur. Iowa, vol. i, p. 223, 1870. Winchell, N. H. and Winchell, H. V.: Bull. vi, Geol. Sur. of Minn., 1891. Waicott, C. D.: Mon. 30, U.S. G. S., pp. 4-20, 1898. Zsigmondy, R.: Colloids and the Ultramicroscope, ist ed., 1909. J. Alex- ander, Trans. Mineralogical Laboratory, University of Missouri. Van Name and Brown— Cadmium Lodide Solutions. 458 Art. XXX VI.—Jonization and Polymerization in Cadmium Todide Solutions ; by R.G. Van Name and W.G. Brown. [Contributions from the Kent Chemical Laboratory of Yale Univ.—cexcyv. ] Ir is a well known fact that water solutions of cadmium iodide show an abnormally low electrical conductivity and freezing point lowering as compared with other salts of like type. This is generally ascribed to the presence in the solu- tion of complex molecules and ions. In a previous article* we have described a method, based on the measurement of distri- bution coefficients of iodine between such a solution and a non- aqueous phase, by which it is possible, by extrapolation, to cal- culate the percentage of simple molecules and ions in a pure solution of cadmium iodide. The application of this method led to the conclusion that in 0-5-molar cadmium iodide at 25° about 6 per cent was present in the form of simple Cdl, molecules, ionized and non-ionized, in 0°25-molar solution 10°6 per cent, in 0°125-molar solution 16°8 per cent, and in 0-01- molar solution 55 per cent. These results may be compared with those of Walton,t who showed in 1904 that the rate of decomposition of hydrogen peroxide in a neutral solution containing an iodide is propor- tional to the concentration of the iodine ion, and used this kinetic method to determine the proportion of iodine ion in a series of rather dilute cadmium iodide solutions. With due allowance for ionization and for differences-in concentration, our results, as will be shown later, are consistent with those of Walton, at least as to order of magnitude. McBaint on the other hand, by a mathematical analysis of the data in the literature concerning conductivity, freezing points, and transport numbers of cadmium iodide solutions, has arrived at quite a different result. McBain concludes that ina 0'1 molar cadmium iodide solution, the only concentration quantitatively dealt with, most of the salt is in the form of simple non-ionized OdI, molecules, and that the complexes make up only about 8 per cent of the whole. : In the present investigation we have attempted to throw further light upon this question by a study of solutions of cad- mium iodide containing dissolved iodine, by means of measure- ments of electromotive force and of freezing point lowering. Electromotive FLorce Measurements. The reversibility and reproducibility of iodine electrodes composed of platinum immersed in an iodide solution contain. * This Journal (4), xliv, 105, 1917. + Zeitschr. phys. Chem., xlvii, 185, 1904. t Zeitschr. f. Klektrochem., xi, 215, 1905. 454 Van Name and Brown—Tonization and ing free iodine, have been proved by the work of a number of investigators.* In its relation to the present problem the work of Lauriet is especially important. Laurie used the electro- motive force of concentration cells composed of two such iodine electrodes as a means of caleulating the iodine ion con- centration in a potassium iodide solution saturated with iodine. The Nernst equation tor the electromotive force of such a cell at 25° may be written ey r= om { (oe AR). — (ve), The electromotive force is thus stated in terms of four concen- trations. Under the conditions of Laurie’s experiments three of these four concentrations were calculable from known data, and the fourth, the desired iodine-ion concentration, was caleu- lated from the observed electromotive force. Concentrated ammonium nitrate solution was used to eliminate diffusion potentials. The method pursued in our own experiments was similar to that of Laurie. The cells measured were of the type Pt : KI + I, (saturated) : NH,NO, : Cdl, + I, : Pt. Each cadmium iodide electrode was measured against two different potassium iodide electrodes of different concentra- tions, designated hereafter as electrodes A, and B, respectively, the former containing 0°1 molar KI, the latter 0-01 molar KI, both saturated with iodine. These two electrodes were then measured against each other, thus furnishing a check upon the results. All necessary data concerning these iodine-potassium iodide solutions have been given by Bray and MacKay.t As intermediate solution to eliminate diffusion potentials, a concentrated solution of ammonium nitrate was employed, to which, following a suggestion due to Luther,§ enough sodium nitrate was added to make the mean cation velocity the same as thatoftheanion. This solution contained 8°3 mols NH,NO, and 1 mol Na NO, per liter. For comparison a few measure- ments were made in which saturated potassium chloride solu- tion was substituted for the mixed nitrates, but the results tailed to show any difference large enough to be of importance in the present work. The nitrate solution was used in all the experiments recorded below. Calculation of the diffusion potential by the Planck or the Henderson equation was practicable only when the ion concen- * See, for example, Maitland, Zeitschr. f. Elektrochem., xii, 263, 1906; also, Jones and Hartmann, Jour. Am. Chem. Soc., xxxvii, 757, 1915. t Zeitschr. phys. Chem., lxvii, 627, 1909. + ¢ Jour. Am. Chem. Soc., xxxii, 914, 1910. $ See Bjerrum, Zeitschr. phys. Chem., liii, 488, 1905. Polymerization in Cadmium Iodide Solutions. 455 trations of both solutions were known, which was true in the case of the cell made up of the two potassium iodide electrodes, but not for a cell with a cadmium iodide electrode. For the potassium iodide concentration cell the value of the diffusion potential was calculated with the aid of the following data taken from the article of Bray and MacKay: Solution 1 Solution 2 Migration 0°1 molar 0:01 molar velocities (K+) 0°0865 0:00941 : 74°8 (I’) 0:0430 0°00484 : Loo (ua 0°0435 0°00457 : 41°5 Using the Planck formula the value of the potential so obtained was 0:0070 volt, while the Henderson formula gave from the same data the value 0:0069 volt. The measured value, as given by the difference between the electromotive force as observed with and without the use of the intermediate nitrate solution, was approximately 0°003 volt. In the opinion of the writers this difference between the measured and calculated values of the diffusion potential is too large to be explained by experimental error, or by uncertainty in the data employed in the caleulation, and is to be ascribed to some peculiarity in the behavior of liquid junctions which involve the tri-iodide equilibrium. Support for this view is given by the fact that the diffusion of iodine in an iodide solution is abnormal in at least one important respect, for the rate of diffusion of iodine im potassium iodide is known to increase with increasing concentration of the latter, while by the rule of Abegg and Bose* we should expect just the reverse. That the discrepancy is in all probability not due to the incomplete elimination of the diffusion potential by the nitrate solution is shown by the fact that the electromotive force of this concentration cell as measured with the nitrate solution as intermediate liquid (mean value of many determina- tions, 0°0560 volt), was in close agreement with the value cal- eulated from Equation 1 (0°0561 volt). Apparutus.—The type ot halt-cell used is shown in the figure. The tube carrying the electrode (of bright platinum foil) was fitted into the neck of the half-cell by a ground joint, so that the iodine solution came in contact with nothing but platinum and glass. Two such cells were clamped in a frame with their siphon tubes dipping into opposite arms of a U tube containing the intermediate solution, and the whole was immersed up to the necks of the half-cells in a thermostat kept at 25°. With stop-cocks closed the cell could be left set up for many days without danger of contamination of the solu- * Zeitschr, phys. Chem., xxx, 551, 1899. 456 Van Name and Brown—Tonization and tion, and could quickly be put into commission again by emptying the siphon arms and renewing the intermediate liquid. The stop-cocks in the siphon arms were generally left open during the actual measurement. In the earlier experi- ments, contact between dissimilar solutions was brought about in the U tube in a layer of sea-sand, as recommended by Bjerrum,* but the use of sand was later abandoned as trouble- some and unnecessary for the present purpose. The measurements were made by the Poggendorf compensa- tion method with the aid of a galvanometer sensitive to 4 xX 10-° amperes. Owing, however, to the high resistance of the cells measured, the accuracy of a single bridge-reading did not much exceed one millivolt. Preparation of the Reference Electrodes.—The complete saturation of the iodide solutions with iodine was accomplished by sealing the solution with powdered iodine in a large glass tube, which was attached to the stirring axle of the thermostat and rotated for a period of at least 24 hours. With solutions so prepared the two reference electrodes, when measured against one another, usually gave a constant potential difference within 12 hours, which retained its value practically unchanged for many days. Occasional shaking of the half-cells (which contained a little solid iodine) was found to favor constant results. The liquid in the siphon arms was emptied periodi- cally, and the half-cell when necessary could be refilled with a portion of the original (iodine saturated) solution without alter- ing the measured electromotive force. All measurements — were at 25° C. Experimental Procedure.—A cell composed of the two reference electrodes was first set up and its electromotive force * Zeitschr. f. Elektrochem., xvii, 58 and 389, 1911. Polymerization in Cadmium Lodide Solutions. 457 measured from time to time until the constant value 0:056 volt was reached. A half-cell was then filled with the iodine-cad- mium iodide solution to be investigated, and measured in turn against each of the reference electrodes. Finally, as a check, the reference electrodes were again combined and measured. Obviously the difference between the electromotive forces of the two cells in which the 1odine-cadmium iodide electrode was used should agree with the electromotive force of the cell made up of the two reference electrodes (0°056 volt). All measure- ments in which the difference above mentioned was within + 0°001 volt of 0°056 volts were considered trustworthy. The cadmium iodide solutions used were of four different concentrations : 0°5, 0°25, 0°125 and 0:01 molar, from each of which a number of electrodes containing varying amounts of dissolved iodine were prepared and measured against electrodes A and B. Since these concentrations of cadmium iodide were the same as employed in the determinations of distribution coefficients described in our previous article* the results of that work were used in calculating (I,), the concentration of uncombined iodine, from the concentration of dissolved iodine as found by direct titration. For electrodes A and B (I,) was obviously equal to 0°00132, the solubility of iodine in pure water, and the values of (I’) for these two solutions were those given on p. 455 of this article. It was, therefore, possible to ealeulate (1’) for each cadmium iodide solution from the observed electromotive force by means of Equation 1. A summary of the results is given in Table I. Except where otherwise stated, concentrations in this and the follow- ing tables are expressed in millimols per liter. In the third column are the potential differences given by each iodine-cad- mium iodide electrode against reference electrode A, while the potential difference for the same electrode against reference electrode B is found in the fourth column on the horizontal line next below. A negative sign prefixed to the recorded potential indicates that the reference electrode formed the neg- ative pole. Usually the reverse was true. The iodine-ion con- centrations calculated from these potential differences are recorded in the fifth and sixth columns, the former giving the two independent values for each solution, derived from the two potentials measured, and the latter the mean of the two. Finally, by graphical extrapolation of the values in column six to zero concentration of iodine, a value has been obtained for each cadmium iodide solution which should represent the con- centration of iodine ion in the given cadmium iodide solution if it contained no dissolved iodine. These four extrapolated values are the important ones for our present purpose. Since * This Journal (4), xliv, 105, 1917. Am. Jour. Sc1.—Fourtu Serigs, Vou. XLIV, No. 264.—DrcremBEr, 19177. 32 458 0°5 molar Cdl, 0°25 molar Cal, Dissolved iodine 1 3448 | 1°641 | 0-792 l 0°0 TABLE I. volt (I2) against against A I-32 0°0423 —0°0134 0°955 0°0504 —0°0046 0°605 0°0536 —0°0029 0°505 0°0589 0:0037 0°315 0°0648 0°0090 0°175 0'0725 0°0178 0°140 0°0751 0°02038 0°0 1°32 0:0470 —0:0099 0:99 0:0490 —0:°0063 0°538 0°0592 0°0027 0°255 0:°0689 0°0124 0°160 0:0745 0°0198 0°090 0°0818 0°0267 0:0 eo, 0°0423 —0'0137 0°648 0°0531 —0°0021 0°250 0°0675 0°0112 0'110 0:0770 0°0213 0°051 0:0870 0°0313 0°022 0:0962 0°0410 0:0 * Solution saturated with iodine. Van Name and Brown—TLonization and Polymerization in Cadmium Lodide Solutions. 459 TABLE I.—Continued. Dissolved volt (1’) iodine (Iz) against against (I’) mean A B | (f *8-16 1°32 0°:0080 6°62 6°79 — 0°'0468 6°95 2°540 0°2967 0°03842 8°69 9°16 —0°0198 9°62 1°397 0°1569 0:°04381 8°95 9°23 0°01 | —0°0114 9°52 molar 4 0°646 O'07138) 00585 8°94 8°94 Cdl, | 0°3104 00843 0:°0654 9°93 10°42 0°0116 10°90 0'1411 0°0168 0:0755 LORSG) AO 77 0°0216 . 11-27 i 0:0 0:0 10°50 * Solution saturated with iodine. all these iodine-ion concentrations are very low we can certainly afford to disregard here the slight error involved in the assump- tion that the results of electromotive force measurements repre- sent concentrations rather than “activities” as defined by Lewis.* Discussion.—In the previous investigation, as has already been stated, values were obtained for the proportion of simple molecules, ionized and non-ionized, in a pure solution of ead- mium iodide at various coneentrations. This quantity we shall designate, as in our former article, by the term ‘active frac- tion.” The rest of the cadmium iodide must consist of asso- ciated cadmium iodide molecules, and of the ions, simple or complex, formed therefrom. These values of the active fraction are calculated upon the assumption that the abnormally low power of cadmium iodide to unite with iodine is due entirely to the presence of com- plexes, a condition which may be only approximately fulfilled. The amount of iodine taken up by an iodide should be practi- cally independent of its degree of ionization, provided that the tri-iodide is ionized to about the same extent, as is no doubt generally the case.t But although it is contrary to experience * For a discussion of the relation between activity and concentration for the iodine ion in potassium iodide solutions, see Bray and MacKay (Jour. Am. Chem. Soc., xxxii, 925, 1910). + This is confirmed by the fact that for all metallic iodides so far investi- gated, with the exception of those of cadmium and mercury, the value of the equilibrium constant K, = (21) (Iz) /(2I3) is the same. The list includes the iodides of di- and trivalent metals, Ba, Sr, Zn, Ni, and La, whose degrees of ionization are certainly somewhat different from those of the univalent iodides. 460 Van Name and Brown—TIonization and that two metallic salts of like type and as closely related as are cadmium iodide and cadmium tri-iodide should show any large difference in ionization, the possibility of slight differences is not excluded, which if present would cause small positive or negative errors in the value of the active fraction, calculated as above. According to McBain* the observed transport numbers indi- cate that the complex ion present in predominating amount is Cdl,’, formed by ionization of (CdI,), according to the reaction: (Cdl,), = Cal a Ca Whether or not this inference is correct it is probable that the above equilibrium is typical, and that the inactive fraction con- sists essentially of (@) associated molecules, (0) complex anions, and (c) simple Cd** cations. It follows, therefore, that the degree of ionization of the simple Cdl, molecules may be eal- culated, at least approximately, by dividing the concentration of the iodine ion, taken from Table I, by the equivalent con- centration of the “active”? cadmium iodide as derived from the data of the previous investigation. Table Il shows the results obtained in this way. The first three columns contain the concentrations of total cadmium iodide, of active cadmium iodide, and of the iodine ion, respec- tively. The fourth column gives the degree of ionization of the simple CdI, molecules as calculated from columns two and three, and the fifth shows, for comparison, the degrees of ionization of cadmium nitrate, a normally ionized salt of like type, at the same concentrations as those of (CdI,)active in col- umn two. The data for cadmium nitrate were obtained by interpolation from values given by Noyes and Falk.t+ TABLE IT. y (1’) for Cd(NOs),2 (Cdley (Coie (I') Te 9( Cigars at 18° 500 30° 30°4 ‘51 rite 250 265 31:2 "59 79 125 Q1° O45 65 Sigg 10 5°5 10°5 95 "87 According to these results the degree of ionization of the Cdl, molecules is considerably lower, “except in the most dilute solution, than that of the cadmium nitrate. A difference in this direction, though smaller in amount, would be expected on account of the presence in the cadmium iodide solution of * Zeitschr. f Elektrochem., xi, 215, 1905. + Jour. Am. Chem. Soc., xxxiv, 475, 1912. Polymerization in Cadmium Lodide Solutions. 461 an excess of Odt+ ions resulting from the ionization of the associated molecules, and it would also be expected that the difference, if due to this cause, would decrease with increasing dilution, as is in fact the case. It is probable, however, that the actual difference between the degrees of ionization of these two salts is considerably smaller than these results would indi- eate, for the error in determining y by the above method may easily be rather large. The work of Walton,* whose kinetic method for determin- ing iodine ion concentrations has already been referred to, is of interest here, since his results yield values of (1’) which can be compared with those in Table I. Unfortunately, Walton’s experiments with cadmium iodide were confined to solutions more dilute than 005 molar. The comparison is shown in Table III, the second horizontal line giving the values of (L’) ealeulated from Walton’s results+ while the third contains our values for the same quantity. Figures enclosed in parentheses were obtained by graphical interpolation. TABLE ITI. (CdI2) 4°9 10 19°4 31°8 . 42°1 (1’) W alton 6°5 (12) 19°3 26°5 31°2 (I') V. N. & B. 10°5 (14) (20) The two sets of results agree in order of magnitude, though Walton’s values are higher and increase more rapidly with the concentration. Since Walton’s method has not yet been very thoroughly studied, particularly as to its sensitiveness toward secondary catalytic influences, the results of the electromotive force measurements deserve the greater weight. Freezing Point Measurements. Since the purpose of these measurements was to determine the effect of successive additions of iodine upon the freezing point of a given cadmium iodide solution, it was necessary to employ a method of the “ undercooling ” type, the very con- venient and accurate method of Roloff,{ as improved by Richards,§ being excluded by the fact that it would not per- mit the concentration of the cadmium iodide to be kept con- stant. After unsuccessfrl attempts to obtain sufficient accuracy with the aid of a modified Beckmann apparatus, using a cry- ohydric mixture for the cooling bath, and other special precau- * Zeitschr. phys. Chem., xlvii, 185, 1904. { These values were obtained by dividing the observed velocity constants in each case by 1°40, the average value of the ratio, Velocity constant / (1'), as found by Walton’s experiments with KI, Nal, and NH,I. { Zeitschr. phys. Chem., xviii, 572, 1895. _ §Jour. Am. Chem. Soc., xxv, 291, 1903. 462 Van Name and Brown—LJonization and tions, an apparatus was finally devised which satisfactorily met the needs of the case. The construction and manipulation of this apparatus have been fully described in a former paper,* so that only the principal points in the procedure need be given here. The different cadmium iodide solutions used were kept in thoroughly stearned bottles of Jena glass. The freezing point of the water was first taken. The water was then replaced by the cadmium iodide solution, which had been prepared by dis- solving the carefully dried and weighed salt in a portion of the same water, and diluting to an exact volume. At least three determinations of the freezing point were made, using varying degrees of undercooling, and no result was accepted unless the final temperature held constant within 0-001° (the maximum sensitiveness of the temperature reading) for at least five min- utes. The volume of solution used in the freezing point apparatus was either 200 or 250°™. The pure cadmium iodide solution was next replaced by a portion of the same solution which had been shaken at room temperature with an excess of iodine in a Jena glass bottle for some hours. At 0° this solution was supersaturated with iodine. ‘To insure equilibrium with both ice and solid iodine the liquid was allowed to partially freeze and was then kept at its freezing point with continual stirring for several hours. Equilibrium was assumed to have been reached when succes- sive determinations of the freezing point agreed, and also suc- cessive titrations of the dissolved iodine. In this way the freezing points were obtained for solutions saturated with iodine, and for those containing no dissolved iodine. Data for the intermediate iodine concentrations were obtained by mix- ing varying amounts of the saturated solution with the original ‘pure cadmium iodide solution, determining the freezing point as before, and, finally, estimating the iodine by titration. In all these measurements care was taken to maintain the room temperature as constant as possible, and in a few cases where appreciable variation occurred the thermometer readings were corrected for change in the length of the projecting mer- cury thread with the aid of the usual formula, using the value 0000156 for the apparent expansion coefficient for mercury in glass. This correction was applied only in working with the more dilute solutions where its importance was obviously greatest. The results of these experiments are given in Table LV. The four concentrations of cadmium iodide studied are the samme as in the measurements of electromotive force at 25°. Iodine concentrations are given in the second column, the starred values being those for solutions approximately saturated with iodine, having iodine present as a solid phase. In the * This Journal (4), xliii, 110, 1917. Polymerization in Cadmium Iodide Solutions. 463 third column are the observed freezing point depressions, which in nearly every case are the mean of two or three separate determinations. The fourth and fifth columns contain the observed values, for the pure cadmium iodide solution, of the van’t Hoff coefficient 7, and of the apparent degree of ioniza- tion y, as calculated in the ordinary way from 2. TABLE IV. n= es Dep. Molec. (CdI.) Dissolved Bp: van’t Hoff y = —— due to dep. for iodine dep. coeff. 2 iodine iodine 500 0:0 1:067° 1°147 7°35% si 9°4 1°089 : 0:022° 2°34(?) a 19°6 1°095 0°028 1°48 - 27:9 LA 0°045 1°61 - *40°4. 1°125 0°058 1°44 250 0:0 OBES 1°145 7°25% ee 9°33 0°546 DeOTS. 1°40 e 19°4 0°558 0°025 1°29 ad 24°9 0°567 0°034 1°37 a #33°6 0°579 0°046 1°37 125 0:0 NY ie 1°193 9°65% } oe 5°93 0°285 0:008° 1°35 7 10°29 0°293 0°016 1°56 3 16°75 0'301 0°024 1°43 i FOLD 0°317 0°040 1°46 10 0°0 0:0365° F962 48°1% SF 2°344 0°0394 0:0029° 1:24 2 4:°266 0°0424 070059 =1°38 re *7°53 00453 00088 1:17 Previous determinations of 2 for cadmium iodide by the freezing point method have been made by Arrhenius,* by H. C. Jones,t and by Chambers and Frazer.t Our values at the two lowest concentrations are in excellent agreement with those - of Jones (whose results only cover concentrations up to 0-1 molar) and at the two higher concentrations they are close to the mean between the results of Arrhenius and those of Cham- bers and Frazer. No explanation is offered for the fact that 2 is slightly larger in 0°5 molar cadmium iodide than in the 0°25 molar solution, but the effect is real as it is even more evident in the results of the other investigators than in our own. Chambers and Frazer ascribe the phenomenon to hydration. The last column of Table V shows the “ molecular depres- sion for iodine ” as obtained by dividing the depression due to iodine (column six) by its total concentration as given in col- * Zeitschr. phys. Chem., ii, 491, 1888. t Ibid., xi, 544, 1893. ¢ Am. Chem. Jour., xxiii, 512, 1900. 464 Van Name and Brown—LTonization and umn two. This quantity proves to be roughly constant irre- spective of the concentration and to have a value between 1:3 and 1:4, or about three-fourths of 1°86, the normal molecular lowering for a non-electrolyte in pure water. The relatively large amount of this increase proves that the cadmium tri-iodide formed is derived ultimately from some source which previously contributed a much smaller number of molecules and ions to the solution. Unless, therefore, we are willing to admit that the cadmium tri-iodide may have in the solution a degree of ionization many times greater than that of the simple cadmium iodide molecules,* these results must ‘be regarded as clear proof of the existence of complexes in these solutions. On the other hand, if the effect is largely or wholly due to complexes, as is probably the case, the relative | constancy of the values in the last column of Table IV is an indication that the complexes are present in considerable quan- tity even in the more dilute solutions. On account of the low solubility of iodine in water at 0° the depression due to the iodine which remains uncombined is so smallt as not to affect the validity of this reasoning. Owing to the presence of complexes it is of course impos- sible to get any accurate measure of the concentration of the iodine ion from the value of 2, but in the present case the error so involved would not necessarily be very large. Of the various kinds of complex ions to be expected here the two simplest and most probable are Cdl,’ and Cdl,’, formed as products of the equilibria 2Cdl, 2 (Cdl) 2 Cas edit and sCdl 2 (Cdl) 222 Cd ee eal= In both of these cases the number of ions produced is the same as the number of CdI, molecules disappearing, so that the net result of the complex formation will be to diminish the freez- ing point depression by a small amount due to that part which remains in the form of non-ionized polymerized molecules, (CdI,), or (CdI,), as the case may be. Hence, if values for the concentrations of the iodine ion are calculated in the usual way from the freezing point lowerings the results will in general be low, but in sufficiently dilute solutions should not be very far from the truth. In Table V the iodine ion concentratiuns, so calculated, are tabulated for comparison with those derived from the measure- ments of electromotive force. For the two lowest concentra- * A rough calculation shows that to account for the results in the absence of complexes the ratio of these two degrees of ionization would have to be over 2 in the 0°01 molar solution, about 8 in the next, and about 25 in the strongest. _ + Its maximum value is 0:0012°, which is reached only when the solution is saturated with iodine. Polymerization in Cadmium Iodide Solutions. 465 tions the agreement is fairly good, but in the 0°25 molar solution the difference is not in the expected direction, and in the strongest solution the discrepancy is surprisingly large, far exceeding the probable experimental error. For this result there is no evident explanation, but it is significant that it TABLE V. (I’) at 0° (L') at 25° (Cdl) From freezing point From EK. M. F. 500 74 30°4 250 36 31:2 125 24 27°5 10 9°6 10°5 coincides with a marked irregularity in the freezing point lowering. As was noted on page 463, the results of all cryo- scopic measurements with cadmium iodide show that above about 0°3 molar the value of 7 apparently rises with the con- centration, although the attendant decrease in ionization and increase in polymerization would both tend to lower it. Nature and Concentration of the Complexes.—Thus far the question of the nature of the complex ions and molecules has been left open. Of the various complex ions which may be present in the solutions, CdI,’, as McBain* has shown, is apparently the most probable one. The high transport num- ber of the anion, which approaches 1:25 in the most concen- trated solutions, cannot be. explained by assuming the pre- dominance of Cdl,” without assigning to that ion an improbably high velocity. This objection, however, would not apply to such ions as Od, I,’ or Cd,I,”, though there would be less reason to expect their presence than that of the less complex ones just mentioned. ~ McBain has calculated the approximate composition of a 0-1 molar solution of cadmium iodide upon the assumption that Cdl,’ is the only complex ion present in appreciable amount. This calculation, which is based entirely upon freezing point, transference, and conductivity data, gives the values (CdI,’) = _0°0084 and (1’) = 0:0126. This would make (Cdt*) = 0:0105, thus accounting for about 19 per cent of the total iodide. McBain concludes that the remaining four-fifths is present in the form of simple, non-ionized CdI, molecules, and that the proportion of complex molecules is negligible. There are several serious objections to these figures: (a) If so large a part of the salt is in the form of simple Cdl, mole- cules the power of the salt to combine with iodine should be but slightly lower than normal, while in reality the “ active * Loe: cit. 466 Van Name and Brown—Tonization and fraction” at this concentration is only about 20 per cent.* (6) The molecular conductivity of the solution as calculated for 18° from the above composition is 22°5,f or less than half the value actually measured, which is 46°7. (c) The value of (1’) calculated by McBain is much lower than that given by the electromotive force method. } In a similar manner it is possible to calenulate the approxi- mate composition of cadmium iodide solutions from the experi- mental data furnished by our measurements of electromotive force, and by our previous study of the iodine-cadmium iodide equilibrium. These calculations will be confined to the 0°01 and 0°125 molar solutions, for which the data are presumably most aceurate. It will be assumed that CdI,’ is the only com- plex ion present in significant amount, and that the degree of ionization of (CdI,), is of about the same order of magnitude as that of the average uni-bivalent electrolyte. Allowing for the effect of the excess of Cdt* ions, which is much larger in the 0:01 molar than in the 0°125 molar solution, we may assume that the degree of ionization is 80 per cent in the former and 75 per cent in the latter. An error of a few per cent in the degree of ionization assumed will not greatly change the results. Using these degrees of ionization, the “active fractions” given on page 453 and the values of (I’) from Table I, we obtain the results recorded in Table V1. That these values differ greatly from those of McBain is evi- TABLE VI. (Cds jr Ca) (I') (CdI.) (CdIz)s 0-01 * -molanzae 2°4 6°45 10°5 0°25 0°3 0°125 molar. _--- 52° 39°7 27°5 72 8°6 dent. Interpolation of these results for 0-1 molar concentra- tion gives, approximately, (CdI,’) = 0-045 and (I’) = 0:021, figures which are respectively 5 and 1-7 times those of McBain. Since the values in Table VI depend upon neither freezing point nor condnetivity measurements, their correctness may properly be tested by calculating the van’t Hoff coefficient 7 and the molecular conductivity for each solution. For 2 we obtain 0°0199 / 0°01 = 1:90, and 0:135 / 0°12 = 1:08 respec- tively, while the measured values (see Table V) are 1°96 and 118. *This low power to unite with iodine cannot be explained by the low ion- ization of the CdI, molecules unless it is assumed that the degree of ioniza- tion of the cadmium tri-iodide is about nine times larger. This is obviously very improbable. + For the method of calculation see p. 467. The ionic conductivities here used were those employed by McBain, namely, 4Cd = 51, Cdl,’ = 41, and I’ = 66°4. If the values given on p. 467 be employed the result is 21°1 instead of 22:5, Polymerization in Cadmium Lodide Solutions. 467 To calculate the molecular conductivity we must obtain a value for the conductivity of CdI,’. If we assume with MeBain that the value 1°25, the limit which the observed. (anion) transport number of a cadmium iodide solution tends to approach with increasing concentration, is that of the anion CdI,’, then the relative velocity of this ion must be — = (0°42, that is, s that of Cdt*. For the equivalent con- ductivity of the Cd*+*+ ion we may use the value 47 for 18°,. which corresponds to 56 at 25°. This gives, for CdlI,’, 34 at 18° and 40°5 at 25°. For iodine ion the values are 66°6 and 76°5 respectively. The molecular conductivity of the 0:01 molar solution at 25° should therefore be { (0:00645)(112) + (0°6024)(40°5) + (0°0105)(76°5)} 0-01 = 162°2,. and for the 0°125 molar solution, calculated in the same way, 69-2. These calculated conductivities are much too high, the meas- ured values being 120 and 57 respectively, a result which seems to be due to some fault in the assumptions made rather than to experimental errors. It is not clear, however, how this discrepancy can be eliminated without introducing some other one. The evidence at hand is in some respects conflict- ing and is obviously insufficient for an exact solution of the problem. In short, though some of the values in Table VI are probably nearly correct, the figures as a whole can repre- sent, at best, no more than a rough approximation to the truth. The lodine-Cadmium Lodide Equilibrium at 0°.—Each of the starred values in Table V represents the solubility of iodine in the given solution at its freezing point, which, as an approx- imation, may be assumed to be the same as the solubility at 0° in the same medium. By subtracting the solubility of iodine in pure water at 0° (0:000638 mols/liter*) we obtain (=I,), the equivalent concentration of the tri-iodide formed, and can therefore calculate the approximate value of the equilibrium. constant: A) == (SVL) /SL). The results so obtained are given in Table VII, which com- pares the values of A, for cadmium iodide at 0° and 25° with those for potassium iodide at the same temperatures and con- centrations. All of these figures refer to solutions saturated with iodine. The values for cadmium iodide at 25° were taken from our previous article ; those for potassium iodide at 0° were calculated in the way just described from data given by Jones and Hartmann’;+ those for potassium iodide at 25° were taken from the article of Bray and MacKay. * Jones and Hartmann, Jour. Am. Chem. Soc., xxxvii, 256, 1915. + Loe. cit., p. 250. | 468 Van Name and Brown—Cadmium Iodide Solutions. TABLE VII. Potassium Iodide Concentra- Cadmium Iodide K, at 0° K, at 25° tion. K, at 0° Ry, at 25° Jones and Bray and equiy./liter (approximate) Hartmaun. MacKay dpe) 0°0154 OO 254s See eee 0°00046 0:5 0°0090 ORO SETAC HRT eee eee 0-00088 ORs 0°0053 O'O0828) 2" io See Aa) er 0°10 NRA er Bae 0:000696 0°00131 0°U02 00011 0°00247 0°000716 0:00137 These figures show that the value of A, for cadmium iodide at.0° is abnormal in the same way as at 25°, being larger throughout than for a normal iodide like potassium iodide, and increasing rapidly with the concentration. The effect upon A, of a change in temperature, however, is practically the same for the cadmium as for the potassium salt. Summary. 1. Measurements have been made: (a) by the electromotive force method, of the iodine ion concentration in cadmium iodide solutions of 0°5, 0°25, 0°125, and 0°01 molar strength, containing various amounts of dissolved iodine; also (0) of the freezing point lowering of each of these cadmium iodide solu- tions, and ot the further lowering produced by the addition of known amounts of iodine. 2. Values of (1’) calculated in the ordinary way from the eryoscopic measurements should be slightly lower, if complexes are present, than those electrically measured. This was found to be the case in the 0°01 and 0°125 molar solutions, but not in the two stronger solutions. 3. The freezing point of a cadmium iodide solution was depressed by the addition of iodine in a nearly constant ratio, which in the stronger solutions was about 1:4° per mol, and only slightly smaller in the weakest solution. This indicates the presence of complexes in considerable quantity even in the 0°01 molar solution. Neither this result nor the abnormally low power of cadmium iodide to unite with iodine can be accounted for, in the absence of complexes, by the assumption that the degree of ionization of .the cadmium iodide is very small, unless this low ionization is accompanied by high ioni- zation of the cadmium tri-iodide, a state of affairs which is decidedly improbable. 4. A tentative calculation of the composition of the two more dilute cadmium iodide solutions, based upon the assump- tion that the ion Cdl,’ and its parent molecule (Cdl,), are the only complexes present, failed to give results in quantitative agreement with other experimental data. EF. V. Shannon—Famatinite from Goldfield, Nevada. 469 ART ee —Famatinite from Goldfield, Nevada; by Hart V. SHANNON. During last year, through correspondence with Mr. Herbert N. Witt, geologist for Goldfield Consolidated Mining Co., the writer obtained a number of specimens of ore minerals from the Goldfield district, with the idea of investigating the mineral goldfieldite, reported by Ransome! from that region. With regard to the speci- mens, Mr. Witt writes as follows: ‘‘T am sending you under separate cover some specimens otf the copper ore that. occurs here. I believe that you will find that this consists principally of famatinite. However, we have found that almost any specimen of this ore will upon analysis give, not only copper, gold, and sulphur, but arsenic, antimony, bismuth, and tellurium. I do not believe that the mineral gold- fieldite exists but is probably a mixture of famatinite, bismuth- inite, and calaverite or sylvanite, with possibly some tetrahedrite. All of these have been recognized here and the one specimen that I have had in polished surface under the reflecting microscope indicates such a mixture. . . . You may be able to detect some of the whitish telluride in the famatinite specimens. This will then, I believe, ame you all the constituents of the so-called ‘ooldfieldite’. # One of the specimens had, on one side, some very minute crystals, of a blackish-gray color and metallic luster, partly embedded in kaolin. These were carefully tested in the hope that they might be goldfieldite but although strong reactions were obtained for antimony, arsenic, copper, and sulphur, no bismuth or tellurium could be detected in the very small amount of materia! available. The crystals therefore seem to be of the same substance as the main mass of the specimen on which they occur, an arsenical famatinite. The crystals vary in greatest diameter from about 1 mm. down to about 0.1 mm., and are so attached to the matrix that they could not be detached without breaking. The larger crystals furthermore had curved faces which gave no dependable reflections. After repeated trials a small crystal was found giving moderately good reflections. In the litera- ture at hand no axial ratios are given for famatinite. Dana? gives the forms observed by Rath, as a(100), c(001), m(110) and 1(130), but gives no axial ratios, nor does he give any angles. Rath’s original paper is not accessible to the writer, hence the angles found on the * Ransome, F. L., U. S. G. S., Prof. Paper, No. 66, 1909. * System of Mineralogy, 6th ed., page 149. 470 E. V. Shannon—Famatinite from Goldfield, Nevada. Goldfield crystal are compared with the observed angies of enargite. The close agreement with enargite in the prism angles is shown below. Observed, famatinite enargite (Dana) [ares 60° 24’ 60° 17’ h:h’ 59° 12’ 59° 43’ This will serve to show the orientation of the crystal and to indicate that the famatinite from the Goldfield district is isomorphous with enargite and the value for the a axis does not differ greatly from that of enargite. With this known orientation the domes observed were plotted on a stereographic projection and gave the indices (104) and (025). The reflections obtained from these faces were poor and the angles obtained were not considered sufficiently trustworthy to serve as a basis for calculating a value for the ¢ axis. Both in the pro- jection and in the accompanying drawing the axial ratios of enargite were used. ‘The figure reproduces the form and appearance of the crystals. The locality is not given more nearly than one of the mines of the Goldfield Consolidated Co. The base upon which the crystals, described above, are implanted is similar to the majority of specimens in the lot received. It consists of fine-grained, pinkish gray famatinite. When polished and examined under the reflecting microscope, the distinctly pink mineral is seen to contain graphic inclusions of a silver-white mineral which in its microchemical reactions agrees with .bis- muthinite. The ore reacts for bismuth but not for tel- lurium. The pink mineral contains both arsenic and antimony, the antimony preponderating. One of the samples consists of a fine clayey gouge containing finely triturated native tellurium. Nothing was observed in the specimens which seemed to correspond to the mineral goldfieldite. Lull— Functions of the ** Sacral Brain” in Dinosaurs. 471 Arr. XXXVITII.—On the Functions of the ‘‘Sacral Brain’’ in Dinosaurs; by RicHarp Swann LULL. [Contributions from the Paleontological Laboratory, Peabody Museum, Yale University, New Haven, Conn., U. S. A. Branca, in a discussion of the fauna of Tendaguru, East Africa,! makes a number of thought-inspiring com- ments upon the huge sauropod dinosaurs which the formation contains. Among other points he is striving to account for the maintenance of their immense bulk upon a possibly meagre diet by assuming digestive powers of extraordinary efficiency. For this he offers the following explanation: ‘One may be inclined to look for the ability to take care of food solely in the stomach, intestines, or liver. However, in the dinosaurs we may take into consideration something else,. 1. é., the ‘sacral brain,’ if we look upon the swelling of the spinal column in the sacrum as a ‘brain.’ According to Waldeyer, it is indeed thinkable that the sacral brain in dinosaurs had a certain independence, and cared: for the functions of nourishment, digestion, and procreation [italics mine], also that through a particularly strong innervation it had become especially power- ful, more powerful than the strongest digestive organs could be without such a sacral brain. In man there appear still to be traces of this, but here the sacral. section of the spinal column is completely surpassed by the brain.’’ What Waldeyer based his argument upon I do not know, but the evidence which I have been able to secure seemingly does not justify such a speculation. This evidence is here presented. Dinosaurian feeding habits. Our assumption of feeding habits based upon the char- acter of dinosaurian dentition } dasities the following conclusions : Iheropoda.—These are the carnivorous dinosaurs in a strict sense, with teeth which were in the main prehen- sile and as such confined to the forward portion of the jaws. They must have been used for rending the prey, for in many instances, such as Allosaurus or Megalo- *W. Branca, Die Riesengrésse sauropoder Dinosaurier vom Tendaguru, ihr Aussterben und die Bedingungen ihrer Entstehung, Archiv fiir Bion- tologie, 3. Bd., 1. Heft, 1914, pp. 71-78. 472 Luli— Functions of the “Sacral Brain” in Dinosaurs. saurus, they are sharp-pointed and compressed, with finely serrated cutting edges. In Tyrannosaurus they become so thick that the knife-like edge is gone, so that these huge beasts must have dismembered their prey by tearing rather than by cutting it. These latter thero- pods are analogous to the crocodiles in dental equipment; Allosaurus, on the other hand, possessed a more efficient dentition. The digestive system of the crocodiles shows the high- est degree of specialization of any living reptiles, as the stomach is very muscular, with lateral tendinous discs forming an organ very suggestive of the gizzard of the eraminivorous birds. In front of this gizzard-like stom- ach is a capacious portion of the esophagus, within which is held the excess of food over the rather small capacity of the stomach. The stomach digestion is highly effici- ent, due not alone to its muscular power, but to the strength of the gastric juice, so that even the bones of the prey. are dissolved and not passed through the intes- tine, as with certain carnivorous birds like the owls. Crocodiles occasionally swallow stones to aid in the tri- turation of their food, just as do the graminivorous birds. To what extent the gizzard was developed in the Theropoda is conjectural, but it would seem as though its need were nearly as great with them as with the crocodiles. One aberrant type of theropod, Struthiom- mus, from the Belly River formation of Canada, has just been the subject of an authoritative paper by Professor Osborn.2 This form is now known to have been abso- lutely toothless, and several theories have been advanced as to its feeding habits—that it was insectivorous, espe- cially ant-eating, or that it fed on small crustaceans or molluses of the seashore, or that it was ostrich-like in habits, browsing upon leaves and buds which its prehen- sile Limbs drew within the reach of the horn-sheathed mouth. Such an assumption as the last, which bears the weight of Osborn’s own opinion, would seem to imply the presence of a more or less efficient gizzard-like stomach functionally comparable to that of the struthious birds. Sauropoda.—The sauropods are clearly of theropod derivation, but it has been pretty generally assumed that they had forsaken the carnivorous habits of their for- “H. F. Osborn, Skeletal adaptations of Ornitholestes, Struthiomimus, Tyrannosaurus, Bull. Amer. Mus. Nat. Hist., vol. xliii, 1917, pp. 733-771. Lull— Functions of the “Sacral Brain” in Dinosaurs. 473 bears for a vegetative diet, and the tremendous growth of certain plants such as the water hyacinth in the Nile or the waters of New Zealand seems to offer an analogy to what might well have been true of certain aquatic vegetation of the Mesozoic upon which these creatures fed. The teeth were now solely prehensile, sufficiently so for their owners’ purpose, but less efficient than those of the Theropoda. The food was in no sense masticated and the inference that a powerful muscular gizzard-like stomach was developed is irresistible, for which the pres- ence of stomach stones, gastroliths, within the ribs of more than one specimen may be taken as added argument. Predentates.—The predentate dinosaurs, on the other hand, had a differentiated mouth armament. ‘The ante- rior or prehensile portion was toothless, except in Hyp- silophodon, but was sheathed with a horny, turtle-like, cropping beak of varying form. The posterior portion of the jaws bore the actual dental battery, consisting of a series of successional teeth which also varied in effici- ency and degree of development in accordance with their owners’ food, as do those of the ungulate mammals. The Jurassic and early Comanchian forms, such as Camptosaurus and Laosaurus, were analogous to the browsing ungulates whose brachiodont teeth are fitted to succulent herbage, while the later trachodonts had a dental battery fully as efficient as that of a horse. These dinosaurs chopped their food into short lengths before swallowing, and it may be that the term mastication, which, however, implies a grinding or crushing rather than chopping, may be properly applied to them. Their need of a gizzard-like organ would seem to be less great than in the sauropods. Stegosaurus, on the other hand, possessed a very imperfect dental battery, as the teeth were both small and relatively few in number—very inadequate appar- ently for their owner’s needs. This genus, however, exhibits a number of characters which, in the Sauropoda, have been taken as indicating an aquatic or at least amphibious life. They are, first, the solid, massive char- acter of the limb bones and the imperfection of their articular ends, those of the stegosaur showing a rugosity fully proportional to those of Brontosaurus. The high position of the ribs, bringing the lungs well toward the dorsal side of the body, and the strongly compressed tail Am. Jour. ae Series, VoL. XLIV, No. 264.—DrEcEmper, 1917. or 474. Lull—Functions of the * Sucral Brain” in Dinosaurs. with its high neural spines and well developed chevrons are also suggestive. Add to these a mouth armament no more effective than that of a sauropod, and the associa- tion of their remains in a common burial, and the infer- ence of similarity of habitus and food is perhaps justi- fied. Just what effect the tall upstanding armor plates would have upon the navigable powers of Stegosaurus is not so clear, but they may have incommoded him under such conditions no more than on land. It may be fairly assumed, therefore, in view of the wide apparent range of feeding habits on the part of dinosaurs, and their relationship to the crocodiles on the one hand and to the birds on the other, that their digestive system was closely comparable both in the development of its parts and in its innervation to that of these living forms. With the birds, the degree of development of the gizzard varies directly with the consistency of the food. Grami- nivorous birds possess the strongest muscular layer and the thickest horny lining, while in the series from the insectivorous birds to the birds of prey this condition becomes gradually less marked and the division of labor between the glandular proventriculus and the mechanical gizzard less noticeable (Newton). The assumption of a similar gradation in the development of this organ in the dinosaurs seems also warranted. Innervation of the alimentary canal. In the reptiles such as the python, crocodile, or turtle, the vagus nerve (Xth cranial) is the principal transmit- ter of stimuli which wnetrate digestive activity, certain of its fibers being distributed to the muscles and mucous membrane of the fauces, the cesophagus, and the stomach, and it finally terminates at the beginning of the intestine at the pancreas. Cranial casts of Tyrannosaurus® and of Stegosaurus and Morosaurus, representing, therefore, the three main dinosaurian groups, all show exits for the IXth to XIth cranial nerves, thus including the vagus, relatively larger if anything than in the crocodile. It is fair to assume, ther efore, ‘that this nerve was at least as well developed in the dinosaurs and that its distribution and function were comparable. *H. F. Osborn, Crania of Tyrannosaurus and Allosaurus, Mem. Amer. Mus. Nat. Hist., new series, vol. i, pt. 1, 1912, pp. 1-30. Lull—Functions of the “ Sacral Brain” in Dinosaurs. 475 Birds.—The vagus (X) of birds arises behind the glos- sopharyngeal (IX) and is connected therewith as well as with the sympathetic system. After receiving branches from the hypoglossal (XII) and taking up the spinal accessory (XI), the vagus runs down the side of the cesophagus to the ventral side of the proventriculus, where, joining its fellow from the other side, it spreads out to supply the stomach. Other branches, leaving the principal stem of each vagus, supply the liver, heart, and lungs, and, as the recurrent laryngeal branch, also supply the distal portions of the trachea and cesophagus. Some fibres of the vagus often extend beyond the stomach, and are connected with the sympathetic nerves of the trunk, supplying parts of the intestinal canal. (Newton.) The approximate agreement in the innervation of both birds and crocodiles is further argument for dinosaurian innervation. | Certain of the spinal nerves (dorso-lumbar) communi- cate with the sympathetic system and thence with the alimentary canal, but their function, in so far as it has been observed, principally in man and certain mammals, is inhibitory, and hence the reverse of a stimulus to digestion. Such sacral nerves as do pass to the alimen- tary canal are distributed to the hinder portion only, beyond the glandular or digestive part. They are stimu- lating, not inhibitory nerves, but their function is merely the elimination of fecal matter and is in no other sense digestive. | The stomach in the mammal at least is largely auto- matic in its movements, as is the heart, and while its activity may be initiated or inhibited by impulses from the vagus or sympathetic nerves, the stimuli which cause the rhythmic movement originate in the muscles them- selves, for this movement will continue after the sever- ance of all nerve connection with the cerebro-spinal or sympathetic centers. The reptilian heart is notorious for its automatic contraction after its excision from the body, and in all probability the heart and stomach of a dinosaur were fully as automatic. All of this seems to show that we have no right in assuming for the dinosaur an innervation or functioning of the alimentary canal at variance with the standardized type of the living amniotic vertebrates. 476 Lull—Functions of the “ Sacral Brain” in Dinosaurs. The spinal canal of Stegosaurus ungulatus has been studied in detail by the author, who finds not only a sacral dilatation but a brachial one as well; that is, from vertebre VIII to XIII, the maximum width, that of 38 mm., as compared with the average of 25 mm., is attained by vertebra XI, which is exactly opposite the shoulder articulations in the Yale mounted specimen. There is also a corresponding heightening of the canal, although this is a less constant feature, for further back (XV and XVII) there is evidence of a hgament or other delimiting structure below the bony roof of the neural arch itself. Brachial and sacral dilatations of the neural canal are most marked in the turtles among existing reptiles, owing to the immobility of the trunk and the consequent reduction of its musculature and associated nerves, the two enlargements being necessary where the nerves depart to the limbs. That this is the whole significance ~ of these two enlargements in Stegosaurus and also in other dinosaurs I have no doubt, and the relative size of each dilatation bears an approximate ratio to that of the limbs innervated, plus in the hinder pair the huge caudo- femoral and other muscles which actuated the tail. I still feel, despite the contention of the German writers, that the ‘‘sacral brain’’—which should not be called by such a term—possessed no unusual function whatever, but only the normal one of transmission and reflex action in an unusual degree, and that to invoke any new and unknown function as a reason for its relatively immense size, especially one connected with digestive efficiency, is not justified by the evidence at hand. Branca further says:* ‘‘We may also think of these animals as sluggish in habit, in consequence of which much less food was required than is the case in an active animal.’’ On the other hand, in warm-blooded animals the largest species occur in cooler climates, because large animals have ‘‘a relatively smaller radiating surface than smaller ones, a factor of the greatest importance in the regulation of body warmth.’’ To the first statement [ can take no exception. The second, however, gives food for thought. In the first place, is it an invariable rule that the largest species of warm-blooded animals occur in cooler climates? The present-day distribution * Loe. cit. Lull—Fruncetions of the “ Sacral Brain” in Dinosaurs. 477 of the elephant, hippopotamus, and rhinoceros does not - bear this out, and even in the Pleistocene the largest elephants, such as Elephas wmperator, were southern forms compared with the smaller, cold- adapted E. primi- genus. With marine creatures Branea’s statement seems more nearly true, for the walrus and huge sea- elephants are both adapted to cold waters, and the same is true of the right whales, Balena mysticetus and B. australis. The sperm whale, on the other hand, 1s trop- ical or subtropical, not occurring, except accidentally, in the polar regions (Flower and Lydekker), while the gereat rorquals (Balenoptera) are found in all seas except the Arctic and probably the Antarctic also. Of the deer, perhaps the largest living form is the Alaskan moose, while no bears in existence can compare in mag- nitude with the great Kadiak bear of the same region. But this argument loses weight if the dinosaurs were not warm-blooded, and though the supposition that they were has been advanced, it is not susceptible of proof. It is within the range of possibility that the temperature of the more agile dinosaurs rose appreciably during the time of their activity, as in many of the so-called cold- blooded (poikilothermous) creatures today; but whether or no any dinosaurs had a mechanism for even a partial maintenance of temperature is unknown. If their bodily heat varied with that of the surrounding air, the greater bulk and hence relatively smaller radiating surface would render them less susceptible to rapid temperature changes, and thus prolong their time of activity by tiding over a brief drop in temperature, but would hardly be available in an extended cooler period. That increase of size in dinosaurs was an adaptation for the conservation of energy, and in this way reduced the relative amount of nourishment necessary for their maintenance, seems hardly probable. 478 3 Scientific Intelligence. SCTENTLE DCy tN bata GaN ee I. CHEMISTRY AND PHYSICS. 1. The Colorimetric Determination of Manganese by Oaxida- tion with Periodate—The rapid determination of small quanti- ties of manganese by converting it into the form of permanganate and comparing the color with a solution of a known amount of the latter was suggested as long ago as 1845 by Crum, and later this principle has been very extensively employed in practical analysis, especially in the examination of iron ores and steels. Several oxidizing agents have been employed for this purpose, lead dioxide, an alkali persulphate in presence of silver nitrate, and sodium bismuthate, all of which are applied in nitric acid solution. H. H. Winuarp and L. H. GreatTHousE have now found a new reagent, periodic acid or its salts, for this purpose and they believe it to be free from all the faults of the previous methods. The periodate is reduced to iodate according to the following equation: 2Mn(NO,), -- 5HIO, + 3H,0 — 2HMn0O, + 5HIO, + 4HNO, Only a small excess of periodate is required, but the success of the reaction depends upon a sufficient concentration of the acid, for otherwise a precipitation occurs, either of manganese - dioxide, or of manganic periodate. The author recommends that the material to be analyzed be brought into a solution containing in 100 ee at least 10 to 15 ee of concentrated sulphuric acid, 20 ce of nitric acid or 5 to 10 ce of syrupy phosphoric acid, or mixtures of two or more of these acids. The solution should have been previously freed from reducing agents by boiling with nitric acid, with the addition of a persulphate if carbon compounds are present, as in the case of steel. Ammonium salts do not interfere, but it is best to remove any hydrochloric acid by evaporating with sulphuric acid, although small quantities may be removed by boiling after adding the periodate. The final reaction is obtained by adding 0.2 to 0.4 g of potassium periodate or of sodium periodate, or an equivalent amount of sodium metaperiodate, boiling for a minute, keeping hot for 5 to 10 minutes and finally cooling. The solution is then diluted to the proper volume and compared in a colorimeter with a standard of known manganese contents, similarly prepared. The solutions used for comparison should not contain more than 1 mg of manganese in 50 ec, for other- wise the color will be too strong—Jour. Am. Chem. Soc., xxxix, 2366. Hi, os Nye 2. The Preparation of Cyanamide from Calcium Cyanam- ide.—li. A. WERNER calls attention to the unsatisfactory yields of cyanamide, CNNH., by the methods heretofore used for the Chemistry and Physics. 479 preparation of this compound from commercial calcium cyanam- ide, CNNCa. This is due to the difficulty of evaporating water solutions of cyanamide, even at low temperatures at diminished pressures, on account of the decomposing influence of water upon the substance. He has, therefore, devised a new method for the operation, which gives very satisfactory yields. In the first place, the amount of pure acetic acid necessary to neutralize a gram of the commercial calcium cyanamide is determined, then to the proper amount of pure acetic acid diluted with nearly its own weight of water in a large mortar a charge of 100 2 of ealclum cyanamide is gradually added in portions of about 15 ¢ with constant stirring while the mortar is kept standing in cold water to avoid much rise in temperature. A pasty mass is finally formed, which is well kneaded and allowed to stand in the air for 24 hours. At the end of this time the mass becomes friable and is coarsely powdered. It is important that a slight excess of acetic acid should have been used, so that the material is faintly acid throughout. The mixture is then extracted with ether in a Soxhlet apparatus, the ether is evaporated at a gentle heat by distillation and finally evaporated to dryness in a desiccator over sodium hydroxide. In this manner a yield of 95.6% of the theoretical cyanamide was obtained.—Jour. Chem. Soe:, erx, 1325. H. L. W. 3. A New Method of Separating Tin and Tungsten.—M. TRAVERS has described a method of analysis which he has applied to wolframites containing tin. The very finely divided substance is fused with anhydrous sodium sulphite in a porcelain crucible at a bright red heat. The decomposition is rapid and perfect, even with minerals containing as much as 50% of tin. The mass is extracted with boiling water, then diluted to 700 to 800 ec, and then slightly acidified with acid. The excess should not be more than 20 ee of normal acid. Brown stannous sulphide is thus precipitated. This carries down a little silica and sul- phides of iron and manganese, but it is entirely free from tung- sten. It is purified by solution in yellow ammonium sulphide, and the tin is determined as oxide in the usual way. The tungsten is determined in a separate sample, starting again with a fusion with anhydrous sodium sulphite. The complete details of the operation need not be given here, as the principal object of this notice is to call attention to the novel method used for decomposing the mineral_—Comptes Rendus, exlvi, 1408. H. Le Ww. 4. Yellow Mercuric Oxide as a Standard in Alkalimetry.— G. Iyoze states that this oxide is a reliable substance for use in standardizing acid solutions, as it is readily obtained in a pure condition, is free from water of crystallization, and is not hygro- scopic. Its use depends upon its reaction with potassium iodide: HgO + 4KI + H,0 — K,Hel, + 2KOH 480 Scientific Intelligence. At least 9 molecules of KI must be added for one of HgO, and in practice it is advisable to use a somewhat larger proportion, -as for example, 10 ce of 60% KI solution for 0.4 g of HgO. As soon as the oxide has dissolved the mixture is titrated with the acid solution to be standardized with use of methyl orange, phenolphthalein, or methyl red as indicator. The yellow oxide can usually be bought pure, but it can be prepared by dissolving 100 g of mercuric chloride in 1 1. of warm water, cooling, and then adding with stirring 625 ¢ of 6.4% NaOH solution. The precipitate is collected, washed until the washings are no longer alkaline to phenolphthalein, air-dried, and then stored in black glass bottles—Zeitschr. analyt. Chem. lvi, 177 (through C.A.). H. L. W. 5. A New Oxychloride of Tun—Harry F. Kenuurr has exam- ined some brilliant tabular or acicular crystals found in cavities in a lenticular piece of metallic tin from an Indian burial mound on Hogtown Bayou, Florida. The crystals were easily crushed to a chalk-white powder, which on heating in the closed tube melted, turned dark, and gave off acrid fumes without a trace of water. The substance gave qualitative tests for tin in the stannous state and for a chloride. A quantitative analysis, made necessarily upon a small quantity, amounting to about 0.2 g, gave results corresponding fairly well with the formula SnCl,.SnO. No reference to the existence of an anhydrous stannous oxychloride was found in chemical literature. No satisfactory explanation of the occurrence of the crystals in the cavities appears to be given, for the suggestion that some chloride solution had access through an opening at the surface does not account satisfactorily for the production of an anhy- drous compound.—Jour. Amer. Chem. Soc., xxxix, 2354. H. i. W. 6. Hquilibrium Temperature of a Body Exposed to Radia- tion.—This problem is discussed in a very lucid and interesting account recently published by Cu. Fapry. The general equa- tion of thermal equlibrium of a small body isolated in free space is derived in the following manner. The size of the body is assumed to be such that its temperature 7 is uniform through- out. S and s denote respectively the area of the surface of the _ body and the cross-section of the beam of intercepted radiation. For brevity, put S/s = 7. The absorbing properties of the sur- face are defined by the absorbing power a which is a function of the wave-length 4:a—=¢(A). The radiation received by the ~ body is supposed to consist of waves that are sensibly plane. It is defined by the curve connecting the energy with the wave- length: H=w(aA). That is, #:-dd means the power which each square centimeter receives from radiations comprised between A and A+dx. It is also necessary to introduce the radiation formula for a black body, since this gives, for each temperature, the curve of energy of radiation for unit surface: R= F (A, T). Chemistry and Physics. 481 Then, the energy absorbed per second by the body is expressed by (oe) sfakar 0 The energy radiated per second is given by ee) SfhaRdxr 0) For thermal equilibrium these two quantities must be equal, hence af $0) Fr, T')dx = [yoda This equation contains but one unknown quantity—the equil- brium temperature 7—and it always gives one, and only one, value for 7. The numerical solution is always easy when the different functions that enter in the equation are given by tabulated data. The radiation equation Ff has a known analyti- eal form. | The special cases outlined below depend upon a simplification of the general equation. The hypothesis is introduced that the incident radiation comes from a black body at a given tempera- ture © and subtending a small solid angle © at the receiving body. The problem is accordingly reduced to that of the thermal equilibrium between two bodies isolated in space, of which one, the emitting body, has a black surface maintained at — a given temperature, while the other, the receiving body, has arbitrary absorbing properties and acquires a temperature which is to be determined. (1) Black or Gray Body.—Then ¢4(A) is a constant that dis- appears from the equation. The integral of F is proportional to T* so that sy T=0M (1) where M = 77/Q. This result can also be derived at once from the law of Stefan. (2) Recewing Body having One Absorption Band.—The wave- length of the center of the band is symbolized by »,. The width of the band must be small as compared with ,, otherwise it is arbitrary. The law of absorption within the band no longer enters into the analysis, and it is not at all necessary for the absorption to be complete for any wave-length. Using Planck’s radiation formula for #(A, 7) Fabry shows that 482 Scientific Intelligence. = . [ log M+ log le a =i) (2) When c/d,® is very large the last equation reduces to teiaa he Tite 0) a C log M (3) This amounts to using Wien’s law instead of Planck’s. 7. Numerical Application: Solar Radiation—To obtain an approximate idea of the order of magnitude of the effects formu- lated the author considers the sun as being a black body at 6000° absolute. The receiving body is assumed spherical (7 —4)and the apparent diameter of the sun is taken as 32’, so that log. M = 12.1 ¢ = 14.350; micron .x degrees umes these conditions equations (1), (2), and (38) lead to the follow- ing results: h, T ste T 0.4 1980°abs Bu 250° 0.5 1700° 10 130° 1 1000° black body 280° 9 550° | It is thus seen that a spherical body having a single absorption band in the violet would attain a temperature approximately equal to the melting point of platinum, when exposed to radia- tion like that which is emitted by the sun and reaches the outer limit of the earth’s atmosphere. This remarkable result is easily explained qualitatively by considering the fact that the body in question can only exchange energy under the form of violet radiation; for, this radiation commences to be emitted to an appreciable extent only at a very high temperature. Until then the sphere absorbs energy without emitting any, and thus its temperature rises. In conclusion it should be remarked that the author touches upon the question of the temperature of space and also gives a tentative explanation of the enhanced brillianey of comets’ tails when near the sun.—Jr. de Phys., v, 207; May- June, 1916. : Hie Sees 8. Colored Flames of High Luminosity—A by-product of an investigation, by G. A. HEMSALECH, on the spectrum of iron was the invention of an assemblage of apparatus which produces flames very intense in color and hence especially suitable for lecture demonstrations. The apparatus comprises four essen- tial parts, which are, the sprayer, the collector, the mixing chamber, and the burner. The sprayer is made of an ordinary glass tumbler the upper open end of which is fitted with a wooden cover sealed in place Chemistry and Physvs. 483 with suitable wax. The cover is pierced by four holes through which two electrodes and two tubes pass. The negative wire is sealed in a glass tube which allows only the lower end of the wire to project a few millimeters. This electrode is placed close to the inner wall of the tumbler and it runs down almost to the bottom of the vessel so as to have the exposed end of the wire immersed in the solution containing the metal, the spectrum of which is desired. The positive electrode has the same form and it coincides with the axis of the tumbler. The tip of the positive wire is adjusted so as to leave a spark-gap of a few millimeters length above the free surface of the liquid. To secure adequate insulation the capillary tube containing the positive wire is sur- rounded by a glass tube of much larger diameter which projects several centimeters both above and below the wooden lid. The wires are composed of iron or aluminium. The inlet tube for alr under pressure runs down near the negative terminal to within a centimeter or two of the liquid surface. The outlet tube is flared at the receiving end which is just below the wooden lid and diametrically opposite to the inlet tube. The collector is simply an inverted glass funnel closed by a wooden disk. Since six sprayers can be used with the same burner, there are six holes in the disk through each of which the outlet tube from a sprayer passes. The upper part of the funnel is joined by a short section of rubber tubing to a glass four- branch tube. The axial extension of the funnel tube passes through a stopper at the lower end of the coaxial mixing cham- ber. The two branches of the glass cross, that are at right angles to this axis, admit oxygen on one side and coal gas on the other. The mixing chamber is made of a brass tube 2.2 em in diameter and 15.2 em high. The top of this tube which constitutes the burner consists of a brass disk 6 mm thick through which one or more holes are drilled. Hach hole should have a diameter not exceeding 2mm. One hole is generally sufficient, but four holes pierced close together at the vertices of a square produce a more brilliant resultant flame. All the inlet tubes of the sprayers branch from one air-supply tube. Since all connections are made with sections of rubber tubing the sprayers can be started, regulated, or stopped by pinch-cocks. The positive electrodes are all permanently con- nected to one coating of a pint-size Leyden-jar. The negative terminals can be successively connected to the other coating of the condenser. An induction-coil giving a 15 em spark is suf- ficient for working the sprayers. The flames should be from 30 to 45 em high and not more than 1.3 ecm in diameter. The cone should not exceed 2.5 em in length. ‘‘The flames obtained in this manner are exceedingly well suited for showing, in a large lecture theatre, the spectra of the more volatile elements, such as Ca, Sr, K, Cu, &c., to an audience 484 Scientific Intelligence. provided with small replica transmission gratings. The rela- tively great length and thinness of the flame obviates the neces- sity for a spectroscope slit.’’—Phil..Mag., xxxiv, 248, October, AON: Hi Si-a 9. X-Ray Band Spectra.—An outline of two papers by DE BroGuiE on this subject was given in the June, 1917, number of this Journal. Essentially the same text has since appeared in another French journal (wde mfra) with the addition of three excellent half-tone plates. The present notice is intended to-call attention to the scientifically beautiful reproductions of the spectrograms of the absorption spectra produced when the X-rays from a tungsten target were allowed to fall upon metallic screens each of about 0.01 mm thickness. The first figure shows the lines of the K and IL series of tungsten together with the dark bands due to the bromine and silver in the photographie emulsion. The remaining figures pertain to the absorption bands (light regions) of molybdenum, cadmium, antimony, barium, tellurium, iodine, mercury, gold, lead, uranium, and thorium. The wave-leneths of the edges of the bands of iodine and tellurium conform to the atomic numbers and chemical sequence of these elements and not to their supposedly anomalous atomie weights.—Journal de Phys., v, 161, May-June, 1916. Hi Sew IJ. Mingeratocy AND GrEOoLoGy. 1. New Mineral Names ; by W. E. Forp (communicated—con- tinued from vol. xlili, pp. 493-494, June, 1917) :— ? Catoptrite. -Aatoptrite. Gustav Flink, Geol. For. Forh., xxxix, 431, 191%7.—Monoclinie, a: 6:3 ¢ = 0:79223 :1 70 4808a, B = 78° 57. Observed forms : @ (100), 6 (010), c(001), #1 ana). d (210), 2 (120), d (012), e (032), 0 (212), p (232), ¢ 272), 7 212). Angles; m:a@= 37 52,°¢-6=67 29. ¢:a= 18 90 70 aoe commonly minute and tabular parallel to 6(010). Cleavage parallel to ¢ (001) very perfect: H.=5°5. G. = 45. = Coloma. black with metallic appearance. In thin splinters, red. Ax. pl. parallel to 0(010). Bx,. makes 14-15° with trace of cleavage. Axial angle small. Inclined dispersion p>v. Strongly pleo- chroic, red-brown to red-yellow. Optically +. Comp.—2Si0O,. Sb,O,.2(Al, Fe),O,.14(Mn,Fe)O. Anal. by Mauzelius, SiO, 7°75, Sb,O, 20°76, Al,O, 9°50, Fe,O, 3°58, FeO. 2:44, MnO 52°61, MgO 3°06, CaO 0°58, H,O 0:11, Total 100°39. Found embedded in calcite with magnetite and other minerals in the Brattsfor mine at Nordmarken, Sweden. Name derived from xarorreov, a mirror, in allusion to its brilliant cleavage surfaces. Ectropite. /ktropite. Gustav Flink. Geol. For. Férh., xxxix, 426, 1917.—Probably monoclinic. In thin crystals, tabular parallel Mineralogy and Geology. 485 to (100) and somewhat. elongated parallel to baxis. Dimensions 2mm. by 1mm. Other forms, rarely observed, are (110), (001), 101). Measured under microscope gave approximate angles ; GO PNO) == "32° 50’, (O0L) (100) = 615" (101) (1200) =56"50"2 Axes @:6:¢=0°74:1:0°34; B= 61° 5’. Cleavage good prob- ably parallel to (001). H.=4. G.=2-46. Luster vitreous to silky. Color light to dark brown. Opaque. In thin section shows yellow color. Non-pleochroic. a=1°62, y =1°63. Opti- eal axial plane parallel to (010). Comp.—Mn,S8i,0,,.7H,O. Anal. by Sahlbom, H,O 8°89, SiO, 35:02, Al,O, 0°75, FeO 5°80, MnO 37°20, CaO 3°59, MgO 7°20, Na,O 0:12, K,O 1°13, other con- stituents 0°19, Total 99°89. Found on garnet crystals associated with barite and calcite in the Norrbotten iron mine at Langbans- hyttan, Sweden. Name derived from éxrpory, evasive, in reference to the difficulty in determining its characters. Flokite. Karen Callisen, Medd. Dansk. Geol. For., v, No. 9, 1917.—Monoclinic. In thin slender prismatic crvstals measuring 1-14 cm. in length by 4mm. in thickness. Observed forms: (110), (100), (010). Faces vertically striated, (100) : (110) = 41°18’. Sections parallel to (010) show twinning on (100). Cleavage perfect parallel to (100) and (010). Conchoidal fracture across prism zone. H. = 5. G. = 2°102. Luster vitreous. Crys- tal transparent and colorless or with faint gold-green tint. At times dark colored from inclusions. Thin sections perpendicular to prism zone show a division into segments with different optical orientation. In the center of crystal the optical axial plane is perpendicular to (010). &6b=y,¢:a=about 5°. Axial angle large. Acute bisectrix probably nearly parallel to ¢ axis and mineral is negative. In Na-light a= 1:4720, y=1:'4736. On warming to 117°-118° sign of double refraction changes. Comp. —A zeolite, H,(Ca, Na,)Al,Si,0,,.2H,O. Anal. by Christensen; SiO, 67°69, AI,O, 12°43, MgO 0°09, CaO 2°65, Na,O 4°36, H,O 13°35, Total 100°57. B. B. fuses easily with intumescence. Insol. in boiling HCl. Found on an old specimen in the Museum at Copenhagen labeled from ‘‘ Eskefjord ? Iceland.” Named after the viking Floki Vilgerdarsen, who gave Iceland its name. Margarosanite. 'This mineral, recently described from Frank- lin, N. J. by Ford and Bradley (this Journal, xlii, 159, 1916), has been found by Gustav Flink at Langbanshyttan, Sweden (Geol. For. Forh., xxxix, 458, 1917). The Swedish occurrence shows slender prismatic crystals, often in radiating groups. Usually the crystal faces are striated or curved. 328. Clark, Wi. 8.247. Mrysdale,C. W338. Hague, Arnold, 73. Hughes, ©. McK., 160. juncersens (Echo. -S86. Kennedy, H. T., 160. Sarasin, E438: Stone, Gi E3) 86: Ontario Basin, Study of, Coleman, Ophi- | 351, 487. | Ophiuroidea, Japanese, Matsumoto, 404. 336. Physical Science, Laws of, North- rup, 79. Physics, Problems in, Masius, 404. Poisson’s Equation, failure of, Prasad: 2333. Poster, -f. E. separation of . gal- lium, 221. Powers, S., granite in Kansas, 146. Priestley memorial, 332. Ourke, 1. cl ., |: radioactivity: 40 meteorites, 237. R Richthofenia in Texan Permian, Bose, 157. Ries, H., dolomitic clay, 316. 492 ROCKS. Granite, Kansas, Powers, 146. Lavas ‘of Morro Hall: Gal. War ing, 98. Magmatic sulphide ores, Tolman and Rogers, 156. Pumicite, Nebraska, Barbour, 83. Rutherford, E., penetrating power of X-rays, 401. Ss Sayre, Materia Medica, 86. Scott, S. E., detection of germa- nium, 313. Scott, W. B., Evolution, 84. Shannon, E. V., famatinite from Goldfield, Nevada, 460. Shuler, E. W., Dinosaur tracks in Glen Rose: limestone, Texas, 204. Simotomai, H., Tarumai dome in Japan vox. Smith, E. F., 70. Sodium vapor, ionizing potential, Wood and Okano, 401. Solar radiation, 482. Solution, Nature of, Jones, 78. Stefanini, G., geological history of Venetia, 200. Sulphur, recovery, Wells, 330. Life of Robert Hare, ak Tarr, W. A., origin of chert in the | Burlington limestone, 400. Temperature, equilibrium iota body exposed to radiation, Fabry, 480. Tashiro, S., Chemical Sign of Life, 4. Texas, Dinosaur tracks, Shuler, 204. — University bulletin, 158. Time, wave work as a measure of, Coleman, 351, 487. Tin, oxychloride of, new, Keller, 480 < , separation, Trav- ers 479. | Waring, G. A. and C. A., INDEX. Tuttle, L., Theory of Measure- ments, 20, Twins, Biology, Newman, 84. U United States, Bureau of Mines, 8o. — geol. survey, 405. Urine, Secretion, Cushny, see V Van Name, R. G., tri-iodide and tri- bromide equilibria, 105' 10ntza- tion of cadmium iodide solutions, 453. Venetia, ¢ nini, 200. Vennes, H. J., retardation of alpha particles by metals, 60. Verwiebe, W. A., Devonian shales of Ohio and Pennsylvania, 2e8 Vogdes, A. W., Notes on Paleozoic Crustacea, 336. Volcanic Eruption, on, Mies Se Helens, Wash.) jillsonmsar Volcano, Tarumai, Simotomai, 87. Volcanologic investigations at Kil- auleéa,. Jagear, anon: geological history, Stefa- WwW lavas of Morro bili Se: California, 98. Wellisch, E. M., motion of ions and electrons through SaAsSes, ae Wickham, H. F., fossil. beetles from Sangamon, Til, 137 Winton, A. L., Food Analysis, 77, x X-Ray band spectra, de Broglie, A84. — from certain metals, tion, Kaye, 334. — penetrating power, Rutherford, . AOT. — relations between Ishiwara, 335. composi- spectra | of, } Warps Naturat Science EstaBlisHMent A Supply-House for Scientific Material. Founded ee Incorporated 1890. A few Oe our recent ecireulars in the various departments: Geology: J-3. Genetic Collection of Rocks and Rock- forming Minerals. J-167. Price List of Rocks. Mineralogy: J-109. Blowpipe Collections. J-74. Meteor- ites, J-150. Collections. J-160, Fine specimens. Paleontology: J-134. Complete Trilobites. J-115. Collec- tions. J-140. Restorations of Extinct Arthropods. Entomology: J-30. Supplies. J-120. Life Histories. . J-128. Live Pupae. . Zoology: J-116. Material for Dissection. J-26. Compara- tive Osteology. J-94. Oasts of Reptiles, etc. Microscope Slides: J-185. Bacteria Slides. Taxidermy: J-138. Bird Skins. J-139. Mammal Skins. Human Anatomy; J-16. Skeletons and Models. General: J-181. List of Catalogues and Circulars. Wards Natural Science Establishment 84-102 College Ave., Rochester, N. Y., U.S. A. Publishers: WILLIAMS & NORGATE, 14 Henrietta Street, Covent Garden, London, W. C. “SCIENTIA” INTERNATIONAL REVIEW OF SCIENTIFIC SYNTHESIS. Jsswed monthly (each number consisting of 100 to 120 pages). , Editor: EUGENIO RIGNANO. ““SCIENTIA”’ continues to realise its program of synthesis. It publishes articles which relate to the various branches of theoretic research, and are all of general in- terest; it thus enables its readers to keep themselves informed of the general course of the contemporary scientific movement. **SCIENTIA’’ appeals to the codperation of the most eminent scientific men of all countries. ° It has published articles by Messrs. Abbot (Washington) - Arrhenius (Stockholm)-Ashley (Birmingham) -Bechterew (Petrograd) - Bohlin (Stockholm)- Bonnesen (openhagen) - Borel (Paris) - Bottazzi (Napoli) - Bragg (Leeds) -Bril- louin (Paris) -Bruani (Padova) - Castelnuovo (Roma) -Caullery (Paris) -Chamberlin (Chicago)- Ciamician (Bologna) - Clark (New York) - Costantin (Paris) -Crommelin (Greenwich) - Daly (Cambridge, U. 8. A.) - Darwin (Cambridge) - Delage (Paris) - De Martonne (Paris) - De Vries (Amsterdam) - Durkheim (Paris) - Eddington (Greenwich) - Edgeworth (Oxford) - Emery (Bologna) - Enriques (Bologna) - Fabry (Marseiile)- Fisher (New-Haven, U.S. A.)- Foa (Torino)-Fowler (Zondon)- Fredericq (Liége)-Galeotti (Vapoli)- Golgi (Pavia) - Gregory (Glasgow)- Guignebert (Paris) - Janet (Paris) - Jespersen (Gentofte) - Kapteyn (Groningen) - Kidd (Oxford)-Langevin (Pavis)-Lebedew (Moscou)-Lodge (Birmingham)-Loisy (Paris) Lorentz (Haarlem) - Loria (Torino)-Lowell (flagstaff, U. S. A.)-Maunder (Green- wich) - Meillet (Paris) - Pareto (Lausanne) - Peano (Torino) - Picard (Paris) - Poincare (Paris) - Puiseux (Paris) - Rabaud (Paris) - Righi (Bolegna) - Rignano (Milano)-Russell (Cambridge)- Rutherford (Manchester)-Sayce (Oxford) -Schiapa= relli (Milano) - Seligman (New York) - Sherrington (Liverpool) - Soddy (Glasgow)- - Svedberg (Upsala) - Tannery (Paris)- Turner (Oxford) - Vinogradoff (Joscow)-Vol= terra (Roma)- Von Zeipel (Upsala) -Westermarck (Helsing/fors)-Willey (Montreal, Canada) -Zeeman (Amsterdam) - Zeuthen (Kopenhagen), and more than a hundred others. ‘“*SCIENTIA’’ publishes, at present, in the section dedicated to sociological articles, a series of studies on the present questions of an international character raised by the war. **SCIENTIA’’ publishes its articles in the language of its authors, and joins to the principal text a supplement containing the French translations of all the articles that are notin French. (Write fora specimen number.) Annual Subscription: 24 sh. post free. Office: Via Aurelio Saffi, 11- MILAN (Italy). CON TEN Ts: #4 Arr. XXXV. —Origin of the Chert in the ah t¢ stones.by WioA ARE. =" ea ee ‘ Solutions; ey R. G. Van Name cad W. . ee XXXVII.—Famatinite from Goldfield, Nevada: i SHANNON Foc U2 |e eee ie XXXVIII.—On the Functions of the “ < Sorel | Dinosaurs; by R. 8. Lunt 2... --- Sp SCIENTIFIC INTELLIGENCE. Chemistry and Physies— Colorimetric Determination of Me Oxidation with Periodate, H. H. Wituarp and L. H. Gz Preparation of Cyanamide from Calcium Cyanamide, E. A. 478.—New Method of Separating Tin and Tungsten, \V Mercurie Oxide as a Standard in Atkalimetry, G chloride of Tin, H. F. Kexturr: Equilibrium > Exposed to Radiation, C. Fasry, 480.—Numerica. Radiation: Colored Flames of High Lumines Gis X-Ray Band Spectra, pz Brosuig, 484. ‘w Mineralogy and Geology—New Mineral Names, W.E. Fon, 48 tive Mineralogy, W. S. Barney, 486. — Wave Work as a Mei A Study of the Ontario Basin, A. P. COLEMAN, 487. Miscellaneous Scientific Intelligence—National joe f Sel can Association for the Advancement of Science: N fox Ki Jonus, 487.—Science and Learning in France, J. H. INDEX, 489. yal ay i ofhian A ‘ SMITHSONIAN INSTITUTION LIBRARIES “UII 3 9088 01298 5966