¥ yes ma Se te Node? 505.73 G > THE AMERICAN JOURNAL OF SCIENCE. Epiror: EDWARD S. DANA. ASSOCIATE EDITORS Proressors GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW ann WM. M. DAVIS, or Camprince, Proressors ADDISON E. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT FE. GREGORY AnD HORACE S. UHLER, or New Haven, Proressor HENRY S. WILLIAMS, or Irnaca, Proressor JOSEPH S. AMES, or Batrimore, Mr. J. S. DILLER, or Wasuineton. FOURTH SERIES VOL. XXXIV—[WHOLE NUMBER, CLXXXIV]. WITH ONE PLATE. NEW HAVEN, CONNECTICUT. OAC D. 22b45° Shr aan? ie ‘seston ch : fag inte bee : whi Aram aan “e ey yee aa ¥ say 4 es ‘ie a § ms CONTENTS TO VOLUME XXXIV. Num J sr°99. : Page Arr. L—Storm King Crossing of the Hudson River, by the New Catskill Aqueduct, of New York City; by J. F. LETS b-day a a Se ee es 1 IIJ.—Lake Parinacochas and the Composition of its Water ; by G. S. Jamimson and H. Bineuam .------.--------- 12 III.—Shell Heaps of Maine; by F. B. Loomis and D. B. VOTES Sa Be AUR ea ype aay ee tee en Wi IV.—Mixtures of Amorphous Sulphur and Selenium as Immersion Media for the Determination of High Refrac- tive Indices with the Microscope ; by H. KE. Merwin Gimli SY LOST 2 CNS 2 ES gee ous ee eet ee Serer: V.—Asymmetry in the Distribution of Secondary Cathode Rays produced by X-rays; and its Dependence on the Penetrating Power of the Exciting Rays; by C. D. CO TIRING p18 LE Sk 8 es IA a Re at ree hier 1 VI.—Derivation of the Fundamental Relations of Electro- dynamics from those of Electrostatics; by L. Pacu... 57 VII.—Hydrolysis of Esters of Substituted Aliphatic Acids ; Die MOR USEML a ge gon eo ese ne, eee OO VIIT.—Some Suggested New Physiographic Terms ; by DEL. IDS GUS piping SE Me 3 Be Ree eC els ergs A epee eC SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Higher Layers of the Atmosphere, A. WEGENER, 88.—Allen’s Commercial Organic Analysis, 90.—Methods for Sugar Analy- sis, A. Given: Laboratory Exercises in Physical Chemistry, J. N. Prine : Entdeckung des Radiums, Mme. Curtin; Uber neuere thermodynamische ~Theorien (Nernstsches Warmetheorem und Quantenhypothese, M. PLANCK : Studies in Terrestrial Magnetism, C. Caren, 91.—Physical Study of the Firefly, W. W. Coptentz: Elements of Hydrostatics, G. W. ParKER, 92. Geology and Natural History—Hffect of Topography and Isostatic Compen- sation upon the Intensity of Gravity, J. T. Hayrorp, 92.—Geological Sur- vey Branch New Zealand Department of Mines, 93.—Physical Geography for South African Schools, A. L. DuTorr: Notes on fossils from limestone of Steeprock Lake, Ontario, C.D. Waucott: Report on some recent collec- tions of fossil Coleoptera from the Miocene shales of Florissant: Glacial Man in England, 94.—Geological and Natural History Survey of Connecti- cut: Report on the Progress and Condition of the Illinois State Museum of Natural History 1909-1910: Mineralien-Sammlungen ; ein Hand- und Hilfsbuch fiir Anlage und Instandhaltung mineralogischer Sammlungen : Der Diamant, 95.—The British Tunicata: Ceylon Marine Biological Reports, 96. Miscellaneous Scientific Intelligence—Carnegie Foundation for the Advance- ment of Learning: Proposed Expedition to Crocker Land, 97.—Expedi- tion Antarctic Frangaise (1903-1905): Publications of the Smithsonian Institution, 98.—British Museum Catalogues: Publications of the Alle- gheny Observatory of the University of Pittsburgh: Bref och Skrifvelsen af och till C. vy. Linn&, 99.—Essais de Synthése Scientifique: Rationalist English Educators, G. E. Hopeson: Zeitschrift fur Gahrungsphysiologie; allgemeine, landwirtschaftliche, und technische Mykologie : Meteorology, A Text-book on the Weather, its Changes, and Weather Forecasting : Chemical Research in its Bearings on National Welfare, 100. iv CONTENTS. Number 200. Page Art. [X.—Interferometry of Air Carrying Electrical Cur- rent ; by C.-BaRuUS’.20s¢2 20.4.0 -c ee 101 X.—Electrolytic-Analysis with Platinum Electrodes of Light Weight ; by F. A. Goocu and W. I. Burpick ..--._.- 107 XI.—Siliceous Odlites of Central Pennsylvania; by V. ZIBGIBR . o20 Sect) ce he sd 225 Ue XII.—Composition, Structure, and Hardness of Some Peru- vian Bronze Axes; by H. W. Foorr and W. H. Burtt 128 XIII.—Photoelectric Effect of Phosphorescent Material ; by C. A. Burman ......-- aad gach. ate eee XIV.—Tertiary Fungus Gnat ; by O. A. JoHANNSEN ------ 140 XV.—Heat of Formation of the Oxides of Vanadium and Uranium, and Eighth Paper on Heat of Combination of Acidic Oxides with Sodium Oxide ; by W. G. Mixrer 141 XVI.—Chattanoogan Series with Special Reference to the Ohio Shale Problem ; by E. O. Unricw .--..--.------ 157 XVII.—Pseudomorphs after Stibnite from San Luis Potosi, Mexico; by W. E. Forp and W. M. Brapiey------.-- 184 XVIUI.—Stratigraphic Relations of the Devonian Shales of Northern Ohio ; by Hin VeeKGN pie eee 187 XIX.—Estimation of Titanium in the Presence of Iron ; by W.. M. THORNTON, Ot sce) ey eee eee eee 214 XX.—Pleistocene Plants from the Blue Ridge in Virginia ; by Ea W. BERRY. 22) eee eee PE Ba See a2 = 218 SCIENTIFIC INTELLIGENCE. Geology and Natural History—Age of the Plant-bearing Shales of the Rich- mond Coal Field, EK. W. Berry, 224.—Geological Survey of New Jersey : Topographic and Geologic Survey of Pennsylvania: State Geological Sur- vey of Wyoming, 225.—Maryland Geological Survey : Cape of Good Hope, Department of Mines, 15th Annual Report of the Geological Commission, 1910, 226.—Monograph of the British Desmidiacee, W. West and G. S. West: Elementary Plant Biology, J. E. Peapopy and A. EH. Hunt: Illus- trated Key, etc., J. F. Conzins and H. W. Preston, 227. Miscellaneous Scientific Intelligence—Introduction to Analytic Mechanics, A. Zrwet 2nd P. Fierp, 228. Obituary—F. ZiRKEL: E. SrRASBURGER: W.M. WoopwortH: M. L. DE BOIsBAUDRAN, 228. CONTENTS. v Number 201. Page Art. XXI.—Ionization by Collision in Gases and Vapors ; DAF NMOS [EZ SR SI ae ati Dubna ale oe aerate Mele MLe ed 229 XXII.— Geology of Arisaig-Antigonish District, Nova Scotia by Mi a, Wimmrdmen = ” re < 5) Le J. F. Kemp—The Storm King Crossing Fie. 5. an Schist Manhattor ntrusien® ) (with * Naw City: Cortlandt St. Hudson & Manhattan Tunnel Co Sandstone Profile of the Hudson River opposite West 32d st., New York City, based upon data obtained by the Pennsylvania Railroad. Fie. 5. which also fails in the anal- ysis of the sea water, No. III. On the other hand, in the two deep waters, Nos. IV and V, the calcium chloride is again high. The magne- sium chloride, the calcium sulphate and the magnesium sulphate are all similar to the analysis of the sea water. Obviously the bore hole yielded a water which in its entirety cannot be referred to river or oceanic sources, although it may have an admixture of them. Its high calcium chloride is like the deep-seated waters of the rocks, which if ever derived from the surface, have un- dergone notable changes in chemical composition. In the earlier paper, Oc- tober, 1908, the writer in closing discussed the char- acter of the lower Hudson Valley in the light of the depth as there determined for the bed-rock at Storm King. The deepest record was 608-6, but of course the middle of the river was then shrouded in mystery. It was assumed further that water was the sole eroding agent. No good data were available regarding bed-rock conditions from Storm King to the sea. The deep off- shore channel of the Hud- son had, however, been well developed by sounding. Since then the data se- cured by the engineers of the Pennsylvania Railroad in connection with the tubes across the Hudson have been made available. A row of wash-borings has been put of the HIudson River. 11 down from the New York shore at West 32d street to New Jersey, and they have been supplemented by several diamond drill holes.* As shown by the accompanying cut (fig. 5), taken from Mr. Rogers’ paper, the diamond drill borings are about 1100 ft. apart at the most important place and 900 ft. apart farther west. It is a very interesting question whether the bed-rock continues across these intervals at a depth of 300 ft., at which depth it appears on either side, or whether there is a deep and relatively narrow notch in the 1100 ft. stretch, so deep indeed as to permit water which had traversed the Storm King pass, 40 miles north at 800 ft. or thereabouts below tide, to reach the sea by a pass, necessarily still lower, opposite New York City. Advocates of water erosion as the cause of the Hudson gorge, would assume the existence of the notch. The bottom of the notch could not be less than between 500 and 600 feet below the nearest known bed-rock. If water erosion is not the cause, then the Storm King pass has been over-deepened by some other agent than water. There is but one other agent and that is ice. Sub-glacial erosions would therefore follow. Opinion in America in recent years has been predominantly against glacial erosion, but various lines of evidence are making this process more and more probable. The Storm King Crossing would seem to be a favorable place for it. Between Storm King and Breakneck Mountain the Hudson valley is very narrow, as is shown by fig. 3. A short distance north the old crystalline rocks end in a sharp fault escarpment which crosses the valley in a northeast direction. The softer Paleozoics have been easier marks for erosion than the hard erystallines and now present a broad valley. The same con- trasts existed at the opening of the Glacial epoch. As the ice-sheet moved southward its only outlet was the narrow pass which was occupied by the river. The ice was crowded into this gorge and certainly was in as favorable a position to deepen its channel as ice under ordinary circumstances ever is placed. It may well be that in this way the channel was over- deepened and that as the ice-sheet approached the terminal moraine just below New York, it became gradually feebler and rose to higher levels, leaving in the end a shallower chan- nel filled with ground moraine. While not denying the possibility of water erosion with a notch at the Pennsylvania Railroad crossing, yet over- deepening by ice has been favored by Dr. Berkey and the writer in their reports as consulting geologists to the Board of Water Supply, but Professor W. O. Crosby, also reporting to the Board as consulting geologist, favors water erosion. *G. S. Rogers, The Character of the Hudson Gorge at New York City, School of Mines Quarterly, xxxiii, 26-42, 1910. Columbia University, New York City. 12 Jamieson and Bingham—Lake Parinacochas. Arr. I].—Lake Parinacochas and the Composition of its Water ; by Guoreer S. Jaminson and Hrram Binenam. A SAMPLE of water from Lake Parinacochas was collected by one of us (H. B.) of the Yale Peruvian Expedition of 1911. The analysis of this sample is discussed in the present paper. Lake Parinacochas is situated in Peru, between 15° and 16° 8. latitude and 73° and 74° longitude, west of Greenwich. In other words, it isabont 150 miles N.W. of Arequipa, and about 170 miles S.W. of Cuzco. Its elevation is about 11,500 feet above sea level. It is fed by half a dozen small streams or large brooks, and at present has no visible outlet. In past geo- logical history it was much larger, and had an outlet which is still visible. In a river valley not many miles from the lake, and at a lower elevation, there are a considerable number of large springs which may possibly be fed by the lake. On the borders of the lake are also numerous small springs, which generally occur in swampy hillocks, three or four feet higher than the surrounding plain. The natives told us that in the wet season the lake was higher than in the dry season. The truth of this is evident from the marks of the salt left by evaporation of the lake as it shrinks during the dry season. In general the surrounding region is at present semi-arid, and is inhabited by a pastoral population. Thousands of sheep and some hundreds of cattle feed in the pastures which were for- merly the bed of the lake. There are ruins of villages and agricultural terraces, indicating that at a previous period there was a much larger population here and that agriculture was more common than at present. We spent eight days (Nov. 5-12, 1911) in three different camps on the shores of the lake and navigated it in a folding boat, the first boat or canoe that had been seen on the lake, according to the natives. We found that the lake was at pres- ent about 18 miles in length, and 7 miles in width, with a nearly uniform depth at the time .of our visit, November, 1911, of 43 feet. Several hundred soundings failed to show more than 5 feet anywhere. Judging by the salt marks on the shore, the lake is probably about a foot deeper in the rainy season. The natives told us that in the dry season the lake sometimes was very much lower than we saw it. The water has had a repu- tation of being brackish for nearly a century,—ever since any one took the trouble to say anything about it in print. The first and only reference to the character of the water that I have been able to find is in the Memoirs of General William Miller (London, 1828), where on a rough sketch-map he notes that the water of Lake Parinacochas is brackish. ‘aSUI[IA JUeIoUR ATOA B FO SUINA OY} YIIM petoaoo ATeQAed [IY e WorT ‘UOTyTpedxe oY Jo TosUTSUE [eoTsojooyore ‘Texon, "Ty “H aw Aq uexye} sv vuvrourd eyy, ‘oye, oyg Jo Aqtmoerj}xe puvy-7y4SIt oy} ye peInoess seM TyIMesey peysiqnd SI SISA[VUe OSOTM 1eyVM FO UouTIOedSs oY], “PUNOISaLOJ 94} UL SeLOYS poteAod-j[vs ‘10}MeD ot} UL vxVl oY} JO JoTJNO " — TOULLOF 94} “YYSII OYJ WO SMOYS VAVSRIVG OULD[OA JOUT}xO oY, “4sSvoyyToU SuLyoo, ‘seqooovulieg syeyT “fT NIA { | | "1 ‘Oly 14 Jamieson and Bingham—Lake Parinacochas and the Parinacochas is a contraction from the Quichua word * Parihuanakocha,” which means the lake of flamingoes. The Quichua or Inca word for flamingoes is ‘“ Parihuana ”; “ Kocha” means lake. There are thousands of pink flamin- goes, but we could find no evidence that they nested here. We found the lake to be the home of a great many birds, which, in the order of the frequency with which we saw them, are as follows: flamingoes, gulls, small divers, large black ducks, sandpipers, black ibis, large teal ducks, large geese, ground owls, and woodpeckers. Owing to the shallowness of the lake and the fact that a sweet marsh grass is abundant in Fie. 2. - Fic. 2. Lake Parinacochas. Flamingoes wading in the lake. it, the cattle are accustomed to wade sometimes as much as three quarters of a mile from shore, in order to get the particu- larly succulent grasses and water weeds. We saw no fish, and were told by the natives that there were no fishes in the water, but there was a great deal of small animal life. Large parts of the lake are covered with alge, making it very difficult to row. In camps, near the banks of the lake, at night the temperature went as low as 22° F. It will be seen at a glance, from the table, that the temper- ature of the water of the lake varies from 61° to 67:5°, depend- Composition of its Water. 15 ing chiefly on the time of day. In the mornings there is no wind, and the water near the surface is warmed by the sun. Shortly after noon a strong wind arises and stirs up the water and cools it near the surface, causing a slight fall in its tem- perature. The specimen of water was taken near the shore, in a location much frequented by birds and cattle. Temperature of Water 10 in, below surface. Water, near shore, 10:00 A. M., 61° F. 24 ft. deep, 4 mile from shore, 10:30 a. M., 63°5° F. a 34 ft. paminlesy cor) bi eal POO WAs Mes ” 64° F. 2 16 whit 150 Vsti ieee MOON Goroeal)s et eo iP mile @ 12-30) 675° Ht es Apatite mccoy: miles, 66° 8s 1:00, 65° F. met te vende ce) "0p ce) 2 1:30, 63° F. Seated sigma (oro e2<00.63"5" Ei. Serra ane ee ch ep CC cates Kee O80. 65° Bi The sample of water, which measured about 300°, had a brackish taste and a slight brown color. The following results -were obtained by analysis: Percentage com- Se position of the Milligrams inorganic per liter. constituents. Chitormen(C\ ets sees eos 2 56.070 46°87 % Dulpmatem(SO)) esos aes oreo cee 12760 10°59 Carbonate (CO ies a2s. abo -e 264:°0 2°16 Borate (By Os) ei iosc at os seen) VOLO 1°36 Naiiate (GNI nia. ess 477 0°40 Phosphate (PO,) . oe aan eeteees 59 0°05 ilica(SiOh eee eee ye oe 8:0. 0:07 Ohm: (ANID) nee ee Cae 3935°0 32°64 eotassiuime (KO) ese en Mees 464°0 3°83 Waller (Cal eases 142.0 118 Magnesium (Mic) 22 see= sc 8 99°0 0°82 hone) vate Se Ce 34 0:03 Sallimiit peters tes teen ed Be EA ORO ORO 100°00 The total residue obtained by evaporation at 100° C. was 12,548 milligrams per liter; the difference between the total residue and the salinity, amounting to 489-0 milligrams, repre- sents in a measure the amount of. organic matter “held in solu- tion. The following table gives the hypothetical combination of the acids and bases : 16 Jamieson and Bingham—Lake Parinacochas. Sodium chloride... j- 95. se Gee 9324°0 podiuno ysulphaten=.< acs: Geena 649°3 podium: borate = 2am cess emo” eile Potassium sulphate ..-..--.---- 965°6 Potassium (nitrates seu 780 Calerm:sulphates = 2.8.54 eee 435°8 Calcium carbonates. 22. .4- aaee 33'S Magnesium carbonate..-..---.-.- 342°8 Herricphosphatescs -\-e vases e. 9°3 Silica 0) See eae ote cae eee 80 12,059°0 In order to compare the percentage composition of Lake Parinacochas with that of the Atlantic Ocean and several other typical saline waters, the following table is given. The analy- ses were taken from Bulletin No. 330 of the United States Geological Survey, entitled “The Data of Geochemistry,” by F. W. Clarke, 1908 : Water C1 | $0. | CO; | PO,|B.0;| Na | K | Ca | Mg|$,0.)NO.| Fe | Dota 1. Lake Parinacochas) 46°86| 10°59) 214) 0:05) 1°36 | 82 63/8:85 118 0-82) 0:07/ 0:40) 003) 12,059 2. Soda Lake, Nevada) 36:51) 10°36 13-78} --.- | 0°25 | 36-63) 2.01) trace) 0:22) 0:°64|___-|____| 113,70 ) 3. Atlantic Ocean__-| 55-48} 7-69, 0-21] -... | ---- |80°60] 1:10) 1:20 3-72] _-..|--_.|.-..| 34,408 4. Great Salt Salt__-| 55°69) 6:52 trace | -__. |trace | 32:92) 1-70) 1:05 | 2-10) __._ |____|_.._| 230,308 | 5. Lagoon of Tamun- : { ica a@hileeses== 50°44) 9:17) ---. | ---- | ---- | 85 35/2:29) 0.01 | 0°60) ____ | 2:14)__._| 285,509 6. Koko Nor, Tibet__| 40:05) 17°84 553) 0°02) --.- | 3060 1:08) 1:77 | 2:90) 009|_..-|-...| 11,008 : Analysts: 1, G.S. Jamieson; 2, T. M. Chatard; 3, G. Dittmaes; 4, E. Waller; 5, F. San Roman; 6, C. Schmidt. * An approximation only. Not given in the Bulletin. On comparing the percentage composition of the several waters given in the table above, it is seen that Lake Parinaco- chas occupies an intermediate position. Its salinity is about one-third that of the Atlantic Ocean, or about one-twentieth that of the Great Salt Lake. It is also noticeable that it contains much more carbonate and much less magnesium than either the Atlantic or the Great Salt Lake. Yale University, New Haven, Conn., April, 1912. — Loomis and Young—Shell Heaps of Maine. 1A Arr. IIT.—On the Shell Heaps of Maine; by F. B. Loomts and D. B. Youne. Durine the summer of 1909 a party of five from Amherst College spent two months working in the shell heaps of the Maine coast, one half of the time being devoted to a careful survey of one heap on Sawyer’s Island, near Boothbay, the second month being spent in a more rapid investigation of sey- eral heaps for comparison. These shell heaps are found all along the coast from Staten Island, Long Island, and the adjacent shore (there being some thirty of them in the neighborhood of New York City), scat- teringly along the Connecticut and Massachusetts shores, and in ever increasing abundance along the coast of Maine, there being in that state a thousand or more in all, when big and little are counted. For over forty years some work has been done in them, mostly sporadic, and without real system ; so that there are considerable collections in the Peabody Museum at Harvard, in the American Museum in New York, in the hands of Mr. Dwight Blaney on Ironbound Island, Maine, and belonging to Prof. J. T. Bowne of Springfield, Mass., not to mention a number of scattered collections. The abundance of the material and the widespread distribution of the heaps make it very desirable that systematic collecting and recording of the collections should be carried on. The vast number of the heaps indicate that they represent an important phase of the life of a large number of Prehistoric Indians. The heaps vary greatly in size, from the one at Damariscotta, covering two or three acres, and some 20 feet thick, to tiny accumulations covering only a few square yards and only inches deep. In general, they are composed of the shells and refuse, which have been thrown away on a camp site, mixed with the ashes of camp- and cooking-fires, the bones of animals eaten, and bits of broken pottery, broken and lost tools, or anything which may be included in camp left-overs. The major part of the heap consists of more or less broken shells of the soft-shelled clam, though other mollusca may occur in great numbers; for instance, quahogs, blue mussels, or oysters, the great heap at Damariscotta being entirely composed of oysters: and there are usually bands in each heap composed entirely of other forms. The heaps are located apparently with reference to camping convenience; nearness to drinking water, food supplies, and protection from storms, all being considered ; so that near any big clam flat a shell heap may be looked for. AM. JouR. Scit.—FourtH Serius, VoL. XXXIV, No. 199.—Juty, 1912. 9) 18 Loomis and Young—Shell Heaps of Maine. One finds the heaps in a variety of ways, the commonest being by observing the white eleam of the shells, where the sea has cut into and exposed the. heap. Or they may be discov- ered as the result of plowing a field, whereby the shells are Fig. 1. 02mile 04. 06 Fic. 1. Diagram of the south end of Sawyer’s Island, showing the posi- tion of the shell heap, and the position and extent of the workings of the Amherst party. Scale, 1/1562. - brought to the surface; or by the bringing up of shells from rodent burrows, ete. The Sawyer’s Island heap was chosen for careful study, because it is of considerable size, and has never been disturbed by previous excavation, and also because it is located in a pas- ture, where the turning of it over caused very little damage. This heap lies on the southeast side of the island and in it two excavations were made, the first uncovering about 400 square Loomis and Young—Shell Heaps of Maine. 19 feet, where the depth of shells was about two feet ; the second making a cut through the center of the heap, and covering over 2800 square feet, where the depth was about five feet. The method used in working the heap was to plot out the surface in sections five feet wide, and as each section was worked, every find (of a tooth, tool, bit of pottery, ete.) was recorded, both as to its horizontal position and vertical depth. Everything with which man had anything to do was saved, except the shells and ashes. In this way the relative as well as the actual abundance of every article in the heaps was recorded.* Sections of the heap were plotted from time to time. The material from Sawyer’s Island heap includes 1040 finds, each with an individal record, representing, of course, several times that number of bones. This material is the basis of the discussion of conditions on Sawyer’s Island, and is sup- plemented by about three times as much from other heaps, where collections were made for comparison, and to give a basis for generalizations. Beside these, the collections above mentioned have been examined in order to make a check on any generalizations. The character of any heap varies at different places and levels, but the following may be taken as a typical section : No. 1. Sod with shells in the grass roots... 3” 2. Shells with small amount of ash... 14 3. Ashes with some shells.______._.- 3 Ae Olean shelliiswen sissies see. eee 4 5. Ashes with some shells .__.___.__- 4. 6. Shells with small amount of ash _._ 14 Het @leaiashesyem a fe bere eG Rotalearthessesn cies aoa ee Rae ae 48 inches of shell heap. No two sections are exactly alike, but everywhere there was the basal layer of ashes, and always the banded character of the material. In layers 2, 3, 5, and 6 the shells were very much broken up, apparently due to the tramping and building of fires on them. Where the shells were but little broken, and free from ashes, they would seem to indicate rapid accumu- lation, and offered but little in the line of finds. Under the heaps we found very few indications of the pits which are described as characteristic of the heaps along Long Island,+ or *JIn each case a recorded find indicates a tool, an animal eaten by the Indians, a pottery vessel, etc. Thus the record of an animal is based on the presence of a jaw, crushed fragments of bone, or single bones, though recorded, not being included in the enumerations. A score of fish vertebree may indicate but a single individual, so in enumerating, the count is based on the number of premaxillz. In the case of pottery fragments each assem- blage of fragments is enumerated as one find. + Alanson Skinner, American Museum Leaflet, No. 29, 1909. 20 Loomis and Young—Shell Heaps of Maine. those which are so characteristic of the Baum Village site in Ohio.* Only under the heap on White Island, and under the one on Calf Island, did anything of the sort occur, and neither of these pits was distinguished by having in it very much more Section of the Sawyer’s Island heap, from «x to y as on map fig. 1, showing the relative layerings of the shells, ashes, ete. Bre. 2. Length, 15 feet. than the neighboring portions of the heap. The bottom layer of pure ash would indicate that those camp sites were used for a long time before the habit of eating molluscs was acquired. Mingled in this layer were numerous bones, especially those of fishes, gen- erally in the most disintegrated con- dition. The layer offered deer bones and all the other articles characteristic of the upper parts of the heaps. It will thus appear that this and similar camp sites are older than the first shells, and it must be then deduced that the original object, of coming to the sea shore, was not clams but rather fishing, and possibly hunting, but espe- cially fishing. Where the layers were made of ashes and finely broken shells, the period of accumulation was longer and agreeing with that, the numbers of articles found in these layers was also greater. Each of the heaps had its own char- acteristics, which can be seen by glancing down the columns of the lists of food-animals, and tools found. Thus Sawyer’s Island is characterized by the tremendous abundance of deer remains, so that it would appear that the hunting must have rivalled the fishing, as doubtless this island was then a part of the mainland. The cod were also abundant and were doubtless the reason for the location of the heap. The heap on White Tsland differed from the others in that at the base of the heap was a con- siderable layer of oyster and quahog shells. On the north end of White Island there was a small heap almost exclusively composed of quahog shells. The same was true of a heap on the north end of Birch * Baum Village Site, Ohio Archeological and Historical Quarterly, vol. xy, pp. 1-95, 1906. Loomis and Young—Shell Heaps of Maine. 21 Island, but in neither case was there indication of long occu- panecy. The Flage Island heap was at once distinguished by the presence of great numbers of large mink and of great auk bones and those of small fishes. It would seem as if these Indians were not the fishermen that those camped at Sawyer’s Island were, for the remains of large fishes were very scarce, while those of flounders, cunners, etc., were abundant; but the apparently easily-caught, and staple food, on this Island was the great auk, the bones of which occurred in large num- Fie. 3. Fic. 3. Diagram of a section, fifteen feet long, along the line between B and Cin fig. 1. All the material found in sections B and C is indicated in this figure, which therefore show the depth at which each find was made in these two sections and the relative abundance and relationship of the differ- ent articles. @ indicates a food animal, bone or jaw: + indicates a bone tool; x indicates a stone tool; — indicates bits of pottery. bers, as did those also of many other birds, many of which we have not been able to identify. In Frenchman’s Bay the heaps were dominantly of soft-shelled clams, but had in them frequent bands of blue mussel shells, a form occurring about the islands in vast numbers to-day. In these northern heaps moose bones become more abundant, and they are especially well character- ized by the much greater proportion of stone implements found in them. The heap near Winter Harbor was perhaps the most prolific im bones and tools of any examined. It was distin- guished by the quantities of cod and bird bones. All these heaps, of course, represent camp sites, occupied for a longer or shorter period, by a smaller or larger group of Indians. Their fortunes varied annually, as is testified by the bones and shells left in the layers. It would seem that at first they did not come to get clams, nor did they even subsist on them. That was a secondary acquirement, but probably, when they began to use the shellfish, it became the staple food supply. The abundance of fish bones, the remains of fish hooks, and the broken harpoon points, indicate that the real business of the camp was fishing, and, presumably, the camp was timed to 22 Loomis and Young—Shell Heaps of Maine. cateh the cod and haddock, when they came in near shore for the spring spawning, and were then in the greatest numbers and most easily caught. That they were spring camps is also confirmed by the condition of the horns on the ecrania of the male deer, which were found in the heap. In the Sawyer’s Island heap, there were 54 male crania, of which only one had an attached horn. The time when the horns are entirely lack- ing is just after the shedding season, which is early spring, and coincides with the spawning time of the cod, ete. From these considerations, it seems almost a certainty that the camps were fishing camps, established in the spring, and only occupied by the Indian at that season of the year. The heaps are found along the shore and on islands adjacent to the shore, often on very small islands, when there does not seem to have been available drinking water. Many or rather most of the heaps are being eaten into by the action of the tides and waves, so that in many cases what remains is only a fraction of the original area. This can only be explained by the fact that the land is gradually subsiding. On this point geologists are in general agreement, the outline of the coast being exactly that typical of a sinking coast line. It is further agreed that the coast, as far to the south as New Jersey, is also sinking, and the fundamental problem is the rate of sub- sidence. At New York City the rate has been determined, by a long set of observations of tide levels, as 1°64 feet per century.* In Maine the rate is probably somewhat more rapid, say three feet a century. Other evidences of sinking are found in the tide mills of early days, which are now so low as to be no longer available as power plants. Then there are numerous places where stumps and peat occur below tide level, the trees having been killed by the rising water. Lastly these heaps are also testimony to the sinking land, for the camps were beyond doubt originally well above the tide’s reach. In none of the heaps explored was any trace of iron imple- ments found, showing that the heaps were completed before the advent of the whites, which began for trading purposes about 1627. The top of the heaps is therefore nearly three hundred years old, at least, and probably somewhat more. The length of time it took to accumulate the four or five feet of many of the heaps (more in some cases) can only be esti- mated, but as the camps were temporary, and the accumulation of ashes and broken up shells must have been comparatively slow, it would seem that not less than 300 to 500 years more should be added to the estimate for the age of the base of the heaps. * This Journal, vol. xvii, p. 333, 1904. Loomis and Young—Shell Heaps of Maine. 23 Considering the age of the heaps, it must be concluded that the level at which the camp was originally established was some ten or more feet above the present level, and therefore many of the heaps now on islands, especially where these are near the shore, were then on the mainland. Turning to the contents of the heaps, we may expect to learn from the contained articles many characteristics of the Indians who made the heaps. From the refuse bones, their food, and their hunting and fishing ability will appear; from the tools, their grade of culture, and something about their mechanical ability will also appear, while from comparisons with like rejects of other tribes, we shall learn something about their relationships. The accompanying table gives all the sorts of food animals which we found and identified from the heaps. Some may not prove to have been food, but most were. Besides these there were a few species of molluses, which occurred only once or twice, and were, therefore, simply accidental. Deer (Odocoileus virginianus borealis (Miller) Allen) occur on Sawyer’s Island in vast numbers, indicating that in this camp deer meat was a staple food supply. In other localities, their remains were comparatively infrequent, though always present. They were doubtless a highly prized food, and the remaining fragments of bones offer mute testimony to the Indian’s pleasure in this meat, in the way every limb bone is split and crushed to get out the marrow, even such small bones as those of the toes being broken open, that none of the marrow should escape. Then the cannon bones, on account of their hardness, are a favorite material for making tools, likewise the antlers, when found. We found, however, no indication of the lower jaw being used for a tool, as was the case among the Indians of the Baum Village Site.* There were 53 crania preserved, of which 52 belonged to males, and only one to a female. Mills concludes, from a similar state of affairs in the Baum Village Site, that the Indians showed a foresight for perpetuating the deer in advance of that now exercised by man generally (loc. cit., p. 27). However, from studying the small fragments of other crania, we feel that the explanation is to be sought in another direction. The crania were always broken open to get out the brain. In the case of males with the heavy frontals, strengthened to support the antlers, the smashing of the brain case was done in the parietal region, the thickened frontals remaining intact: while in the case of females, the frontal bones being thin, the cranium was broken through this region, or they were ut least also broken in getting the brain * Ohio Arch. and Hist. Soc. Quart., vol. xv, p. 79, 1906. \ DISTRIBUTION OF ANIMAL’ REMAINS TO SHOW THEIR RELATIVE ABUNDANCE IN THE VartouS Hmaps. The figures indicate the number of individuals found in each case. eet Bl eae eel os aia rete ees esl|SSl\ES\B4E| ws) S| eS | 25 Ee | oe im, a| Oeil am a4 a fates na |e ol Bie |o la |e Deer (jaws). - 225 2252 708 4 937s a ee eae eee i Moasges $800. 27 a DS CR ee aT te eA ea eoraat fnee 1 5 Garbomse 2. 2. 2.5 Sees 2 Sie C2 OD KSRS Chai ecaae mera 1 Dow type! Loy ee Sy ie ein ee Bie | andl chan tie Dog, type... oe 525 AG) 5 al ee a am Ta ee Dope, type Sscasew ase: TW, TG eae Rae Re a ieee OWN fhe oho Sys ees = Dimer z Dal ANE. olka Ee eae ER SUV HOE AS osu 0h pce Oe Dy ceases uf abet Eon Whe es ete AA ee aa FVaCcoOn: Seas oie eee 6 Dt eeeR Poe cfieeeey reef ee 3 pear lacks eee 3 Deal bcaly Teall eee age Often ee ake wee eee 4 Degen 1 el ens 1s Ips he Dyn Spee i a SS AT 2G) be Ae mt een ee Mink 6 sp; MOV. sa.coee Shy cue tiles cal eae all OMe) oer de Mink ¢ pea Aiea) 2) aloe we (yt ee al Renee Le | ee ni eal harbor. 2.202 uce: 3 Pees ocala 4 1 1 1 Seal, Greenland._..._-- ie DU eee oe (Shae eee NBS ee Whales ee oc vere au a ee ae Ds SARS UR Pin I Beaver fet Ur Jaen oe 20 | 2 Shela SP Gail ee 5 Muskratet 22 seth eee et es Pcs r Ei 1 1 Mouse; northern) pime i250 | N23 4) 2a ee eee eae ies Vole, meadow ._-.._--- ig ager ees l(t es PIMA Aukes ereatie. (asta | « Sia] A EOD | OR eats Deke hfs cae oh ee re HWeseer Weed eee i 29] ola sl nee Al ae Woon... Oa =e ee Dey line Sree 3 1 pee Hage 2 Sere) ee es Piper cece ani r ti iheltre tiie ei. 1 GOOse 255 4325 ee Nas Go aye lope hae 1 1 2 1 Birds, undetermined _.--|| 5 |) 59) LON) 22) Ab ian asain PETE «Seen cs ee ee aera Wel be Be res Tp Rees eee ay || OY or yee a eee Neen Beh ae ulhedeste Vat Ss See ei Codes ec tes Nansen 104 | 10 1 | 18) 12 2 1 |103 Maddock vee ese cS. Baa 11 aes DO) Viste eee a aan Culp = eee ee =) 1 a fc Desa Com eaes Mloumders pee oe eee alee AS Peete 7 | 10 Guiles Cunner 2 So) ee 5)6| Veer ee OAS 2a Ma He ac Sblurgeonian. 02. 3u22! Qa 2 Mal ree een Ded a Wochisheet yah ote tee LC [eee eek Bi 0S Sy ill rata eae Clam, soft shelled __-_.-_- eam eb clio Me Seulh Scuiihaeste MPS @lami, quahogy so. As << a Ss & | a a a [| : [| a | [| [| LIZZ i VTA i i oe Ya AT la PIZZA AV i i | [a = : N VATA TATA N \ q SN NENAERENG SHEAR RAH BBO SONINS NENANARS \ | N CANADIEN NI N BB PAA AA se) AVY TT N Sn) —e 46 Merwin and Larsen—Mixtures of Amorphous The mixtures are prepared by weighing into a 3-inch test- tube the required amount of Se, heating till the Se is thor- oughly fused, allowing to cool, adding the S and heating over a low flame just sufficiently hot to allow thorough mixing with a glass rod. As the material cools it may be gathered on the rod, removed, and cut into smal! fragments, which may be kept permanently in the stoppered tube. A gram or two is sufficient for examining a hundred preparations. For use with the microscope a small piece of the mixture and a little of the mineral powder are heated on a glass slip under a cover-glass over a small flame till the mixture is fluid. The powder and liquid are then mixed and pressed into a thin film. The film is then heated during 15 to 30 seconds till bubbles begin to appear, again pressed thin and cooled. If care is taken no appreciable amonnt of sulphur will be vapor- ized from the film. In the preparation of the mixtures, artificially crystallized sulphur should be used. Flowers of sulphur is satisfactory if it contains no mechanical impurities. The powdered or fused selenium furnished for this investigation by three dealers has been found to be sufficiently (better than 99°7 per cent) pure. It was tested by determining the refractive index of the glass made from the samples, and of that made from different frac- tions of the distilled selenium, and from precipitated selenium. The extreme values obtained in lithium light were 2°712 and 2°718. Mechanical impurities were sought for under the microscope in thin films of the glass. They were removed from one sample by distillation in a bent test-tube. In case it is not convenient to determine the refractive index of the selenium directly, its purity may be tested under the micro- scope. A mixture of 73 per cent Se, 27 per cent S, has the refractive index of pure sphalerite for light of wave-length 6354p. This is the equivalent wave-length of light trans- mitted by a thin film of this mixture. The test is made then by embedding the powder of colorless or amber-colored sphal- erite* in the mixture and determining the relative refractive indices through a screen of the same mixture. If the slide has a temperature of 30°-35° ©, or if it is studied through a screen of 85 per cent Se, the sphalerite should have the higher refractive index. The refractive index of a mixture, correct to +°001, may be found on the goniometer by cautiously melting the mixture in the angle between two narrow strips of glass joined firmly at * One per cent Cd or 1g per cent Fe raises the refractive index about ‘001. The presence of 1 per cent or more of Fe produces a deeper color. The presence of disturbing amounts of Cd is very rare and can be known by analysis. Sulphur and Selenium as Immersion Media. 47 the ends by melting. The glass strips should be about 4 or 5™™ wide, 2°" long, and 1™™ thick, and should inelude an angle of 35° to 45°. Microscope object glass cut crosswise is suit- able provided its sides are sufficiently nearly plane and parallel, in which ease the refiections of a wire or cord 3™™ in diam- eter in front of a window at a distance of 10 feet are not dis- torted nor entirely separated when the reflections are seen parallel to the direction the glass is to be cut. If. the refractive index of the mixture in the prism is to equal that of the film in which the mineral is embedded, it is essential that the mixture be heated to 250° or more before it is put into the prism. This is especially true of a mixture that has been standing some time and is partly crystallized. The refractive indices in Table I and on the chart were found in this manner. The probable error does not exceed +005 except in the case of wave-lengths less than 625up in mixtures containing more than 80 per cent Se, where absorp- tion is very strong. Likewise, the probable error in the dis- persion of a given mixture does not exceed + ‘002. The mixtures containing less than 15 per cent Se crystallize so readily that they are not well adapted to accurate work. Certain liquids, having refractive indices from 1°80 to 2°10, which fill the gap between the sulphur-selenium mixtures and liquids already well-known will be considered in a subsequent paper. Geophysical Laboratory and Geological Survey, Washington, D. C., April 18, 1912. 48 C. D. Cooksey—Secondary Cathode Rays. Art. V.—On the Asymmetry in the Distribution of See- ondary Cathode Rays produced by X-rays ; and its Depend- ence on the Penetrating Power of the Exciting Rays; by C. D. Cooxsry. Ir has been shown by Bragg and Madsen* that the amount of secondary §-radiation excited in various solids by y-rays is not symmetrically distributed about a plane perpendicular to the direction of propagation of the y-rays, but that the @-radia- tion which comes from the side from which the y-rays emerge is greater than that coming from the side on which the y-rays are incident. The same effect is observed, though to_a less degree, in the case of secondary X-rays,t ’and the secondary cathode rays produced by X-rays,t and those produced by ultra- violet light. This lack of symmetry is less for soft y-rays than for hard and still less for X-rays, which are usually con- sidered as very soft y-rays. The order of magnitude of the ratio of emergence to incidence radiation ranges all the way from 20:1 down to unity, depending on the nature of the radiations used and the substance in which the secondary rays are excited. Since I first showed that this effect was true for the cathode rays produced by X-rays, I have been experimenting with a view to find how the ratio of emergence to incidence radia- tion depends on the hardness of the exciting rays. But with these rays the effect is so small at best, that the variations which might be produced by the widest possible variation in hardness of the primary are not likely to be much greater than the experimental errors always inherent in X-ray measure- ments. Some results which I at first obtained|) seemed to indicate that there was a slight increase in the ratio with an increase of hardness of the primary. But owing to the hetero- geneity of the rays from an ordinary tube, and the difficulty, at that time, of sorting out and using rays of a single penetra- ting power, these results were not very convincing. Since these results were obtained, however, the work of Prof. Barkla and others on “‘ Fluorescent” X-radiations, a sum- mary of which is to be found in the Phil. Mag. for Sept., 1911, page 396, has afforded a convenient means of. obtaining homo. geneous beams of X-rays of known penetrating powers over a wide range. It was by this means that the experiments described in the present paper were performed. * Trans. Roy. Soe. South Australia, vol. xxxii, May, 1908. + Bragg and Classon, loc. cit., Oct.,; 1908. t Cooksey, Nature, vol. Ixxvii, Dp: 509, 1908. § StuhImann, Phil. Mag.. vol. xxii, p. 804, 1911. || Nature, vol. Ixxxii, p. 128, 1909. 0. D. Cooksey—Secondary Cathode Rays. 49 Apparatus and Method. The apparatus used is shown in the accompanying diagram and was a modified form of the apparatus used in previous experiments.* The X-ray tube was completely enclosed in a thick lead box, and the rays, proceeding from the anti-cathode, K, passed through a window in the box and fell on the radiator, R,, 7 Electroscope Fy which consisted of a metal known to give off fluorescent X-rays. The fluorescent rays from R, passed through windows in a lead screen, S, into the ionization chambers, A and B, which were maintained at potentials of +240 and —240 volts respectively. A wire, O, passing into A and B through insu- lating plugs was connected to the gold leaf of a sensitive elec- troscope, E,, of the type described by R. T. Beatty.t The two quadrants of this electroscope were connected to A and B respectively, and the wire, C, was normally connected to earth. A third ionization chamber, D, received some of the rays from KR, and was always connected to earth. A wire running from the inside of D and insulated from it con- nected with the gold leaf of a second electroscope, E,, of * This Journal, vol. xxiv, p. 285, 1907. +Phil. Mag., vol. xiv, p. 604, 1907. Am. JOUR. Poa Ue Series, Vou. XXXIV, No. 199.—Juny, 1912. 2) 50 . ©. D. Cooksey—Secondary Cathode Rays. the ordinary type. This wire and gold leaf could be charged to any convenient potential; and the rate of fall of the gold leaf when the rays were turned on, served as a measure of the intensity of the rays from R,. Each electroscope was fitted with a microscope and scale for observing the motion of the gold leaf. A lead screen, F, could be slid across the window in front of B by means of a screw, giving a very fine adjust- ment of the amount of the radiation entering this chamber. The chambers, A and B, consisted of brass cylinders about 10s in diameter and 2°2°"* long, covered with lead on the out- side. The ends of B were covered with aluminium foil. The end of the wire, C, which entered A, terminated in a wire ring of a diameter slightly less than the inside diameter of A, and lying in a plane perpendicular to the axis of A. The opening in the screen, 8, opposite A was so adjusted that no part of the beam from Rk, could fall on the ring or on the walls of A. Brass plate holders, E and I, covered the ends of A. These plate holders were made exactly alike to be interchangeable. Each one consisted of two square brass plates, held together with screws, and having a circular hole bored through their center of about the same diameter as the inside of A. These plate holders could be slid on and off the ends of A in some- what the same manner as a camera plate holder, and stops were so placed that they would always come to the same position, with the center of the hole on the axis of A. In order to produce the emergence cathode rays, it was necessary to pass the rays from KR, through a plate of some metal in which cathode rays are produced in sufficient numbers to be easily measured. As these metals are relatively opaque to X-rays it was necessary to use very thin sheets. Gold and silver were fixed upon for this purpose. The gold leaf was 0-810 °°™® thick and the silver leaf 1:°810-°°™*. Circular disks of aluminium about 0:012°S thick were made with a diameter slightly larger than the holes in the plate holders, and on one side of each was laid a thin foil of one of the above metals, stuck on with a thin coat of vaseline. When thicker sheets of metal were wanted more foils could be added by first holding the disks, already covered with one foil, in the vapor of heated vaseline, and then laying them on the foil to be added. One of these disks covered with one! of the metals, say gold, was then screwed between the two parts of each plate hoider and the holders placed over the ends of A with the aluminium sides facing in. When the tube was turned on, and the wire, C, insulated from the earth, there would be a motion of the gold leaf in the electroscope, E,, toward one or other of the quadrants, depend- C. D. Cooksey—Secondary Cathode Rays. 51 ing upon whether the ionization in A or B was more intense. By a suitable adjustment of the screen, IF’, the ionization in the two chambers could be made equal, when there would be no motion of the gold leaf. After this adjustment was made the tube was stopped, and the front holder, E, reversed so that the gold side faced into A. The leaf in the electroscope, E,, was then charged to a potential of about 200 volts, and the tube again excited. When the leaf reached a certain division on the scale the wire, C, was again insulated, and after the leaf in K, had passed over a convenient number of scale divisions the tube was stopped. The position of the leaf in E, was then observed, and © connected to earth through a potentiometer, and a sufficient potential applied to keep the leaf at its observed position. Calling this potential a, we can take it as a measure of the excess of ionization in A over that in B. If e denotes the ionization produced by the emergence secondary rays from the gold, and e’ by the emergence secondary rays from aluminium, we shall have e—e' =ka where & is a constant. After this measurement was made the holder, E, was turned back with the aluminium face in, and the other holder, I, reversed so that the gold faced in. The tube was again excited, and the potential, b, acquired by the leaf in E, while the leaf in E, passed over the same interval as before, was measured. Denoting the ionization produced by the incidence secondary rays from the gold by z, and that by the incidence secondary rays from the aluminium by 2’, we can write as before i—a=kb Letting e’/e = m, and 7’/i = 7 we shall have e/i = a/b (1 —n)/(1 — m) where e/2 is the ratio of the emergence secondary radiations produced in the gold by a certain beam of X-rays from R, to the incidence secondary radiations produced by the same beam after traversing the ionization chamber. In the light of the results obtamed in a previous paper,* and owing to the fact that the ionization chambers used in the present work were only two-fifths as long as those used before, we may assume that the secondary X-rays from the metal produced a negli- gible amount of ionization in the chamber compared to that produced by the cathode rays from the metal. No data are readily obtainable in regard to m, but it is known that the amount of cathode rays given out by aluminium *This Journal, xxiv, p. 285, 1907. 52 C. D. Cooksey—Secondary Cathode Rays. is very small compared with the amount given out by metals of high atomic weight, such as gold and silver. Unless the ratio of emergence to incidence radiation is far different in aluminium from what it is in gold and silver, which is unlikely, the difference between m and x will be a quantity which we may neglect, and we may write for a sufficient approximation efi = afb The determination of e/2, as above described, was always repeated with the two plate holders reversed with respect to the ends of A, which should give the same result if both plates were the same. The values obtained, however, in the two cases differed very widely; much more than could be accounted for by the experimental errors. The only explana- tion seemed to be that, owing to dirt on the surface or varia- tions in the thickness of the metal foils, one plate was more efficient in giving out cathode rays than the other. If this were the case both the incidence and emergence rays should be effected in the same ratio, and the geometric mean of the two determinations of ¢/2 should give the same result. To test this, different pairs of plates were made and also more foils were added to the same plates, but no matter what the discord- ance was between the two values of e/¢ for a given pair of plates, the geometric mean of the value found when one plate was on the front of A and the other on the back, and the value when these positions were reversed, always gave the same result within the limits of error of the experiment. This ratio, ¢/2, does not, however, give the true ratio of emergence to incidence effect for the same intensity of exciting rays owing to the fact that these were absorbed in the air of the ionization chamber and to some extent in the layer of the metal from which the cathode rays come. The ratio of the number of emergence cathode rays coming from a layer of the metal so thin that the exciting rays suffer no absorption in it to the number of incidence cathode rays coming from the same layer is the true value sought. Let this be designated by R. Let 8 be the coefficient of absorption of the cathode rays in the metal from which they. come, i, the coefficient of the exciting X-rays in the same metal, and X, their coefficient in air. The number of emergence cathode rays produced in a layer of thickness dx at a depth « of the metal by X-rays of intensity I is equal to Kidz where K is some constant. If the intensity of the X-rays on entering the metal is I), and the thickness of the plate ¢, the number of the emergence cathode rays which get out of this layer into the ionization chamber is equal to \ C. D. Cooksey—Secondary Cathode Rays. 53 Tie ee and the total number getting out of the whole plate is —pt t —A; Wl a= Kl 8 Me Mi da 0 = Be Shee (8—Ax)E —1| The X-rays after being absorbed in the plate on the front of A are absorbed by the air in A and by the plate on the back of A. Therefore the intensity of these rays after reaching a depth « of the second plate will be —A —A,t' —A,x The BU at 1x where ¢’ is the leneth of the ionization chamber. Therefore the number of incidence cathode rays getting into the ioniza- tion chamber from the second plate of thickness ¢ will be equal to _ Khe mee nie [ 1- € aoa K(B+ 2,) The ratio of the number of emergence cathode rays entering A to the number of incidence cathode rays is therefore Dats R Bay Qr,t+A, 0" KOEN: os 0 = gaa “gehE— € = Since 2, is small iompared to 8, the most important part of this equation is Re” Designating the other factors by S we may write e/é R=—;} Ge Aal’ If we assume that the absorption coefficient of the cathode rays is proportional to the density of the absorbing material we can calculate the values of @ from the values given by Sadler* for their coefficient in air. Sadler} has measured the coefficients of absorption in air of the fluorescent radiations characteristic of copper and arsenic, * Phil. Mag., vol, xxii, p. 447 1911, + Loe. cit. 54+ C. D. Cooksey—Secondary Cathode Rays. and he and Barkla* have shown that the ratio of the coefl- cients of any two fluorescent beams is constant for all absorb- ing substances in which these beams do not excite a fluorescent radiation. We can, therefore, calculate the values of 2, in air from those in aluminium and in air given by Barkla and Sadler. The fluorescent rays used were those from cliromium, iron, zine, and tin, and with the éxception of tin Barkla and Sadlert have measured their absorption coefficients in gold and silver. Their value for the coefficient of the rays from chromium in gold is unreliable according to their statement, but it will be seen that , is always so small with respect to 8 that an exact knowledge of this quantity is not important. The absorption coefficients of the rays from tin in gold and silver are not given by the above writers, but may be obtained approximately from the constant ratio between absorption coefficients cited above. Tin rays, being more penetrating than those characteristic of silver and gold, will excite their fluorescent radiations and, therefore, will probably be less penetrating to these metals than the above calculation would indicate, but an inspection of the curves given by Barkla will show that the increase of absorption is not very great in this part of the spectrum, and it will be seen that the effect of X, is negligible in the case of the rays from tin. The accompanying tables give the value of the quantities ‘occurring in the formula as calculated from the data given by Barkla and Sadler and the observed values of ¢/2 in gold and silver. The last column gives the values of R as calculated from these results. Most of the values of ¢/2 were obtained using only one sheet of metal foil, but in some cases more than one sheet was used. As the terms in the formula containing ¢ drop out when ¢ is equal to or greater than two sheets of the metal, and as no consistent variation with the thickness of the sheets was observed, the mean of all the values of e/z using more than one sheet and the resulting values of R are given separately in the tables following the greater value of 7. It is at once apparent from the tables that the ratio of emergence to incidence cathode rays does not vary with the penetrating power of the exciting X-rays in the range of pene- trating powers used; that is, for an increase of absorbability as measured in aluminium of about 8000 per cent there is no measurable variation in the ratio. It is interesting to note in this connection the values of the ratio compared to those found by Stuhlmann for the cathode rays produced by ultra-violet light. The value he gives for * Phil. Mag., vol. xvii, p. 739, 1909. + Loe. cit. Or 0. D. Cooksey—Secondary Cathode Rays. De GOLD. Radiator | —4|¢x 10° Aot’| a Aat’'| 2/2 X,in Al} r, pies) uiltonisehare leslie Obst (ras | 08 1:03 1°10 1°30/ 1°18 Chromium _|9780° |0°0321) 70:0 |> 1:07 367 = 1°6| 1:03 1°10 | 1°34| 1°22 ~ 0°8] 1:01 1:06 | 1°31] 1-24 Jbgayse ho Se 7070° |0:0209| 66:0 |> 1:05 239 = 1:6) 1:02 TO ae | 0-8) 1°01 102 | 1°21] 1-19 Tine ee 3438 |0:00938) 57°5 |> 1:01 106°3 = 16 1:01 1:02 | 1°25] 1:23 0-8] 1:00], | 1:00 | 1°15} 1°15 RT yeep e eeu se 106°4 |0°0004, 107 {> 1:00 4°33 = 1°6' 1:00 1:00 | 1:16] 1°16 SILVER. 1:8| 1:02 1:09 | 1°33] 1°22 Chromium -_| 6095: |0°0321) 38:0 |> 1:07 367 = 3°6| 1:03 1:10 | 1:40] 1:27 bij 1°8| 1:03 1:08 | 1:31] 1°21 rome tse 28 000 10:0209))8 35:8). >> 1:05 239 — 3°6| 1:02 1:07 | | 1:8) 1:02 1:08 | 1-24) 1:20 VAG te wel 1830: |0°00938] 31:4 |> 1:01 106°3 | | = 3-6] 1:01 1:02 | | 1°8| 1:00 1:00 | 1°23, 1°23 ine ea ee 23-02 0.0004, 5°75 |> 1:00 4°33 ) | = 3°6/ 1:00 1:00 | 1:23 1:23 silver is 1:07, and for platinum, which seems to behave, for X-rays at least, about the same as gold, 117. This is about the same as the mean of all the values which I have obtained for gold. His value for silver is somewhat less than the mean of my determinations. If the mechanism of production of cathode rays is the same with ultra-violet light as it is with X-rays, the ratio of emergence to incidence effect does not seem to vary much over a range of absorbability corresponding e We X-rays from tin up to that corresponding to ultra-violet ight. Summary. The ratio of emergence to incidence cathode rays produced : in gold and silver by beams of fluorescent secondary X-rays has 56 C. D. Cooksey—Secondary Cathode Rays. been measured with the object of finding the dependence of the ratio on the penetrating power of the exciting rays. The fluorescent secondary X-rays from tin, zine, iron, and chromium were used as exciting rays, representing an increase in absorbability as measured in aluminium of about 8000 per cent between tin and chromium. After allowing for the absorption of the exciting rays in the layer of the metal from which the cathode rays come and in the air of the ionization chamber, it was found that there was no definite variation in the ratio of emergence to incidence effect. In closing I wish to express my thanks to Professor H. A. Bumstead for the interest he has taken in the work, 2 Sheffield Scientific School, Yale University, New Haven, Conn., March, 1912. LI. Page—Fundamental Relations of Electrodynamics. 57 Art. VI.—A Derivation of the Fundamental Felations of Electrodynamics from those of Hlectrostatics ; by Lricx Page. Maxwe v’s electrodynamic equations are based upon three experimental laws : (1) the inverse square law for the electric force between two point charges relatively at rest; (2) Ampere’s law for the force between current elements, or its equivalent; (3) Faraday’s law of current induction. Helm- holtz gave a derivation of Faraday’s Jaw from Ampere’s law by means of the principle of conservation of energy, which, however, has been shown to be erroneous.* Indeed, it has been impossible by any of the methods heretofore used to derive the electrodynamic equations without making use of all three of these experimental laws. The object of this paper is to show, that if the principle of relativity had been enunciated before the date of Oersted’s discovery, the fundamental relations of electrodynamics could have been predicted on theoretical grounds as a direct conse- quence of the fundamental laws of electrostatics, extended so as to apply to charges relatively in motion as well as to charges relatively at rest. Of course, only that part of the theory derived from the principle of relativity that is independent of any a priori knowledge of the electrodynamic equations, will be made use of. That is to say, we will use only the kine- matics of relativity :—to use the dynamics of relativity, which is derived from the electrodynamic equations, would be to reason in a circle. A material system is defined as an aggregate of material bodies having no relative motion, and showing no linear accel- eration or angular velocity as a whole. Suppose now that we have any number of these systems moving in various directions and with various velocities relative to one another. The principle of relativity states that there are no experimental methods, practical or ideal, of distinguishing one such system as being marked out as different from all the others. In other words, if there is an ether, there exist no experimental methods by which we can find out which of these various sys- tems is at rest relative to the ether. One of the most obvious consequences of this principle is that the velocity of light, as measured in any one system, must be the same as measured in any other system. Otherwise there would be accessible to us an experimental method of locating the luminiferous ether, which is in contradiction to * Maxwell’s ‘‘ Hlectricity and Magnetism,”’ 3d edition, vol. ii, p. 192. 58 L. Page—Fundamental Relations of Electrodynamics. the principle of relativity. As a mathematical consequence of the fact that the velocity of light must be the same as observed from different systems, Kinstein, in his celebrated paper* in the Annalen der Physik, has derived a set of space time trans- formations, which, because they were first deduced by Lorentz from entirely different considerations, usually go by his name. Einstein starts off by a consideration of the meaning that can be attached to time simultaneity at two different points in any one system. Suppose A and B to be two widely separated places in the same system. An observer at A is watching cer- tain phenomena in his immediate neighborhood, while an observer at B is watching certain other phenomena in his (B’s) immediate neighborhood. They wish to compare the times of their observations. Obviously they must be provided with synchronous clocks. How are these clocks to be set synehron- ously? Let A send a light wave toward B when A’s clock indicates the time ¢,. This light wave reaches Bata time ¢, on B’s clock, and. is returned to A by instantaneous reflection, reaching A at the time ¢’, as indicated on A’s clock. Since the measured value of the velocity of light is the same in all systems, and the same in all directions in any one system, the clocks at A and B will be synchronous when, and only when, ty = 4(¢, +74). Applying this definition of synchronism to two systems in motion relative to one another, Einstein is led to a set of transformations which show that the time at a point P in one system is a function not only of the time at a point Q” in the other system, but also of the relative positions of the points P and Q. When applied to the measurement of distances, these trans- formations show that a bar which is fixed in the first system with its axis parallel to the direction of relative motion of the two systems, and which has a length 7 as measured by an observer in the first system, will appear to have a shorter length when measured by an observer in the second system. This apparent shortening is not surprismg when we consider the method used in measuring a body which is in motion rela- tive to the observer. Let AB bea bar which has a velocity relative to the observer in the direction AB. In order to measure the length of the bar, the observer must mark the positions of the two ends of the bar at the same instant, and then measure the distance between these two marks. If he marks the position of the end B a little earlier than he marks the position of the end A, his measurement will be too short. Hence we see that space measurements as well as time meas- urements on moving systems, depend on the definition of simultaneity at different points of the same system. * Annalen der Physik, xvii, 891, 1905. L. Page—Fundamental Relations of Electrodynamics. 59 Let K (0) denote the earth’s system at any instant. Then K (v) denotes a system with velocity v relative to the earth. Let XYZ be a set of orthogonal right-handed axes fixed in the earth’s system, and so oriented that K(v) has a velocity v in the positive 2 direction. Let X’Y’Z’ be a set of orthogonal right-handed axes fixed in system K(v) and mutually parallel to XYZ. Unprimed letters denote quantities as measured in the earth’s system, and primed letters denote the same quantities as meas- ured in the system K(v). Then the space time transformations between K(o) and K(wv) take the form: (2) ’ v li : bp he buna © = —— pS 2 en oe 1 = : 2 Pi c ae’ oe ae y=y' a z— vb an a se a! Oh vy Cc where ¢ denotes the velocity of light. and where the time epochs are so chosen that the times at the respective origins of the two systems are zero when these origins coincide. Let a particle have the velocity V relative to K(o), and V’ relative to K(v). Let V,, V,, V,, be the components of V, and V,’, V,’, V,/ the components of V’. Then the following kinematical transformations follow at once by taking the time derivatives of the space time transformations, with consider- ation of the relation dé 4/1 = Ne = Tinian ; Cc (6 if Ve p F f 1S 1 vV, a 2 We a Wie “ a i ah e V ————— We V, / ite Vv : AV Tee v (ees vV, Cc Ce ¢ V,-—»v Wile [veaes r Z balis Me vV, Ee at vV/ d= 2 1+ g 60 L. Page—Fundamental Relations of Electrodynamics. Vy! Mee Ee Vee ey c vy" 3 = : 3; a —_— ( saa a) (: 7, ub i V1 Fr =e Cc e ve wes ( Se ae Ve (: a We ve c Cy Moving Charges. - We can represent the field due to a charged particle which is at rest relative to the observer by radial lines of force so drawn that equal solid angles contain the same number of lines of force. Then we can define the intensity at any point as having the direction of the line of force at that point and as Fig. 1. 0 0 é being proportional, in magnitude, to the density of the lines of force at that point. Now let us extend this definition of intensity to charges which are moving relative to the observer. Consider a charge ¢ at’ the point O’ (fig. 1) in K (vw). Let dS’ be an elementary surface fixed in K (wv) at P’, and perpen- dicular to O’P’. Let O’P’= 7’, and the angle between O’P’ and the Z’ axis be @’. Then E’, the force at P’ as measured in K (v), will be a We wish to find the force EK due to e, at a point P in K (0), when P coincides with P’. On account of the different definitions of simultaneity in the two systems K (v) and K (0), when P’ and P coincide the charge e as viewed from K (0) will be at some point O not coincident with O’. Let OP =7, and let the angle between OP and the Z axis be 0. The space time transformations give the relations L. Page—Fundamental Relations of Electrodynamics. 61 = oe 7 ay? eee aes oe P sin f = ee sin 6 V1 we ci sin’ 6 1 cos 6’ = —— cos 6 1 — — sin’ 0 The direction of the lines of force, as viewed from K (0), and hence the direction of the intensity, will be OP, and noé O’P’. Now dS’ as viewed from K (0) will not be perpen- dicular to OP. Let dS be the component of d§8’, as viewed from K (0), which is perpendicular to OP. Then a short calculation gives 1 LS = YS AS Uv : We sin? 6 ¢ Now the density of the lines of force at Pin K(0) is to the density of the lines of force at P’ in K (v), at the instant when P and P’ coincide, as dS’ is to dS; that is to say in the 5 ON ie ratio 1: "A 1— — sin’ 0. Hence we have G — Therefore K II 7% aS e | Z 5, DS SS wo 62 L. Page—Fundamental Relations of Electrodynamics. The force E, as already noted, has the direction of the line of force through P, as observed in K (0); that is, the direction OP, where O is the apparent position of the charge to an observer in K (0) at the instant considered. Thus, by means of the principle of relativity we have been able to derive from the laws of electrostatics, with considerable ease, an expression which Heaviside has derived from the electromagnetic equations only by the use of somewhat compli- cated mathematical processes. The relations between the components of E at P and EK’ at P’ follow at once from the expressions we have already derived. - 1 ed is. = —— 1 a @ 1 EK, = —> oe @ 1 == 18,’ Force between Ourrent Elements. We can consider a current as made up of a given quantity of positive electricity moving with a given drift velocity along the wire in the direction of the current, and some other given quantity of negative electricity moving with some other given drift velocity in the direction opposite to that of the current. Let u, be the velocity of the positive electricity, and wu, that of the negative electricity. Let A, be the linear density, or the quantity of moving positive electricity per unit length of wire, and 2», the quantity of moving negative electricity per unit length of wire. Consider an element of the wire of length ds. Then we can define a current element as (A,w,+A,u,) ds. Now this element of wire is as a whole uncharged. So there must be a quantity of positive electricity (k—2,) ds, and a quantity of negative electricity (k—2,) ds at rest in the element, & being some constant. As the current is due to that part of the charge in the wire which is in motion, our problem reduces to a consideration of the forces between two charges both of which are moving relative to the observer. In order to make our reasoning as simple as possible, we shall confine ourselves to currents lying in the same plane. There is no difficulty in extending the reasoning to currents which do not lie in the same plane, but in that case the demon- stration becomes a little more complicated. L. Page—Fundamental Relations of Etectrodynanvics. 63 At a given instant in K (0), two charged bodies (fig. 2), one at A and the other at B, have velocities relative to K (0) of v and w em./see. respectively. AB=~v. Choose axes XZ so that z is parallel to w. Let the origin be at B. We wish té find the force on the charged body at B, due to the other charged body. To find this foree we must observe from the Fie. 2. system K (w). But according to the time synchronism of K (wz), when the one charged body is at B, the other will not be at A. It will be at O, a point whose codrdinates are found to be } Uv, r sin ¢ — 7 — sin 6 c fina 1? COS @ pds tae __ uv cos (6—¢) 1 We eos (6—¢) C. e These distances, as measured in K (w), are (the primes refer to K (wu) ), uw ,2 ‘ Ud. r cos @ 4/1 — r sin 6 —?r—, sin 0 A== z C= : 5 1 — vv cos (O@—¢) 1 Ue £98 (06—¢) eee! ee Cc c and the distance between the charges is 2,42 ine 8 = Sin* 6— c iP U -! — 1—2 eit sin 6 pore Ce er uv COS = 3 c wu f o cos > l 2 SSS ee ee 64 L. Page—Fundamental Relations of Electrodynamics. Applying I and reducing, we get v* ; UC ene ( — a) (sin o — Gr sin a) a ‘ 1 Let F, and F, be the forces as measured in K (0) that must be applied to the charge at B in order to produce the same effect as F,’ and F,’. Then vw ; Ui) eyes (2 _ =) (sin ¢——a sin 6) = = z ie (2 — ax sin® 0)? on Now replace ¢, by a current element (A,w,+),v,) ds. In this current element there is at rest the positive electricity (k—2,) ds, and the negative electricity (4—),) ds. Consider the positive electricity 4, ds which is moving, and a portion r, ds of the negative electricity which is at rest. Then the components of the force due to ¢, on the negative electricity r, ds at rest will be vy? — Xr, 2, ( _ 2) sin ¢ ds F, = : 3 2 ( Fe sin? a)? ion Oh — , 2, a cos d ds b= Z r vy ie 4 rv? (1 — —,> sin’ 6 c But the components of the force due to ¢, on the positive electricity X, ds in motion is, as we have just found, L. Page—Fundamental Relations of Electrodynamics. 65 A, e, (1 — =) (sin o — oh sin 6) ds : c c x = 2 ee (1 — sin? 6) ¢ vy? en (1 _ =) cos @ ds ee % Qe 3 a ee sim s66)) 2 Cc Combining, we have left the force telco vy? Ae (ei —--) sin @ ds 2 2 aes 3 2? ( — —; sin’ 0)? c in the X direction. Proceeding in the same manner, we find the total force on the current element at B due to the moving charge at A is ae hr ae v ——_+— sin @ ds (1 — a) e / i — 5 2 a Oo 3 a (1 — % sin’ 0)? c where 7, is the current in electromagnetic units. As the drift velocity of the charges constituting a conduction current in a wire is certainly smail compared with the velocity of light, we can place the factor equal to unity. If we replace ¢, by a current element, we will find for the total force exerted by the current element 7, ds, at A on the current element 7, ds, at B, as measured on the earth (system K (0) ), the expression 7,7, sin @ ds,ds, KF, = — 2s 2 ” where 7, and 2, are measured in electromagnetic units. Am. Jour, Sct.—Fourts Srries, Vout. XXXIV, No. 199.—Juxy, 1912. 5 SS 66 L. Page—Fundamental Relations of Electrodynamics. This expression gives all the forces between currents, and also the induced current phenomena due to moving a closed circuit through a so-called magnetic field. The induced cur- rent effects produced in a secondary circuit by variation of the current in the primary are very simply treated as follows : Fauraday’s Law. Whenever a charged body is accelerated, it is obvious that the lines of force will be kinked. If the charged body is accelerated only for a very short time, these kinks will travel outwards in the form of a pulse. Now this pulse must have the same velocity relative to the system of the field inside the Fie. 3. pulse as it has relative to the system of the field outside the pulse. These two systems, however, may be chosen arbitrarily. Therefore the pulse must have the same velocity relative to all systems. The only velocity to satisfy this condition is the velocity of light. Hence the velocity of the pulse must be ¢.* *This reasoning may be objected to on the ground that the pulse may expand as it moves outward: i. e., the outside of the pulse may have a greater velocity than the inside. But if this was true under certain conditions, it would be necessary to assume that the reverse was true under certain other conditions. So we would be forced to the most improbable conclusion that the inside of the pulse might outstrip and pass through the outside of the pulse. L. Page—Fundamental Relations of Electrodynamics. 67 Consider two charges e, and e, at A and B (fig. 3) respec- tively. Let the charge at B be at rest relative to the observer in K (0), and the charge at A be moving to the right with the ' velocity v. While e, is at A the acceleration 7 is applied to it in the direction of its velocity v Let AB=r=ct. BE is an are with A as center and cé as radius, CD an are with A as center and ¢ (¢—dt) as radius. If, as before, we define the intensity as proportional to the density of the lines of force at the point considered, the force just to the left of B will be provided v is small compared toc. So the intensity at the same point due to e, will be é, (: = =) c 1 2 . v ; 3 sin a ir (1 — —, sin’ 0)? c If we denote by E, and KH, the components of E parallel to and perpendicular to the radius AB, K= 2 v a ) en -) oe F Cc C = cot a “ S 2 / 2 3 ” ( Tones sain? a)? ihe (2 _ oe sin® 0)? G e So we see that the component of the force parallel to the radius AB is continuous through the pulse. tsin 6 . Now cot meee =i f ls is small. c If we replace e, by a current element ids a) he di sin 6 e a u dt r Diet Fe 3 ( a ee sin? 6)* in electromagnetic units. 68 L. Page—Fundamental Relations of Electrodynamics. ure , If — is small, this reduces to ce which is the expression for the induced electromotive force in one wire due to a variation of the current in another. Conclusions. Our object was to deduce the fundamental laws of electro- dynamics,—the law for the force between currents, and the law governing current induction,—from those of electrostatics. We assumed that part of the theory derived from the principle of relativity which depends only upon the fact that the velocity of light must be the same as measured in different systems, and which depends in no way upon the electrodynamic equa- tions. Then we extended the following conceptions of electro- statics to moving charges : (1) To an observer at rest relative to a charge, the charge can be replaced by a field of lines of force radiating from ihe charge in such a way that equal solid angles contain equal numbers of lines of force. (2) To an observer relative to whom the charge is in motion, as well as to an observer at rest relative to the charge, the electric intensity due to the charge is proportional to the instantaneous density of the lines of the force at the point considered. By the means of these extensions of electrostatic concep- tions to moving charges, we were able to deduce (a) the expression for the electric intensity due to a charge moving relative to the observer; (6) Ampere’s law, or its equivalent ; (¢) Faraday’s law, or its equivalent. Viewed from another standpoint, the fact that we have been able, by means of the principle of relativity, to deduce the fundamental relations of electrodynamics from those of electro- statics, may be considered as some confirmation of the principle of relativity. I want to express my thanks to Professor H. A. Bumstead for several valuable suggestions, and to Dr. H. M. Dadourian for his help and encouragement. Sheffield Scientific School, March 7th, 1912. W. A. Drushel—Hydrolysis of Esters. 69 Arr. VII.—On the Hydrolysis of Esters of Substituted Ali- phatic Acids; by W. A. DrusuEt. {Contributions from the Kent Chemical Laboratory of Yale Univ.—ccxxxi.] 8, Ethyl a- and B-monochlorpropionates, ethyl a- and B-mono- brompropionates and ethyl aa-dibrompropionate. Irv was shown by Walden* that the electrolytic dissociation of the brompropionic acids is very much influenced by the position of the halogen with respect to the carboxy] group. DeBarr,+ in his investigation of the hydrolytic action of water upon halogen substituted fatty acids, found that the monohalo- gen substituted propionic acids are fairly stable in the presence of water at temperatures below 100° C., and that when decom- position takes place at a higher temperature (150° C.) the posi- tion of the halogen with respect to the carboxyl group has a marked effect upon the velocity of the reaction. The purpose of the present investigation was to study the influence of the position of the halogen on the velocity of the hydrolytic decomposition of the ethyl ester of chlor and brompropionic acids. Preparation of Hsters.—The ethyl a-chlorpropionate was obtained from Kahlbaum. The ethyl a-brompropionate was prepared by the method of Zelinsky.{ The boiling points of the esters having indicated a fair degree of purity, they were analyzed for halogen. Small portions of the esters were weighed into flasks and decomposed with pure potassium hydroxide by warming on the steam bath for twenty-four hours. The solutions were then neutralized with nitric acid, diluted to definite volumes and the halogen estimated gravi- metrically in aliquot portions. Chlorine found, I 26-28 per cent, II 25°91 per cent, mean 26°10 per cent; chlorine calcu- lated for CH,CHC1.COOO,H,, 25:96 per cent. Bromine found, I 44-37 per cent, II 44°29 per cent, mean 44:33 per cent; brom- ine calculated for CH,CHBr.COOC,H,, 44:18 per cent. The 8-halogen substituted propionic esters were prepared by a modification of Richter’s method§ for the preparation of B-halogen substituted propionic acids, which apparently had not heretofore been used for the preparation of the esters. The process for the preparation of the acids depends upon the direct replacement of iodine in @-iodopropionic acid in water solution by chlorine or bromine. In this process the iodine is said to separate out as freeiodine. To prepare the correspond- * Walden, Zeitschr. f. phys. Chem., x, 650. + DeBarr, Am. Chem. Jour., xxii, 388. t Zelinsky, Ber., xx, 2026. § Richter, Zeitschr. f. Chem., 1868, 449-451. 70 W. A. Drushel—Hydrolysis of Esters of ing ethyl esters, pure ethyl B-iodopropionate was dissolved in about five times its volume of chloroform and chlorine or brom- ine added in excess, keeping the temperature of the mixture below 50° C. by means of a water bath. Under these condi- tions the iodine is readily replaced by chlorine or bromine with the formation of considerable quantities of chloride or bromide of iodine and the liberation of much free iodine. The chlor- ide or bromide of iodine was removed by shaking the chloro- form solution of ester with water, and the free iodine was then removed by washing with dilute cold sodium carbonate solu- tion and again with water. The chloroform solution of the new ester was dried over calcium chloride and the chloroform removed by warming on the steam bath. The esters so pre- pared were further purified by fractional distillation under diminished pressure. The §-chlorpropionic ester boiled at 162° C. at atmospheric pressure, and the 8-brompropionic ester at 85°C. under a pressure of 25". These esters on analysis by the previously described method gave the follow- ing results: Chlorine found, I 26-07 per cent, II 25°93 per cent, III 26-00 per cent, mean 26:00 per cent; calculated for CH,C!.CH,COOC,H,, 25°96 per cent. Bromine found, I 44:10 per cent, I 44:20 per cent, III 44:10 per cent, mean 44:13 per cent; calculated for CH,Br.CH,COOO,H,, 44:18 per cent. Ethyl aa-dibrompropionate was prepared by a modification of the method of Philippi, Tollens.* In this method it is recommended that 9 parts of bromine be heated with 4 parts of propionic acid in a sealed tube at 190° ©. to 220°C. for 24 hours, that the hydrobromic acid formed be removed, 9 more parts of bromine be added and the mixture heated for 48 hours at 190°C. to 220°C. A satisfactory yield of aa-dibrompro- pionic acid was obtained by heating at 160° C. for two inter- vals of 6 hours each. Subsequently it was found that the first atom of hydrogen in propionic acid is completely replaced by bromine by heating at 160°C. for two hours. The purified dibrompropionic acid was esterified with absolute alcohol and dry hydrochloric acid in the usual manner. The ester obtained by this method boiled between 102° C. and 103° C. at 38™, and on analysis for bromiue gave the following results: Brom- ine found, I 61:57 per cent, I] 61:54 per cent; bromine calcu- lated for CH,CBr,COOO,H,, 61°51 per cent. Hydrolysis of Esters.—The ethyl esters of the monohalogen propionic acids were found to be soluble in water only to the extent of about 4° per liter, and the ethyl ester of aa-dibrom- propionic acid was found to be much less soluble, making it impossible to dissolve enough of this ester in decinormal acid to make satisfactory velocity measurements even at 60°C. In * Philippi, Tollens, Ann. d. Chem. u. Pharm., elxxi, 315. Substituted Aliphatic Acids. (a 50 per cent acetone or alcohol the solubility of this ester is no greater than that of the monohalogen propionic acid esters in water. The presence of this amount of acetone or alcohol was found to retard the velocity of the hydrolysis so that no satis-. factory results could be obtained. In the case of the mono- halogen propionic esters 1°™" of each ester was dissolved in 250° of decinormal hydrochloric or hydrobromic acid previously warmed in the thermostat to the required temperature. As soon as the ester was completely dissolved 25°™ of the reaction mixture were titrated with decinormal barium hydroxide, using phenolphthalein as an indicator. Subsequent tritrations were made at the time intervals indicated in the tables, and the final titrations usually after the lapse of 12 to 14 days. In order to ascertain if any halogen was liberated in the form of free halogen acids from the esters or the halogen propionic acids resulting from the hydrolysis of the esters, titrations were made with decinormal silver nitrate at the time when the equilibrium was apparently reached. It was found that at 25° C. and 35° C. there was no apparent decomposition in this direction in the time required for all except the final titrations, and in these there was only a little decomposition observed at 35° C. and none at all at 25°C. The necessary correction was applied where there appeared an appreciable increase in the concentra- tion of the halogen acid. At 50°C. the last titration, preced- ing the equilibrium titration, was made in each case within 10 hours of the beginning of the reaction. In the ethyl @-chlor- propionate and the ethyl a- and 8-brompropionates as much as 2 per cent or 3 per cent of the halogen was set free as halogen acid, for which corrections were made accordingly. From these titrations the velocity constants recorded in Table I were calculated by using the well-known titration formula for mono- molecular reactions: K = ae loo @, — 7.) —loe 7, — T,) 1, where T, is the initial titration, T, the final titration, and T, the intermediate titrations, all expressed in cubic centimeters of decinormal barium hydroxide, and ¢ is time in minutes between the titrations T, and T,. It was thought desirable also to make hydrolysis measure- ments in the absence of any added catalyzing acid. In this case the reactions proceed very slowly at first but with increas- ing velocity as the concentration of the acid liberated becomes greater. The velocity of the hydrolytic decomposition of the esters is accelerated by the halogen substituted propionic acids resulting from the hydrolysis of the esters and by the halogen acids subsequently set free by the hydrolysis of the halogen substituted propionic acids. This second accelerating effect becomes very prominent after the hydrolysis has proceeded for 72 W. A. Drushel—Hydrolysis of Esters of some time, but when the hydrolysis is made in the presence of an initial decinormal concentration of halogen acid the accel- erating effect of the halogen substituted propionic acid and of the subsequently liberated halogen acid is negligible in com- parison to the accelerating effect of the added catalyzing halo- gen acid. By making occasional titrations with decinormal silver nitrate, the amount of halogen acid liberated in the sec- ondary reaction was ascertained. The results obtained in the absence of any added catalyzing acid are recorded in Tables II and III. TABLE I. Ethel esters of halogen substituted propionic acid in N/10 hydrochloric or hydrobromic acid. z Ethel Ethyl Ethyl Ethyl Ethyl a-chlor- a-brom- B-chlor- 6-brom- propionate propionate propionate propionate proprionate Time Time Time inmin. 10°K inmin. 10°K 10°K in min. 10°K 10°K A. At 28°C. 60 70°3 420 28°8 21°38 9°6 (8°40) * 360 71°8 1330 30°0 21°2 ; 9°9 8:18" 600 70°2 1900 29°3 20°6 (10°6))*~ S212 1380 70°5 2790 29°3 20°3 9°8 8°12 2220 69°2 4215 ~- 28-1 21°6 9°8 8°16 2800 70°7 4800 27:2 21°5 7 8°16 5700 27°0 20°9 9°7 8°17 70°45 28°5 21-06 9°76 15 B. At 35°C. 200 177-9 180 572 39°4 25°0 22°2 260 177-4 360 56°5 38°3 23°3 21°4 380 177°4 510 58°3 41°0 24°3 22°4 DOOM i 72 1380 574 39'3 23°8 21:0 ODT 1560 042 39°6 23°8 21°4 1410 176°8 1800 55'8 38°3 24°4 21°3 1980 56°3 37°8 23°1 21°5 177°3 56°9 38°9 23°9 21°6 Gi. YAtrb0: C 20 534 20 186°1 92°5 110 86°5 52°3 35 86541 50 192°5 94°8 180 85°2 53°4 50 529 110 181°2 96°0 300 89-0 56°3 80 541 180 185°7 90°0 420 86°0 56°7 110 537 300 195°8 90°3 580 83°8 53°8 180 527 420 183°4 91°5 SS SS 300 8528 ——- 861 54°5 = 187°4 92°5 533°8 * Not included in the average. Substituted Aliphatic Acids. 73 Tasie II. Esters in water alone. Ethyl Ethyl Ethyl Ethyl Ethyl a-chlor- a-brom- B-chlor- B-brom- propionate propionate propionate propionate propionate Time (a) Time (a) (a) Time (a) (a) inhr. @hydr. inhr. @hydr. @hydr inhr. @% hydr. % hydr. AY Ate say CO: 163 0 49.5 3°0 2°6 51 9 2°83 .281 0°3 112°5 725) 5°9 100 0°4 5°2 385 0-7 163°5 10°9 10°9 166 I'l 9°7 475 1:2 281 201 20°4 270 2°] 18°0 573 1°8 385 28°8 S07) 360 WoT) 26°3 475 36°2 45°5 458 3°4 36°2 B. At 50° C. 24 O'l 24 orn 6°4 52 21) 22 50 0°4 50 lil 16°3 96 4°] 49 121 1°9 70 18°2 33°71 120 5°3 65 175° 4:2 120 32 67'6 150 en 85 200°5 =5°7 144 40°5 85°4 175 10°1 119 241 8:0 175 50°1 107 216 13°2 141 336 061 9°2 200 57 122 240 2, 168 310 67°4* 151+ 310 22°47 175§ TasLeE IIJ.— Summary. Hydrolysis in N/10 HCl or HBr. Ethyl Ethel Ethyl Ethel Ethel a-chlor- a-brom- B-chlor- 8-brom- propionate propionate propionate propionate propionate 10°K 10°K 10°K 10°K 10°K At 25°C. 70°4 28°5 21:06 9°75 8°15 CG BOL 177°3 56°9 38°9 23°9 21°6 50°C. 533°8 187°4 92°5 86°1 54°5 Hydrolysis in water alone. Time (0) Time (0) (0) (b) (6) inhr. %hydr. inhr. %hydr. @hydr. Ghydr. 4% hydr. At35°C. 458 1-1 458 28°3 45°5 3°4 36°2 “50°C. 385 19°2 310 85 100 13 99 (a) % hydr. denotes total acidity referred to the initial concentration of ester. (6) 4 hydr. denotes ester hydrolyzed, corrected for halogen acid set free. * Tn 310 hours 6°3% of the total chlorine of the ester was set free as HCl. + In 310 hours 66% of the total bromine of the ester wasset free as HBr. t In 310 hours 9°44 of the total chlorine of the ester was set free as HCl. § In 310 hours 756% of the total bromine of the ester was set free as HBr. 74 W. A. Drushel—Hydrolysis of Esters. Summary.—In the presence of decinormal hydrochloric or hydrobromie acid at temperatures not exceeding 35° C., the ethyl esters of the halogen substituted propionic acids decom- pose almost quantitatively according to the equation: O,H,X. COOC,H,+HOH= C,H,X. COOH+(, H,O Below 35° C. _ the halogen substituted propionic acids decompose very slowly according to the equation: O,H,X.COOH + HOH > HX + Ofte Be OH.COOH. When decomposition takes place in this direction the @ position of the halogen favors the reaction. When hydrogen is replaced by halogen in the acid group of ethyl propionate the velocity of hydrolysis in the presence of an added catalyzing acid is much less than that of ethyl propi- onate, but in the absence of any added catalyzer it is much greater than that of ethyl propionate. The position of the hal- ogen with respect to the carboxyl group has a marked effect on the velocity of hydrolysis. Esters with halogen in the a posi- tion hydrolyze more rapidly than esters with the halogen in the 8 position when the hydrolysis is made in the presence of an added catalyzing acid ; the same is true when the hydrolysis is made in the absence of any added catalyzing acid if correc- - tions are applied for the halogen acids set free by the hydrolysis of the halogen substituted propionic acids. Cairnes—Some Suggested New Physiographic Terms. 75 Arr. VIII.—Some Suggested New Physiographic Terms ;* by DeLorme D. Carrnes. TABLE OF CQNTENTS, Introductory. Definition of Terms. Classification of Physiographic Processes. Equiplanation. Deplanation. Applanation. Conclusion. — Introductory. Tux introduction of a new scientific term is only justified when there is good reason to believe that its use will promote the special branch of science to which the term has reference. There is a somewhat widespread tendency to-day to over- burden our technical vocabularies with long words of uncouth sound which are difficult not only to remember but also to pronounce, and which apparently serve mainly to complicate rather than simplify the study of the subject into which they are admitted. Instead of rendering the study of any particu- lar branch of science unduly difficult, our endeavor should be to promote simplicity and clearness both in thought and expression. It is desirable, however, that a sufficient number of technical terms should be employed to allow of the concise and accurate description of the various phenomena to be con- sidered. Whenever a fundamental distinction is recognized between processes, forces, or agents, which may be very dif- ferent in their nature, and still may or may not tend to pro- duce results that are easily confused, it promotes clearness of conception to employ definite terms ‘to indicate the nice dis- tinctions that have been found to exist. Physiography is one of the most rapidly growing of the more youthful branches of science, and has, so far, fortunately escaped any considerable overcrowding of its vocabulary. In fact the greater number, if not all of the terms that have been suggested within the past few years, have been greatly bene- ficial in that they have added to our knowledge of the subject, and have materially helped to clarify our ideas concerning the various phenomena with which they are concerned. Among the more important of the terms that have been recently introduced might be mentioned,—nivation,+ topographic uncon- * By permission of the Director of the Geological Survey branch of the Department of Mines of Canada. { The term nivation was proposed by F. EH. Matthes in 1899, and has since been described by Hobbs and others. See :—Matthes, F. E., ‘‘ Glacial Sculp- ture in the Bighorn Mts., Wyo.” ; Twenty-first Ann. Rep. U. S. Geol. Sur- vey, Pt. Il, 1899, pp. 173-190. Hobbs, W. H., ‘‘ Characteristics of existing glaciers,” 1911, pp. 18-23. 76 Cairnes—Some Suggested New Physiographie Terms. formity,* topographic adjustment,* superimposed youth,* and solifluction.t All these have not only proved particularly con-. venient, but have given us more clear-cut conceptions and have also had a decided influence in promoting closer field observa- tions. The suggestion of the definite name peneplain by Prof. Davis in 1889 for a thing previously considered in a general way for a number of years, vastly promoted its consideration, and greatly stimulated physiographic research. Investigations, discussions, and writings followed rapidly after the christening of the peneplain idea. At the annual meeting of the Geological Society of America in Washington, D. C., during December, 1911, the writer pro- posed the term equiplanationt to include certain physiographic processes. Since that time a considerable number of inquiries, concerning this subject, have been received. As a consequence of the nature of these inquiries it has seemed advisable to write this essay, which is intended to be somewhat in the nature of a supplement to the original. In addition to containing some additional data concerning equiplanation, two companion terms “applanation,” and “ deplanation ” are also suggested in the hope of presenting in a more definite and comprehensive manner certain points concerning the various plain-forming processes and the more important forces and agents they involve. It is, of course, not anticipated that these names will prove of the same importance as those above mentioned, but it is hoped that they also may serve a useful purpose. In the preparation of this paper, the writer has been very kindly advised by Professor Isaiah Bowman of Yale University, to whom he wishes to express his sincere gratitude. Definitions of Terms. Before defining deplanation and applanation, it is thought advisable again to define equiplanation, in order that the rela- tive significance of the three companion terms may be the more apparent. Liquiplanation (L. aeguus, equal; L. planus, a plain) includes all physiographic processes which tend to reduce the relief of a region and so cause the topography eventually to become more and more plain-like in contour, without involving any loss or gain of inaterial, i. e., the amounts of material * Salisbury, R. D., ‘‘ Three new physiographic terms,” Jour. of Geol., vol. xii, 1904, pp. 707-715. + Andersson, J. G., ‘‘ Solifluction—a component of subaerial denudation,” Jour. of Geology, vol. xiv, 1906, pp. 91-112. t The title of the paper read was: ‘‘ Differential erosion and equiplana- tion in portions of Yukon and Alaska.” Cairnes—Some Suggested New Physiographic Terms. 77 remain apparently equal, or are not increased or decreased by the plain-producing process or processes. Material may be exported from certain districts during the time equiplanation is in progress. but this export takes place quite independent of the equiplanating. Deplanation (L. de, from; L. planus, a plain) includes all physiographic processes which tend to reduce the relief of a district, and so cause the topography eventually to become more and more plain-like in contour, dominantly by subtract- ing material from the area or areas affected. Applanation (L. ad to or upon; L. planus, a plain) in- eludes all physiographic processes which tend to produce land-forms having plain-like surfaces, or tend to reduce the relief of a district and so cause the topography eventually to become more and more plain-like in contour, dominantly by adding material to the area or areas affected. Classification of Physiographie Processes. All physiographic processes, forces, or agencies that tend to alter the topography of a district by reducing the relief and so eventually cause the surface to become more plain-like in con- tour, may thus be considered as deplanating, applanating, or equiplanating in character, depending upon whether their activities result dominantly in a loss, a gain, or neither loss nor gain, respectively, of material, within the areas affected. Many processes or forces, however, such as certain forms of voleanic activity and some phases of diastrophism, instead of producing plain-like surfaces, tend primarily to destroy them. All physiographic processes, that do not tend ultimately to reduce the relief of the region in which they are engaged, may thus be regarded as “‘nonplanating.” It is thought that it may be conducive to a clearer understanding of various physio- graphic processes, if each be considered under the division of this suggested classification to which it belongs. Normal stream action, for instance, has a deplanating effect, but in dis- tricts having interior drainage, the results of stream action are to produce equiplanation. Similarly, wind action may be either deplanating, applanating, or equiplanating. Frost action may be strongly deplanating in character, or may be an impor- tant factor in causing equiplanation. Most modern writers in describing plain-like surfaces, which here include both plains and plateaus, consider these all to belong to one of two main genetic types. For example, Geikie* classifies plains as : *Geikie, James, ‘‘ Karth sculpture,” pp. 335-339, 1898. ~ ‘ 8 Cairnes—Some Suggested New Physiographic Terms. 1. Plains of accumulation. 2. Plains of erosion. Other writers divide plains into: 1. Peneplains. 2. Constructional plains. These are typical of the various modern classifications of plain-like surfaces where such are based upon genesis. Many writers thus appear to consider that plain-like surfaces are pro- duced entirely either by erosion or deposition, i. e. by either subtraction of material from or addition of material to the areas affected, that is by deplanation or applanation. No eredit is given to equiplanating processes, although these, especially in desert countries, have frequently been described. In such regions, the eroding or abrasive forces have been quite as effective as those of deposition. Eyuiplanation. In the writer’s original paper dealing with equiplanation, this term is defined and the process is considered in a general way. Also a special phase of equiplanation that was found to be active in portions of Yukon and Alaska is also described. In portions of the Yukon plateau which is generally con- sidered to represent an uplifted and subsequently dissected peneplaned surface,* extensive portions of but gently undulat- ing upland from 3,000 to 3,500 feet above sea-level are still preserved that are now dissected by deep, steep-walled, depres- sions representing the main drainage-ways of the areas. On the upland, forces are engaged in dissolving and disintegrating the rock material constituting the residuals or monadnocks, and are transporting this to the intervening depressions, where it is deposited. On account of the arctic climate there prevail- ing, the material, wherever deposited, is soon frozen or becomes associated with other frozen debris, and is consequently held, and added to the flat or nearly flat surface accumulations of superficial deposits which extend over considerable portions of the plateau that are but very imperfectly drained on account of the perpetual frost in the ground. In this way the bedrock depressions of the upland are becoming filled with material derived from the adjoining higher elevations. Thus, the *Dawson, G. M., Trans. Roy. Soc. of Can., vol. viii, sec. 4, 1890, p. 13. Brooks, A. H., ‘‘Geography and geology of Alaska,” U.S. Geol. Surv., 1906, Prof. Paper, No. 45, pp. 36-41, 286-290. Spurr, J. E., ‘‘ Geology of Yukon gold district, Alaska,” Eighteenth Ann. Rep. U. S. Geol. Surv., Pt. III, 1898, p. 260. Spencer, Arthur C., ‘‘ Pacific mountain system in British Columbia and Alaska,” Bull. Geol. Soc. of Am., vol. xiv, pp. 117- 132, 1908. Hayes, C. E., ‘‘ Expedition through the Yukon district,” Nat. Geog. Mag., vol. iv, p. 129, 1893. Cairnes—Some Suggested New Physiographic Terms. 79 tendency is for this elevated surface to become gradually more and more plain-like in contour, without experiencing any regional loss or gain of material. A prolonged rise of temperature, however, would thaw the accumulations of debris on the upland, and would result in the greater amount, at least, of this material being rapidly removed by water action to the main valleys. While equiplanation is engaged in planating the upland, normal erosion is slowly destroying the remaining portions of Fie. 1. Fic. 1. General view of the dissected upland surface of the Yukon pla- teau along the 141st meridian (the Yukon—Alaska International boundary) at north latitude 66° 50’. This upland has an average elevation of about 3,500 feet above sea-level. this surface by more completely dissecting it, and by widening the intersecting drainage depressions and causing the valley- walls of neighboring streams to gradually approach one another. The equiplanating process here described is not thought to be limited to the particular district under consideration, but is believed to be more or less active throughout extensive areas subjected to an arctic climate. Since, however, these areas have been but little explored, the phenomenon has not been described. 80 Cairnes—Some Suggested New Physiographic Terms. In regions having interior drainage, equiplanation is also nearly everywhere in evidence. There, wind action, stream and water action, and the various other subaérial destructive and constructive physiographic forces, slowly diminish the relief by constantly removing material from the higher, gener- ally peripheral areas, and de epositing it on the basin floors of the various centripetal systems, thus causing all the local base- levels to rise. This is a true equiplanating process, as by it the relief is reduced and the surface of each region so affected tends to become more and more plain-like in contour, and in so Fie. 2. a as, YOTO. Y DD CA : PRIA OLOGICA Fic. 2. A view of the upland surface, 3,500 feet above sea-level, along the 141st meridian (the Yukon—Alaska International boundary) at north lati- tude 66° 51’. _Equiplanation is here active; the material composing the abrupt-edged limestone residual is being disintegrated and dissolved and subsequently added to the accumulations of debris that fill the adjoining bed-rock depressions. doing no loss or gain of material is involved. In many such regions having interior drainage, however, more or less fine material is exported by the wind, as explained under “ depla- nation,’ causing the general level to become reduced. The gradational agents are, nevertheless, truly equiplanating and operate, to a ‘large extent, independently of the deplanating wind action. Equiplanation is particularly effective, in regions having Cairnes—Some Suggested New Physiographic Terms. 81 interior drainage, during the early stages of the physiographic eycle and under conditions just sufficiently arid to prevent the streams from reaching the sea. As the climate becomes increasingly arid, and in arid regions that are approaching a late mature or old stage in the physiographic development, more and more fine material tends to be exported by the wind. However, in many regions possessing interior drainage, the exportation of wind-borne dust is but a minor process, and equiplanation dominates the situation. Passarge has described various such districts, notable among which are certain regions such as the Kordofan type of desert, of which the climate has Fie. 3. Fic. 3. ) i Ze eg == ga 588 y ‘emer SENG PONY 2) = = 5a Bo si SEAN 4 aig is I) on —————— J Ges = a Neo y MH HB Nee z ee = ate: hey [ SB EN 3 °o == Re sia | Bee Sy Pe > sg 6¢ Neco fos N°) > S= — . see 3; Os 7 2 = ces @) by al» oh Cole a SS 2S) .&8 ie: GF Oc —==— Fy =) aig é a Sd eM ——<——— Qos a = Ss 5 2 ll | ap ‘ Hon\ Sea ae ———_ \| {) + a . Bae ————————— Doe RQ oD - Ere Llo| Bh ate AY posed that no exposure of the odlite existed, but that the loose pieces were the result of the erosion of the overlying strata. In the fall of 1910, G. E. Anderson located the first exposure in a railroad cut north of State College. Subsequently, while * Journal of Geology, vol. xx, p. 259, 1912. ’ Liegler—Siliceous Odlites of Central Pennsylvania. 115 engaged in field work during the spring and summer of 1911, Mr. Anderson and the author found a number of outerop- ping beds of siliceous odlite in place. Hence, now for the first time, the exact relation of the siliceous odlite to the rocks of the area can be determined accurately. Similar siliceous odlites oceur near Tyrone on the Little Juniata River, and have also been noted by Stoset in the Mercersburg—Chambersburg quad- rangle in the southern part of the state. The eeologic con- ditions both at Tyrone and at Chambersburg are essentially the saine as in Center County. Occurrence and Distribution.—Both calcareous and siliceous odlite occur in Center County. The calcareous odlites, as a rule, are massive oOlitic limestones of dark gray to blue color, with small, well-rounded spherules. Occasionally they are also dark gray, slightly crystalline, and have small dark green odlites. The siliceous odlite oceurs interbedded in thin irregular layers in the limestone, and also in irregular chert concretions and nodules in some rare limestone layers. As far as observed, the odlite beds are confined to the Ordovician and Cambrian strata, which may be subdivided as follows in Center County : 7. Utica and Lorraine Shale, 1200 feet. Gray, sandy shale, and intercalated, gray, resistant sand- stone layers in upper part. Unfossiliferous. Black, fissile, and calcareous shale below, fossiliferous. 6. Trenton Limestone, 600 feet. Dark black, and blue, blocky limestones with interbedded shale layers and crystalline gray limestones, usually fossil- iferous. 5. Black River Limestones, 100 feet. Compact, very pure, dove-colored limestone, with few fossils at top, below almost black even-fracturing limestones, fossiliferous. Shaly layers at bottom. 4. Chazyan, 2300 feet. Compact, dense, light gray, dolomitic limestone with chert nodules at top, unfossiliferous in part. 1200 feet shaly lime- stones, interbedded with massive, pure, and crystalline, blue limestones at base. Occasional odlitic limestone beds. Fos- siliferous. . Beekmantown, 2400 feet. Series of highly shaly, laminated limestones, with many shale partings. The limestones, as a rule, are clastic “ Edge- wise” and other conglomerates. Shale partings are often ripple-marked and. show suncracks. Fossils rare in upper portion. At base interbedded sandstone lenses. Some odlitic limestone in lower portion, and siliceous odlite layers throughout. co *U.S. Geol. Survey, Folio No. 170. 116 Ziegler—Siliceous Odlites of Central Pennsylvania. 2. Upper Cambrian Sandstone, 500-600 feet. Pure, white sandstone, poorly cemented and friable, rarely quartzitic. Shows distinct, thin, and even bedding. The grains are pure, well-rounded sand. In lower portion, gray, and occasional red to brown, tough quartzite, highly and irregularly cross-bedded. 1. Upper Cambrian Limestones (base not exposed). Odlitic limestones; dark gray to blue, sandy, and shaly limestones. Bedding planes, distorted, and irregular. Layers of Cryptozoon proliferum abundant, also trilobites and brach- iopods (Dikellocephalus newtoni and Lingulepis sp. ?). The exact number of horizons of siliceous odlite it is difficult to state. In one section of the Beekmantown limestones eleven definite layers were exposed. This section represents only about 139 feet near the top of the formation. Outcrops of the calcareous odlites are not at all rare, but good exposures of the siliceous occur only in a few localities (see fig. 1). As a rule the siliceous odlite occurs in loose fragments in mantle rock, especially in the area indicated, while most of the Beekman- town limestones have been removed by erosion, and only the resistant layers of odlite remain. The sandy area underlain directly by the Upper Cambrian sandstone furnished all the odlite specimens from this locality, which in all cases were fragments of loose bowlders. Lhe Structure of the Oolite-bearing Rocks.—As will be evi- dent from the accompanying map, the rocks are folded into an anticline pitch-down toward the northeast. The oldest rocks on the area are hence shown in the extreme northwest corner of the map where Upper Cambrian limestones have been brought up bya thrust fault. From this core northwestward we get the parallel frmges of the younger formations. Nittany Mountain in the southeast corner of the map is the result of a synclinal pitch towards the east. Typical Detgiled Section of Ooblite Horizon.—The finest section of siliceous odlite occur in the gorge of Spring Creek, four or five miles south of Belle Fonte. The following is a detailed section of the best exposure : 31. 50 feet. Gray, thin, and irregularly bedded limestone, with edgewise limestone conglomerate, carrying in upper portion Rastrites, near 2. barrandi. In the basal portion we have a thin layer of sili- ceous oblite, underlain by dense sandy limestone. 30. 2 feet. Dark gray, slightly odlitic limestone. 29, 2feet. Light gray, very dense limestone. 28. 7 feet. Dense, blue limestone, with conchoidal fracture, carrying black chert concretions. 27. 4 feet. Impure, dark gray limestone. Liegler—Siliceous Odlites of Central Pennsylvania. 117 26 bo bo 10. ge bo bo bo lO co He OK .3 feet. 8 feet. 2 feet. 6 1n. 5d ft. . 10 in. 15 in. 10 in. 8 in. 8 feet. 3 in. . 18 in. 2 feet. oe ft. 34 ft. 3 feet. 5 feet. 4 feet. 2 feet. 6 feet. 3 in. 16 in. 10 in. 6 feet. 74 ft. 4 feet. Dense gray limestone, carrying five distinct bands of siliceous odlite. The odlites are small. Dense, gray, shaly, and laminated limestone. Oolitic limestone of mottled appearance. Layer of siliceous odlite. Dense, black, shaly limestone, interbedded with gray, crystalline, dolomitic limestone. Dark blue, dense l1.s. in upper part with white, odlitic chart concretions. Dense, black crystalline I.s. with chert concre- tions, odlitic in thin sections. Dark blue, dense l.s. with gray chert concretions. Siliceous odlite, made up of white quartz spherules with a calcareous cement, acting as a matrix of pebbles, of dense gray |.s. and of pure white sandstone at base. Alternating, very dense, dark, and light gray, lam- inated limestone, with occasional chert nodules. Black, odlitic chert. Black, coarsely crystalline, l.s., with vugs, and small calcite seams. Dense, gray, dolomitic limestone, with conchoidal fracture. Mottled shaly limestone. Dense, finely banded, light gray, dolomitic lime- stone. Dark blue, odlitic, finely laminated limestone, containing in its lower portion pebbles of the un- derlying light gray limestone. The line on con- tact is irregular. Light gray, dolomitic limestone with conchoidal fracture. f As above, but blocky, and smooth breaking. Dense, black limestone. Dark gray, crystalline limestone with chert con- cretions. At topa breccia of limestone fragments cemented by well-crystallized calcite, probably a collapsed cave. White-grained, black, siliceous odlite carrying enclosed dense, black chert concretions which are odlitic under crossed nicols. Dense, gray, limestone. Light gray, siliceous odlite, carrying also frag- ments of an older, darker odlite-horizon as pebbles. Dense, gray, slightly crystalline limestone and dolomite. Covered. Light gray limestone with irregular and angular pieces of odlitic chert. Total, 139 ft. 3 in. Am. Jour. Sct1.—FourtH Series, Vou. XXXIV, No. 200.—Aveust, 1912. 9 118 = Ziegler—WSiliceous Odlites of Central Pennsylvania. All of the oGlite layers in this horizon are fairly definite, and are exposed for about ten or fifteen feet along the bedding planes. Over this distance they maintain a fairly uniform character and thickness. Layers Nos. 1, 4, and 6 show con- temporaneous erosion of some odlite horizon, and its redepo- sition in a limestone matrix in layer No. 1, and-in an odlite Hie. 2. Fie. 3. Fic. 2. Siliceous odlite layer on Spring Creek. Fic. 3. Fragment of siliceous oGlite, coarse, showing included fragment of limestone and of fine-grained siliceous odlite. horizon in the other two. No. 11 also shows contemporaneous erosion, in this case, of one of the limestone layers. Fig. 2 shows the occurrence of odlite in No. 31, and fig. 3 a specimen from layer No. 4, and shows the association of the two differ- ent oOlites. Petrographic description of the Oolites.—Several distinct types of siliceous odlite exist in this area, and they will be de- scribed in order. As may be noticed from the accompanying map, the name of the locality is applied to the occurrence. Type One: the Filmore phase.—The specimens were col- lected about one-and-one-half miles south of Filmore, and occur here as loose fragments, and in one doubtful exposure. The odlites are here fairly variable in size and in shape, and as a rule are mixed with many flat and lense-like grains, which appear essentially like lentils, are somewhat imperfectly cemented, and consist of quartz. The nucleus is in nearly all cases a grain of quartz, usually of irregularly, rounded, and elongated shape. Sphericity of grains is notably absent, and Liegler—Siliceous Oolites of Central Pennsylwania. 119 many grains are angular. At times the nucleus is made up of several grains of quartz. The quartz enlargement of the sand- grain nuclei is fibrous chalcedony, and also the fine chert mosaic deposited in concentric rings, as many as twelve being faintly visible. A few of the quartzes show resorption, presumably by solution which subsequently served to enlarge them, partly replacing the original grain with chalcedony. The same pro- cess is indicated by the absence of definite outlines in some sand grains and their tendency to pass by insensible gradations into the fibrous chalcedony. Many of the odlites are entirely chert or chalcedony, and have completely replaced the nucleus; but even in this case, under crossed nicols the outline of the quartz grain which served as a center of deposition is faintly visible. The quartz-grains appear to be igneous in origin and carry fine trails of dust, and very small inclusions of a mineral Fie. 4. Fie. 5. Fic. 4. Filmore type of siliceous odlite. x 72. Fic. 5. Kvumrine type, at contact between calcareous and siliceous odlite. x 382. apparently biotite. Fibrous chalcedony also fills most of the interstices between the odlites, and is usually in indistinet bands parallel to their general outline. Some of the quartzes show a tendency toward secondary enlargement. Very little calcite is present, and that seems to be a secondary infiltration subse- quent to the odlite formation. The sand nuclei vary in size from minute fragments up to 627" in diameter. The longer diameter often exceeds the shorter four or five times. Type Two: the Altro phase.—The siliceous odlite occurring in the vicinity of Altro is the handsomest in Center County. It is practically pure quartzite in composition, as shown by the following analysis by Wieland :* * This Journal (4), iv, p. 262. 120 Liegler—Siliceous Odlites of Central Pennsylvania. BIO) . lonization Current. Barss—lonization by Collision in Gases and Vapors. 235 ics ie cS ee a ia i) rz i=) @ oO po > Oo — cw) oS i ine) 0 Fig. 2. Air. Fie. 3. Hydrogen. 0 400 See oo Tone CCE Sea Ty So eee ies COC Co Co ae>/Anne eee 7Anmey a8) ie at 0 400 800 1200 ah 800 Fie. 4. Sulphur dioxide. 0 400 800 0 200 400 600 0 200 400 600 > Volts per centimeter, Fie. 5. Alcohol. 236 =Barss—Lonization by Collision in Gases and Vapors. As a further test the equation a/p = f(X/p) was shown to hold. Since this equation is a general one, if we plot a curve for each pressure, taking as co6rdinates a/p and X /p, the curves for different pressures should coincide. This was tried for four different pressures in hydrogen. Values of a/p and X/p are given in Table III, and the curve connecting a/p and X/p is shown in fig. 8. Points on the curve marked TABLE IJJI. eeprom xX a a/p X/p 960 0°80 0°05 61°9 1100 0°93 0:06 72°0 1260 1:08 0:07 81°3 1660 2°28 0°14 107°1 1980 4°88 0°31 Nor 20 P= 10cm 700 0°60 0°06 70:0 960 1:08 0°10 96°2 1080 1°60 0°16 108°1 1260 3°28 0°32 126°'0 3. P=4:2em 400 0°46 O11 98-0 450 0°63 0°15 107°4 500 1°0 0°24 119°1 580 2°0 0°47 138°2 4, P=19°6™ 1660 esi 0:07 84:7 1800 1:76 0:09" 92°0 1980 2°30 0°12 101°0 2240 3°81 0719 114°1 1, 2, 3, and 4 correspond to pressures of 155°", 10°, 4:2", and 19-6 respectively. It will be noticed that all the points cal- culated are close to a common curve, which agrees with the theory. From these results in air and hydrogen, we may conclude that the negative ions generated in a gas by collisions with alpha particles are identical with those generated by Réntgen rays and ultra-violet light. In other words, the collision con- stants are independent of the nature of the ionizing agent. Vapors. The method of experimenting was the same for all vapors employed. The apparatus was evacuated to a low pressure (01"™) and vapor was admitted to a pressure depending on the Tonization Current. > Barss—TIonization by Collision in Gases and Vapors. 287 400 800 1200 COC EE ao a eee ar Ae ie | Ee A We Nb Sey pees Ee lt TR 30 Fic. 6. Ethyl chloride. SS ea ee Pl aS ee a a ao) 2 ee oe IS a a BRE eee ig 40 0 400 800 0 200 400 600 0 200 400 600 Fic. 7. Hther. 400 800 HEREC ott scant Hp > Volts per centimeter. Fic. 8. Methyl iodide. 238 Barss—lonization by Collision in Gases and Vapors. vapor pressure of the particular substance being used. This pressure was reduced by means of a water pump and the vessel was refilled with vapor. This process was repeated several times so that when readings were made a very small percentage of air was present with the vapor. Three sets of curves were obtained in each vapor corresponding to three different pressures, each set containing three curves corre- sponding to three different distances between the plates. These curves are shown in figs. 4 to 8 inclusive. In the vapor curves as in the air curves, ordinates represent ionization cur- rents and abscisse represent the field strength in volts per centimeter. In figs. 4, 6 and 8, the values of field strength for the middle set of curves are given at the top of the figures. A characteristic series of readings is given in Table IV. It was found that, for high pressures (9°9%, 5°3™ and 2°9™ in SO,), the ionization corresponding to saturating field strength was proportional to the pressure. This fact was observed for all vapors; and the result enabled smaller pressures, which could not be measured accurately by the open manometer, to be cal- culated. The vapors were not admitted into the chamber at pressures high enough to produce any appreciable permanent condensa- tion on the walls of the apparatus; for after each vapor was used, the ionization in air was tested, the results varying only as the sensitiveness of the electrometer varied. In every case there was no change in pressure during the time taken for a Tasre IV. SO2 (pressure 1°6™). See fig. 4. x: 0G, — A) 6 (L=0°8:") 6 (L=0 5") 50 6°30 80 7-98 2°85 100 6°30 160 7°98 2°9 200 6°31 280 2°9 420 8°03 525 780 6°30 3°35 795 6°41 845 8°40 1045 9°50 6°4 1056 aA 1200 11°10 13:0 1300 0°50 Barss—LIonization by Collision in Gases and Vapors. 239 series of observations, that could be detected by the mano- meter or the direct deflections of the electrometer. In the above table, L is the distance between the plates and 6 the corresponding electrometer deflection in centimeters per second. Discussion of Curves. The curves at the higher pressures in all the vapors employed give evidence of some extraordinary but characteristic behavior. The curve for a smaller distance between the plates rises more abruptly than that for a greater distance; so that the curves for two different distances cross each other. As the pressure is diminished, this process is somehow reversed, the curves becoming approximately parallel and finally diverging as the pressure becomes still further reduced. This is contrary to the behavior of gases in so far as we have experimental evidence, and is contrary to the facts that would be expected from Townsend’s theory. According to this theory, for any pres- sure, provided it is kept constant, the curves corresponding to different distances between the plates should diverge, as they do at the lower pressure. We have, therefore, experimental evidence which points to a distinet difference between gases and vapors, in so far as ionization by collision is concerned. Values of a in sulphur dioxide at a pressure of 0°5™ and in © ether at a pressure of 0-6 (pressures at which the curves diverge) were calculated from the formula stated on page 230 and are given in Table V. The values of a for different dis- tances between the plates agree sufficiently well to lead to the conclusion that Townsend’s theory, as developed for gases, holds in vapors at low pressures. TABLE V. SOs. pressure 0°5°™. x a (L=1™) a(L=0:8@) a«(L= 0:5) 280 0°17 370 0°45 0-48 500 Ue 1-75 1:90 580 P 3°41 3°25 3°44 Ether, pressure 0°6™., ¥ a(L=t5™) a(lL=im) a(L=0-5) 360 - O38 0°31 530 1-75 2-02 1-8 610 ~ 4°13 4°3 240 Barss Lonization by Collision in Gases and Vapors. The ionization in vapors was much more intense than in air under similar conditions of pressure, distance between plates and field strength. The thing which next suggested itself was to take air at a pressure that “would give about the same ioni- zation as that obtained with vapors (under saturating field strength) and to investigate for evidence of the crossing effect for different distances between the plates. Unfortunately when the experiment was tried, sufficient potential was not available to produce ionization "by collision at the pressure required. Reversing the electric field gave the same general results, the curves crossing for both positive and negative applied potentials. Experiments with a Cylindrical Chamber. It was thought that the apparent contradiction in the case of vapors might be due in some way to the collision effect of the positive ion; for, as stated in an earlier part of the paper, the ionic mobilities in vapors have given evidence of a characteris- tic behavior, that of the positive ion in some, cases exceeding that of the negative ion. To test for the relative ionizing powers of the positive and negative ions, a cylindrical ionization vessel was employed. A brass vessel 12°75 high and 6™ in diameter was coated on the inside with powdered uranium oxide. A central brass elec- trode 0°3°" in diameter, provided with an earthed guard ring, was connected to one pair of quadrants of the electrometer and an electromotive force was applied to the case of the chamber. It has heen shown that when the potential difference is small, the current through a gas in such a vessel is approximately the same whether the case has a positive or negative potential ; but when the potential difference is large, the currents are no longer equal. When the case is negative, a large increase in current is obtained when the electromotive force is increased. In this case all the negative ions traverse the field of strong electric force near the central electrode, and consequently acquire a velocity sufficient to ionize by collision the molecules of the gas. When the case of the chamber is positive, the increase in cur- rent is relatively very small when similar i increases in the electro- motive force are made. ‘This time the positive ions traverse the strong field near the central electrode and experiment has shown that they do not acquire sufficient velocity (no doubt due to their greater mass) to ionize by collision the molecules of the gas. This effect was tried in air at a pressure of 3°8"™, the result- ing curves agreeing very well with those obtained by Kirkby.* * Phil. Mag., vol. iii, p. 212, 1902. Barss—Ionization by Collision in Gases and Vapors. 241 Investigations were next made with sulphur dioxide and alcohol vapors for two different pressures in each. The results were of the same general nature as’ those obtained in air, thus showing that the positive ion in vapors plays a very small part in ionization by collision until the sparking potential is almost reached. So far, the anomaly presented by the behavior of vapors has not been definitely explained. It may be due to some instabil- ity in the molecular structure of the vapors when they are acted upon by alpha radiation; in other words, the alpha radiation may be effective in producing a chemical change. It is more probable that the anomaly may be explained by supposing the formation of aggregate molecules which act in the capacity of a thin metal wire in carrying a conduction cur- rent through the vapor. The continuity of these aggregate molecules would be more stable when the plates are closer together than when they are farther removed; and hence we would expect the current to increase as the distance between the plates becomes less. This fact experiment shows to betrue. It is also possible that the formation of these aggregate molecules will be a function of the pressure, the number formed decreas- ing as the pressure becomes less until finally a pressure is reached when the current is not increased by this conduction effect, but is entirely due to ionization proper. Summary. The collision constants are shown to be independent of the nature of the ionizing agent; therefore the negative ions gener- ated in a gas by collisions with alpha particles are identical with those generated by Rontgen rays and ultra-violet light. The theory of ionization by collision as developed for gases holds for vapors ionized by alpha particles provided the pres- sure is not too great. As the pressure is increased, evidence of irregularity is obtained which seems to point to some instability in the molecular structure of vapors, either due to the forma- tion of aggregate molecules or to some chemical change. In vapors, as in gases, the negative ion alone is effective in producing new ions by collisions with neutral molecules, until the sparking potential is approached. In concluding this paper, I wish to thank Professor E. M. Wellisch of the Sloane Laboratory for his many valuable sug- gestions and for his kindly interest throughout the experiment. 242 Williams—Ceology of Arisaig-Antigonish, District. Arr. XXI1.— Geology of Arisaig-Antigonish District, Nova Scotia ;* by Murron Y. WititaMs. (Contributions from the Paleontological Laboratory, Peabody Museum, Yale University. — Abstract of a thesis presented to the Faculty of Yale Uni- versity for the degree of Doctor of Philosophy, June, 1912.) Purposes of the investigation and scope of the article. Dvurine the summer of 1910 the writer and his assistant, Mr. M. H. McLeod of Northeast Margaree, Cape Breton Island, Nova Scotia, were engaged by the Geological Survey of Canada to make a detailed examination and survey of an area in northeastern Nova Scotia which it was hoped would furnish critical data for the unravelling of the Paleozoic strati- graphy of that part of the province. Pending the publication by the Geological Survey of the complete report with maps, the following summarized statement is offered as covering the most important conclusions. Acknowledgments. The writer wishes to acknowledge his indebtedness to Pro- fessor Charles Schuchert of Yale University and Professor W. H. Twenhofel of the University of Kansas for advice and assistance during the early part of the field work. Thanks are also due Professors L. V. Pirsson, Joseph Barrell, Isaiah Bow- man, and J. D. Irving of the Geological Department of Yale University, for assistance in the preparation of the report here summarized. Location and extent of the district. The Arisaig-Antigonish district of northeastern Nova Scotia fronts on Northumberland Strait about one-third of the way from Cape George to Pictou Harbor. From a water-front 10 miles long with Arisaig Point at its center, the district extends inland 11 or 12 miles to the southeast, including the gypsum deposits south of the Intercolonial Railway. The approximate area studied is 115 square miles. Previous work. Geologically the region is classic as a result of the labors of Sir J. William Dawsont and the Rev. D. Honeymant{ in * Published by permission of the Director of the Geological Survey of Canada. + Dawson, J. W., Acadian Geology, editions 1-4, 1855, 1868, 1878, 1891. ¢ Honeyman, D., Quart. Jour. Geol. Soc., London, xx, 1864; xxvi, 1870. Trans. Nova Scotian Inst. Nat. Sci., 1866, 1882, 1886, 1887. Williams—Geology of Arisaig-Antigonish District. 243 unravelling the stratigraphic sequence of the Silurian section exposed along the coast at Arisaig—a section unique because of its completeness, fossil contents, and decided European aftini- ties. Later workers on the Silurian rocks have been Dr. H. M. Ami* of the Geological Survey of Canada, and Professors Schuchert and Twenhofel.t In 1886 Hugh Fletchert of the Geological Survey of Canada published his final report on Pictou and Antigonish Counties, which gives the best general account of the district so far printed. Dr. Ami carefully described the Devonian strata of the area and published an account of the lower vertebrates obtained from them. Numer- ous other geologists have directly or indirectly added to the information relating to the region about Arisaig, which accord- ing to Honeyman became during his later life a household word in the homes of Canadian geologists. Physiography. Not only to the geologist and physiographer but to the casual traveler as well, the Arisaig-Antigonish district is a region of interest and delight. Possessing for Nova Scotia a maximum difference of relief, the area presents to the visitor during the summer months a delightfully green expanse of lowlands and rolling uplands, which flank a steep-scarped and generally wooded plateau. Trout streams, often several miles in length, occupy picturesque gorges in the plateau and uplands and flow in gently graded valleys across the lowlands. The plateau in places attains a height of 1000 feet and has an aver- age elevation of about 800 feet. The uplands vary from 200 to 400 feet or more in height and the lowlands occupy the lower elevations down to about 50 feet above sea-level, The plateau quite detinitely belongs to the land forms recog- nized in the Maritime Provinces by R. A. Daly§ as being the remnants of a Cretaceous erosion surface of low relief, and the uplands belong to the secondary erosion surface of Tertiary time recognized by Daly in Nova Scotia. However, because of the resistance offered by the Silurian and Devonian rocks to erosion, the land surface above them was never reduced very nearly to a plain. The lowlands, consisting of stream valleys and seaward slopes, are the product of the earlier erosion cycles plus the differential erosion of glacial and recent time. They ave best developed on the soft Carboniferous strata. * Ami, H. M., Trans. Nova Scotian Inst. Nat. Sci., i, new ser., pp. 185-192, 1892. Bull. Geol. Soc. America, xii. pp. 301-312, 1901. + Twenhofel, W. H., and Schuchert, Charles, this Journal (4), xxviii, pp. 143-169, 1909. { Fletcher, Hugh, Geol. Surv. Canada, II, p. 128P, 1887. § Daly, R. A., Bull. Mus. Comp. Zoology, Harvard University, xxxviii, pp. 73-108, 1901. 244 Williams—Geology of Arisaig-Antigonish District. Glaciation.—Throughout the region mounds of unsorted clay and gravel are common, and in places are capped by sorted sands. On the plateau the ice of glacial times moved, as inferred from the direction of strize on the rock surfaces, in a direction 10° east of south, but on the lower lands its course was somewhat varied as a result of the diversity of the topog- raphy at that time. Recent movements.—Post-glacial rejuvenation is thought to be indicated by the presence of elevated terraces along the Morthumberland Strait between 10 and 145 feet above the present sea-level. These terraces appear to be old sea beaches which have been modified only by the agents of erosion at work at the present time. The geological record. d The geological history of the district is written in sedimen- tary records representing, with interruptions, the time from Upper Cambrian to Pennsylvanian. The following table will help to make the sequence clear. SEDIMENTARY FORMATIONS. Cenozoic. Quaternary. 1. Recent.—Stream gravels and residual soils, modified glacial gravels. 2. Pleistocene or Glacial.—Unstratified clay-gravel deposits, red clayey marl. Paleozoic. ? Pennsylvanian or Upper Carboniferous. _ 1. Listmore formation (Millstone Grit of Fletch- er).—Light gray and red-brown sandstones, thin argillaceous shale, thin green conglomer- ate, ete, Thickness), (Mletcher) ee ese = 982 feet. Mississippian or Lower Carboniferous. 1. Ardness formation (Carboniferous Limestone of Fletcher).—Brown and green sandy shale, ripple-marked sandstone and shale, gypsum (along the I. C. R.), and a compact bed of gray limestone. Thickness (Fletcher and corrected), 2045 “ 2. McAra’s Brook formation (Carboniferous Con- glomerate of Fletcher).—Limy gray shaie, green shale, cross bedded conglomerate, brec- cia and basal conglomerate. Cut by intrusive diabase sheets and dikes. Thickness (letcher),1145 ‘ Williams— Geology of Arisaig-Antigonish District. Devonian (Lower). 1, Knoydart formation(Upper Devonian of Fletch- er).—Hard, fine-grained, red, sandy slate, and hard, gray sandstone, cut by small diabase dikes. Thickness (outcrops measured by Fletcher ; probably should be doubled), ---- Silurian. Arisaig series (Silurian of Fletcher). 1. Stonchouse formation (= more or less of Lud- low of England).—Red shale and limestones, argillaceous limestone and gray shales. Thick- messn(@luenhotel) 28 see oes Sek eee wes . Moydart formation (approximates the Louis- ville of United States, or Wenlock of Eng- land).—The red stratum or red shale, argilla- ceous limestone and shale. Thickness (T'wen- YORI oo Sp RN SSS ie See een ere, a ee 3. McAdam formation (Rochester of United States or upper Llandovery of England).— Black shales and argillaceous limestone. Thickness (Twenhofel) + iron-zone probably WO) ses iS tea d 21 Se Me eee eee 4. Ross Brook formation (=Clinton of the United States, or lower Llandovery of England).— Green shale with thin sandstones, dark papery slates, etc. Obscure basaltic? intrusive. hinckmessy(liwenhiotell) ee ee 5, Beechhill Cove formation (= lower Clinton).— Sandstones, limestones and shales, resting on aporhyolite flow and volcanic breccia. Thick- ness (estimated from width of outcrop), .-.- bo ? Ordovician. 1. Malignant Cove formation.—Coarse, cross bed- ded, silicified conglomerates and grits ; irreg- ular dikes of basalt. Thickness observed (original thickness probably much greater), - - Upper Cambrian (Ozarkic). Brown’s Mountain group (included in Cambro-Silu- rian of Fletcher). 1. Baxter's Brook formation.— Red and gray sandstones and schists, red and green slates ; intruded by rhyolite necks and dikes, quartz porphyry neck, diabase and basalt intrusives. Thickness represented estimated at.---.----- 2. James River formation.—Flinty graywacke and grits, silicified banded slates; intruded by granite, rhyolite, diabase, basalt, and monzonite. 245 683 feet. 379 1120 833-4 200 20+ 66 (73 (73 ce Thickness represented probably -.-.-.--.-.- 1 mile + Am. Jour. Sc1.—FourtH Series, Vou, XXXIV, No. 201.—Sruprempmr, 1912. Ne 246 = Williams— Geology of Arisaig-Antigonish District. Upper Cambrian deposits.—The oldest rocks known in the district are metamorphosed graywacke and slate of the Brown’s Mountain group, which underlie the plateau areas and form the base for the younger formations. All of the younger sedi- ments at one place or another rest directly upon the metamor- phie rocks. The Brown’s Mountain group may be divided lithologically into two divisions,—a thick lower formation of silicified graywacke, impure quartzite and gray banded slate known as the James River for mation; and an upper division of crumpled red slate with some sandstone and schist, known as the Baxter’s Brook formation. Qdlitic hematite beds are found in the James River rocks near the base of the Baxter's Brook division and again at a lower horizon. The sedimentary origin of the iron ore is most probable from the consideration of the odlitie and sparingly fossiliferous character of the ore, its longitudinal extent, and its close association with definite rock horizons.* Some secondary concentration or transference of material may, however, have taken place. So far as could be observed, the two formations have entirely conformable relations to each other, and on the evidence of Obolus (Lingulobolus) spissus and Lingulella (?) obtained from the upper iron-ore horizon (both from the ore itself and the associated schist), these rocks are proven to be of Upper Cam- brian or Ozarkie age. The iron ore is likewise correlated with the Wabana ore of Belle Isle, Conception Bay, Newfoundland ; but because of low grade and faulted condition it has not yet been commercially developed, although portions of it will prob- ably be profitably mined sooner or later. The characters of the Brown’s Mountain sediments are thought to indicate that they were deposited in a shallow transgressing sea; the upper red slates were formed, perhaps, by the inwash of fine oxidized sediments from a land already reduced to moderate relief. The present structure of the Upper Cambrian rocks consists of broad folds extending in a northeast direction, crossed by a few closed northwest secondary folds. The upper red slates are also intimately crumpled and folded as a result of the many intrusive bodies which have penetrated them. 4 Ordovician deposits.—Erosion agencies appear to have worn wide channels out of the Brown’s Mountain rocks before the later sedimentary formations were laid down. ‘The coarse cross- bedded conglomerates and grits of the Malignant Cove forma- tion, occurring at Malignant Cove and to the south, were deposited upon the cleavage surfaces of the James River slates. Their deposition was evidently influenced by strong current * For particulars see Woodman, J. E., Canada, Dept. Mines, Mines Branch, Report on Iron Ore Deposits of Nova Scotia, Pt. I, 1909. Williams— Geology of Arisaig-Antigonish District. 247 action and their sedimentation characters, together with their general red color, suggest for these poorly sorted but well-worn deposits a continental origin. ecause of the orogenic disturbances suggested and because of the silicified character of its rocks, which are similar to those of the Brown’s Mountain group, the Malignant Cove formation is thought to be a remnant of early Ordovician sedimentation. Silurian deposits.—On the shore about one mile west of Malignant Cove, rocks of Silurian age rest unconformably upon an old rhyolite flow. As the Malignant Cove conglom- erate contains fragments similar to, if not identical with, this rhyolite, it seems probable that the rhyolite flow was earlier than the conglomerate deposition and may have been at one time covered by the Malignant Cove formation. Be this as it may, but small isolated deposits of conglomerate now remain between the rhyolite flow-breccia and the sandy shales at the base of the Arisaig series. The Silurian formations occupy an area about 13 miles wide, extending from Malignant Oove about 6 miles to the southwest along the shore. The 3,500 feet of sediments here represented consist in a generalized ascending order of argillaceous sand- stone, black carbonaceous shales, arenaceous and argillaceous shales, a 24 foot bed of fossiliferous hematite, argillaceous limestones, and red shales. As previously shown by Schuchert and Twenhofel, the Arisaig series represent a period of time in_ Europe between the lower Llandovery and the Ludlow, and so far as they can be correlated with other American occurrences they represent the time interval between the Clinton of eastern New York and the Guelph of interior America. The sedi- ments are thought to be the deposits of a shallow sea during varying conditions of clear and muddy waters. For the detailed description of the formations and list of their fossils the reader is referred to the work of Twenhofel and Schuchert cited above. The Silurian formations are separated from the older rocks by a great fault having a probable throw of 3000-4000 feet. During the down-faulting, readjustment within the younger strata took place, resulting in crumpling, overturning, and many small faults which have divided the area into a number of blocks. Because of the soft, yielding nature of the strata the structure is but poorly expressed in the surface exposures. Devonian deposits.—Red sandy slates containing some gray impure sandstone rest unconformably upon the Silurian strata to the southwest. They, too, have been downfaulted to the north- westward by the major dislocation of the region, and have suf- fered readjustment in the form of a synclinal flexure and minor faulting. Ami, who gave these the name of Knoydart 248 Williams—Geology of Arisaig-Antigonish District. formation, has shown from Ostracoderm fishes obtained from the lower strata that the deposits may be correlated directly with the lower Old Red Sandstone of Europe. In origin the Knoy- dart formation is evidently coutinental and probably originated in large measure along the estuary of a Lower Devonian river. Mississippian deposits. —Mississippian sedimentation is rep- resented in the Arisaig-Antigonish district by the McAra’s Brook and Ardness formations. The former consists of red conglomerates, sandstones, and sandy shales, and on the east of the area includes considerable thickness of micaceous eray sandstone and oil-shale. Remains of Calamites and fern-like impressions have been found in the oil-shale. Near the top the sandstone is limy and is apparently conformably overlain by the basal limestone of the Ardness formation. “ The limestone which forms the lowest stratum of the Ard- ness formation is about 20 feet in thickness, and is succeeded by red sandstone and sandy shales, with some similar inter- bedded deposits of gray or greenish gray color. Along the south of the district, particularly in the valley occupied by the Intercolonial Railway, gypsum deposits probably 200 feet thick succeed about 200 feet of red sandstone and shale which rest upon the basal limestone. In the gray beds, particularly in those exposed along Northumberland Strait, fossil plants and carbonized wood occur in small amounts. The horizon-mark- ing fossils are brachiopods obtained from the basal limestone exposed west of McAra’s Brook. These are: Productus doub- lett Beede, very common; P. dawsoni Beede ; Pugnawx sp. undet.; Martinia glabra (Martin); and Beecheria davidsoni Hall and Clarke, rare. The limestone is thus shown on faunal evidence to be the same as that occurring at Windsor in the Windsor series. On the basis cf the age determination thus made and the apparent conformability of the McAra’s Brook and Ardness formations, they are both considered to be of Mississippian age. t Pennsylvanian deposits.—W estward along the Northum- berland Strait the Ardness formation is overlain by strata con- sisting of red and gray sandstone and sandy shale. Fletcher has termed these rocks the Millstone Grit formation ; to avoid possible confusion with other areas in Nova Scotia, also thought to be equivalent to the so-calied Millstone Grit, the present writer has distinguished the deposits in the Arisaig region as the Listmore formation. So far as evidence goes in the Arisaig-Antigonish district, the Listmore formation overlies the Ardness contormably. In def- erence, however, to the determination of Fletcher, based on observations made over a wide area, the age of these strata is provisionally considered as Pennsylvanian, and a disconformity vr) Ae Williams—Geology of Arisaig-Antigonish District. 249 is thus presupposed between the Listmore and the Ardness formations. summary of the Carboniferous deposits.—The three forma- tions of Carboniferous age are much alike in their general char- acters. Although exhibiting minor flexing and ‘faulting, the strata have not suffered er reat disturbances and overlap the great fault zone which affected the older formations. They generally dip with gentle gradients away from the rocks repre- senting the older land. The many highly inclined contacts and the unsorted, breccia nature of the basal conglomerate, partic- ularly along the scarps of the plateau consisting of Upper Cambrian rocks, suggest deposition of material near its source in previous troughs of erosion. The McAra’s Brook conglom- erates probably represent a phase of continental sedimentation which was later characterized by swampy conditions or, at any rate, non-oxidizing conditions, which favored the deposition of gray sandstone and beds of oil-shale and impure coal. Shallow marine or littoral conditions followed, culminating in the lay- ing down of the basal Ardness limestone. Shallower waters again prevailed, and in isolated pans gypsum deposits collected as a result of excessive evaporation. As the Appalachian rev- olution began to affect the area, continental conditions finally superseded the littoral and continued not only during the depo- sition of the upper beds of the Ardness formation and the whole thickness of the Listmore formation, but according to the work _ of Fletcher lasted through the formation of the coal measures and the deposition of the Permian in the vicinity of Pictou and westward. The Windsor submergence was evidently the last of marine conditions for northern Nova Scotia. Igneous Geology. A number of eruptive and irruptive rocks are associated with the Upper Cambrian formations of the Arisaig-Antigonish dis- trict. The largest individual intrusion is that of fine-grained pink granite north of James River station. This is in the form of a stock, and as seen in surface exposures is dense and evidently represents conditions not far removed from the con- tact with the former cover. East of Malignant Cove, what appears to be an irregular monzonite stock is exposed along the shore. A neck of rhyolite forms the center of the Sugar Loaf Hall south of Malignant Cove, and rhyolite exposures in the vicinity indicate that erosion has laid bare considerable masses of an eruptive rock which had never quite reached the surface. The rhyolite grades into quartz porphyry containing large phe- nocrysts of orthoclase feldspar. The irregular porphyry bodies evidently represent intrusions similar to those of the rhyolite but more deeply eroded. 250 Williams—Geology of Arisaig-Antigonish District. Through the surrounding rocks there are many small rhyo- lite dikes which appear to be connected with’ the general rhyolite intrusion. An aporhyolite or devitrified rhy olite flow probably 200 feet thick rests at the base of the Silurian section and is of historical interest because it was long mistaken for metamorphosed sediments on account of its banded structure. “ Kozoon” forms were at one time reported from it. None of the above intrusives is known to cut any rocks younger than those of Upper Cambrian age, and excepting the monzonite all are perhaps of related origin. Fragments of rhyolite are common in the Malignant Cove conglomerate, which is supposed to be of early Ordovician age, and fragments representing the James River granite and the monzonite are also thought to have been recognized in the conglomerate deposits. The available evidence thus favors the supposition that the intrusion of rhyolite, granite, quartz porphyry, and monzonite, and the extrusion of the aporhyolite flow took place during late Cambrian or early Ordovician time. Obscure tuff beds associated with the Upper Cambrian iron-ore zone indicate voleanic activity at a still earlier date, but the deposits are too much altered to shed much light upon the characters of such remote eruptions. At a number of localities in the district diabase occurs either as irregular necks, which is the case at the Sugar Loaf Hill north of Antigonish town, or else as dikes, generally but a few feet across. In a number of cases the basic intrusives are of a basaltic rather than of a diabase nature. Unlike the igneous rocks already described, the diabase cuts rocks of all ages from Upper Cambrian to early Mississippian. Apparently the intru- sion took place during one general activity, and so the diabase dikes and necks are probably of early Mississippian age. Intimately associated with the diabase intrusives along the shore east of Arisaig Point is a long red dike of soft fissile character. It cuts the diabase dikes and the aporhyolite flow. This dike has been traced for nearly 3 miles, although numer- ous breaks occur in it, and part of the way two dikes are pres- ent instead of one. Studied microscopically, the red dike is seen to contain much iron oxide, but it evidently was originally composed of clastic material. It is thought to have been of a fine breccia nature, which may have originated during pulsations of material which reached near ly to the surface. In places the red dike is associated with basalt which it cuts. It is possible that the basalt and the red dike represent late phases of the diabase intrusion and are essentially of the same age. Thus igneous activity represented in the Arisaig-Antigonish district is thought to have been confined for the most part to late Cambrian or early Ordovician time, and to the early part of the Mississippian period. Schuchert—Jackson on the Phylogeny of the Echini. 251 Arr. XXIII.—Jackson on the Phylogeny of the Echini ;* A synopsis by CHARLES SCHUCHERT. Iy this monumental monograph are established the phylo- -geny and classification of the Echini, including young and adult, fossil and living types, and “based on the sums of the characters and not on single characters.” The volume also contains a revision of all Paleozoic Echini. The splendid and fully illustrated work is dedicated to the great echinologist, Alexander Agassiz, and to Alpheus Hyatt, “my beloved master and friend, whose principles of research are the keynote of this memoir.” Hyatt’s principles are the stages in development, senescence, acceleration, and parallel- ism, and it has been Jackson’s constant aim to compare these stages with the characters of more or less closely associated types. coon began to study Echini in 1896 and during the past seven years he has devoted most of his time to a detailed study of the species and genera of this class of Echinoderma. He assembled in his private collection more than 40,000 specimens of Itecent and Mesozoic Kehini in all stages of growth, actually studying more than 50,000 specimens, so that he might thoroughly understand the Paleozoic species and their phylo- genetic relations to the later forms. That he has succeeded . the volume bears abundant evidence, for no class of inverte- brates, as a class, has been wrought ont with more care and philosophic insight. One of the most important features of the work is a new method of determining ontogenetic stages of growth by noting how the plates are introduced ventrally, and in the localized stages among the plates dorsally. Echini are a_ particularly good class to study phylogenetically, because they have so many parts, all of which must be taken into consideration. This mass of detail furnishes constant checks and when all are in accord proves the accuracy of the resulting phylogenetic scheme. Geological Occurrence. Aldrovanus in 1618 was the first to figure a fossil echinoid from the Paleozoic and curiously one of the rarest of species and the oldest geologically, Bothriocidaris globulus. The author recognizes 24 genera of Paleozoic Echini and of these but 4 are new to paleontology (Hyattechinus, Lovene- * Phylogeny of the Echini, with a revision of Palaeozoic species; by Rob- ert Tracy Jackson. Mem. Boston Soc. Nat. Hist., vii, quarto, 491 pages, 76 plates, and 258 text figures, Jan., 1912. 252 Schuchert—Jackson on the Phylogeny of the Echini. chinus, Meekechinus, and Pholidechinus). Of good species there are 119 and of these 23 are new. Of ¢neertae sedis and nomina nuda there are 3 genera and 34 named forms. Ordovician.—The oldest and most primitive Echini oceur in the Middle Ordovician of Esthonia, Russia, where the genus Bothriocidaris is found with 3 exceedingly rare species (13 in the Jewe and 2 in the Lyckholm formations). Silurian.—The oldest American representative of the class was recently found in the Rochester shale of New York (Koninckocidaris silurica, n.sp.). In the Llandovery of Eng- land occurs Maccoya phillipsiae, while the lower Ludlow has furnished Palaeodiscus ferox and Echinocystites pomum. Devonian.—Germany has in the Middle and Upper Devon- ian Xenocidaris (3 species), Hocidaris laevispina, aud Lepido- centrus (8). In the Upper Devonian of New York is found L. drydenensis. In England occur Lepidesthes devonicans and Pholidocidaris ucuaria. Lower Carboniferous.—The Lower Carboniferous is the period of greatest development of Paleozoic Echini and in America alone there are 50 good species, with 31 more in Europe. In the Millsap formation of Colorado occurs dLio- cidaris cannoni, a new species and the oldest stratigraphically of the Cidaridze, the stock that gave rise to Mesozoic and later Echini. Archaeocidaris has 10 American species and 12 other forms occur in Europe. Other genera in America are Lepido- cidaris (1 species), Lepidocentrus (1), Hyattechinus (3), Pholide- chinus (1), Palaeechinus (1), Maccoya (2), Loeeneeaiee (4), Oligoporus (5), Melonechinus (11), Lepidechinus (3), Perischo- domus (1), Lepidesthes (6), Pholidocidaris (1). Upper Carboniferous.—* In the Upper Carboniferous the Palaeozoic Echini have dropped out with extreme suddenness and relatively few species are known.” Archaeocidaris has 17 species in America and 2 in Europe. The only other form is the American Lepidesthes extremis, 0. sp. Permian.—The cidaroid Miocidaris keyserlingy occurs in Germany and England. Of Archaeocidaris there is 1 species in America, 1 in India, and 1 in Australia. The only other form is the American Meekechinus elegans, n. sp. Types of variation defined. Jackson states that next to stages in development variation is an extremely important subject as a basis in phylogenetic determinations. Echini are especially valuable on which to study variation, because in them variation can be so definitely expressed. It is seen in the introduction of columns, number vf plates in a row, number of ocnlars that reach the periproct, ete. Schuchert—Jackson on the Phylogeny of the Echini. 253 “In order to appreciate variation it is of fundamental importance to be familiar with the characters of the associated species and genera of a case in hand, and also the develop- mental characters of the same. Variation may be fairly clas- sified under five more or less distinct heads : “1. Arrested variation, in which the variant retains char- acters seen in its own young and typical of the adults of more primitive allies, but characters which are usually eliminated in development. . “9. Progressive eariation, in which the variant has char- acters not typical of the species, but which are further evolved on the direct line of differential development, and are seen typically in more evolved nearly allied species or genera. . . “3. Regressive variation, in which the variant takes on characters of the adult of some simple and more primitive type of the group. Such characters are not necessarily a repe- tition of youthful characters but may go back to a remote ancestry. An arrested variant in a sense is one form of regres- sive variation, but a regressive variant includes much more than arrested variation. To distinguish them, an arrested variant is one that has developed to a certain point as usual, and then failed to take on the later added characters typical of the species, so that, although an adult, it has immature char- acters. A regressive variant is one that has attained full char- acters and then in later life has reverted to youthful or primitive characters as an individual variation, or it is a variant that from youth has primitive characters not normally seen in the development of tle species. “4. Parallel variation is where a character is taken on exceptionally which may be compared with characters nor- mally occurring in some type of the group not closely con- nected, so that it cannot be genetically compared. «5. Aberrant variation is where a character is taken on which is quite abnormal, not to be correlated with the typical condition in associated forms” (pp. 18, 19). Comparative Morphology. Significance of abnormal symmetry.—Kchini are remarkably constant in their pentamerous system, but Jackson found 71 variant individuals or on an average ‘‘a little more than one to a thousand. The variants are partially or completely tri- merous, tetramerous, and hexamerous. . . . The ocular plates seem to exert a controlling influence in the building up of the corona, as below and in immediate contact with the oculars originate the coronal plates, both ambulacral and inter- ambulaeral. In -connection with each ocular is developed a whole ambulacrum, and, in addition, a half-interambulacrum 254 = Schuchert—Jackson on the Phylogeny of the Echini. on either side. That is, while an ambulacrum originates on the ventral border of an ocular, each interambulacrum may be considered as composed of two halves, the plates of which originated on the left or right of the area in contact with the adjacent oculars. If this is true , then the loss of an ocular would cause a failure to develop of the plates that normally went with it, also an abnormal position of an ocular should cause an abnormal distribution of the associated coronal plates” (35, 36). The “ variations from the pentamerous symmetry can all be considered as monstrosities” (50). Ambulacral areas.—The ambulacrum ‘is the most essen- tial feature of a sea-urchin, and has a first importance in clas- sification and morphology, ‘on account of the varied structure that it presents” (53). Interambulacral areas.—* The interambulacrum in Echini functions chiefly as a space filler and a bearer of spines and pedicellariae. The spines serve for protection and more or less in locomotion, and pedicellariae as grasping, cleansing, and protective organs. In spite of this secondary physiological importance, the interambulacrum forms a large part of the test of the sea-urchin in most types, and is of very great interest, especially in Palaeozoic genera. The interambulacral plates originate in direct contact with the ocular plates and quite independently of the genitals. “The full differential characters of the interambulacrum as of the ambulacrum are expressed at the mid-zone of the adult. Here are usually found the full number of columns of plates characteristic of the species, also the typical tubercles, spines, imbrication, or other characters which go to make up the specific description. The ventral border in the basicoronal zone represents the earliest formed plates and the youth of the individual, as far as it can be gathered from the study of an adult specimen, though the actually first formed plates may have been resorbed in development. Passing dorsally, with later added plates, new characters may come in until we get the full differential features developed at or about the mde zone. Dorsal to the mid-zone we pass into the area of young last formed plates which have not yet acquired the full char- acters. Or again dorsally, we may find senescent features in the loss of columns of plates. Passing from the basicoronal row dorsally, we find in most Palaeozoic types, and many post- Palaeozoic as well, stages in development strongly marked, which stages can be correlated with the adult condition of simpler genera or simpler species within the genus. The interambulacrum in Echini has from one to fourteen vertical columns of plates in each of the five areas, which represents the least and greatest number known at present. There are Schuchert—Jackson on the Phylogeny of the Echini. 255 intermediate grades representing every step between this least and greatest specialization of the area, and it is a matter of great interest to follow the progressive series as represented by stages in development, and by adult types, to see how the pro- gressive differential structure is built up. As the plates of the ventral border are the oldest or first formed of any plates seen in an individual specimen, and as the later added _ plates succeed one another as we pass dorsally, it might be thought that we could read stages in development as expressed by rows and columns of plates with ease and certainty, and such can be done in many types . . . Complications may come in, however, especially resorption of the base of the corona by encroachment of the peristome cutting off part of the ventral plates, and also rarely resorption within the corona, as excep- tionally in Arachnoides, or differential growth of associated plates, which may separate plates originally in contact (Echin- arachuius)” (62-4). Base of the corona.— The characters of basicoronal inter- ambulacral plates are the more striking and may be stated in brief. Where no plates haye been removed by resorption, there is a single plate at the ventral border of the interambu- lacrum. The primitive type of this character is Bothrioci- daris, which continues to build a single column. This same character of a single plate ventrally, but succeeded by two plates in the second row, is characteristic of the young of all modern regular Echini. . . . In the adult of most regular Eehini the single plate and probably more have been resorbed by the advance of the peristome (Kucidaris). In the Palae echinidae with many columns of plates, apparently only one plate has been resorbed, when we find two plates in the basicoronal row, . . . or in the Archaeocidaridae, several rows of plates may have been resorbed, and we find four plates in the basicoronal row. ri “In Bothriocidaris the basicoronal row consists of two high hexagonal ambulacral plates with pores superposed in each ambulacral area and one interambulacral plate in each interam- bulacral area. This same character is seen in young cidarids, young Strongylocentrotus, and LEchinus, young Salenia, Arbacia, and Phormosoma. It is, I think, fair to call this a primitive character, and it represents what I (1896) described as the protechinus stage. The protechinus stage is comparable, in other groups of animals to the protoconch of cephalous Mollusea, what I (1890) described as the prodissoconch of Pelecypoda, and to Beecher’s (1901) protegulum of Brachio- poda and protaspis of Trilobita. All are referable to what I termed (1890) the phylembryonic stage in development, a stage in which the differential characters of the class are estab- lished in ontogeny” (69-71). 256 Schuchert—Jackson on the Phylogeny of the Echini. Oculars and genitals.—* An ocular plate in Eehini overlies an ambulacrum wholly and the two adjacent interambulaera in part on either side. Immediately on the ventral borders of the oculars all coronal plates originate. It seems that at this point the tissues exist which give rise to new plates. The five oculars are always present barring the excepted Pour- talesia. The genitals overlie the interambulaera in part, but not the lateral borders of the same, and never reach the ambu- lacra. In some cases the genitals may not reach the interam- bulacra. Five genitals are ‘ty pically present, but the posterior genital may be wanting (spatangoids) or one absent as an aber- rant variation. “Tn the ancient Bothriocidaris the oculars are exceptionally large, relatively to the size of the animal; on the other hand, the ‘genitals are exceptionally small, relatively the smallest of all known Echini. “No pores have been observed in genital plates in Bothrio- cidaris. It is possible they did not have genital pores, as such are wanting in the young of Recent Echini ; more likely they were present, but do not show in external view . . . Also no pores have been observed in ocular plates of Bothrio- cidaris. “ Genital and ocular plates are rare in Palaeozoic types, yet excepting the Echinocystoida I am able to show them in all families other than the Archaeocidaridae and in most genera. After Bothriocidaris just considered, the leading character in the Palaeozoic is for all the oculars to reach the periproct, and to cover the ambulacra and in part the interambulacra on either side. Also the genitals reach the periproct, are larger than the oculars, and cover the interambulacra in part, but not wholly, because the lateral borders of the interambulacra abut against the next adjacent ocular on either side.” In Paleozoic Echini, as a rule, “the ocular plates are imperforate. . . . In post- Palaeozoic Echini ocular plates have one pore not always visible externally (Salenia, Arbacia) and very rarely a second pore may exist as a variant. I have seen only two or three such” (86-89). Systematic value of oculars.—* A close study reveals char- acters of importance to general morphology, to the evolution of the group, to the relation of the species in the genus and related genera, and to geographical distribution. Ocular plates present an excellent systematic character which has been largely overlooked. “ Early in my studies of these plates it was seen that they had an important bearing, and observations were made on all available specimens of regular Echini, Mesozoic and Recent. In the fossils this is not always easy, as for purposes of study, all five oculars and genitals must be observed, and they are Schuchert—Jackson on the Phylogeny of the Echini. 257 frequently lost in fossils. I have succeeded, however, in mak- ing observations on something over 50,000 regular Recent and Mesozoic specimens representing 133 species.... The reason for making so many observations was that while the character of a species is usually gathered correctly from five or ten specimens, the variations seen in a large number present inter- esting data for comparative study... “In Mesozoie regular Echini the dominant character is for all the oculars to be exsert, or excluded from the periproct. In the Recent regular Echini the young also have all the oculars exsert. In the adult all the oculars may be exsert or one or more be insert. While the exsert character of the young is like the Mesozoic, the becoming insert in development is the taking on of a character which in this respect is directly com- parable to the dominant character of the Palaeozoic. . . “‘ As becoming insert is a progressive character with develop- ment, species in a genus that have the greatest number of ocular plates insert may be considered in this respect more evolved than other species which have a less number (Arbacia, Kchinometra). Also, as a matter of variation, individuals that have fewer oculars insert than is characteristic of the species may be considered arrested variants, and those that have more plates insert than is typical may be considered progressive variants. Such variants can frequently be compared directly with related species or genera where the fewer or greater num- ber of oculars insert is a typical specific character (Cen- trechinus). Specimens of a given species from different local- ities present often quite striking differences as regards the number of plates which are insert, those from one locality having typically fewer oculars insert than those from a differ- ent locality. Such variation with locality may well be con- sidered as indicating incipient species, as, where there is a dit- ference, specimens from one locality must be more progressive or less progressive than those from another. . . “The number of oculars insert has been spoken of by pre- vious writers as if it were a concurrent of age, and the largest specimens had the most oculars insert. My observations are directly opposed to this view. All the evidence goes to show that the full number of oculars that are to become insert are developed early in the life of the individual, and apparently later no change in this respect takes place. A series of specimens half the mature size or larger may in most species be safely accepted as showing the mature characters as regards oculars. This is on the basis of observations on 11,500 specimens of Strongylocentrotus drébachiensis, all from one locality, Dumpling Islands, the specimens varying from very young to adult, and all measured and tabulated as later 258 Schuchert—Jackson on the Phylogeny of the Echini. described. With few exceptions it was found that the larger individuals in a species are typical as regards ocular plates, and that variations, both arrested and pr ogressive, are more frequent in smaller individuals, often half grown as regards size” (90-1). Genital pores.—“ In very young Echini genital pores do not exist... Typically, in post-Palaeozoic regular Echini there is a single genital pore within the confines of each genital plate. In the Ordovician Bothriocidaris genital pores are unknown. ... In other Palaeozoic Echini genital plates typically have more than one pore to a plate. There may be two or three . . or there may be three to five ina plate... . Instead of a few pores there may be numerous genital pores to a plate, even as many as ten or eleven.... It is possible that in types where fine madreporic pores are unknown, some of the larger pores served as madreporic openings. Otherwise all the pores in genital plates doubtless connected with genital glands, as in Recent Echini with accessory pores” (170-172). Secondary value of genitals in classification.—“‘As seen from these studies, the genital plates have nothing to do with the interambulacrum, which develops on either side of the oculars. The genitals typically possess genital pores, and one of them possesses madreporic pores, but both of these structures may pierce other parts of the test. Genital plates may, therefore, be considered as structures of secondary importance, of much less morphological value than are the oculars” (173). The lantern.—“ It is believed that the structure of the lantern is of great value in systematic classification, and that the structure of its several parts presents characters that are of ordinal or subordinal value. As Dr. Mortensen pointed out (1904), the structure of the teeth, keeled or unkeeled, is ‘a very important character, though it has hitherto received very little attention.’ Besides the teeth there are other features of value. Briefly stated, the essential points are: teeth grooved or keeled ; epiphyses narrow, or wide and united by suture; the top of the pyramids, as seen when the epiphyses are removed, a smooth floor, or pitted; foramen magnum deep, or shallow; angle of outline of the lantern depressed or erect ; compasses present or absent” (177) Classification. The class Echinoidea Jackson defines as follows: “The Echini, though possessing a wide range of structure, may be described as animals possessing alimentary, reproduc- tive, nerve, and water vascular systems within an enclosing superficial pentamer ous skeleton which bears movable spines. There are from two to twenty columns of plates in each of the Schuchert—Jackson on the Phylogeny of the Echini. 259 five ambulacral areas and from one to fourteen columns of plates in each interambulacral area. New coronal plates are formed at the ventral border of the five ocular plates, ambu- lacral pores pass through ambulacral plates, rarely (clypeas- troids) in part between plates. The peristome in all but the Exocycloida bears from one to many rows of ambu- lacral plates, with or without non-ambulacral plates. There are five oculars (apparently in part or wholly wanting in some of the Pourtalesiidae), and five genitals or fewer, the whole being fused into a mass in certain types of Exoeyeloida. The genitals typically have each one or more pores as exits of the five interradially situated reproductive glands. In addition, typically, madreporic pores exist in gen- ital 2, but are not recognizable in most Palaeozoic forms. The periproet is more or less plated, situated within the oculo- genital ring, or in irregular types outside of that area; the anus is in the periproct. The masticatory lantern is composed of forty pieces (or clypeastroids thirty pieces); it is wanting in adult spatangoids. Respiratory organs consist of Stewart’s organs, peristomal or ambulacral gills. Locomotion is effected by ambulacral feet or by spines, or both” (200). A key to the classification of the Echinoidea is given on pages 201 to 208. Other keys to the species of Paleozoic Echini are given under the systematic descriptions. Ancestors of Echint.—The author states that Echini “make no close approach to other classes of the Echinodermata. .. . What the ancestor of the Echini as a class was is unknown, but it might fairly be sought amongst the Cystoidea” (200). Basis of classification.—All Echini recent and fossil are classified by Jackson on the basis of “the structure of the adult and the development of the same... . no single character has been followed.” The characters taken into consideration are : the ambulacrum, interambulacrum, coronal imbrication, basi- coronal plates, ventral resorption of corona, ocular and genital plates, periproct, peristome, Aristotle’s lantern, perignathic girdle, spines and tubercles, gills and sphaeridia. ‘‘ The relative value of these parts naturally differs in different groups of the Kehini ” (199). The protechinus.—* The most primitive type of Echini, I believe emphatically, is Bothriocidaris. This view is based on the simplicity of its structure, and especially on the close comparison of this structure with that seen in the very young of all geologically later Echini known and the youthful charac- ters retained at the ventral border in the adults of many types” (208). ‘“‘Kach interambulacrum of Bothriocidaris consists of a single column of plates, which is represented by a single plate at the ventral border of the interambulacra in the young of all other 260 Schuchert—Jackson on the Phylogeny of the Echini. Eehini. . . . there is ample proof that the interambulacrum begins with a single plate, as shown by Lovén (1874), and Mortensen (1903). ... This structure with less evidence I cor- related (1896) as a stage in development with the single column of plates in Bothriocidaris, naming it the protechinus stage. As Palaeozoic types with many columns of interambulacral plates begin at the ventral border, the young, with a single plate rep- resenting a single column, and later add their several columns during development, it seems that Bothriocidaris throws great light on the numerous columns there existent ” (210). Order Bothriocidaroida.—Oft Echini the oldest and most primitive order is the Bothriocidaroida, found in the Middle Ordovician of Esthonia. The only genus, Bothriocidaris, has 3 very small species, with 10 columns of hexagonal ambulacral plates, each with a pair of centrally placed podial openings, and but 5 columns of interambulacrals which may have small spines. Plates not imbricate. Periproct within the oculo-genital ring, which consists of 5 very large oculars and 5 very small gen- itals. Jaws present. It is out of this stock, the protechinus stage, that all regular Echini have evoluted as follows : Later Echini.—* The feature of Palaeozvic Echini is that they have more than two columns of plates in each interambu- lacral area. This is true of all known forms excepting Bothri- ocidaris and Miocidaris as far as the latter occurs in the upper Palaeozoic. Gregory (1887), Sollas (1899), and others have assumed that the most primitive form of Echini had many columns of interambulacral plates in an area, and several authors have considered Palaeodiscus as the most primitive known type. On this basis evolution would entail a loss of such parts, as our modern types all have two columns of inter- ambulacral plates in an area. The evidence of development and adult structure is opposed to this view. At the ventral border of the young of all known modern types, and at the ventral border of the adult where not removed by resorption, we find a single primordial plate in each interambulacral area succeeded in the second row by two plates. There is no evidence in development of a larger number of columns dropping out to two in any known living form, or indeed, in any fossil form except- ing as seen in senescence (Perischocidaris), and in the little known Tetracidaris of the Cretaceous. I, therefore, consider the Echini usually classed as the Euechinoida, with a geolog- ical range from the Lower Carboniferous to Recent inclusive, and comprising the orders Cidaroida, Centrechinoida, and Exo- cycloida as next related to Bothriocidaris. This view is based on structure and development. Iam well aware of the inter- vening [great] geological gap, but can only appeal to the rarity of all forms in the Silurian and Devonian to account for the absence of intermediate types. Schuchert—Jackson on the Phylogeny of the Echini. 201 “The order Cidaroida is placed as directly derived from the Bothriocidaroida without known intermediate forms. The Cidaridae as regards the structure of the young and adult are the least removed from Bothriocidaris of any known echinoid, living or fossil. The young have high hexagonal ambulacral plates with the pores of the pore-pairs superposed. Hach inter- ambulacrum has a single plate ventrally, sueceeded by two plates in the next row. The peristome has a single row of pri- mordial ambulacral plates which are like those of Bothrioci- daris excepting that in that type there are two peristomal rows. The base of the corona has not yet been resorbed, exactly like adult Bothriocidaris. In young cidarids the genitals are large and oculars small and exsert, unlike Bothriocidaris. : The Cidaroida present distinctly a combination of Palaeozoic with modern characters” (211-2). Order Cidaroida with 10 columns of simple ambulacral plates and 10 of interambulacrals. Coronal plates rarely imbri- cate. Represented in the Paleozoic by Miocidaris (1 species in the Permian Zechstein) of Germany, and another in the Millsap of Colorado. Order well represented from early Meso- zoic time to Recent. Out of the Cidaroida was developed the Order Centrechinoida, where the ambulacral plates are usu- ally compounded of demi-plates. The stock arose in the Tri- assic and continued to Recent. This order divides into 3 new suborders: (1) Awlodonta (Triassic to Recent), with teeth of . the lantern grooved, and with epiphyses narrow and not meet- ing in suture over the foramen magnum (Hemicidaride, Aspidodiadematidee, Centrechinide, and Echinothuriide) ; (2) Stirodonta (Jurassic to Recent), with the teeth keeled and with narrow epiphyses (Saleniide, Phymosomatide, Stomopneu- stide, and Arbaciide); (8) Camarodonta (Cretaceous to Recent) with keeled teeth and wide epiphyses meeting in suture over the foramen magnum (Temnopleuride, Echinidee, Strongylocentrotidz, and Echinometride). The last named suborder is the most specialized of modern regular Echini. Order Exocycloida, or the irregular Echini, developed out of the Stirodonta in the Jurassic and persists to Recent. Here the periproct is always outside of the oculo-genital ring and lies in interambulacrum 5. “Assuming a monophyletic origin for the group, the three suborders present a striking series of structural departures from the regular Echini from which they doubtless originated. Considering the characters of the group as a whole in brief, the compound ambulacral plates and peri- stomal gills of the Holectypina and the auricles of that group and the Clypeastrina, the existence of keeled teeth, where teeth are known, and the presence of sphaeridia, are all characters which unquestionably associate the Exocycloida with the Cen- trechinoida and not with the Cidaroida, where these structures AM, Joe coms guers SERIES, VOL. XXXIV, No. 201.—Srpremper, 1912. 262 Schuchert—Jackson on the Phylogeny of the Echini. are non-existent. Mr. Agassiz (1909) has shown that in the young of the spatangoid Echinoneus a well developed lantern exists. This discovery is of the greatest interest and impor- tance, as previously teeth were unknown in this group. . Looking back to the Centrechinoida, we find that this type of lantern exists only in the suborder ‘Stirodonta. Further, the attachment of muscles, as stated, occurs only in Arbacia and prob- ably other members of its family. . . . Itherefore consider the Exocycloida as connected with the Arbaciidae, prebably through some early common ancestral stock” (217-8).' The Exoey- cloida have 3 suborders: Holectypina (Jurassic to Eocene) with the ambulacral plates compound or simple and with the ambulacral areas not petaloid dorsally ; Clypeastrina (Oreta- ceous to Recent) with more or less flattened tests, ambulacral plates simple and the areas petaloid dorsally, while the lantern is highly modified ; Spatangina (Jurassic to Recent) with the ambulacral plates simple and the areas commonly petaloid dor- sally but with no lantern nor perignathic girdle in the adults. Order Plesiocidaroida, an aberrant and imperfectly known stock restricted to the Triassic (Tiarechinus), in which the peri- proct is central but the genitals are large and occupy most of the dorsal surface. There are 2 columns of ambulacrals and 3 of interambulacrals. Plates not imbricate. Base of corona not resorbed. “It is not closely affiliated with any other group” (220). Order Perischoechinoida, arose in the Bothriocidaroida at least as early as the Silurian and persisted into the Permian. Corona and periproct regular in form and position, with from 2 to 20 columns of plates in each ambulacral area and from 3 to 14 in each interambulacral area. No perignathic girdle, the lantern muscles attaching directly to the base of the interam- bulaeral plates. Embraces the families Archaeocidaride (EKocidaris, Archaeocidaris, Lepidocidaris), with 2 columns of ambulacrals and 4 to 8 of interambulacrals, plates thin and imbricating, base of corona resorbed, and pri- mary spines large ; Devonian to Permian. Lepidocentride (Koninckocidaris, Lepidocentrus, Hyatte- chinus, Pholidechinus), with 2 columns of ambulacrals and 5 to 14 of interambulacrals, plates thin and imbricating, base of corona not resorbed, and all of the spines small; Silurian to Mississippian. Palacechinide or Melonitide (Palaeechinus, Maccoya, Loven- echinus, Oligoporus, Melonechinus), with 2 to 12 columns of ambulacrals and 3 to 11 of interambulacrals, plates not imbri- cate, some resorption of base of corona, only small secondary spines ; Silurian to Mississippian. In this family the genealogical relations of the genera are Schuchert—Jackson on the Phylogeny of the Echini. 268 clearly evinced by the structure and development of the ambulacra, an entirely new method of getting ontogenetic stages of growth (231). Lepidesthide (Lepidechinus, Perischodomus, Perischocidaris, Proterocidaris, Lepidesthes, Pholidocidaris, Meekechinus), “ one of the most specialized of all groups of Kchini,” with 2 to 20 columns of ambulacrals and 3 to 13 of interambulacrals, plates imbricating and no resorption of base of corona, primary*spines small; Devonian to Permian. Order Echinocystoida (new), arose in the same stock that gave rise to Perischoechinoida but is an offshoot from a com- mon early stock. Irregular in form with the periproct appar- ently eccentric in an interambulacrum. From 2 to + columns of ambulacrals and 8 to 9 of interambulacrals. Plates thin and imbricating, with the spinessmall. Lantern typically echi- noid, but no perignathic girdle. Families Palaeodiscide ( Palae- odiscus) and Boney stidee (Echinocystites). Art. XXIV.—The Belt and Pelona Series; by Oscar H. HersHey. Tue sedimentary rocks of the Coeur d’ Alene Mountains in. Northern Idaho were first systematically studied in 1903 and 1904 by Mr. F. C. Calkins, assisted by Messrs. W. A. Williams and D. F. MacDonald.* Although consisting chiefly of variable proportions of only two minerals, quartz and sericite, it was found Ber. to subdivide them into six formations as in the following table Generalized Tabular Section in Coeur d’ Alene District. Approximate Name Description thickness in feet Striped Sandstones, siliceous, generally flaggy to shaly; Peak color mostly green and purple: characterized by shallow-water features, as ripple marks, sun cracks, etc. Top removed by erosion. 1,000 + Wallace Thin-bedded, bluish and greenish, more or less calcareous shales, underlain by rapidly alter- nating thin beds of argillite, calcareous sandstone, impure limestone ; these under- lain in turn by gray-green siliceous argillites. Ripple marks, sun cracks, etc., throughout. 4,000 *The Geology and Ore Deposits of the Coeur d’Alene District, Idaho; Prof. Paper U. 8S. Geol. Survey No. 62, 1908. 264 O. H. Hershey—Belt and Pelona Series. Approximate Name Description : thickness in feet St. Regis Indurated shales and more or less flaggy sand- stones ; colors mostly green and purple; characterized by shallow-water features. 1,000 Revett White quartzites, generally rather thick- bedded; interstratified with subordinate quantities of micaceous sandstone. 1,200 Burke —_ Light-gray, flaegy, fine-grained sandstones and shales, mostly greenish, with a variable amount of purple quartzitic sandstone and white quartzite. Shallow-water features throughout. 2,000 Prichard Mostly argillite, blue-black to blue-gray, gen- ‘ erally showing distinct and regular banding. Considerable interbedded gray indurated sandstone, upper portion characterized by numerous alternations of argillaceous and arenaceous layers, and by shallow water. features. Base not exposed. 8,000 + 17,200 My study of these rocks has been most extensive in the Wardner District, where I have mapped an area about 44 miles long and 3 miles wide. Here the Burke and Revett formations are thicker and more complex than elsewhere in the Coeur d’Alene region and I have found it practicable, in order to bring out the structure better, to make local subdivi- sions of the formations. I have also added another formation at the bottom of the series. This, the Cataldo formation, consists of at least 1,000 feet thickness of heavy beds, many of which are finely laminated and cross-bedded, of medium- grained quartzite of light lilac color, alternating with thinner- bedded strata of dull greenish sericitic quartzite. It is apparently the basal member of the Belt series and probably corresponds to the Creston quartzite of Daly. It does not appear at the surface nor in the mines of the Wardner District, but must underlie them at great depth unless cut out by intru- sive rock. That Calkins did not intend to include it in the Prichard is proved by his mapping a small area of it as Burke. The Prichard formation in the Wardner District is largely a dark gray to black argillaceous material in thin regular beds. Where a slaty structure is well developed, the rock resembles the ordinary black slates of other regions. Another type is a hard greenish-gray siliceous shale. Scattered through the formation, but most abundant in the upper part, are beds of O. H. Hershey—Belt and Pelona Series. 265 white quartzite, generally of coarser grain than quartzites in higher formations. I have no means of determining the thick- ness of the Prichard in this district, but I am not disposed to criticise Calkins’ estimate of 8,000 feet for the Coeur d’ Alene District as a whole. At the top of the Prichard formation there is a rock made up of dark bluish-gray argillite and whitish arenaceous mate- rial in very thin alternating lamine. This passes by transition strata into the Lower Burke division, the best section of which is on the southern side of Kellogg Peak. The predominating type isa greenish, thin-bedded sericitic quartzite, the darker green bands of which appear quite strongly chloritic. Layers of hard, white, heavy-bedded, poorly sericitic quartzite, 2 to 6 feet thick, alternate with greenish sericitic quartzite and are quite numerous, though they constitute a comparatively small part of the Lower Burke member. In this section there is not much purplish material, but elsewhere in the district, particu- larly in the Deadwood Gulch region, there is much purplish gray, thin-bedded quartzite and shale which locally may attain a thickness of several hundred feet, though generally inter- bedded with greenish layers. My impression is that they characterize the upper portion of the Lower Burke. Litho- logicaily they are indistinguishable from the purplish gray quartzites and shales of the St. Regis formation except that, I think, they are harder and less shaly than the latter. These purplish gray horizons have given me more trouble in the’ mapping of the Wardner District than has any other type of rock. The thickness of the Lower Burke member is about 1500 feet. The Upper Burke member consists of an alternation of thin-bedded, greenish, sericitic quartzites and heavy-bedded, white, nearly pure quartzites. The first type of rock is gener- ally softer, thinner bedded, and of a lighter green than the typical greenish gray Lower Burke quartzite. The heavy- bedded white quartzites microscopically resemble the white quartzites of the Revett formation, but they break down rap- idly under weathering agencies and give rise to much fine yellowish quartzite debris. Underground, they are in places hard and flinty and in others have lost the cementing material and become an incoherent pure white sand. I draw the base of the Upper Burke at the bottom of the first heavy bed of such quartzite; this is usually 30 to 50 feet thick, though in one section it is 300. feet thick. The thickness of the Upper Burke member is about 2,000 feet, making 3,500 feet for the Burke formation as against 2,000 feet for the formation in the eastern part of the Cceur d’Alene District as estimated by Calkins. My impression is that the thinning eastward is at 266 O. H. Hershey—Belt and Pelona Series. the loss chiefly of the Upper Burke, which I could recognize in the vicinity of Burke, but not near Lower Glidden and tevett lakes. The Revett formation at the type section in the cirque walls overlooking Revett Lake consists of heavy-bedded, hard white quartzites alternating with much thinner beds of oveenish seri- citie quartzite, of a thickness, west of Lower Glidden Lake and on the west side of Twenty-two Mile Creek, of 1,000.to 1,200 feet. In the Wardner District the series is thicker and the greenish sericitic quartzites form a more important constituent. The Lower Revett is harder and more resistant to weathering than either member of the Burke. Its areas are characterized by a very stony soil, in many places taluses bare of vegetation, in which material there is an angularity not present in the debris from any other formation. In a good section in Big Oreek Valley the thickness is 1,300 feet. The Middle Revett divi- sion is simply the highest prominent band of white, nearly pure quartzite; it is heavy-bedded and cross-bedded, weathers either white or yellowish, and has a habit of outcropping as a persist- ent precipitous ledge. It is generally 75 to 100 feet thick, but thins perceptibly in short distances, and may be practically absent from portions of the district. The characteristic rock of the Upper Revett is a rather thin-bedded, greenish, sericitic quartzite, similar to the thinner-bedded portion of the Lower Revett, though near the top it has several white quartzite beds and locally a few thin purplish gray bands; in fact, it resem- bles the Lower Burke. The thickness is about 900 feet. This gives the Revett formation a thickness of about 2,300 feet in the Wardner District. Much of the St. Regis mapped by me in the district is the basal member, a rather hard, thin-bedded quartzite of gray and lavender colors. There is an alternation of green and purple- gray shales and thin-bedded quartzites. No complete section is exposed in the district, but I have adopted Calkins’ estimate of 1,600 feet for the Coeur d’Alene District in general. The Wallace formation is confined chiefly to a large block between two great faults, the Osburn and Alhambra, and probably includes nearly all of the 4,000 feet thickness estimated by Calkins for the formation. The lithologie features are similar to those near Wallace and eastward, except that the quartzite beds are thicker and purer and there is less limestone. Although much of the formation is somewhat calcareous, the only lime- stone observed is a thin bed of blue-gray, nearly pure limestone in the Elk Creek basin. The blue-gray argillites weather to an orange color and the quartzites to a light red color. The Striped Peak formation has long since been eroded from all parts of the district. O. H. Hershey—Belt and Pelona Series. 267 Thus we may have a total thickness (including 1,000 feet of Cataldo quartzite) in the Belt group of strata remaining in the Wardner District of 19,800 feet. A geologist, whose past expe- riences with other alleged thick formations entitle him to take a critical view of the subject, has expressed the conviction that Calkins? estimates of the thickness of the Coeur d’ Alene rocks, which were evidently derived by the measuring of sections at right angles to the bedding planes, are far too high. My esti- mates are based on a knowledge of the value that must be given to individual formations in making up the mass of certain fault blocks and are not dependent entirely on the original attitude of the bedding planes. I am not certain that the Prichard formation attains the 8,000 feet thickness assigned to it, but I am confident that the estimates of thickness of the other forma- tions will in a general way stand the test of the most intensive study. lec aatiote in the Belt area higher members of the series occur, and in the Belt Mountains and Mission Range of Mon- tana they are nonconformably overlaid by Cambrian sediments. In the “ Geologic Map of North America,” * besides the main area lying chiefly in Idaho, Montana, and. British Columbia, a small area of Belt rocks is indicated in the Wasatch Range north of Ogden, Utah. The central portion of the Uinta Range is also mapped as Belt. The Wasatch area is presum- ably that which has been discussed recently by Mr. E. Black- welder.t The prevailing rocks are variously colored quartz- ites or quartzitic saridstones. There are many thin beds of conglomerate and at several horizons hard shales and slates, chiefly dark purplish brown and bright green, though some are distinctly black and others rich maroon. A few thin beds of brown dolomite occur. Cross-bedding, ripple marks and sun- cracks are prevalent. These strata are referred to the Algon- kian and are presumed to be overlaid nonconformably by Lower Cambrian quartzite 1,000 to 1,500 feet thick, and this by Olenellus-bearing shales and limestones. The base of the Cambrian quartzite | is placed at a well-marked conglomerate. During the first six months of 1905, I made an extensive geological reconnaissance of a portion of Eureka, White Pine and Elko counties in Eastern Nevada, particularly of the ranges on both sides of Steptoe Valley. . I find in my notebook under date of May 16, 1905, the following generalized section : * Prof. Paper U. S. Geol. Survey No. 71, plate 1, 1911. + New Light on the Geology of the Wasatch Mountains, Utah ; Bull. Geol. Soc. Am., vol, xxi, 1910. 268 O. H. Hershey—Belt and Pelona Series. Paleozoic Section in Eastern Nevada. Estimated thickness 1. Blue, heavy-bedded, very fossiliferous, black-cherty limestone (Ruth) ------. 1,000 feet . Drab and light brown, massive cherty limestone with brown, coarse crinoidal limestone locally prominent in lower portion: (Arcturus) esa eee 700-to 1,000 * 3. Yellow and red, very fine-grained, soft sandstone alternating with light gray limestone layers ; Upper Carboniferous : (Areturus); 2. eee ee Saami 200 to 1,000 * 4. Light gray, drab and dark gray limestone, not very cherty, but with locally promi- nent red and yellow sandstone bed 10 Lo tor200feet thick 22 22554 eae 500 to 2,500 “ 5. Brown and olive quar tzite, (Diamond Peal) sree ter aal toe hin i ee eee 0to 100 « 6. Black, soft shale (White Pine) --_---.--- 50) tho) BOD. 7. Light gray, massive limestone.---__--- SOMO AKO) S. Black shale and red and white quartzite - 100 to 200 *“ 9. Dark brown, dark lead-gray and black limestone ; upper part Upper Devonian (Bisy) 2 rasa ae an 1, i o00te 2.0008 LD White; hard, fine-gr ained quartzite (Eureka) SOULE ere try ee ee Me ey! HO Ome 11. Blue, thin-bedded limestone; Lower Ordovician....._-- 200 “ 12. Dark brown, very fossiliferous and black, soft shales ; ; Lower Ordovician -_---. 200 to 500 * ‘13. Blue limestone ; Lower Ordovician ----- OOpecs 14. Light gray, imperfectly fossiliferous, thin- bedded, in part brecciated or conglom- eratic limestone ; Lower Ordoyvician_. 2,000 “ 15. Massive, blue-gray limestone with pecu- liar wavy bands in shades of blue---- 500 “ 16. Blue, thin-bedded cherty limestone - --- - TOO" 17. Black, fossiliferous shale and blue shaly limestone ; Lower Cambrian -_-_-- ---- 300 “ 18. Blue and gray, heavy-bedded limestone - 1,000 “ 19 Darkblue imestone= 22225 eee 50'to 100) 20. Brown quartzite, green mica slate and blue limestone interbanded .--. -- be 300 “ 21. Pink and white quartzite._.- -.2...5-22 2,000 <“ 22. Black slate weathering purple----.----- 150 to 1,500 “ 23. White conglomerate (bearing jasper pebbles) and quartzite... --= 9-35 e2 100 to 200 “ DA) Green eS late rare an eA ete = eee eee 500 to 1,000 “ 25. White and pink quartzite.........-_-- 500 to 2,000 “ O. H. Hershey—Belt and Pelona Series. 269 Estimated thickness 26. Green and black slate._-..--.--.-- aes 400 feet 97. brown and pink quartzite. .2.2.+..---. 300 “ 28. Dark gray slate weathering dark brown- 500 “ 29. White, hard, fine conglomerate - -_- ---- - BOs * 30. Reddish brown, weathering dark gray, micaceous schistose slate ..-.---..----- PLO Sl black sehistoseuslatere 8 sane 500 “« (Underlaid by intrusive granite) Motaleee sue ce ne 15,550 to 23,850 feet Small collections of fossils were sent to the United States Saale Survey, but I am so far separated from the reports on them, that I will merely make the statement that Olenellus was found in the limestones at a considerable height above the base, particularly in No. 17, if I remember rightly. At any rate, there is over 1,000 feet of undoubted Lower Cam- brian limestones and shales. These pass by the transition formation, No. 20, into a great series of quartzites and slates, which appeared to me to be conformable throughout, though several conglomerate bands were observed. The lithologie characters are similar from the base of the limestones down, strongly suggesting a single series of sediments. There is no sharp break in the apparent degree of alteration, but in a gen- eral way a progressive increase in the metamorphism down- ward. Thus, Nos. 30 and 31 mark the approach to the stage of metamorphism reached by the crystalline schists. I was of the impression that the entire series represents continuous sedimentation and I classed all the strata under the Lower Cambrian limestone as also Lower Cambrian, though I recog- nized that this would give the Lower Cambrian period dispro- portionate length. Since becoming acquainted with the Belt rocks of Idaho, I have been impressed by a marked resem- blance in a general way between them and this Nevada quartz- ite-slate series, not only in character of sediments but in degree of alteration. In both regions, quartzites predominate above and black slaty rocks below. The alteration (by regional metamorphic action) is slightly greater in the Nevada section than in that of Idaho. This pre-Olenellus quartzite-slate series in Nevada may, doubtless, be correlated in a general way with the similar series in the Wasatch Mountains. Walcott regarded the latter as conformable and arbitrarily drew a line beneath which he considered the rocks as Algonkian. Blackwelder, however, has found evidence of a nonconformity that may be taken as a convenient plane separating the Lower Cambrian and Belt series. The same nonconformity may be marked by the white 270 O. IH. Hershey—Belt and Pelona Series. conglomerate, No. 23, of my Nevada section. This would place about 6,250 feet of sediments in the Belt series. The lower members occur in a small area on the eastern slope of the Egan Range about 10 miles north-northeast of Ely. On the eastern slope of the same range about 4 miles north-northwest of Warm Spring, the section extends down to No. 26. In a belt extending from 5 miles north-northeast of Cherry Creek to 9 miles south of that town, the section extends down to No. 30. At Aurum, on the eastern slope of the Schell Creek Range, No. 24 is exposed, but 15 miles farther south the sec- tion extends down to No. 30. Nowhere is the base of the Belt sediments exposed. All the granite seen in Eastern Nevada is intrusive and probably post-Carboniferous in age, except in the range west of Clover Valley, where, under a limestone series, there were observed the following rocks : 1. Light gray caleareous mica schist intruded by white mus- covite pegmatitic granite. 2. Gray mica schist. 3. White, coarse-grained micaceous quartzite 4, White micaceous quartz schist. 5. Dark gray, coarse biotite gneiss and schist. 6. Light grey granitic gneiss, varying to a slightly sheared, fine-grained granite. 7. Gray, medium-grained sheared granite. I have no doubt that these rocks are Archean in age. The Belt series and a large part of the Paleozoic section are absent. Blackwelder, in the paper already cited, argues that the Algonkian sediments in the Wasatch region were depos- ited “chiefly by rivers on extensive plains in a climate which was semi-arid or at least subject to dry seasons.” He points out that “the regularity of the dip and the small amplitude of the cross-bedded structures implies currents moving in a single general direction, and currents of water rather than of wind.” The ripple marks and tension cracks admittedly imply shallow- water conditions, with frequent exposure of the mud flats. He considers that “the imperfect assortment of the materials appears to indicate that the sediments were deposited rapidly, little time being given for that complete sifting of fine from coarse debris which is characteristic of the work of waves upon an open beach.” He recognizes that the Algonkian (Belt) sed- iments of Idaho were probably deposited under conditions similar to those which obtained in the Wasatch region. Since reading his paper, I have been endeavoring to consider the rocks of the Coeur d’Alene District in the light of a possible origin from meandering rivers upon a great delta plain,* but * Relative Geological Importance of Continental, Littoral and Marine Sedi- mentation, part 3; Jour. Geol., vol. xiv, pp. 524-566, 1906. O. H. Hershey—Belt and Pelona Series. 271 they do not impress me as typical river sediments. Rivers meandering on a plain leave channels which become filled with sediments and preserved. Nothing even remotely suggesting such a fossil channel has ever come to my attention in the Coeur d’Alene District. Furthermore, in my description of the Belt formations in the Wardner District I have taken pains to mention the existence in some of the formations of frequent bands of practically pure, white quartzite. These were origi- nally beds of nearly pure fine-grained sand comparable with the St. Peter sandstone in the Mississippi basin. They prove a nearly perfect assortment of materials at frequent intervals ranging from late Prichard time to the close of the Revett period. They are not typical river sediments. It is difficult to understand how a river might deposit over a considerable area a bed of nearly pure quartz sand 100 feet thick. For these reasons I consider the fluviatile origin of the Belt terranes not yet satisfactorily proved. Indeed, I am inclined to support the hypothesis that the sediments were deposited in a vast inland sea or lake into which large rivers carried great quan- tities of fine sediment, some of which was precipitated without perfect sorting, while some was brought under the action of currents that formed the white sand beds. The Oro Grande series of marble, quartzites and slates in the Mohave Desert in Southeastern California* was correlated on lithologic grounds with the Lower Cambrian strata of Inyo County, California.t At that time I had not seen the latter, but I have since visited the White Mountain Range and con- tinue to hold the impression that the Oro Grande series is the correlative of part of the Lower Cambrian of Inyo County. I want to call attention to the fact that the Oro Grande sediments have been altered by regional metamorphism to the same degree as the Lower Cambrian and Belt quartzites and slates of Eastern Nevada, excluding the lower two members of my section. They have a somewhat more altered appearance than _the Belt sediments of the Coeur d’Alene District of Idaho; for instance, the slates are, macroscopically, more evidently micaceous. At the Yellow Butte, about 12 miles by the California North- eastern railroad from Weed, and hence north-northwest of Mt. Shasta in Northern California, there is an “island” of older rocks rising through the Tertiary lavas. It consists mainly of a fine-grained hard white and gray quartzite in which are thin micaceous schists, the whole being unlike anything in the Kla- * Notes on the Later Cenozoic History of the Mohave Desert Region in BRN oe California; Uniy. Calif. Publ., Bull. Dept. Geol., vol. vi, pp. + Some Crystalline Rocks of Southern California ; Am. Geol., vol. xxix, pp. 286-287, 1902. 272 O. H. Hershey—Belt and Pelona Series. math region on the west, but resembling the Lower Cambrian and Belt quartzite-slate series of Nevada, and the quartzites and slates of the Oro Grande series in Southeastern California. With it is a granite, probably intruded in the quartzite. In the Klamath region the Devonian rocks, including much igueous material, rest nonconformably on a series of crystalline schists that have been described by me under the names of the Abrams mica schist and Salmon hornblende schist.* In the summer of 1907 I greatly extended my acquaintance with this series. I found the hornblende schist member (which also con- tains fine-grained hornblende and micaceous gneisses) exposed in a narrow belt along Chinech Creek southeast of Orleans, and at Elk Creek near Happy Camp in Siskiyou County; in the latter area it attains a maximum width of several miles, with a prevailing eastern dip and an exposed maximum thickness of at least 5,000 feet. Near Seiad post-office there is an acid, white, pegmatitic granite, apparently very old (probably Arch- ean), abounding in large inclusions of hornblende schist and gneiss from the Salmon (Chinech) formation, but chiefly a fine- grained, thin-bedded, hard, vitreous quartzite of pink, white, aud light green colors, with some mica schist and traces of marble. The quartzites seemed to represent a formation not heretofore discriminated in the Klamath region. It may extend south to Marble Mountain. Farther up the Klamath River is another belt of Archean schists. Under the Salmon hornblende schist, much of it rather coarse-textured, there is about 2,000 feet thickness of dark gray mica schists of the Abrams formation. Under this is 1,000 feet of light greenish (chloritic and, in places, actino- litic) coarse-textured mica schist, part of which appears to have originally been a granite, and somewhat resembles the lower member of the Pelona schists of Los Angeles County, California.t It is my impression that the various phases of schists do not maintain a regular sequence throughout the area of development of the Abrams formation. The chloritie and actinolitic schists probably represent igneous rock intruded at various horizons in the original sediments. By following the old Kelsey trail to Marble Mountain, I crossed the mica schist, then probably 5,000 feet of hornblende schist and hornblende gneiss, then intrusive granite and basic rocks, then fine-textured mica schists and micaceous and hornblendic gneisses accom- panied by the peculiar, thin-bedded, white quartzite and white marble of the Seiad area. Higher there is a great series of * Metamorphic Formations of Northwestern California ; Am. Geol., vol. XXvii, pp. 225-245, 1901. acre Crystalline Rocks of Southern California; Am. Geol., vol. xxix, 1962. O. H. Hershey—Belt and Pelona Series. 273 dark-colored, in part black, mica schists, mica slates and fine- textured eneisses accompanied by marbles that extend to and constitute the prominent peak, Marble Mountain. Finally, the series is capped by 100 feet of the Salmon hornblende schist. At Scott tar there is an area, probably a mile long and a quarter of a mile in maximum width, of light gray, coarse- textured eneiss that was originally an Archean granite, prob- ably intruded into the Abrams sediments before the development of schistusity. Thence I traced the Abrams schists ina great belt through the Siskiyou Mountains into Oregon. In discussing the Pelona schists, which I had mapped in an area about 20 miles long in Los Angeles County,* I suggested that they are of the same age as the Abrams mica schists of the Klamath region. Last year I became acquainted with another area of the Pelona series. It constitutes the greater portion of the Rand Mountains near Randsburg in Kern County, California. It has recently been described by Mr. F. L. Hesst as largely a gray mica-albite schist. In the Atolia scheelite district it is bounded on the southeast by a complex of Archean granitic and dioritic rocks. Hess says the main granite mass seems to be under and may be older than the schists of the Rand Mountains. Iam inclined to agree with him, though it involves an overturn of the strata along the border. A narrow band of marble and limestone follows this border, except where the contact is faulted. In the gray schist area I noted bands of quartz schist, hornblende schist, chloritic schist, and actinolitic schist, characteristic features of the Pelona series. As the two areas are only about 60 miles apart, I correlate them with considerable confidence. I propose, further, to extend the name Pelona series over the Abrams and Salmon formations of the Klamath region. I believe this Pelona series has a definite time position in the geology of the Pacific Coast country, comparable with that occupied by the Belt series of sediments. It is the youngest important Archean series. Further, so far as my observation goes, it is the last sedimentary series preceding the Belt series. Kellogg, Idaho. * Univ. Calif., Bull. Dept. Geol., vol. iii, No. 1, pl. 1, 1902. + Gold Mining in the Randsburg Quadvangle, California; Bull. 430, U.S. Geol. Survey. 274 Hutchins—Absorption and Thickness of thin Films. Arr. XX V.—Absorption and Thickness of thin Films ; by C. O. Hutcums. Tur here recorded observations were made in the course of an attempt to find a window for radiometric apparatus that would effectually close such apparatus against all convective air currents, and at the same time introduce less absorption than rock salt or fluorite. It was argued that since the absorp- tion of a body is an exponential function of its thickness, the body being made very thin, the absorption would nearly van- ish even if the coefliicient of absorption were high. As is shown below, it is an easy matter to prepare permanent films of comparatively large area having a thickness of One- tenth to one-twentieth of a wave-length of light, and an extremely small absorption even for very long waves. Preparation of Films. Make a collodion by dissolving gun-cotton in amyl acetate. Instead of gun-cotton I have used shavings from the transpar- ent celluloid handle of a discarded toothbrush, and found the result almost as satisfactory. The collodion should be rather thick at first, then gradually diluted until the proper consist- ency is found by making trial films; then filter through cotton until it is perfectly pellucid. For practice one may bend up a ring, say 5°" in diameter, at- the end of a stout iron wire, soldering it so that it may be smooth and flat ; later, any desired support may be used. Fill a large evaporating dish with clean water. Dip the rounded end of a glass rod in the collodion and touch the adher- ent drop to the surface of the water in the middle of the dish. The drop instantly spreads to a large disc on the surface of the water. The film quickly sets, by the evaporation of the solvent, so that at the end of a minute or soit may be removed to the support. Slip the wire ring under the margin of the floating film ; bring the ring into contact with the film at the margin, then very slowly and steadily raise the ring to a vertical posi- tion, at the same time withdrawing it from the water. The portion of the film not on the support gathers to a tail at the lower edge. The film is then hung up to dry, and as it does so it will be seen to show the Newton black all over its whole surface, being too thin to reflect much light. A little practice enables one to spread a film upon any desired support. When dry they are very strong and do not break from shock. Hutchins—Absorption and Thickness of thin Films. 275 The films are transparent to long waves. The transmission of the films for very long waves was tested as tollows: A radiomicrometer was set up. Before the open- ing of this was placed a thin blackened copper tank containing water at room temperature. This tank could be raised and lowered in a frame by pulling and releasing a string. Behind the movable tank was a large cubical tank containing ice. The radiation was therefore from a body at about 22° to one at 0°, consequently the radiation maximum lies far out in the infra- red at wave-length 10 w. Exposures were made by raising the movable tank, as noted above, and the ratio of the mean radio- micrometer deflection with the film interposed to the mean without the film gives the transmission. The following is a typical observation : Mempaot movable) tanks2- 220 2502222 DYE, 6c of stationary tank...-.._..-.-- 0°-0 Mean deflection with film ._..__.___-- 138°9+0°23 Ke aS without film ..__._.-- 139°6 40-21 Transmission...._------- 0°995 Several films have been tested in like manner, all showing an absorption of less than one per cent; whereas a plate of rock salt 0-4 thick, newly and perfectly polished by Brashear’s method, gave an absorption of 22 per cent under the same con- ditions. It would appear, therefore, that a nearly perfect win- dow for radiometric instruments has been found, provided only that the instrument is not to be exhausted. A thin coating of collodion may be of use in protecting the surfaces of salt prisms and lenses trom moisture. I polished two plates of salt and coated one by flowing it with very dilute collodion. The two have been exposed side by side for two weeks. The coated plate is still perfect, whereas the uncoated clouded in a day or two. It is too early, however, to know if the protection is permanent. A thin collodion film does furnish a permanent protection of silvered mirrors against tarnish. A silvered mirror in this laboratory turns yellow in a week owing to sulphur gases from a nearby railroad, but when flowed with a very dilute collodion it remains untarnished. I have such a mirror that has been exposed for three years that is as perfect as when new. The films are so thin that they in no wise impair the optical qualities of the surfaces to which they are applied. Thickness of the films. It is of interest to know the thickness of these black films. One film where it is doubled back upon itself over the wire ring used as a support, thus giving a double thickness, showed 276 LHutchins—Absorption and Thickness of thin Films. first order interference color of about wave-length 000057, eae ; 4 ; : Taking the index of refraction of the material as 1:5, we have for the film thickness : 000057 : = ‘0000059°™ 4X2xX1°5 The thickness was also found thus: A loop was made at the end of a fine platinum wire. Three films whose diameters were 12, 14, and 11:5°™ respectively were gathered from the water on which they were formed into a small ball on the wire loop, the water pressed out, and the remaining material dried to constant weight. The wire alone weighed .-.........-.- 0-02188" Wire with film ee ae SOME a ey 0:0231 Weightiofilimuy sy «as Seo eee eee ae 0°0013 ‘Areaof film t2 Salts AES BU ane 37Q¢m2 The density of a sample of the toothbrush handle from which the film was made was 1136. From the above we get for the mean thickness of the film: 070013" a a a 0000002652 372 X 1:136 equal to one-twentieth the wave-length of green light. Bowdoin College, July 1st. R. L. Moodie—Mazon Oreck Shales. Oy. Arr. XX VI.—The Mazon Creek, Lllinois, Shales and their Amphibian Fauna; by Roy L. Moovtr. Tuer Mazon Oreek shales have been noted for many years for the excellence of the fossils which are found in them. These fossils are, for the most part, plant remains, but occasionally animal fossils are found, usually in nearly perfect preservation. Among the vertebrates which have been recorded from these beds are about twenty-five species of elasmobranch, dipnoan, crossopterygian and acanthopterygian fishes. Nine of these species are founded on scales and will doubtless be subjected to revision later. Other vertebrates are represented in the amphibian fossils, of which there are, at present, ten species known. These ten species are distributed among eight genera, five families and four orders; thus showing the amphibian fauna of the Mazon Creek shales to be a diverse one. This diversity undoubtedly indicates a long antecedent history for the group. The genera and species, even nearly all of the families, are exclusively Mazon Creek forms, not being known from elsewhere in the world. From this we know that in the Pennsylvanian the Amphibia were already a sufficiently old group to have established themselves into distinct, widely sepa- rate geographic groups. Just how long it may have taken for them to so establish themselves is, of course, a matter for con- jecture. The larger amount of the material representing the amphib- ian fauna of the Mazon shales is in the Yale University Mu- seum, where it had been gathered by Professor Marsh. The writer is indebted to the authorities of the Museum for the privilege of studying this interesting collection. Seven of the species of the Mazon Creek Amphibia are known only in this collection at Yale. Seven of the species are founded on con- siderable portions of the skeleton, and some of them show many soft parts as previously described by the writer.* Two species are known from nearly complete remains of three indi- viduals. The ten species are as follows: 1. Amphibamus grundiceps, Cope, 1865. Three nearly com- plete specimens. 2. Amphibamus thoracatus, Moodie, 1911. One incomplete Specimen. 3. Micrenpeton caudatum, Moodie, 1909. One complete indi- vidual. ‘ 4, Eumicrerpeton parvum, Moodie, 1910. Three nearly perfect orms. * American Naturalist, xliv, p. 367, 1910. Am. JOUR. laren SERIES, VOL. XXXIV, No. 201.—SrrremBer, 1912. 278 PR. L. Moodie—Mazon Creek Shales. 5. Mazonerpeton longicaudatum, Moodie, 1912. One nearly perfect skeleton. 6. Mazonerpeton costatum, Moodie, 1912. One imperfect skele- ton. 7. Cephalerpeton ventriarmatum, Moodie, 1912. Skull and anterior part of body. 8. Erpetobrachium mazonensis, Moodie, 1912. Humerus, radius and ulna. 9. Spondylerpeton spinatwm, Moodie, 1912. 12 dorsal (caudal ?) vertebra. 10. Erierpeton branchialis, Moodie, 1912. Impression of man- dibles, parts of body and hyobranchial bars. These species have been arranged zoologically according to the following plan: Class—Amphibia, Linné, 1758. Subclass—Euamphibia, Moodie, 1909. Order—Branchiosauria, Lydekker, 1889. Family—Branchiosauridae, Fritsch, 1879. Micrerpeton caudatum, Moodie. EHumicrerpeton parvum, Moodie. Mazonerpeton longicaudatum, Moodie. Mazonerpeton costatum, Moodie. Order—(?) Caudata, Dumeril, 1806. Family—Cocytinidae, Cope. Erierpeton branchiulis, Moodie. Subclass—Lepospondylia, Zittel, 1887. Order—Microsauria, Dawson, 1863. Family—Amphibamidae, Cope, 1875. Amphibamus grandiceps, Cope. Amphibamus thoracatus, Moodie. ' Cephalerpeton ventriarmatum, Moodie. Family—Molgophidae, Cope, 1875. Erpetobrachium mazonensis, Moodie. Subclass—Stegocephala, Cope, 1868. Order—Temnospondylia, Zittel, 1887. Suborder—Embolomeri, Cope, 1885. Family—Cricotidae, Cope, 1884. Spondylerpeton spinatum, Moodie. - It will be seen from the above arrangement that nearly all of the orders of extinct Amphibia are represented in the Mazon Creek fauna. Since these animals are the oldest known land vertebrates of which there are skeletons, and since the beds in which they occur have never received adequate RR. L. Moodie—Mazon Creek Shales. 279 description, it was thought that it might be of interest to pre- sent a short discussion of these remarkable fossil-bearing beds. The writer was enabled to spend a week studying the fossil beds of Mazon Creek last summer with the aid of funds from the Department of Zoology of the University of Kansas and the Elizabeth Thompson Science Fund. It is fitting to express, in this place, to Mr. J. O. Carr of Morris, Illinois, my appre- ciation of the favors shown me while collecting at Mazon Creek. My purpose in visiting the locality was primarily to collect Amphibia, but although several thousand nodules were examined, none contained amphibian remains. Mr. Carr has collected on Mazon Creek for more than thirty years and knows more of the conditions of fossilization and location of the var- ious beds than any one else. It is, however, a significant fact that during all these years of assiduous collecting he has never found an amphibian nor a fragment of one. It was this fact, together with the further one that the Amphibia referred to above represent over sixty years collecting from these beds, which interested me in making the following comparative table of the rarity of the various kinds of organic remains found in these beds. : ' If we take 100,000 nodules as a basis for computation of the rarity of the various forms, something like the following will be the approximate result: . Of 100,000 nodules, 20,000 will be barren or contain only indeterminate fragments. 68,500 will contain plant remains. 7,500 will contain insects, Crustacea, myriapods, scor- ions, spiders and other Arthropoda. 3,900 will contain fish coprolites or scales. 95 may contain fish or fragments of fish. 4 may contain molluscs. 1 may contain an amphibian. Total 100,000 nodules. Perhaps even 100,000 is a little low as a basis of estimate. Mr. Carr was of the opinion that one nodule in every 500,000 might contain an amphibian. The table given above will, at least, be serviceable in giving an insight into the relative abundance of the various forms. The beds where the nodules are usually collected occur along both banks and in the bottom of the creek (figs. 1 to 4) in two localities. One locality, known as the Bartlett place, where I camped (fig. 1), is situated eight miles southeast of Morris, in Grundy county, Illinois, Wauponsee township; N. W. quarter section Fre. 1. The ‘‘Upper Beds.” Fic. 1. Nodules weathering out of the soft blue clay shale. Two quite long ones may be seen near the lower left hand corner of the figure, one of ” them still in place. r Fie. 2. The ‘‘Upper Beds.” Fic. 2. ‘‘The stream bed is filled with glacial bowlders, which form an excellent base on which to crack the nodules.” Mr. J. C. Carr looking for nodules in the creek bed. He usually carries a basket and fills this, then sits down by a good bowlder and cracks the nodules collected. R. L. Moodie—Mazon Creek Shales. 281 30; Twp. 33; Range 8; the land now being owned by Mrs. Emma Akerly of Wilmington, Tllinois. The fossil-bearing nodules oceur throughout six to eight feet of shale, just above the coal, along both banks of the creek at the “ upper beds,” as the Bartlett, place is called. They may be also seen in the creek bed, when the water is low, still embedded in the shale. With a common potato fork the shale is easily turned and the nodules come out for all the world like potatoes. Once in a while “ pockets” are struck from which one may secure a peck or more of nodules. Nearly every nodule has a fossil at the “‘ upper beds ” but all of the fossils are not well preserved, possibly only one or two out of every eight or ten being worth carrying to the museum. The stream bed at the Bartlett place is filled with glacial bowlders (fig. 2) which form an excellent base on which to crack the nodules. The nodules crack best when wet and it requires some skill and practice to crack them evenly. The nodules vary in shape and form from perfectly round ones, one-half an inch in diameter, to oval, elongate ones 17 inches or more in length. Many are quite irregular and it is soon noticed that the irregu- lar nodules seldom have good fossils, often none at all. The nodules seem quite light, and in one place, where the stream curves, they are piled in a long windrow. On this pile were found several good specimens of Crustacea and many good plants, in nodules cracked open by the frost. The fossils at the “upper beds” are localized into special strata. At one place in the upper part of the deposit in a red- dish shale one finds that imsects are more abundant than they are lower down. The Crustacea appear to come from a definite locality in apparently the same shale. At the lower end of the deposits certain definite species of Pecopteris are localized. It is an interesting fact that one seldom finds a Weuropteris at the “upper beds.” The most abundant fossils are various species of Pecopteris and Annularia. When specimens of WVewrop- teris are found they are usually discovered at the lower end of the exposures. In one place behind the “island” very blue nodules, hard and flinty, with sometimes well-preserved fronds of Pecopteris, are found quite definitely localized. These nodules are more likely to assume an irregular shape. These localizations of the various species of fossils are, of course, what we would expect from our knowledge of the manner in which the recent fauna and flora are distributed. There is, to be sure, more or less intermingling of species. The myriapods, as far as they have been found, are localized. Mr. Carr found three within a space of a few feet. But again these are found widely scattered. The exposures at the “ upper beds” are about a quarter of a mile long. They disappear under a heavy ledge of sandstone. The ‘‘ Lower Beds.” Fic. 3. Looking south up Mazon Creek. The pebbly banks are cov- ered with nodules and the stream bed is filled with them. They are easily collected by wading. Just above the rushes in the upper right corner the nodules are non-fossiliferous. Note the high bluff to the back. This continues along the west bank for some distance. Fie. 4. The ‘‘ Lower Beds.” Fie. 4. A nodule just weathering out of the shale (at the head of the hammer). Others may be seen at the base of the slope. This shows the shales in position. The nodules contain, for the most part, species of Neuropteris. R. L. Moodie—Muzon Creek Shales. 983 At the “lower beds” (figs. 3, 4), so-called because farther down the creek, conditions are quite different from those just described. The west bank of the creek is higher and almost perpendicular, and the east bank is low and flat, the bluff being a quarter of a mile away, so that the chances for collect- ing from the shales are fewer. The bed of the creek, however, is wider and there are more nodules washed out. The most abundant fossil at this place is Vewropteris. The nodules at the upper end of the exposure are all, almost without exception, barren of fossils. The exposures here are of about the same thickness and extent as the “upper beds,” though the species, contained in the nodules, are not so varied. Judging from the collections made while there, the Arthropoda are the more abundant in the lower beds. This is, however, a matter which needs further investigation. Besides the place near Morris mentioned by Bradley there are no other localities known where the nodules are collected. Bradley* says of the Mazon Creek beds: ‘“ The outcrop (i. e., the Coal Measures) along Mazon Oreek appears nearly continu- ous, but still I have not been able to satisfy myself as to the connection of the above beds with.those of the lower part of the stream. The strata there developed consist of very variable sandy clay shales and sandstones, in some places becoming nearly pure clay shales, but containing many nodules of ear- bonate of iron. Pine Bluff, at the lowermost crossing of the Mazon, is composed of about forty feet of heavily bedded but. rather fissile sandstone, partly nearly white, partly nearly ferruginous. Less than a mile up the creek, the lower part of this bed changes to highly argillaceous sandy shales, with oceasional streaks and nodules of sandstone. ‘The section is not quite continuous, but there is no distinct line of demarka- tion to separate these latter beds from the ferruginous sandy shales, twenty or thirty feet, of section 24, of township 33 north, range 7 east, which contain large numbers of the fossil- iferous nodules of carbonate of iron, for which this locality has become famous. * * * * These nodules range from about two to ten feet above the main coal seam of all this region, the intervening space being occupied by the soft, blue clay shales, filled with fossil plants, which, at most points, overlie this seam. About a mile farther up the stream, coal has been dug in the bed and banks of the stream but is now abandoned. * * * * On the north side of the [linois river, in the neigh- borhood of Morris, the coal outcrops in the banks of the canal, and in the stretch of lowland about one mile to the northward. The overlying beds are here mostly blue clay shales, with occasional irregular layers of sandstone. The iron nodules *Geol. Surv. Illinois, IV, pp. 196-7, 1870. 284 R. L. Moodie— Mazon Creek Shales. above mentioned occur at the same level, but not in so great numbers as at the Mazon locality. The shales immediately above the coal frequently yielded magnificent specimens of fossil ferns and other plants.” The same nodules are thrown out of a coal mine at Braid- wood, Illinois, and doubtless close search would reveal other localities where the shale is cut through in mining. The beds at both localities, along Mazon Creek, : are slightly ‘folded. This is especially true of the “upper beds,” where a conspicuous fold causes the beds to disappear, to reappear in the bed of the creek a mile and a half north. This is directly across the large “ox-bow” bend of the creek. Mr. Carr said that he had followed the creek around the bend without discovering any new outcrops of the shales. The beds at Mazon Creek were first explored in 1857 by “M. Joseph Evans, who sent his specimens to Berlin, Germany, where they excited great interest. It was he who collected the type specimen Amphibamus grandiceps, Cope. Since the time of Mr. Evans many have collected at Mazon Creek, and without doubt the fossil-bearing nodules from the locality are more widely scattered in the museums of the world than are organic remains from any other one bed. The most eager and faithful collector at these beds has been Mr. J. OC. Carr. He has presented many collections to schools and individuals, as well as furnishing material for many paleontologists. The writer is indebted to him for the presentation of an excellent series of nodules. The nodules at Mazon Creek will always be abundant, and the collecting will always be good so long as the creek continues to carry away the draimage of the region. We may thus hope to learn much, in the future, of the animals and plants of this wonderful locality. It is not possible, on account of the value of the land, to do extensive excavating, and this is not necessary, for the waters of the creek will, in time, make all the necessary excavations. It is quite interesting to note in this place the discovery of a similar bed of fossil-bearing nodules in the banks and bed of Rock Creek near Twin Mounds Post-office, some 22 miles southwest of Lawrence, Kansas. The nodules have been known and collected for some time, and their similarity to those of Mazon Creek has been noted. The nodules at Twin Mounds contain identically the same genera and species of plants, insects, spiders, Crustacea, and a Prestwichia (Eupro- ops) dane, M. & W., as are found at Mazon Creek. There have been, so far, no evidences of vertebrates ; even fish copro- lites and scales are wanting. Further search may reveal these, as well as an interesting amphibian fauna. The beds at Twin Mounds will be fully described elsewhere, and further mention need not be made of them here. R. L. Moodie—Mazon Creek Shales. 285 The paleogeographic conditions of the beds containing the Amphibia at Mazon Creek, as well as those at Twin Mounds, are well represented in Doctor Schuchert’s map.* A refer- ence to this map, that of the Upper Pennsylvanian times, will show that the Mazon Creek deposits and the Twin Mounds deposits occur on the margin of the heavily shaded portion of the Upper Pennsylvanie sea. Doctor Schuchert suggests that the Twin Mounds beds represent a peninsula or a small island near the shore during Pennsylvanic times. So far as our knowledge goes there is no evidence of verte- brate life of the uplands at this time. It was confined to the waters and to the borders of the waters. To be sure, we know very little of upland deposits, but there should be some sug- gestion of the vertebrates did they occur. It is a matter of the profoundest interest to witness here in the Upper Pennsylvanian the appearance of the earliest types of that branch of the ani- mal kingdom which, in later epochs, was to dominate the entire world, some of these types returning to the water as a second- ary adaptation. It is not possible for us to examine these lowly organized creatures without thinking that in them, or in creatures like them, lay the possibilities for the development of that race of animals to which we ourselves belong. Dr. Jennings has recently+ expressed this idea, approaching the subject from an entirely different point of view when he says: “T was in actual material existence as a living organism, and indeed thousands or millions years old, when the pyramids were built, ...” Is it possible for us to conceive that the habits of these amphibian creatures of the Mazon Creek region have left an impress on our own characters? It so, are they not worthy of our very careful attention ? Just what the details of the mode of evolution from these small creatures may have been we do not yet know, but patient and faithful search will reveal many new facts of the profound- est importance. The University of Kansas, Lawrence, Kansas. * Plate 84, Paleogeography of North America, Bull. Amer. Geol. Soc., xx, 1910. + Science, N. S., xxxiv, p. 904. 286 R. L. Moodie—American Jurassic Frog. Arr. XXVII.—An American Jurassic Frog ; by Roy L. Moopir. Tue remains of Amphibia between the close of the Triassic and the opening of the Tertiary are among the rarest if not the rarest of fossil vertebrates. One can almost count on the fingers of one hand the species of Amphibia known in the post- Triassic, pre-Tertiary times. These are: Hylwobatrachus croyi Dollo of the Belyian Wealden; Scapherpeton excisum Cope, S. favosum Cope, S. laticolle Cope, 8. tectum Cope ; Hemitrypus jordanianus Cope, all from the upper Cretaceous (Laramie) of Montana, and Canada. So far as I know, other amphibian remains have not been described elsewhere from the formations mentioned. % Professor Marsh, in 1887, referred to some amphibian bones from the Como Beds* of Wyoming, calling attention to the previous mention of the species. + Again, in discussing the fauna of the Denver Basin, + he mentioned the amphibian where he says: “A batrachian (Hobatrachus agilis) and a pecu- liar fish ( Ceratodus yiintheri_) have likewise been found in this horizon.” The material has not been fully described, since all Professor Marsh’s mention of them in 1887 is as follows: “More recently, various bones of small anourous amphibians (Hobatrachus agilis) have been found, the first detected in any Mesozoic formation. ” \ Thus it is that, up to the present, the Hobatrachus agilis of Marsh has been a nomen nudum, and the discovery has been discredited. Recently through the kindness of Professors Schuchert and Lull of Yale University Ihave been permitted to examine the remains referred to by Professor Marsh, and find on examination that they are undoubted remains of Salientia, and of the modern type. There is no distinction, so far as I am able to observe by the most careful comparisons, between this Jurassic frog and the frogs and’ toads which are around us to-day. This is the more remarkable on account of the great age of the species. The Salientia have been suggested in the Pennsylvanian by Pelion lyelli Wyman, from the Coal Measures of Ohio, and attention has been directed to the salientian char- acters of the species by various writers. No intermediate forms are known and both the Jurassic and the Carboniferous forms are far too indefinite for any conclusions to be based on them as to phylogenetic descent. The transition took place in the Permian or late Pennsylvanian since this species Hobatrachus agilis Marsh has every identical character of the modern Sali- entia. The origin of the Salientia, like that of nearly all of * This Journal (8), xxxiii, p. 328, 1887. + Proc. Brit. Assn. Science, Aberdeen Meeting, 1885, p. 1033. t Monograph U.S. G. S., xxvii, p. 508, 1897. R. L. Moodie—American Jurassic Frog. 287 our great vertebrate groups, is still involved in mystery. The mystery 1s so great that it leads some to doubt the validity of the theory of evolution by gradual development. No distinct ancestral forms are known for any of the Amphibia save the Caudata. Smith Woodward has made the same statement in regard to the fishes and has emphasized the importance of the Mutation theory of DeVries as the paleontological record has thus far been read. The record is however still imperfect. The present specimens seem to indicate a bufonid nature for the species. In fact I think we will be safe in locating the species in the family Bufonidee and possibly even in the genus Bufo. It is entirely too soon for the species to be cer- tainly placed in Bufo, since such a determination will have to await future discoveries of more complete material, especially of the pectoral girdle and the skull. The reasons for placing the species in the Bufonide are simply on account of the well-developed condition of the lower end of the humerus and its apparently calcified condition. This is hardly more than an inference but it is an inference which has long been justified in Paleontology. Certainly the ulno- radial articular end of the humerus of ZHobatrachus agilis is not the same as that of the modern Rana pipiens, or Rana catesbiana and it does resemble the epiphysial structures of cal- citied cartilage described by Parsons for some of the toads. The specimens represent two individuals and by the follow- ing parts: One specimen, a lower end of a left humerus, somewhat smaller than the type; the other or type humerus; the lower end of a tibio-fibula; the entire left (¢) femur; the entire right ilium, all, apparently, of a single individual except- ing the humerus first referred to, which indicates a second frog though possibly of the same species. All of the speci- mens are trom Quarry 9 of the Como Bluff in Wyoming. The humerus (No. 1862 Yale University Museum) of the type as above stated is represented by the lower end only, this portion measuring 6™ in length, by 2™™ in distal width, by slightly more than one half a millimeter in shaft diameter. The well-developed characters of the bone indicate a bufonid nature for the species as indicated above. The head, that is, the ulno-radial articular surface, is as distinctly marked as in all modern Salientia with which I am acquainted. The ball is apparently capped with calcified cartilage. Above the ball is: a distinet pit for muscular attachments, precisely as in modern frogs. The shaft is quite slender and nearly circular. The dium (No. 1568 Yale University Museum) is quite peculiar and will possibly be sufficiently characteristic to sus- tain the validity of Professor Marsh’s genus, Hobatrachus. The element is of the right side. It measures 10™™ in greatest length, by 3"™ in greatest width, by 2™™ in greatest thickness on the articular surface. The element is a slender rod, like the 288 R. L. Moodie—American Jurassic Frog. modern salientian ilium, with the anterior end greatly narrowed and pointed; the pointed portion occupying one and one half millimeters. The shaft of the ilium is flattened laterally. It expands in width from a little less than one-half a millimeter to slightly more than three millimeters. The articular surface is marked by four pits which are the broken surfaces indicating the firm union of the elements of the pelvic girdle. The element is greatly thickened posteriorly, with a slightly devel- oped, posterior, dorsal crest. The femur (No. 1862 Yale University Museum) is quite distinctly amphibian of the salientian type. It is a slender rod of bone from which the epiphyses have been lost, leaving in their place pits occupying the ends of the bone; indicating the slight development of the endochondrium, as in all Am- phibia. The lower end of the femur is divided into two sur- faces by an imperfect partition, much as in modern frogs. The upper end is peculiar in having a well-developed crest which, in life, was undoubtedly capped bya large amount of cartilage. In the fossil state it has been preserved asa spine. The femur measures 12™™ in length, by 8™" in distal width, by 1™™ in diam- eter of shaft, by 2°5"™ in proximal width. The ¢2bi0-jibula (No. 1894 Yale University Museum) is repre- sented by a portion of the lower end including 8™™ of the ele- ment. Its characters are so clearly those of the modern Salien- tia that a description is hardly necessary. The lower end is divided by grooves, one on either side, indicating the previous separation of the tibial and fibular elements, thus plainly show- ing that the frogs have had a long pre-Jurassic history. The humerus of the other individual (No. 1863 The Yale University Museum) is similar to the one described for the type, though somewhat smaller. Like the type there is only the lower half preserved, measuring but slightly more than 4"™. The Jurassic Frog thus indicated by these imperfect remains was an animal about the size of Bufo debits Girard of south- western Kansas, and Texas; though possibly shorter in its bodily proportions as indicated by the short ilium. The above description will indicate, without a doubt, that the Eobatrachus agilis of Marsh is the oldest known frog. It is hardly necessary to figure these imperfect remains since the above description has been written with the skeletons of sev- eral species of Salientia at hand and the comparisons are so exact, the characters so identical and the frog skeleton has been figured so many times that it hardly seems necessary. The specimens are in Yale Museum, where they may be seen at any time by anyone interested in a more direct generic and speci- . fie comparison. This has not been attempted in this short account chiefly on account of the lack of skeletal material of Eobatrachus agilis and on account of the lack of sufficient recent skeletal material. Linhart—Hydrolysis of Metallic Alkyl Sulphates. 289 Arr. XXVIII.—On the Hydrolysis of Metallic Alkyl Sul- phates; by G. A. Linnarr. [Contributions from the Kent Chemical Laboratory of Yale Univ. —cexxxiv. | Il. Methyl and Propyl Barium Sulphates. In a previous article* it was shown both theoretically and experimentally that the rate of decomposition of ethyl barium sulphate in dilute hydrochloric acid solution is perfectly normal, and that it is approximately proportional to the strength of the catalyzer used. The purpose of this paper is to show the influence on the rate of decomposition of the salt of substituting a methyl or propyl group for the ethyl group. It was anticipated that the rate of decomposition would decrease with the increase in the molecular weight of the alkyl group, and this was confirmed by experiment, as is shown in Table II. Preparation and Analysis of the Salts—The salts were prepared as described in the first paper,t except that mechan- ical stirring was used during the preparation of the methyl and propyl sulphuric acids, as well as in their neutralization with barium carbonate; in the first case to prevent charring, in the second, to hasten the neutralization, as well as to pre- vent overflowing. A large crystallizing dish was used for the purpose. The propyl salt thus prepared contains varying amounts of water of crystallization (one crop of crystals contained one molecule of water, another contained only one-half molecule of water) while the methyl salt thus prepared proved to be anhydrous. In fact all three salts, methyl-, ethyl-, and propyl barium sulphate, tend to lose their water of crystallization on standing. In one case the mother liquor from the methyl barium sulphate was allowed to stand over night, and on the next day it was found that the salt had crystallized in large rectangular plates (1x2 "2 centimeters). All three salts may, however, be prepared with definite amounts of water of crys- tallization by allowing the crystals to remain in a Buchner funnel which is connected with the aspirator for several hours until the salt is fairly dry. It is then spread out on filter paper and kept in a dry place for a few hours. The salts thus prepared correspond to the formule: Ba(CH,SO,),.2H,O, Ba(C,H,SO,),.2H,O, Ba(C,H,S0,),,H,0. Method of Hydrolysis.—The hydrolyses were made as described in the first paper.t Attention has been called to the fact that the precipitated barium sulphate chars on igni- * This Journal, xxxii, 53. + Loe. cit. t Loe. cit. 290 Linhart—Hydrolysis of Metallie Alkyl Sulphates. tion, due to the inclusion of either alkyl barium sulphate, alkyl sulphuric acid, or possibly both. In the case of the methyl salt very little charring was observed, whereas in the propyl the whole mass turned black, so that the crucible had to be inclined and the lid placed in such a way as to allow a current of air to pass through during the ignition, until the barium sulphate became white. Theory.—F¥rom theoretical considerations, which have been confirmed by experimental results, it has been shown that for every two gram equivalents of alkyl barium sulphate decom- osed or of barium sulphate formed, one gram equivalent of alkyl sulphuric acid is formed, so that the concentration of the total ester [Ba(RSO,),+HRSO,] undergoing hydrolysis is, at any time ¢, not (A—«) but (A—w#+1/2«), while the concen- tration of the acid accelerating the reaction is (B+1/2 @), as is shown in the following equation :* Ba(RSO,), + HOH = BaSO, + HRSO,+ ROH from which was derived the mathematical expression for the velocity constant, dx 2X 2°3 A(B+1/2 2) —— — 7) 5 . — a = K(A—1/2 x) (B+1/22) or K (A+B) 08 B(A—1/2.2) where A stands for the initial concentration of the alkyl barium sulphate and B that of the hydrochloric acid used, both expressed in gram equivalents per liter. Experimental Results. TaBce I. Methyl Bariwm Sulphate T = 60° BaSO, t aaa a a) in hours in grams in grm. equiv. K B = 0:25 N HCl 98°1 0°1352 0:0579 0:00782 191°2 0°2597 0°1113 0:00772 268°8 0°3637 0°1558 0:00772 335°7 0°4447 0°1907 0°00763 404°7 0°5285 0°2264 0:00761 481°0 0°6236 02672 0:00768 527°5 0°6752 _ 0°2893 0:00768 or 0°7002 0°3000 =A B=05N HCl 47°5 0°1270 0:0547 0:00791 95°5 0°2510 01075 0:00791 168°5 0°4228 071812 000791 265°0 0°6262 0°2683 0:00793 oc 0°6932 0:2970= A * This Journal, xxxii, 53. Linhart—Hydrolysis of Metallic Alkyl Sulphates. 23°0 47°5 95°5 143°0 log TaBLeE I (cont.) B=10 N HCl 0°1312 0°0562 0°2610 0°1118 0°4840 0°2072 0°6646 0°2848 0°6932 02970 =A Ethyl Barium Sulphate t in hours in grams 0°1150 0°2490 0°3630 0°4596 0°5700 0°6670 0°7002 Propyl Barium Sulphate t in hours 72:0 211°0 379°0 545°5 783°0 930°0 1170:0 1435°0 1651°0. in grm. equiv. B=10 N HCl 0°04938 0°1067 0°1555 0°1969 0°2442 0°2858 0°3000 = A Gs see in grams in grm. equiv. B = 0:0625 N HCl 0:0204 0:0088 0-0590 0°0253 071194 0-0512 071704 0°0730 0°2796 071198 0°3424 071467 074356 0°1867 0°5740 072459 0°6794 0°2919 0°7002 03000 = A B = 0125 NHCl 0-0560 0°0217 O-1114 0:0477 0°3498 071456 074982 0°2134 075961 0°2554 0°6720 0°2879 0-7002 0°3000 = A B = 0:25 N HCl 0°0252 0:0108 0-0949 0-0407 071906 0:0817 0°2196 0-0949 0°3886 071665 0°5586 0-246] 0°7002 0:3000 = A 0°00852 0°00852 000852 0°00849 T = 60° K 0°00828 000828 0°00828 0:00835 0:00828 0:00832 T = 60° K 0:00634 0°00597 0°00630 0°00631 0°00630 0°00627 0:00607 0:00621 0°00625 0°00605 0:00576 0°00589 0:00602 0°00608 0°00598 0:00574 0°00594 0°00567 0°00564 0:00588 0°00598 291 292 Linhart— Hydrolysis of Metallic Alkyl Sulphates. B= 0) NACI 23°0 0°0482 0:0207 0:00604 66°5 0°1344 0°0576 0°00592 1 WSs) 0°2200 0°0948 0°00575 169°0 0°3208 0°1374 0°00574 200°0 0°38 744 0'1604 0°00575 250°0 0°4575 0°1960 0:00574 326°5 0°5734 0°2457 0:00572 fod 0°7002 0°3000 ‘ B=1:0N HCl 44:0 0°2316 0:0992 0°00684 92°0 0°4574 0°1960 0:00687 138°0 0°6452 0°2764 0°00686 164°0 0°7394 0°3168 0°00685 a 0°8100 03470 = A a TABLE IT. K for Esters in 1:0 N HCl Ba(CH3S0O.)2 Ba(C2H;S0O.)2 Ba(C3H,SO.)2 0°00852 0:00828 0:00684 (00852 0:00828 0°00687 0°00852 . 0°00828 : 0°00686 0:00849 0°00835 0°00685 (A= 0°2970) 0:00828 (A = 0°3470) 0°00832 (A = 0°3000) Summary.—The following conclusions are drawn from the experimental results : 1. Alkyl barium sulphates and alkyl sulphuric acids decom- pose extremely slowly in water solution even at moderately high temperatures. : 2. The rate of decomposition of the salts is decreased with the increase in the molecular weight of the alkyl group, as is shown in Table II. 3. The inclusion of impurities is greater the larger the molecular weight of the alkyl group. 4, Fair velocity constants are obtained on the hypothesis that alkyl barium sulphates and the corresponding alkyl sul- phuric acids are esters of similar stability in the presence of aqueous hydrochlori¢ acid. Experimental results have been obtained on the hydrolysis in acid solution of the calcium and strontium esters and also of the alkyl sulphuric acids, and in alkaline solution of the sodium, strontium and barium esters. These results will be published in subsequent papers. Drushel and Dean—HHydrotysis of Esters. 293 Art. XXIX.—On the Hydrolysis of Esters of Substituted Aliphatic Acids ; by W. A. Drusuer and EK. W. Dean. [Contributions from the Kent Chemical Laboratory of Yale Univ.—ccxxxy. | 4, Ethyl Esters of Glycollic Acid and of Methyloxy, Ethyloxy and Propyloxy Acetic Acids. Previous work on the hydrolysis of the esters of aliphatic acids has been reviewed in the first paper* of this series. The substances which had been previously studied were the methy], ethyl and propyl acetates, the methy! and ethyl esters of chlor- acetic acid, and the ethyl esters of di- and trichloracetic acids. The velocity of hydrolysis of isobutyl acetate had also been measured. It was shown in this paper that the introduction of halogen into the acetyl radical lowers the rate of hydrolysis. It was also shown that the esters of different alcohols and any one acid differ but little in their velocities of hydrolysis in acid solution. The substances studied were the methyl, ethyl, propyl and isobutyl chlor- and bromacetates. In a second papert it was shown that ethyl eyanacetate hydrolyzes more slowly than the chloracetate. From these results it would seem probable that the rate of hydrolysis bears some relation to the strength of the acids in esters and that this is in accord with the rulet stated by Nernst for the degree of hydrolysis of salts in general. Oyanacetic acid is about two and a half times as strongly dissociated as chloracetic, which in turn is about eighty times as strong an acid as acetic. It was shown, how- ever, that this factor does not entirely account for the ‘differ- ence in reaction velocity of the esters. In a third paper§ the results of work thi alpha and beta chlor- and brompropioniec esters are published. It was found, here that the ester of the beta acid hydrolyzes more slowly than that of the alpha acid in the cases of both bromine and chlorine substituted derivatives. This is apparently an excep- tion to the theory mentioned above, as the beta acids are weaker than the corresponding alpha acids. A study has been undertaken recently to show the effect of the presence of hydroxyl and alkyloxy groups in fatty acid radicals on the rate of hydrolysis. The following esters are considered in this paper: ethyl glycollate, CH,OH.COOC,H, ; ethyl methyloxy-acetate, CH,OCH,.COOC H,; ethyl ethyloxy- acetate, CH,OO,H, COOC H, and ethyl propyloxy-acetate, CH,OC,H, .COOC, H,. * This Journal, xxx, 72. + This Journal, xxxiii, 27. { Nernst, Theoretical Chemistry, p. 521. § W. A. Drushel, this Journal, xxxiv, 69. Am. JouR. Sct. —FoOuRTH SERIES, VOL. XXXIV, No. 201.—SrpremsBer, 1912. 20 294 Drushel and Dean—Hydrolysis of Esters of Preparation of the Esters.—The preparation of these esters was attended with some difhiculty, largely due to the impossi- bility of employing any very effective methods of purification. The glycollic and methyloxy esters in particular are very soluble, rendering it impracticable to wash them with water. The last two esters of the series are more satisfactory in their preparation, although small yields were obtained by the methods employed. The glycollie ester was prepared according to the method of Schreiner, * which depends upon heating ~ together sodium gly collate, ethyl chloracetate and absolute alcohol i in a sealed tube at 150° to 160° for a day. The tubes were then opened and the ester purified by fractional distillation. The first attempts to prepare glycollic ester by this method proved unsuccessful. There seemed to be a great tendency on the part of the ester to break down into glycollic acid and this was very difficult to remove, although its boiling point is 30° above that of the ester. The presence of small quantities of water or free acid in the glycollate appears to prevent almost entirely the formation of the ester. By the use of absolutely neutral and carefully dried sodium glycollate we were able to prepare a small quantity of the ester, which boiled at a con- stant temperature of about 155° and was proven free of halogen. The boiling point given in the literature is 160°, but the fractions we obtained boiling at this temperature gen- erally contained free acid. It does not seem probable, how- ever, that the presence of traces of glycollic acid can noticeably, affect the hydrolysis, as it is one of the products formed in the reaction and is a very weak acid. The alkyloxy acetic esters were all prepared by the same general method, the action of sodium alcoholate on ethy] chlor- acetate. The reaction proceeds according to the following equation : CH,C].COOC,H, + NaOR > CH,OR.COOC,H, + NaCl. In one case the mixture was allowed to react in ethereal solution, in the others an excess of absolute alcohol was used. The yields were rather poor, varying from 30 per cent to 35 per cent in the different cases. It seems possible that this may have been due to loss of ester while the excess of alcohol or ether was being distilled off, as the reaction is one which should go to completion. The boiling points of the esters are given as follows: ethyl glycollate, 160° ; ethyl methyloxy-acetate, 130°; ethyl ethyl- oxy-acetate, 152°, and ethyl propyloxy-acetate, 184:5°. The solubilities are not given in the literature. All but the propyl- oxy ester seem to be very soluble; the latter dissolves with a little ditiiculty in the quantities used in the hydrolysis. * Ann, d. Chem, u. Pharm., exevii, 5. Substituted Aliphatic Acids. 295 Procedure.—The hydrolysis experiments were carried on in the thermostat described in the first paper* of the series. The temperature was regulated so that variations were never more than a tenth of adegree. The esters were hydrolyzed in 250° flasks, which were filied to the mark with decinormal hydro- chlorie acid and warmed to the temperature of the thermostat before starting the reaction. The esters were measured out in quantities of about 2°5°"* from a graduated pipette. At regular intervals 25°™* portions of the reaction mixture were removed by means of a pipette, run into cold water, and immediately titrated with decinormal barium hydroxide solution. Phenol- phthalein was used as an indicator. Experiments were carried on at 25°2°, 35°, and 45°, duplicates being obtained in all the cases. Calculations were made by using the titration formula for reactions of the first order. Figures are given below for the series which gave, the most regular constants. Averages of duplicate series are also recorded. TABLE I. Hydrolysis at 25°2°. N/10 HCl. Ethyl Ethyl Hthyl Alkyloxy Acetates Time Acetate Glycollate Methyloxy Hthyloxy Propyloxy (minutes) 10°K 10°K 10°K 10°K+ 10°K+ 120 63°6 71:0 ee 37°4 34:1 240 67-0 69°5 36°3 36°4 36:0 360 64°9 69°6 39°9 36°5 35°6 600 64:2 (79°8) 41°8 35°4 35°7 840 64:9 (Ls) 36°7 36°3 36°3 1380 65°6 69'9 38°2 36°5 36°9 2040 65°0 70°6 37°7 35°2 35°9 3180 sate eS 36°0 3 Bie Averages... 64:7 70°4 38°1 36°2 35'8 Averages... 64:9 68°5 38°9 36°1 35°8 (duplicate) Hydrolysis at 35° 90 (155°5) W711 89°5 (80°5) (71:8) 180 158°8 184-0 89°5 85°3 (78-9) S00t maori 183°1 88°4 87-2 82-1 600 162°1 176°1 90°9f 88°3f 85-4 900 162°8 168°0 90°4 87:2 85'3 1440 166°4 166°9 90°5 87°5 86°8 2160 (174°0) 160°5 92°3 87°9 86°8 Averages.. 162-7 - 172°8 90°2 87°2 85°3 Averages __ 162°6 166°2 90:0 87°5 84-0 (duplicate) * Loe. cit. + These constants for 25°2° are calculated from series run at 25°0°. Time intervals are slightly different. ¢ Time interval, 700 minutes. 296 Drushel and Dean—Hydrolysis of Esters. TaBLE I (continued), Hydrolysis at 45°. 30 372 (366) (252) 227 (174) 60 377 (360) 232 218 205 90 378 391 230 214 205 150 375 390 225 218 215 270 (338) 391 227 rally 215 450 376 394 226 215 217 630 366 393 225 JAG 216 Averages.. 374 392 227 * 218 212 Averages.. 374 385 226 217 212 (duplicate) The constants in parentheses are not counted in the averages. - Taste II. Summary. Temperature______---- 25°2° 35° 45° 10°K 10°K 10°K 10°K 107K 10°K I II I II I II Ethyl acetate -__-_- 64:7 64:9 162-7 162°6 374 374 Ethyl glycollate_ os 70'4 + 68°5 172°8 166°2 392 385 Ethyl methyloxy- 38:1 38°9 90°2 90:0 227 296 acetate 3 Ethyl ethyloxy- 36:2 36'1 S720 "ra MORSE ON, acetate Ethyl propyloxy- 35°8 35:8 85°3 84:0 BUA Bie acetate The temperature coefficient varies between 2°3 and 2:5 for an increase of ten degrees. Average, 2°4. Summary. (1) Ethyl glycollate has a greater reaction velocity than ethyl acetate when hydrolyzed in acid solution. This would indicate that the presence of the hydroxyl group in an ester accelerates its decomposition. (2) The methyloxy, ethyloxy, and propyloxy esters hydro- lyze more slowly than the acetate, indicating a retardation caused by the presence of an alkyloxy group. This retarda- tion increases with the size of the alkyl radical in the substi- tuted group, the difference, however, being less marked between the ethyloxy and propyloxy than between the methyloxy and ethyloxy esters. In a later paper we expect to show the effect of the presence of a second hydroxyl group in the acyl radical, and also the result of changing the substituted group from the alpha to the beta position. This work will involve the hydrolysis of esters of glyceric, lactic, and hydraerylic acids. S. R. Williams—Electromagnetic Effect. 297 Arr. XXX.—An Electromagnetic Hffect; by S. R. WILLIAMS. Unper the above caption Bowden* in 1895 described some interesting experiments on the behavior of a column of mercury in a magnetic field when carrying an electrical current. I have repeated his experiments with several variations, and inasmuch as Bowden closes his short report by remarking that some of the effects discovered by him are “ difficult to understand,” I may be permitted to offer an explanation of the effects and at the same time describe a very convenient method for measuring magnetic field strengths. Bowden's Experiments. Fig. 1 shows a sketch of the arrangement of Bowden’s appa- ratus. A horizontal glass tube connected two troughs, A and B. Midway between A and B a vertical tube, V, was sealed into the horizontal tube. This was then filled with mercury as indicated and placed between the poles of an electromagnet. The dotted line a, d, c, d, shows the position of the pole pieces. The horizontal tube is normal to the magnetic field. If a current flows along the horizontal tube of mereury from A to B and the magnetic field is in toward the paper, then the mer- cury will rise in the vertical tube, i. e., in the same direction as the mechanical force acting upon a wire carrying a current in the same direction, AB, when placed in a similar magnetic field. The glass tube is clamped in place and cannot move, hence the mercury moves with respect to the tube. Why it does is, perhaps, worthy of consideration. It has been observedt that, if a heavy electric current is passed along a horizontal column of mercury, the column will break in two, then reunite to repeat the process of breaking again. This so-called “ pinch-effect” has been explained from the viewpoint that the column of mercury is composed of a number of conducting filaments, each of which carrying a current in the same direction would have a mutual attraction for the other and, if this force is great enough, will crowd the filaments together and pinch off the column of mercury. If these filaments exist then, when placed in a magnetic field, as indicated in fig. 1, they would be acted upon as a flexible wire would be when carrying a current and placed in a magnetic field. Hence if this flexible wire were placed in the horizontal * Bowden, Phil. Mag., vol. xl, p. 200, 1895. + Northrup, Phys. Rey., vol. xxiv, p. 474, 1907. 298 S. R. Williams—Llectromagnetic Hffect. tube it would be pressed up against the top of the bore of the tube and at the opening of the vertical tube would be bent up, making the wire to appear to rise in the vertical tube. If either current or field is reversed the wire would fall for the same reason. A good illustration of these filaments is found in the effect of an electric field upon some finely broken kernels of = Q Se ee REE fy) AE DDSUILLS IS ST ES PEPELIPELES TL LES FIG. 4 FI@.5 halloysite, which is a compound of aluminium silicate and a very small amount of iron, nickel, cobalt, and manganese. I found that the small particles of halloysite (size of pinheads) when placed in an electric field such as one gets between the poles of a static machine, arrange themselves in filaments and, as one pulls the poles farther and farther apart, a chain of these S. R. Williams—Electromagnetic Hffect. 299 particles several centimeters long can be drawn along with the pole. Along these chains conduction takes place without any sparking. This, to my mind, is entirely different from the experiment described in Deschanel’s Natural Philosophy, p. 67, Pt. III, by Professor Everett, in which the poles of a static machine are immersed in a bath of oil of turpentine. In this bath are thrown filaments of silk, and when the electric field is applied, the filaments arrange themselves parallel to the lines of tension, but do not take part in conduction. This a FIG. 2 last experiment simply shows the tendency of an elongated mass of a dielectric to set itself parallel to an electric field. In fig. 2 is shown a photograph of the filaments of halloysite between the poles of an electric machine. A piece of photo- graphic paper (velox) was placed directly under the poles and the broken halloysite piled around them on top of the paper. Keeping the electric machine excited, the negative pole was slowly separated from the positive. As the negative pole moved farther and farther away it dragged a chain of the particles of the halloysite with it, along which the machine was discharging without any sparking. When the poles were separated by a distance of about 10°™ an incandescent lamp was held over the poles and sensitized paper and an impression obtained as shown in fig. 2, From the standpoint of the fila- ments, all of Bowden’s effects become intelligible. If one looks at the horizontal tube end on and places the tube between 300 S. R. Williams—Electromagnetic Effect. the poles as shown in fig. 3, there is a rise of the mereury in the vertical tube no matter what the direction of the current and the magnetic field may be. Here the force acting upon the current in the mereury is nermal to both the vertical and horizontal columns no matter what the direction of the current and the magnetic field. This lateral pressure on the sides of the horizonal tube acts hydrostatically and forces the mercury up in the vertical tube, as that is the only place that the pres- sure may manifest itself in a movement of the mercury. In another form of tube he had four lateral tubes attached to the horizontal one. Fig. 4 shows the arrangement with horizontal tube end on. When the current flowed toward the reader and the field was in the direction indicated, the mereury rose in B, fell in C and momentarily rose in A and D. This momentary change in A and D was simply an inertia effect due to the changes i in height of mereury in B and C. In connection with this behavior of a mereury or liquid con- ductor carrying a current in a magnetic field, it is well to keep in mind the distribution of the lines of flow through the hori- zontal tube past the openings in the lateral tubes. This point was investigated in the following way: A conducting sheet of an electrolyte (acidified water) was arranged in the form which was the cross-section of the tube used by Bowden and shown in fig.5. This cell was formed by sticking strips of hard rubber to a plate of glass. The lines of flow were obtained by the usual method employed in the laboratory, viz., using the sec- ondary of an induction coil for the source of E. M. F. and a telephone receiver to locate the equipotential points. The lines of flow curve out into the opening of the lateral tubes as shown in fig. 5. After trying out the apparatus, similar in form to that used by Bowden, my attention was called to Lippmann’s galvanometer* and its application by Leduct and DuBois¢ in measuring magnetic fields. In principle their apparatus and Bowden’s are one and the same. Fig. 6 is a schematic view of Ledue’s apparatus. It differs from “Bowden’s in this respect, that the vertical central tube, C, extends below the horizontal tube, and the current is sent through the vertical column instead of the horizontal one. In Ledue’s apparatus the col- umn of mercury carrying the current was inclosed by two par- allel plates, placed very close together with lateral openings for the tubes A and B, fig. 6. Bowden made his apparatus from glass tubing, and as a cross of the form in fig. 6 is easily made by sealing ‘glass tubing together, it seemed worth while to see how satisfactorily a tube might be made for measuring mag- netic tield strengths instead of the flat cell used by Ledue and * Lippmann, Jour. d. Phys. (2), vol. iii, p. 384, 1884. + Leduc, Jour. d. Phys. (2), vol. vi, p. 184, 1887 t+ DuBois, Wied. Annal., vol. xxv, p. 142, 1888. S. R. Williams—EHlectromagnetic Effect. 301 ! more difficult to build. The measurements of the motion of the mercury column due to electromagnetic action were carried on by Mr. William Lyman, an advanced student in the department. Mr. Lyman’s Experiments. Five tubes were made by ourselves in the laboratory and numbered 1, 2, 3, 4, and 5, respectively. The diameters in millimeters of the vertical central tubes and the side arms were as follows : TABLE I. No. 1 2 3 4 5 Vertical central i tulbey eee 12°9 6°8 6°8 3°55 9°33 Side tubes _.-_-- -- 6°8 6°8 12°9 3°55 P3535 These tubes were placed between the poles of a small elec- tromagnet, whose pole faces were 101 76™" and the distance between them 46"". The arrangement is shown in fig. 7. In fig. 8 is shown the relation between the change in height of 302 S. 2. Williams—Electromagnetic Effect. the mereury column in one of the side tubes, and the field strength when the direction of the field was changed. ake first the results were very erratic. This was found to be due largely to the adhesion of the mercury to the walls of the tube. This trouble was remedied by placing a small quantity of glycerine on top of each meniscus. The glycerine behaved as 4.0: nartth i | i t a Sere ieee & : L. 3,0 : 2.0 _—- Z f= 1 H+ oy | i t +t mao tt 4 at t a ti 1.0=e E + AH eo ido ieoet snes HEHE toe ry 400 800 1200 1600 2000 FIG.8 a lubricant and the mercury moved along the tube like a well- oiled piston. With the glycerine covering all of the exposed mercury surfaces, the tubes could be left for weeks and would still yield consistent results. As a precaution the tubes were tilted in various directions each time they were used in order to get the glycerine down around the edges of the meniscus. The elevation and depression of the mercury column was read by means Os a microscope and micrometer eyepiece in the case of tubes 1, 2, and 3, while a telescope and micrometer eyepiece were used foe tubes 4 and 5. Discussion of Results. From the curves 1, 2, 5, and 4 it would appear that the smaller the tubes the more the mercury was displaced. Curve 5, however, contradicts this, and all of the curves point to the * Fifteen amperes flowed through the vertical central column of mercury for all the values given. S. R. Williams—EHlectromagnetic Hffect. 303 reason why. As the tubes get smaller and smaller there is a greater departure from a str aight- line variation of the displace- ment of the mercury with tield strength. It is evident that the friction of flow of the mereury in the tubes is the cause of this variation from the straight-lme law because the frictional forces become larger the smaller the bore of the tube. In curve 5 the frictional force has become so large that it reduces the amount of the displacements for the different field strengths to values lower than that in curve 4. One may conclude that some tube with dimensions between those of 4 and 5 would give a maximum displacement for the field strengths used in this work. If the frictional force is a constant, then all of the curves shown in fig. 8 will become practically straight lines beyond certain field strengths. The curves indicate this to be true. The results obtained from these five tubes show that this is a simple and accurate means for measuring magnetic field strengths over wide ranges. No difficulty was experienced in measuring fields as low as 50 gauss. In actual practice we cali- brated the tubes with a solenoid whose field was known, and then used the tube between the poles of the electromagnet whose field was to be determined. This eliminated the difti- culty spoken of by those who have used Ledue’s form of appa- ratus, viz., that the thickness of the conducting sheet of mercury must be known with great accuracy if one is to calculate the field strength from the amount of displacement of mercury, the current strength, and the dimensions of the tube. If one can do sufficient glass-blowing to put together glass tubing as shown in fig. 6, an instrument can be easily and cheaply made which, when calibrated, may be used for accu- rately measuring unknown fields, and this without knowing any of its dimensions. Summary. 1. This paper has attempted to explain the mechanism of this electromagnetic effect by assuming that in a liquid conductor we have a bundle of conducting filaments. The current flow- ing through these in a magnetic field will behave as it does in a flexible conductor in a magnetic field. In any case, whether we are dealing with conduction through gases, liquids, or solids, the effect obtained is due to the reaction between the magnetic field produced by the moving charges and the field in which the char e@e 1s moving. 2. This work has developed a simple and accurate method for measuring magnetic field strengths. Physical Laboratory, Oberlin College, Oberlin, Ohio, March, 1912. 304 Scientific Intelligence. SCIENTIFIC INTELLIGENCE, I. Puysics. 1. The Growth of Air Bubbles at the Walls of a Beaker con- taining a Liquid, when the Gas at the free Surface is not air ; by C. Barus (communicated).— When water is poured into a beaker in air and an artificial atmosphere of hydrogen is then allowed to rest on the surface, it is surprising that the invisible air bubbles, which abound at the surface of all solid parts under the liquid, gradually become inflated to a relatively enormous size. They may be finally lifted off by their buoyancy. All submerged objects, notably wire gauze, become coarsely jeweled in the lapse of time ; but the bubbles do not reappear when shaken off. This, however, is the second stage of the phenomenon. If the bubbles are not forcibly removed, they will of their own accord ultimately vanish by contraction. Clearly the phenomenon as a whole is a case of the diffusion of hydrogen, through water, into the space indicated by the original microscopic air adhesions ; but it is not at once evident why hydrogen should diffuse from top to bottom ; 7. ¢., against the hydrostatic pressure gradient of the liquid. Nor does it appear why it should afterwards reverse the process ; for the phenom- enon is quite as evident for columns of water a foot or more high, or for large bulks of air under hydrogen. The explanation, which I have given, is as follows: Let A be the head of water above a given small air bubble, and 6 the barometric pressure. At the outset, therefore, the pressure of the hydrogen is 6 and that of the air underneath, b + Apg, in the usual notation. The pressure urging diffusion is thus initially from hydrogen into pure air 4, and for pure air into hydrogen B+ hApg. But as the diffusion proceeds, the gas within the bubble is no longer pure air, but a mixture of hydrogen and air, corresponding to the partial pressures p and p’; so that throughout the experiment B+hpg=p+p'. Now p’ continually diminishes from above atmospheric pressure, as the ‘air escapes from the bubble, while p continually increases from the influx of hydrogen into it. The time must, therefore, come when p’=hpg or B=p, at which the influx of hydrogen must cease altogether. When p’ dimin- ishes further, owing to the escape of air, it follows that since p'B. In other words, both hydrogen and air after- wards escape from the enormously inflated bubble, until it van- ishes. In this way the very curious result is brought about of the apparent diffusion of a gas against hydrostatic pressure into an infinitesimal.air bubble, until the latter is inflated to a bead, and thereafter of the reversal of diffusion and deflation—all in a con- tinuous sequence of phenomena. Brown University, Providence, R. I. Geology and Mineralogy. 305 II. Gronogy anp MINERALOGY. 1. First Annual Report of the Director of the Bureau of Mines, Joseru A. Hotmes, for the fiscal year ended June 30, 1910. Pp. 57. Washington, 1912.—The first annual report from the Bureau of Mines gives an occasion for reviewing briefly what has been accomplished since it was established by act of Congress in July, 1910. Prior to that time the work in this field had been carried on under the auspices of the U. S. Geological Survey. The desirability, however, of increasing the health, safety, econ- omy and efficiency of work in the various lines of mining and in the metallurgical industry has led to the separate development of the Bureau as now constituted. From the start the work has been carried forward with energy and efficiency and has only been limited by too small pecuniary support. The scope of the work may be partially appreciated from the amount of money expended. This was a little more than $500,000 in all for the year ending June 30, 1911. Of this sum $310,000 was devoted to the investigation of accidents in mines, $100,000 to the testing of fuels, and $79,000 went to the expenses of administration, laboratories, etc.; while smaller sums were devoted to making public reports and to mine inspections. The work of the Bureau is three-fold, including investigations as to mine accidents, of fuels and those in special technologic lines. The investigations relating to mine accidents have had a wide scope and have already accomplished important results, although with more liberal support they can be much extended, They look to the development of conditions which shall make work in mines as safe and healthful as possible and reduce accidents to a minimum. The Bureau has been also called to do much rescue work when bad accidents have occurred. The magnitude of the interests involved will be seen from the statement that there are 700,000 coal miners in the country connected with some 15,000 mines, out of which 500,000,000 tons of coal are produced annually. It is not to the credit of the country that the number of men killed per thousand in 1910 was more than four times what it was in Belgium, double what it was in Prussia and more than twice what it was in England. In metal mines accidents are nearly as frequent while health conditions are worse; the latter is also true in metallurgical plants. The fuel investigations have been conducted with the object on the one hand of ascertaining whether the supplies purchased by the Government conformed to the contract specifications and in addition extended to the general character of coals, lignites and other mineral fuels belonging to the Government. Connected with this subject is that of the general prevention of mineral waste in all lines, which greatly needs scientific treatment. Besides the administrative headquarters in Washington, the Bureau has also an experiment station in Pittsburgh with good 306 Scientific Intelligence. facilities for carrying on the various lines of experimental work, both physical and chemical, in testing fuels, explosives, etc. The publications of the Bureau include an extended series of bulletins, also a number of technical papers and finally a series of miners’ circulars. 2. The Platinum and Platiniferous deposits of the Ural.— Prof. L. Duparc has given (Arch. Sci. phys. nat., vol. xxxi) a full and valuable account of the deposits of platinum in the Ural, the single region upon which the world depends now for substantially all its supply of that metal. His researches on the subject have extended over a period of more than ten years. Briefly stated, he shows that the platinum is immediately associated with a band of basic eruptive rocks following more or less closely the water- shed between Europe and Asia, particularly on the eastern side of the mountains. This band extends from the northern to the central Ural, while farther south it is continued by a series of isolated occurrences having the same general trend. In this, the western, zone occur the important deposits of platiniferous dunite. To the east les a second band, shorter and not continuous, in which serpentine rocks prevail; just to the south of Ekaterinen- burg it disappears or becomes lost in the first zone. These bands are flanked on both sides by metamorphic crystalline schists. With respect to the individual occurrences, described in detail by the author, they show in general a mass of dunite at the center, more or less elliptical in form, with its axis approximately north and south. A border of more or less developed pyrox- enite —also carrying platinum-—is observed accompanied by certain melanocratic rocks and, in addition, an external zone of feldspathic rocks, including gabbros, diorites, ete. The platinum occurs in the dunite both directly crystallized with the olivine and also with the chromic iron; in the latter case the platinum sometimes forms a spongy mass enclosing chromite grains. Both in the dunite and the pyroxenite the platinum is regarded as a magmatic mineral, a product of the differentiation of the magma. It is interesting to note that by careful search it has been found possible to observe the platinum in place in the dunite, although such an occurrence is rare. The author has also earlier described (with P. Pamphil) the various peculiar types of rock associated with the platinum (koswite, issite, etc.) and has also (with H. C. Holtz) given anal- yses of many samples of the platinum, 3. The Gabbros and Associated Rocks at Preston, Connecti- cut ; by G. F. Loveutin, Bull. 492, U. 8S. Geol. Surv., 1912.— This bulletin describes an area of about one hundred square miles in eastern Connecticut. The bed rock formations comprise (1) Metamorphic sedimentary rocks—quartzite, quartz-biotite schist, hornblende schist, black pseudoporphyritic schist (Kinsigite), and dolomite ; these are assigned provisionally to Cambrian and Car- boniferous ages. (2) Gabbro—two principal and several minor varieties. (3) Granite—three varieties. The gabbro was the Geology and Mineralogy. 307 first intrusive ; it preceded and was of sufficient mass to control the large regional metamorphism which followed. The secondary lines so developed determined the direction of the granite intru- sion that accompanied the regional movement. The great Lantern Hill quartz is considered to have originated from a more or less complete replacement of alaskite by quartz during the pneumatolytic stage of granite intrusion. The bulletin as a whole is distinctly petrographic in character. F. W. 4. Danas Manual of Mineralogy; for the Student of Elementary Mineralogy, the Mining Engineer, the Geologist, the Prospector, the Collector, etc.; by Witt1am E. Forp. New edition, entirely revised and rewritten. Pp. 460; with 357 text figures and 10 half-tone plates. 1912. New York (John Wiley & Sons) and London (Chapman & Hall).—Dana’s Manual has played an important part in the elementary instruction in Miner- alogy since its first publication in 1848. The work has been revised several times and in 1878 it was rewritten and an extended chapter on petrography added. A fourth edition, further revised, was issued in 1887, but though many times reprinted, twenty- five years have now passed since any change has been made in the text. There has been, therefore, a pressing call for the thorough working over of the volume, which has been ably accomplished by Professor Ford. In fact, the book as now issued is from beginning to end a new one although carried out on the same lines as its predecessors. The wide experience of the present author in teaching elementary classes in mineralogy has enabled him to present the whole subject in a thoroughly fresh, clear, and concise form. Thus, starting from a book of well-recognized merits in its earliest forms, he has produced a work which should be widely useful not only in elementary instruction but also meet the needs of those interested in the science on the practical side. In the arrangement of species, the chemical order has been followed instead of bringing the compounds of a given metal together as in earlier editions. But the end aimed at in that method has been accomplished by the addition of a chapter detailing for each metal the prominent species and giving the essential facts in regard to their occurrence as ores. The ex- tensive chapter on “ Rocks” of the two preceding editions has been omitted since several books on this special subject are now available ; a brief chapter is included, however, giving the chief types of rocks and the minerals prominent in their formation. The concluding section of the volume is given to a series of determinative tables based upon physical characters. In this the prominence of the individual species is indicated by the character of the type employed. In an appendix a summary is given of the mineral production for the United States in 1910; it is stated that these tables are to be revised from time to time so as to keep the information given as nearly as possible up to date. The illustrations in the text are for the most part new, and a series of half-tone plates give reproductions of typical specimens. 308 Scientific Intelligence. III. Muiscriiannous Screntiric INTELLIGENCE. 1. On a Period of 33°33 years in the Earth’s Climate, and its connection with Sun Spots ; by Jonas Zixius. Reprinted from Das Weltall, the astronomical periodical of the Treptow Observ- atory. Pp. 10, 4to.—This is one of a series of papers on the subject of the periodicity of terrestrial phenomena in general and their connection with sun spots. Since Fabricius discovered sun spots in 1611 and the apothecary Schwab detected with his “imperturbable telescope” their periodicity in 1843, and Wolf in 1879 fixed the length of the period at 11 years, the idea of the influence of these solar disturbances on terrestrial affairs has con- tinually grown until, in the investigations of Dr. Zilius, it has reached formidable proportions. How many new cases of this connection he has discovered and investigated does not appear from this article. He presents tables of deficiency and excess of rainfall at Paris, Prague, Madras, and Bernaul from 1815 to 1884. The wettest and dryest years do not fall exactly together at these stations, but they average up to show a maximum of dryness in 1883-4 and 1866-7. These agree with minima of sun spots at 1833°3 and 1866°8, and these in turn coincide with the displays of the Leonids or November meteors. One cannot but be impressed with the author’s power of mar- shalling statistics, but there also comes to mind the saying that “ fioures cannot lie but they are prone to prevaricate.” —w.._B. 2. The Elements of Statistical Method ; by Wit¥orp I. Kine, M.A., University of Wisconsin. Pp. xvi, 250. New York, 1912 (The Macmillan Company).—The author claims, and no doubt correctly, that no book published in America has attempted to cover the field of statistical method in its present state of advance- ment, while those published abroad are either partial or technical. This volume presents the subject simply and free from all intri- cate mathematical arguments. It is in form suitable for use as a text-book, but at the same time it is convenient and attractive for general reading. : W. B. 3. Archiv fiir Zellforschung, edited by Dr. Ricuarp Goxp- scHMIDT, Professor in the University of Munich.—Part I of the eighth volume of this well-known journal, devoted to cytological investigation, is really a botanical number. In addition to four- teen brief reviews of recent botanical papers, it includes the fol- lowing original contributions : Kernstudien an Pflanzen, by H. A. C. Miiller ; Etudes sur le développement du sac embryonnaire et sur la fécondation du Gunnera macrophylla, by J. A. Samuels; and Cell Structure, Growth, and Division in the Antheridia of Polytrichum juniperinum, by Charles E. Allen. ‘ES TE We New Circulars. 84: Eighth Mineral List: A descriptive list of new arrivals, rare and showy minerals. 85: Minerals for Sale by Weight: Price list of minerals for blowpipe and laboratory work. 86: Minerals and Rocks for Working Collections: List of common minerals and rocks for study specimens; prices from 1% cents up. Catalogue 26: Biological Supplies: New illustrated price list of material for dissection; study and display specimens; special dissections; models, etc. Sixth edition. Any or all of the above lists will be sent free on request. We are constantly acquiring new material and publishing new lists. It pays to be on our mailing list. Ward’s Natural Science Establishment 76-104 Cotiecr AVE., Rocuester, N. Y. Waro’s Natura Science EstastisHMent A Supply-House for Scientific Material. Founded 1862. Incorporated 1890. DEPARTMENTS: Geology, including Phenomenal and Physiographic, Mineralogy, including also Rocks, Meteorites, etc. Palaeontology. Archaeology and Ethnology. Invertebrates, including Biology, Conchology, ete. Zoology, including Osteology and Taxidermy. Human Anatomy, including Craniology, Odontology, ete. Models, Plaster Casts and Wall-Charts in all departments. Circulars in any department free on request; address Ward's Natural Science Establishment, 76-104 College Ave., Rochester, New York, U.S. A. CONT ENTS: Arr. XXI.—Ionization by Collision in Gases and Vapors ; by W. A, Bagss? co ole e222 e oer, eee een XXII.— Geology of Arisaig-Antigonish District, Nova Scotia ;.. by. Mas Yo owas Miss <2 a ae eee 242 XXIII.—Jackson on the Phylogeny of the Echini; A synopsis by 0. ScmUCREET )) 22232. 2. - ewe es ae ea XXIV.—The Belt and Pelona Series ; by O. H. Hersuny .- 263 XXV.—Absorption and Thickness of thin Films ; by C. C. HUTCHINGS < 200.35. Sa .o8 2S eee eee XX VI.—The Mazon Creek, Illinois, Shales and their Amphib- jan Fauna; by Ri: Moopimc 7) oe ee age ae ee XXVII.—An American Jurassic Frog; by R. L. Moopiz 286 XXVIIJ.—On the Hydrolysis of Metallic Alkyl Sulphates ; by: SG, oA rine ager ooh 2152's Sate ed oe ee 289 XXIX.—On the Hydrolysis of Esters of Substituted Aliphatic Acids; by W. A. DrusuEx and E, W. Dean ...------ 293. XXX.—An Electromagnetic Effect; by 8S. R. Witiiams... 297 SCIEN TIFIG INTELLIGENCE. Physics—Growth of Air Bubbles at the Walls of a Beaker containing a Liquid, when the Gas at the free Surface is not air, C. Barus, 304. Geology and Mineralogy—Firsi Annual Report of the Director of the Bureau of Mines, J. A. Hotmus, 305.—Platinum and Platiniferous Deposits of the Ural, L. Duparc: Gabbros and Associated Rock at Preston, Connecticut, G. F. Loveuutn, 306.—Dana’s Manual of Mineralogy; for the Student of Elementary Mineralogy, the Mining Engineer, the Geologist, the Prospec- tor, the Collector, etc., W. E. Forp, 307. Miscellaneous Scientific Intelligence—On a Period of 33°33 in the Harth’s Climate, and its connection with Sun Spots, J. Zinrus: The Elements of Statistical Method, W. I. Kine: Archiv fiir Zellforschung, 308. ‘ Library, Bureau of Ethnology. SO te VOhe xX XTVi , OCTOBER, 1912. Established by BENJAMIN SILLIMAN in 1818. THE AMERICAN JOURNAL OF SCIENCE. Evitor: EDWARD S. DANA. ASSOCIATE EDITORS Proressors GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or Camsrince, Proressors ADDISON E. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT E. GREGORY AND HORACE S. UHLER, or New Haven, Prorressor HENRY S. WILLIAMS, or Iruaca, Proressor JOSEPH S. AMES, or Batrimorz, Mr. J. S. DILLER, or Wasuineron. ‘ FOURTH SERIES VOL. XXXIV—[ WHOLE NUMBER, CLXXXIV]. No. 202—OCTOBER, 1912. WITH PLATE I. RSThl OO Fh my ao Aol NEW HAVEN, CONNECTIC ELD Jétiana! MuceWee —lenai Muse 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, Remittances should be made either by money orders, registered letters, or bank checks (preferably on New York banks). REMOVAL NOTICE. It is with pleasure that I announce the fact that I have removed to more commodious and pleasant quarters. For years | have been handicapped by lack of room to carry on my fast increasing business and properly display the ever- arriving new lots of goods, but the task of removal was so great, the time necessary to do this great task and yet keep my customers satisfied, nonplussed me; but I have at last “taken the bull by the horns” and made time and am now . at last situated where | can properly show my vast stock as it should have been shown years ago. In going over my stock, some of which has been packed away for years, I was astonished myself at the many rare and beautiful minerals and other objects | came across. You will certainly miss a rare treat if you cannot come and see me, but if this is impossible, send me a list or an idea of what you want and | will ship you a box on approval, express prepaid, and you can return me what you do not desire. In my new quarters | am gomg to make a specialty of the following: Minerals, Gems, Curios, Mosaics, Antiques, Jewels, Synthetic Gems, Art Objects, etc. Send for my new Catalogue, just out. It will pay you to be on my mailing list. A. H. PETEREIT, 261 West 71st St., New York City. re THE AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES.] +6. Arr. XX XI.—On the Emission of Electrons by Metals under the Influence of Alpha Rays; by H. A. Bumsteap and A. G. McGovean. I Introduction. Iy a previous paper under the same title by one of the pres- ent authors,* an account was given of some experiments upon the so-called 6-rays which are emitted by metals when struck by a-rays. The emission is known to consist of electrons mov- ing with comparatively small velocities. Maximum estimates of their velocity, based upon the potential difference necessary to cause saturation of the current carried through a high : ¢c vacuum by these electrons, give about 3° x 10° —, correspond- ing to a potential difference of about 25 volts; minimum esti- mates, obtained by measuring the positive potential which a source of 6-rays will attain if insulated in a high vacuum, give velocities corresponding to 1 to 3 volts. Campbellt+ has recently brought forward some evidence for believing that the electrons have considerably smaller velocities even than this, if indeed they have any measurable velocity at all. The experiments to be described in §3 of this paper, however, render this conclusion improbable. In the previous paper it was shown that there was a close analogy between the emission of 6-electrons by a metal and the ionization of a gas by a-rays. The number of electrons ‘emitted by the metal varies with the speed of the a-particles in the same manner as the number of ions produced in a gas; so that, by interposing various thicknesses of aluminium foil * This Journal, xxxii, 403, 1911; Phil. Mag., xxii, 907, 1911. + Phil. Mag., xxiii, 481, 1912. Am. Jour, Sci.—Fourts Srrizs, Vou. XXXIV, No. 202.—Ocropmr, 1912. 21 Os a rs 310 Bumstead and MeGougan—Emission of Electrons by between the source of a-rays and the metal, one can plot a curve entirely similar to the ionization curves first obtained by Bragg. The number of electrons emitted by the metal increases with the number of foils interposed until their com- bined thickness is nearly equal to the range of the a-rays in aluminium, after which the emission of electrons falls off rapidly.* The increase in the number of electrons with dimin- ishing speed of the a-rays, however, is not so great as the increase in the number of ions ina gas. The ionization curve of the metal lies within, or to the left of, the corresponding curve for gaseous ionization and has a less pronounced maxi- mum or “knee” just before the end of the range is reached. This result was anticipated before the experiments were under- taken, for reasons given in the previous paper. The same arguments which led to this conclusion also gave reason to believe that the curve for a metal of high atomic weight, such as gold, would lie within that of a metal of lower atomic weight, such as aluminium, just as the latter would lie within the curve of gaseous ionization. This expectation, however, was not fulfilled by the results of the experiments. The curves for gold and for aluminium coincided within the limits of accuracy of the somewhat rough method of determining them which was used. Quite apart from the theoretical reasons discussed in the former paper, it was somewhat surprising to find that two metals which differ so much as aluminium and gold gave, nevertheless, the same ionization curves. For the ionization curves of gases and vapors differ considerably among them- selves, not only in the area enclosed (total ionization) but also in their shape.t The close similarity between the curves for aluminium and gold gave rise to the suspicion that the elec- trons which had been producing the effects observed came not from the metals but perhaps from a layer of adsorbed gas which was the same in both cases. In order to test this possi- bility the following experiment was made. $1. Attempt to Remove Adsorbed Gases by Heating. A strip of thin platinum foil, 6°" long, 3-7" wide, and 24 x10-*™ thick was stretched horizontally between two heavy brass clamps; the clamps were carried each on a vertical copper rod which passed through the cover plate of the evacu- ated chamber in which the 6-ray effects took place; the rods / were insulated from the plate by ebonite, an earthed guard * The increase in the emission as the speed of the a-particles decreases has also been observed by Campbell, Phil. Mag., xxi, 276, 1911. + Taylor, Phil. Mag., xxi, pp. 573, 575, 1911. Metals under the Influence of Alpha Rays. 311 tube, and amber, and the joints made tight with sealing wax. The platinum foil was arranged so that it could be exposed to a pencil of a-rays from polonium deposited on the end of a copper plug 4™™ in diameter. Between the polonium and the platinum, one could interpose aluminium foils, without inter- fering with the vacuum, in the manner described in the former paper. Two, three, four, five, or six layers of foil could be interposed, each 3°2 x 10-*™ thick. The case was exhausted to about -0001™ with the help of charcoal and liquid air; one of the insulated copper rods which carried the platinum foil was connected with a quadrant electrometer, the case surround- ing the exhausted chamber was charged positively, and meas- urements of the negative current leaving the platinum were taken in the usual manner. After an “ionization curve ” had been determined in this way, the platinum strip could be heated by a current sent through it and the two copper rods, and the curve could be again determined, when the liberated gas had been removed. Even before the platinum strip was heated its behavior gave evidence that the occluded gases had some effect upon the phenomena, if only a temporary one. With the metals used previously, the saturation value of the current had been obtained with + 40 volts on the case. With the platinum, an hour after the liquid air had been applied to the charcoal, it required + 160 volts to cause saturation and the current at this potential was 20 per cent greater than at 40 volts. Four hours later, 120 volts was sufficient to cause saturation, and the current was only 12 per cent greater than at 40 volts. After an interval of 24 hours, saturation was reached at 80 volts with a 5 per cent inerease over 40 volts. During the same time the magnitude of the current (taken under similar conditions) fell off about 30 per cent. But the shape of the ionization curve, obtained by using a saturating potential and interposing aluminium foils, changed very little, if at all, while these very considerable changes were going on in the conditions of satura- tion and in the actual magnitude of the current. The results are given in Table 1. The first line of the table gives the time after the liquid air was applied to the charcoal bulb, the second line gives the value of the currents (with two foils inter- posed) corresponding to these times; the last five lines of the table give the values of the current when different numbers of foils are interposed, the current with two foils being taken as 100 in each case, to facilitate comparison. It will be seen that the relative va.ues for 2, 3, and 4 foils show no progressive change with the time; the differences between them are of the order of magnitude of the experi- mental errors. The values for 5 and 6 foils, however, appear 312 Bumstead and MceGougan—Emission of Electrons by TABLE 1. Time | ahr. 2 hrs 4 hrs 6 hrs. 24 hrs. Cy | 186 174 150 146°5 129 eee pea = Foils | 2 100 100 100 100 100 3 112°0 2, al 23s33 113°0 111°8 4 103°8 103 105 103°3 104°7 5 59°5 60°3 64 63°7 66'2 6 Doral 93: 25°3 25'6 26°6 to increase with the time; these values are on the decreasing portion of the ionization curve, where a small change in the range of the a-rays makes a large difference in the current. The observed increase can be explained by supposing that the progressive removal of occluded gas from the platinum foil slightly diminishes its stopping power for a-rays and thus the 6-radiation from the emergence side is increased. After making the measurements which are recorded in the last column of Table 1, a current of 12 amperes was passed through the platinum foil. With this current a bright red heat was obtained in the middle of the foil, fading away grad- ually to the ends, which were cooled by conduction through the clamps and copper rods.* The current was continued for ten minutes, during which time the pressure rose from less than "00017" to 004". The charcoal bulb did not absorb the gas, although the liquid air was left on over night; this is doubt- less due to the fact that the gas emitted by the platinum con- tained considerable hydrogen which is not readily absorbed by the charcoal. The liquid air was then removed, and the Toepler pump was operated, while the charcoal was re-heated to aid in sweeping out the hydrogen. When a pressure of about ‘001™™ had been reached the liquid air was again applied, and the pressure soon fell to less than -0001™". After two hours the value of C, was 134 and the variation for the differ- ent foils interposed did not differ appreciably from the last column of Table 1. This test, however, was not very satisfactory on account of the failure of the charcoal to remove promptly the gas emitted by the heated platmum. Accordingly the charcoal bulb was removed and a Gaede pump substituted for the Toepler pump. When a vacuum of 0001" had been maintained for an hour, * A glass window in the brass case permitted observation of the foil, Metals under the Influence of Alpha Rays. 313 measurements were taken as before. Then the strip was heated five times, for ten minutes at a time, with intervals of fifteen minutes between, to avoid too great heating of the clamps and copper rods: the pump was kept running contin- uously. Measurements of the 6-ray current were then made as before. The results are given in Table 2. TABLE 2. Before | After Foils C, = 150 Cy 34 2 100 100 3 110° 10872 4 102°7 101°0 5 62°5 61:0 6 24-7 24°92, Here there does appear to be a slight alteration in the form of the ionization curve, and in the direction expected. But the differences are so slight that no great confidence can be placed in the result. Further experience with the method did not give any reason for hoping that the question as to the effect of adsorbed gases could be definitely settled in this way. The results of this experiment are susceptible of two quite different interpretations. We may say that, since the heating of the platinum strip did not alter the shape of its ionization curve, we may conclude that this shape is due to the properties of the metal itself and not to adsorbed or occluded gases. On the other hand, we may put the emphasis on the fact that, under certain conditions, at least 30 per cent of the d-ray effect is due to such gases, and that when these are removed there is no change in the shape of the curve. If the residual effect is really due to the metal we must suppose that platinum and the adsorbed gases have ionization curves of very nearly the same shape, which is not altogether probable considering the variations which are met with in the curves of gaseous ionization. From this point of view, therefore, it seems more probable that the d-ray emission (slow electrons) is mainly due to such a layer of gas, which may be reduced but not entirely removed by the methods which we have employed.* *Since the present investigation was completed, a paper has appeared (Pound, Phil. Mag., xxiii, 813. May, 1912) in which the effect of occluded gases upon the magnitude of the d-radiation is clearly brought out. 314 Bumstead and McGougan— Emission of Electrons by $2. Determination of the Ionization Curves of Various Metals. In the determination of the ionization curve of platinum in the preceding section and of those of gold and aluminium in the previous paper, the metals were so thin that the a-rays passed through them, except when they were near the end of their range. This has the advantage that, at least until the Fie. 1. LLL ar N \ \ = \ \ N \ \ N \ N \ N \ \ Wit, WY MMM CENTIMETERS. 1 2 3 4 L top of the curve is reached, no correction for the charge of the a-rays 1s necessary and one ‘need not apply a magnetic field in order to determine this charge. On the other hand, a difficulty results if one wishes to make a careful comparison ‘of different metals. It is impossible to obtain foils of different metals Metals under the Influence of Alpha Rays. 315 which have the same retarding effect on the a-rays; and hence the emergence 6-radiation from the different metals correspond to different speeds of the a-rays even when the same number of aluminium foils are interposed. This is plainly shown by a comparison of the results for platinum with those for gold and aluminium given in the previous paper. The platinum foil is so much thicker than the others that the entire form of the curve is distorted. Moreover, in the foregoing experiments, very few points on the curve were taken, so that, indeed, it is only by courtesy that it can be called a curve at all. For these reasons a much more careful determination was under- taken by means of the apparatus shown in fig. 1. A copper plug, P, 4™™ in diameter, whose lower surface is covered with polonium, is surrounded by a brass cylinder, C, which limits the cone of the rays so that they fall completely within the brass ring, E, which supports the sheet of metal under investigation. In order to obtain more points upon the ionization curves, two dises, D, and D,, are used instead of the one which was used in the previous experiments. In order to make the drawing clearer, the rod supporting E is shown in the same plane with the axes of the dises; in the actual appa- ratus it is in the plane perpendicular to this. The dises are divided into eight equal sectors. D, has a hole 15°" in diam- eter cut in each sector. One hole is left open and the others are covered with 1, 2, 3, 4, 5, 6 and 7 layers of aluminium foil of thickness 3°2 x 10~-*™, having a retarding effect upon the a-rays, according to Taylor’s results, equivalent to that of 0-58 of air (air equivalent). The dise D, has one sector without a hole so that the a-rays can be stopped completely ; the other sectors have holes, one of which is left open, others being covered with 1, 2, 3, and 4 layers of thinner aluminium foil, 0°64 x 10-*™ thick. Thus five of the thin foils are equivalent to one of the thicker. The dials, S, and S,, outside the evacuated chamber enable one to set the dises D, and D, so that any combination of the thick and thin foils may be interposed in the path of the a-rays, or the brass sector may stop them entirely, or the two holes allow them an uninter- rupted passage to the electrode. The metal plate attached to the ring, E, was in every case chosen of sufficient thickness to absorb completely the a-rays so that the 6-electrons were emitted only from the side on which the a-rays were incident. The electrode is insulated from the case by amber, guard-tube and ebonite, and is con- nected to a sensitive gold leaf electroscope of the Hankel type.* A key is connected to a potentiometer arrangement so that * This instrument is described in the previous paper, this Journal, xxxii, 405, 1911; Phil. Mag., xxii, 909, 1911. 316 Bumstead and MeGougan—FKmission of Electrons by the leaf can be insulated, grounded, or charged to any desired potential ; the volt-sensitiveness was thus taken after each reading. In the following measurements, the sensitiveness was adjusted to give a deflection of about 25 divisions on the scale of the microscope for 0:2 volts. The tube T is con- nected to pump, gauge, and charcoal bulb. When the a-rays fall on the electrode, E, they carry over to it their positive charges; the d-rays which they excite are negatively charged and these, leaving the electrode, add to its positive charge. In order to insure the removal of all the emitted electrons from the electrode and to prevent the 6-rays emitted by other parts of the apparatus from reaching it, a positive potential of 40 volts is applied to the cuse. If a sufficient magnetic field is applied witli its lines of force parallel to the electrode, the electrons emitted will be turned back to the electrode, and in this way the charge due to the a-rays alone may be determined. For this purpose an electromagnet was constructed of Swedish iron, 2 inches square in section. It was forged into the shape of a rectangle 30° by 25°" and a gap left in one side 15-2°" long which was just sufficient to embrace the exhausted chamber. It was wound with about 1000 turns of No. 14 cotton-insulated, paraftined wire. The field at various points between the poles was measured with a Grassot flux-meter; a current of 5 am- peres produced a field, midway between the poles, of 95 gausses. It was found that, with the case earthed, this field reduced the current received by the electrode to a minimum ; no further diminution occurred when the current through the magnet-coil was increased to 9 amperes. On account of the large air-gap the field was very nearly proportional to the current. | The results obtained when the electrode was a sheet of aluminium are shown in fig. 2, in which the ordinates represent the number of aluminium foils between the poloninm and the electrode, and the abscissee are the currents measured by the electroscope. Curve Ia gives the results with no magnetic field, and thus represents the total effect due to both a- and 6-rays. Curve II shows the currents observed when the magnetic field was on, due to the charge carried by the a-rays alone. Ourve Id is obtained by subtracting the abscissee of IL from Ia and represents the 6-ray effect, or “ionization” of the metal. It will be observed from an inspection of Curve II that the number of a-particles which reach the electrode apparently decreases as more foils are interposed. Up to five thick foils this decrease is approximately linear, and amounts to about 20 per cent of the total. This is too great a falling off to be attributed to the scattering of the a-rays according to the Metals under the Influence of Alpha Rays. 317 results obtained by Geiger.* It is possible that an explanation of this effect may be found in the phenomena to be discussed in the following section. Beyond five thick foils, the number of a-particles diminishes rapidly; this is doubtless due to their absorption in the aluminium foils, the more oblique rays being the first to be stopped. By dividing the abscissee of Id by those of II, a curve could be obtained which would represent the ionization or 6-ray Fie. 2. 12 CURRENT effect produced by a fixed number of a-particles, which is, strictly speaking, what should be given by an ionization curve. Nevertheless, it does not seem advisable to use results thus obtained in the present investigation ; the a-ray currents are small and the errors introduced by using them as factors might be considerable. Moreover our purpose is primarily to com- pare the effects with different metals. The ionization curves of gases and vapors have also usually been obtained without allowance for the decrease in the number of a-particles, so that a more direct comparison with them is possible by using the eurve Id. An estimate of the number of 8-electrons due to one a-particle may be obtained, however, by dividing the abscissee of Id by those of II. Allowing for the fact that the charge on an a-particle is twice the electronic charge, we find that the number of 6-electrons per a-particle emitted by an * Proc. Roy. Soc., Ixxxiii, 492, 1910. 318 Bumstead and MeGougan—Emission of Electrons by aluminium plate from the incidence side only, varies from 7 to 17 as the speed of the a-rays is gradually reduced. One very striking result appeared in the course of these experiments which was quite unexpected from anything pre- viously known as to the effects of a-rays. As the number of aluminium foils is decreased, the ionization follows a perfectly regular Bragg curve until only one thin foil is left. When, however, this is removed so that there is no obstacle between the polonium and the electrode, a very large increase is ob- served in the é-ray current. Thus in the series represented by fig. 2, the d-ray current for one thin foil is 7-82 while for no foils it is 16°27, an increase of 107 per cent. It has been shown that this is due to a very absorbable radiation consisting partly of electrons moving with considerably higher velocities than the hitherto recognized 6-rays. An investigation of this absorbable radiation will be described in the next section. Experiments similar to those which have been described at length in the case of aluminium were made also with copper, gold, lead, and platinum. In all cases the surfaces of the metals were made clean and bright by fine sandpaper. The magnitude of the é6-ray currents obtained from the various metals under similar conditions were not very different from each other, when correction was made for the decay of the polonium in the intervals between the experiments. The main purpose of the present investigation was to ascertain the varia- tions in the form of the ionization curves for different metals and not their absolute magnitudes ; for this reason no attempt was made to get an accurate determination as to the latter point. It is rendered difficult by the fact, discussed in § 1, that the d-ray current falls off with the lapse of time after the liquid air has been applied to the charcoal. This effect was observed in all the metals studied, but it was not quite so marked as in the case of the thin platinum foil described in §1. However, the diminution in the current sometimes amounted to as much as 20 per cent and continued to be noticeable for two or three days.* As in the case of the platinum foil, the relative values at different points of the range of the a-particles were not affected by this variation, the ratio of the ionizations at any two points remaining practically constant. The results for the different metals are given in Table III and plotted in fig. 83. The values used are the currents due to the 6-electrons alone, corresponding to Curve Id in fig. 2. In order to make the comparison easier they have been reduced to the same scale by making the current for one thick foil the * Our observations in regard to the variations in the magnitude of the d-ray current with different metals is in substantial agreement with those of Campbell, Phil. Mag., xxi, 298, 1911. i Metals under the Influence of Alpha Rays. 319 same for all the metals ; the other currents are then altered in the same ratio. TABLE IIT. aeramEeeae Al Cu Au Pb Pt 0 (0) 16°27 14°67 15°47 13°57 15°87 0 ] 7°82 Nate) 7-92 7°87 ene! 1 0 8:07 8:07 8°07 8:07 8:07 2 0 8:53 8°58 8°43 8°53 8°41 3 0 9-17 9:02 8°82 9°32 9°08 4 0 9:77 9°67 9°62 9°57 9°60 4 1 9°83 9°82 4 2 10°12 9°92 9°82 9°85 9°90 4. 3 10°24 9°99 9°99 10°04 9°94 4 4 10°05 10°00 5 0 10°31 10°16 10°19 10°26 10°11 5 1 10°45 10°25 10°25 10°35 10°25 5 2 10°21 10°00 10°31 10°41 10°26 5 3 9°71 9°56 10°11 10°06 9°96 5 4 6°35 7:95 8°30 8°30 8:15 6 0 2°45 2°69 2°29 2°41 2°51 6 1 1°43 173 1°56 1°25 1°61 6 2 “14 WE °06 7/3} 83 X-ALUMINIUM +-GOLD ©-COPPE! O-~PLATINUM e-LEAD 10 CURRENT 320 Bumstead and McGougan—LKimission of Electrons by The points for the different metals lie so closely together that only one curve has been drawn. _ Anyone who has had experience with such measurements will recognize that the dif- ferences are too small to have any significance. Even in the ease of aluminium, which appears to differ somewhat from the others, the differences are not at least more than 2, or 3 per cent; and differences of this order, which are obviously acci- dental, occur in all the curves. We are forced to conclude, therefore, that the ionization curves as observed, for all metals, have the same form. This is in agreement with the results of the less accurate experiments upon gold and aluminium described in the previous paper. Whether or not the curves so obtained really represent the ionization of the metals is by no means certain. As has been said, the fact that there is no change in the form of the curve, when its magnitude is con- siderably decreased by the removal of a surface film of gas from the metal, makes it not improbable that the whole effect may be due to such a film, The probability of this explana- tion is increased by the fact that the ionization curves of gases and vapors do vary considerably; and it seems, therefore, unlikely that metals, so different in all their properties as those used above, should show such complete similarity in this respect. § 3. Investigation of an Absorbable Radiation accompanying Alpha Rays. When the last thin aluminium foil is removed from the path of the a-rays the number of electrons leaving the metal plate is greatly increased; this is shown by the first line of Table III, §2. There appears to be, therefore, a very absorbable or “soft” radiation emitted by the polonium, which is completely stopped by 0°64 x 10-*™ of aluminium. To obtain an idea of the nature of this radiation some experiments were made, in which we used the apparatus described in the previous paper.* It is substantially the same as that shown in fig. 1 of the last section, except tor slight differences in dimensions, and the fact that there is only one dise to carry the aluminium foils instead of two. A much stronger preparation of polonium, which we owe to the kindness of Professor Boltwood, permitted the use of an electrometer instead of an electroscope. A brass plate was used as the source of 6-rays, and the negative current from it was measured, with one thin foil inter posed and with none, when various positive potentials were applied to the case rang- ing from 40 to 1000 volts. The results of these measurements are shown in fig. 4, where the abscissee are the potentials ap- * Phil. Mag., xxii, 917, 1911; this Journal, xxxii, 413, 1911. Metals under the Influence of Alpha Rays. 321 plied to the case and the ordinates of Curves I and III are, respectively, the currents observed with one foil, and with no foils, interposed. The difference between the ordinates of the two curves shows the effect of the assumed soft radiation. This Fic. 4. 1,00 - - : 000 200 400 600 800 VOLTS 1 effect is greatly diminished as higher positive potentials are applied to the case, so that at 1000 volts it is reduced to about one-fourth of its value at 40 volts. This indicates that the radiation consists of electrons with much higher velocities than those attributed to the d-rays, and with a wide range of 322 Bumstead and McGougan—Emission of Electrons by velocities. As the positive potential on the case is increased, more and more of these electrons are withheld from reaching the brass plate. The fact that this radiation communicates a positive charge to the plate is not an obstacle to the hypoth- esis that it consists of electrons. Since the early experiments of Lenard it has been known that electrons, moving with velocities corresponding to several hundred volts, when they . fall upon a metal, cause the latter to emit secondary electrons, which may carry a larger negative charge away from the metal than it receives from the incident stream.* It seemed possible that the diminution in the current with an increasing field might be due to an effect upon the electrons emitted by the plate rather than upon the radiation itself—for example, by increasing the reflection, or the number of secondary electrons from the case. In order to test this possibility the brass plate (corresponding to E in fig. 3) was enclosed in a tin box, whose top was made of wire gauze in order to permit the a-rays and the new radiation to reach the plate. It was insulated from the case and could be charged by means of an external electrode. The box was kept charged to 40 volts, while positive potentials up to 1000 volts were applied to the case ; in this way the field in the vicinity of the plate, E, remained practically constant while the soft radiation had to pass through a variable field. The results were not essentially different from those shown in fig. 4. “In order to determine whether or not this soft radiation was peculiar to polonium, experiments were made in which the active deposit of thorium (obtained from a preparation of meso- thorium) was used. It was much less active than the polonium and more time was necessary for each reading; on account of the decay of the activity it was not practicable to wait until the changes due to the removal of the gas layer had ceased, before beginning the readings. It was thus impossible to obtain as satisfactory numerical results as with the polonium, but there could be no doubt about the existence of the soft radiation. It produced a greater effect, in proportion to that due to the a-rays, than in the case of polonium ; the ratio was about twice as great. On the other hand, the diminution pro- duced by an opposing electrical field was not as great as with the polonium ; with 940 volts on the case, the effect of the soft radiation was about one-half as great as with 80 volts. A sim- ilar change of potential with the polonium reduced the effect to one-third. * See also Gehrts (Ann. d. Phys., xxxvi, 1001, 1911), where it is shown that, when electrons with a velocity corresponding to 200 volts fall upon a copper plate, the secondary electrons carry away from the plate more than twice the charge brought to it by the incident electrons. A | Metals under the Influence of Alpha Rays. 323 Returning to fig. 4, it will be observed that Curve I, which was supposed to be due to the a-rays alone, shows a diminution of the current as the potential on the case is increased, which is similar to that of Curve III but much less in amount. This was at first difficult to explain; it is quite evident that the fields used were too small to accelerate the a-rays sufficiently to cause an appreciable decrease in the 6-ray current. The fact that the soft radiation accompanied the a-rays in the thorium deposit as well as in the polonium suggested that it might be a secondary effect; if this were so, then the currents plotted in Curve I (fig. 4) would be due not to the a-rays alone, but there would be a small admixture of the assumed secondary rays from the lower side of the aluminium foil on the disc.* A small cirele of the thicker aluminium foil, of the same diameter as the opening in the brass cylinder, C (fig. 1), was pushed up against the polonium. In this way any soft radia- tion coming directly from the polonium would be stopped ; but a secondary radiation due to the impact of the a-particles on the inner walls of the brass cylinder, C, would not be stopped unless a foil were interposed below the cylinder by means of the wheel. The experiment was made as before by applying various positive potentials to the case and taking alternate readings with and without the foil below the cylinder. The results are shown in Curve II of fig. 4. The sensitive- ness of the electrometer had changed by about 5 per cent since the experiments with the uncovered polonium ; the values of all the currents were reduced in the same ratio so that the measurements with the interposed foil should agree. In Curve I, the crosses represent the measurements when the polonium was covered, the circles those obtained when it was not covered, a thin foil being between the cylinder and elec- © trode in both cases. A comparison of Curves III and II shows that the direct radiation contains a component which is much less affected by the retarding field than the secondary radiation alone. This agrees with the results of Wertenstein (1. ¢.), who, however, worked with a magnetic instead of an electric field. If we assume that the soft radiation is made up of two portions, one coming directly from the polonium, and not retarded by the field, while the other is secondary and con- sists of electrons, we may show by a simple calculation that the experimental results are accounted for in a very satisfac- tory manner. * At this stage of our experiments the very interesting paper of Werten- stein (Le Radium, ix, p. 6, 1912) came to hand. By measuring the ionization due to RaC in gases at low pressures, he has demonstrated the existence of two soft radiations : one is secondary, and deviable in a magnetic field, and doubtless consists of electrons, while the other is not appreciably deflected in a magnetic field of 1100 units. The remainder of our work was done with a knowledge of Wertenstein’s results, and the next experiment was directly suggested by his paper. a. ee os < 324 Bumstead and MceGougan—Emission of Electrons by Let @ be the current of 6-electrons leaving the electrode due to the pencil of a-rays which strike it; this is present in all the curves. Let > be the current due to a soft radiation from the polonium which is unaffected by the field; this is present in III but absent in Iand II. Lets be the current due to a secondary radiation (consisting of electrons) when there is no obstacle between the polonium and the electrode (Curve III); this will vary with the electric field. The secondary radiation which produces s in due to nearly all the a-particles liberated by the polonium, through the complete solid angle, 4, but the secondary rays from the plug which carries the polonium and from the deeper parts of the cylinder are cut down by the limited aperture. When the polonium is covered, the secondary rays are due to the a-par- ticles which emerge through a solid angle approximately equal to 27r; their effect will be equalto ms wherem<1. (Curve II.) Finally when the thin foil is interposed below the cylinder, the secondary rays from the lower side of the foil will be due to the a-rays which get out of the cylinder and pass through the foil; the solid angle in this case is about 0-147, but the beam of secondary rays is not limited by any diaphragm. We may write 2s, (x = “AY AVAYAVAY anata "ASTANA AY AVAVAYAY AY ANAYAYAYAYAVAYAYATAYATAVATAVAY VAT ATAVANAY FATANAYAY TAY AVAAYAVAYAYAVAYAVAYAYAYAVAV SY we POROONK een PERO 6A ATAVATAYATAYAYANAVAYAVAYAVLVATATAYATAYA ESA AT LAS sVATAWAVAVAVAVAVAVAVAVAVAY aa POP DACAMAMAVAMad = Wattatat aTuNATAU CtaMtANPAN GTA PSR Seen eaten PY ANAY STANAAV RATAVATAO YANN EERE Waren ie 100% Fic. 5. solutions of varying composition. per cent solutions. All the solutions stant NT ATAYAAYANAVAY, NAVAN APATAYA a} cVAYAYARAYAVAVAVAYAN? 1ANAYATAY A SaTAVAVAVAY VAN fn ERP R RRR ERAN RETIRE ON SAT ATANAVAYACATAVAY MAVAPATATAVAT ATATAY £7 -TANATANA ATA TATA SSS I ATAY ANA LTA ATANAVATAY STAUAVATAVAT AMAA) LAPUA ANTAL AAO AT ATANATACAN Sa TLNAAAWAYAY WWAVAYAVAVAVAVAVAYAY AVAVAVATAVA 7AM aaa aaa at Ararararatararar sa ANCA ata ATANAI STAY AAAI A ATA SOSA aVAPAN ANSP ANATANAAOAVAOAUAYAVAVAPAY A" OYA AOAN YAMATO Ba AOSAAOOA CELL Y RATAN = TAMAS oa “ATTATAS ANATAYAYANA aS AYAYATAYAUANATAYATAYA AANA ATAU AVA AANA TANANATATATA FAVAVAVETAtATAAVAVANAAV ATAVAVAVATAVAVATAOA AtaTAtat aS ATaT AVATAATATATATATATAN TAK VAVAYAYAYAVAYs Xs AAA TAVAVAN AVAVA. ©. /AN0A¥A¥ SVAYAY AAV AVAY AYAYAVAVAVANAVAVATA? .VAVAVATAAVAY AVA AYA AVAYAYAVAVAY AD RA NAATAY TAYATATA LS a5 Beers Ta a TAWAa ANAT aN, YAY AYAVAYAVAVAVAY ATATAVAYAYAVAAVAVAS ma PSAVAYs Way CTATANAN A A ANAVATA AANA AVA AVAVAVAVATA LAV ATAVAVAVAYAVATAPAYAY A ATAYAPATATAN YAY AVAY AVAYAVANAYAY AVAV AVANAVAY AY => SG AANA A AAA ATATAVAYAVAYAY LWAVATAVANAVACAVANA AVAVAVAYAVATATAPAVAV AVAAVAVAYATAMAA SAV raya Sy ASANAVAYATAY AVANAVAYAVAYAVAYAYAY AVAYAY 27 YAY ATAVAYAVAYANAY AVAAVATANATAY ACA A ara PAPAY STANT ARASAVATAPAVAY ACLU ATAT ATAU OX LOGS bee OOOO OOOT OOK PAYAVAYAVATAVAYAVA ee TODO OOOO OOOO) OY AVAVAVAYAWAY, AvATAVAY UAYANAVAVATAYA VAYAY LY AVATAVANAVATAVAVAY STAVATATAYAAVANS Terie SYA YAY ATAYAYAY AY AVATANA YAY VAY AVA AV AYATAVAYAY ATAVAYAYAYAY AV aT AYAYAYAYAVAVAVAYAYAVAYAYATANAVAVAY =! TRAMANAYAYAN i IPA VAVARAVAATAVAVAVAV .TAVAVATAVAMA EEX ITT ARK IVATATAVAYAYAYAY AVAVAVAVATAVAVAVAVAV AV AVATAVAVAYAVAYAY refers only to 10 aa? AVATATAT IN AVAVAVAYAYAY, TAVAVAVAY AYAVATANAN NVANAYLYAY , \\WAYAY a ATAVATAVAYAVANAYAVAYA TAVAVACAYAVAVAVANAYAYA vara LAS SV AVAVAVAVATAY ATAY AVAWAW “a aAVAMAS TAVAY SYANAN AY ANAVAVAY ANAT AVAYAVAVANAVAVAVAYAY AV AYA PAY AY AVATANAVAYAVAY AVAY AY AVAYAY AVAVAYAYAYAYAYAY AY AYA’ NaS.0Os. 5H.O 10% The action of sodium thiosulphate on mercuric chloride in plotted contain a con- quantity of sodium chloride (2 g. per 100°), it having been found that the decomposition of sodium mercuric chloride proceeds much more slowly than that of mercuric chloride and better crystals appear to be obtained by the use of it. 1. All solutions lying in the field above the line L,I precipi- tate white mercuric chlorosulphide, HgCl,.2H¢S8. It ei be observed that the line I,II ineludes all elven in which the ratio of HgCl, to Na,8,0, is as 38 mol 2 mol while all solu- Pm, Cadmium, and Mercury. 371 tions in the field alluded to contain more mercuric chloride than this ratio. The chemical reaction which will be proved below is: ge +3HeCl,+2Na,$,0,—HeCl, . 2H¢8+4NaCl+2H,S80, All solutions lying in the field included between I,II and LItt precipitate mixtures of HgOl,.2HgS and black Hes, apparently amorphous. All solutions on the line LI contain Hel, and Na,§,O, in the ratio 1 mol: 1 mol. The chemical reaction is H, 0+ Na, 8,0, + HgCl, = 2NaCl + H,SO,+ HgsS (black). 3. All solutions lying between the lines IIIT and I,IV precipi- tate, i the beginning, black HgS, though later a red precipi- tate may be obtained. 4, The line I,1V ineludes all solutions in which the ratio of HeCl, to Na,S, O, is 1 mol: 4 mol. The principal reaction is HeCl, +: 4Na,8,0, = 2NaCl + 3Na,SO, +485 + HgS. The mercuric sulphide precipitated here is, in the beginning, a deep red modification which the microscope shows is not cinnabar but anew form, appearing transparent and orange-colored under the microscope and having lower indices of refraction than cinnabar. It should be stated here that the precipitation of this phase continues until the dilution of the supernatant liquid has fallen to about 1 per cent of mercuric chloride, when the black form in erystalline condition begins to come down with it. Proof of the above reactions.—1. The reaction 3H¢O0), + 2Na,S,0, + 2H,0 = HeCl, . 2HeS + 4NaCl + 2H, SO,. A solution ‘containing 20 g. NaCl, 1:64 ¢. Hel, and 1-000 2. Na,S,O, . 5H,O in 100 to 200% water was boiled for a short time and filtered. No gas escaped during the process. The curdy yellowish precipitate was carefully ~ washed, and the fil- trate and washings were diluted to 250°. This solution was free from mercury, gave a strong acid reaction with litmus and a decided reaction for sulphate with barium chloride. 02011 g. pure dry sodium carbonate required of this solution 120°6° for neutralization, using methyl orange as an indicator. 0°386 g. H,SO, found. 0:394 g. H,SO, cal. from the above equation.* In a second experiment 20 g. NaCl, 1°65 g. HgCl, and 1:000 g. Na,S,O,.5H,O in about 200% water were boiled and filtered as before.+ The filtrate was diluted to 500°. It contained no mercury, no thiosulphate and was strongly acid as before. No sulphur dioxide was evolved during the reaction. 0°3006 g. Na,CO, required for neutralization 348-5° solution, using as indicator phenol phthalein at boiling temperature. Total H »O, found 0°3991. Cal. from equation 3955. Thor- BEC be remembered that the thiosulphate has the formula Na2S20s. {It may be the slight excess of mercury was absorbed by the precipitate. 372 Allen and Crenshaw—Sulphides of Zinc, oughly washed precipitates made in the above way were dried and analyzed. In all of them a small excess only of mercuric chloride was used, since a large excess gave finely divided pre- cipitates dificult to wash. Found. Cal: for HgCly.2 HgS. al 2. 3. Hg 81°51 a ee 81°48 Cl eft bes 9°37 =e 9°41 H,O "44 ae “48 ae 2. The reaction H,O + HeCl, + Na,S,O,=2NaCl + HeS + H,SO,. 1:000 g. HgCl, and 2:0 g NaCl were dissolved in water and added to a solution of 0°920 g. Na,S,O,.5H,O, the total volume after all was added being 150°. A curdy yellow-pre- cipitate was obtained in the cold, probably mercuric thiosul- phate, which on boiling turned black. A careful test of the precipitate for chlorine revealed not more than a trace. In a second experiment a solution containing 1:10 g. HgCl,, 2:0 g. NaCl and 1-000 g. Na,8,0,.5H,O in 150° water was boiled for a short time and filtered. The filtrate and washings were diluted to 250°. The resulting solution contained no mercury but was strongly acid. 02024 g. pure dry Na,CO, required 119-4° of solution for neutralization, using as indicator phenol phthalein at boiling temperature. Total H,SO, found =~ 03921. Cal. from the equation 0°3951. Ina duplicate experi- ment a solution identical with the first was prepared, washed and filtered. The filtrate and washings were diluted to 250° as before. 0°2029 g. pure dry Na,CO, required 119°5°° solution for complete neutralization. Total H,SO, found = 0°3927. Cal. from the equation 0°3951. 3. The reaction HgCl, + 4Na,S,0, =HeS + 48 + 2NaCl + 3Na,SO,. This is the principal reaction at 100° in sealed tubes when the thiosulphate is present in sufficient excess. The solution is generally, however, slightly acid, and when it is boiled in an open vessel considerable sulphur dioxide is evolved. All the mercury is precipitated, and in two distinct layers. The bottom layer, which of course, precipitated first, is bright red and consists of rhe new hexagonal mercurie sulphide which we will call 6’-HgS;* the upper layer consists of a mixture of this form with metacinnabar, a’-HgS. The total preci- pitate when mixed has a puce color almost exactly like lead peroxide. It contains the a’-HgS and #’-HgS in approxi- mately equal quantities, as was estimated by matching the color with various mixtures of the pure substances. *We will use the symbol o to designate the stable form and for mono- tropic forms the symbols a’, 3’ to distinguish them from enantiotropic forms, for which the unaccented letters are commonly used. Cadmium, and Mercury. 373 As evidence of the above reaction we give the following data: TABLE X, Action of HgCl. on a large excess of NaeS.03. NagS2Os. NaS.20s. NaeS20s. HegCl, 5H,0) NaCl | H.O 5H.O0 5H.O|)/HgS+S/HgS+S} HgS | Hgs taken| taken |taken| taken} used Cal. found Cal. | found | Cal. 1.)1:000g¢| 5:000g | 20g | Tae 3°66 3°66 1°32 1:32 86 8d 2./1:000g|} 5°000g | 20g | Tre 3°56 3°66 1:29 1°32 82 85 The data were obtained in the manner described on p. 352, i. e., the excess of thiosulphate was obtained by titrating an aliquot part of the filtrate with iodine; the total weight of the precipitate, washed with alcohol and dried, was then taken, and finally the weight of the sulphide was found after extract - ing the precipitate with carbon disulphide, washing out the excess with ether, and drying. Metacinnabar, a'-HgS. The metacinnabar of nature, when pure, is entirely black and crystalline and is generally re- garded as an isometric mineral, though crystallographic data are meager. We have obtained black crystalline mercuric sulphide by only one method, viz., by the action of an excess of sodium thiosulphate on sodium mercuric chloride in dilute solution. By reference to the previous pages (pp. 371-372) it will be seen that moderately concentrated solutions (10Z%—257) of total salts containing the substances in the ratio HgCl,:4Na,S,O, give first a red precipitate, which continues to form until the con- centration of the mercuric chloride is approximately 1 per cent, when the black crystalline sulphide separates with the red. The solutions a, 4, ¢, d, e (fig. 5) precipitated a mixture of the two torms from the outset. The formation of the red modifi- cation may be entirely inhibited, however, by the addition of sulphuric acid to the solution before boiling. experiments where the concentration of the mercuric chloride remained constant, 1g. : 700° H,O, and the thiosulphate varied from 2g. to 20 g., the addition of 4 drops 80% H,SO, entirely prevented the formation of the transparent red sulphide. This was proved by microscopic examination of the products. On the other hand, if the acid was omitted, all the other conditions remaining constant, a large quantity of the red form (8’-Hgs) For the preparation of the black a’-HgS in was intermixed. In six different Am. Jour. Sci.—FourtTH SERIES, VOL. XXXIV, No. 202.—Ocrosmr, 1912. 25 ; 374 Allen and Crenshaw—Sulphides of Zine, portions large enough fora specific gravity determination, we used the following solution: 6 o, HeCl, ;60e, NaCl; 12 ¢. Na,§,0, . 5H,O in 41. of water. 10" 30 per cent sulphuric acid was added and the solution boiled till precipitation was complete. More than a proportional quantity of acid was thus required for the larger volume of solution. The precipitate, after thorough Ww ashing, was dried with alcohol and ether; then the sulphur was removed by repeated digestion with carbon disulphide, washed with ether, and dried by heating in vacuo to 250°. It contained only a trace of occluded salt. A weighed portion dissolved in aqua regia and changed to chloride with excess of hydrochloric acid was precipitated with hydrogen sulphide. The filtrate was evaporated to dryness and finally heated to redness in a tarred platinum dish and weighed. Three-tenths of a milligram of sodium sulphate in one gram substance = 0°03 per cent was thus obtained. Of course this does not prove the original form of the sodium to have been sulphate, though it probably was. The sulphide was tested for water by heating i in an atmosphere of carbon dioxide, till it was largely volatilized, the water being absorbed by calcium chloride. Of course precautions were ‘taken to guard against driving the sulphide vapor over into the absorption tube, and when the apparatus was cold, the carbon dioxide was removed by dry air. No water was found. Under the microscope the product was apparently all crystalline, though it was impossible to say that there was no intermixture of amorphous sulphide. The slow precipitation of mercuric sulphate or chloride in strongly acid solution by hydrogen sulphide was also tried for the preparation of this form. The device referred to on p. 355 was used for the purpose. The products obtained in this way showed no crystal faces, and being opaque it was impossible to say whether they were crystalline or not. Similar products were also obtained by the action of sulphur and sulphuric acid on metallic mercury in closed tubes containing an atmosphere of hydrogen sulphide. At high temperatures, 200° and 300°, respectively, some cinnabar was obtained by both the last two methods. Properties of a'-HgS. The opacity of the black mercuric sulphide made optical tests on it impossible. The attempt to prepare measurable crystals of it was futile and therefore our product could not be certainly identified with metacinnabar. | The following facts, however, make the identity highly prob- able. Both minerals are entirely black when pure, even in a powdered condition. Natural metacinnabar is regarded as tisometric ; the laboratory product was obtained from thiosul- phate solutions in the form of bars crossed at right angles like the principal axes of a cube, suggesting skeleton crystals of the Cadmium, and Mercury. 375 regular system. The specific gravities of several products made in this way at 25° were, I II Ill IV Vv Average 7592 7°586 7°642 77588 7°568 7°60 The variation in these determinations is, perhaps, to be accounted for by the presence of some amorphous sulphide in varying quantity. The microscope was unable to decide this oint. The best determinations of the specific gravity of the nat- ural mineral are close to 7°'7. G.E. Moore,* who first described and named the mineral, said that the specific gravity varied from 7-701 to 7-748 “Sowing to intermixed cinnabar.” Genth and Penfield,+ who obtained the number 7:706 on a specimen of metacinnabar from San Joaquin, Orange County, California, say: “Color iron-black, but many pieces show already a change into ordinary cinnabarite both by a good lens and the reddish black powder which some of the particles yield on pulverizing.” It is probable, therefore, that these values are somewhat high ; on the other hand, it is possible that the values obtained on the laboratory product are somewhat low, since it was impossible to prove that the latter was free from amorphous material. True, Moore ealls the mineral amorphous, but no other investigator confirms him. On heating alone or more rapidly with ammo- nium sulphide or dilute sulphuric acid, the metacinnabar passes over into cinnabar. This is true for both the natural and the synthetic mineral (see p. 377). 8’-HgS. It has already been stated that the precipitation of solutions of sodium mercuric chloride by sodium thiosul- phate in the proportion HgCl,:4Na,S,O, yields a beautiful red sulphide differing from cinnabar. If the product is to be free from a’-Hg§, it should be filtered before precipitation is com- plete. For this reason it is wise to begin with pretty concen- trated solutions, 10 to 25 per cent of total salts, and interrupt the operation while the product is still bright red. It has been found by actual test that the addition of a fraction of 1 per cent of black HgS to the first precipitate which forms (the two being thoroughly mixed in a dry condition), can read- ily be detected by the eye. A microscopic examination in this case was futile. Judging by the color test, the following con- ditions gave a product about 99°5 percent pure. Two solutions of 400° each were prepared ; one containing 40 g. NaCl and 344 g. HgCl,, the other 125°6 g. Na,S,0,5H,O. The solutions should be filtered if not perfectly clear, and the sodium mercuric chloride solution is then added gradually, with * This Journal, iii, 36, 1872. + Ibid., xliv, 383, 1892. 376 Allen and Crenshaw—Sulphides of Zine, constant mixing, to the thiosulphate solution. The mixed solu- tion should not contain a trace of precipitate. The product is now brought as rapidly as possible to boiling. After about 12 minntes the precipitate is rapidly filtered on a hot water filter and washed thoroughly with hot water. The temperature of the solution should not be allowed to fall materially during the process. In that case a further precipitate containing some black HgS is formed, as may be seen by the color. When the filtrate has reached room temperature, the further precipitation becomes very slow. If now the supernatant liquid is raised again to boiling, the 8’-HgS is again formed. After the pr ecipitate has been dried by alcohol and ether, the free sulphur is removed by carbon disulphide. That the substance is really a sulphide of mercury will be seen from the following analysis : Hgs = 98:49 Clery = 10 IME SKO} = 3} lel iw) = 1:14 99°96 A weighed portion of the substance was dissolved in caustic alkali, and the solution treated for a short time with hydrogen sulphide. A slight excess of dilute sulphuric acid was then added, and finally a slight excess of ammonia. The mercuric sulphide was then collected, freed from sulphur, and weighed according to Treadwell.* The water and sodium sulphate were determined by the methods described for the a’-compound. The chlorine was doubtless present as sodium chloride, and the remainder of the sodium was probably in the form of sodium sulphate. If we calculate the chlorine and the sodium in this way it would change the above results only a trifle. The greater quantity of impurity found here than in the a’-HgS is due, no doubt, to the higher concentration of salts in the solution which one must usein order to obtain the 8’-form. ‘he determinations of density and refractive indices were made on preparations which had been heated to 250° in vacuo for some hours. Determi- nations of the water and sodium as sulphate in two of them were as follows: I II = * Toa ae oe H,0. Na,SOx4 HO. Na2SO, 0°37 & 0°46 0°49 ee 0°33% These powders were apparently homogeneous as, observed by the microscope, but they were quite fine and somewhat vari- * Quantitative Analysis, translation by Hall, p. 139. Cadmium, and Mercury. 377 able in specific gravity. The following are the determinations of this property at 25°: I I Til IV Average 7°221 7°215 7179 7199 7°20 The refractive indices in lithium light were found to be 2°58 and 2°82 (Merwin). There can be no doubt that this is a new form of mercuric sulphide with properties quite different from Fie. 6. i 605° ca feats tales ea all : | | ; o D or ° Be Temperature. o1 for) or ° 545° Fie. 6. Sublimation curves of o-HgS. cinnabar, though the color of the powder is not to be distin- guished from ordinary vermilion. Relation of the mercuric sulphides to one another. Like zine and cadmium sulphides, mercuric sulphide volatil- izes without melting at atmospheric pressure. The pressure reaches one atmosphere at about 580°, as may be seen from the curves in fig. 6. In these experiments the sulphide was heated in an atmosphere of hydrogen sulphide. Cinnabar is the stable form over the whole temperature range up to the volatilization point. That the other two forms are both monotropic is proved by the following facts. At 100° both a’-HgS and 8'-HgS are changed into cinnabar with ammonium sulphide solution, while at 200° the same change takes place very slowly in sealed tubes with 380 per cent sulphuric acid. At 400°, 450°, 500°, and 550°, both are transformed into cinnabar when heated alone in evacuated glass tubes. The heating was done in a stirred nitrate bath. Table XI shows the results. Allen and Orenshaw—Sulphides of Zinc, | reqeuutd 0} pesuvyo poe ea Er x iy 266 WLYY eLO] » P& o0LG otis sre ~— a requuuta 0} Ieqeuuts 0} : | pesueyo | pestueyo | 266 WEY ELON ” PE | .00€ |%66 UeGA eXo}y ” ve | .00¢ | S8H-»2, > ” Pe | .008 Ieqeuuld Yeg + | S5H-» 2CT | JO 91N4XIt v JO | IQeUuLd 07 [Me 1o]Oo 04} PYH » pg | .0c7 | Pesavyo 246 » Pe | Sh |SSH- P21 ynoqe ” pg | .097 : / | reqeuut eg + | Ssy p wce IeqeuuLd of JO 9IN} XIU B JO) Onova UL pesurya sor ONOvRA UT ONOvA UL | IO[OO ay} pep | ouoCTR paywey=] | pe | OOF |-}Ueo1ed j[eus; euo[e pozvopT | pe .00F SsH- ps] > | ouoje poyweyy | pS | OOP / | TeqvuUpo oF | pesneyo %OT 1OS*H Aa. Paha a as, qnoge 208 WFLA PEP | 002 eas es eras Rake | | | Ieqraur 04 | OSH IBqeauL 07 esuRvyo MOS BOS YA ie 2006 pesuryo 266 ” PG 0006 eee Seiad a ae IeqeRauto 04 | pesuvyo | IVqBUULd 09 ie Apusawddy ¢7("HN) WM | po | OOT | pedueyo te | SAN) yy | ~~ 001 pesueqoun | SPHN) WH | pt | 001 q[nsery SUOlIpuo. § lem] eange 4 Useyy, | SUOTAIPUOD = /oumty,| oange 4[NSeyy Suontpaoo lca eInye | 18440 -1edui J, | TOTO -1odutay, T0440 | -1ad wa J, S3H-¢ IWQRUULOBIOIY-Gop{—/p IeQVUULN-Ss-9 ‘Oplyd[ns olmMo1em JO TOIAVYE [euULeyy,, [XK AAV], Cadmium, and Mercury. 379 The thermal behavior of these substances is confusing. In the first place, cinnabar turns black as the temperature rises. This is not due to any transformation, but merely to a vari- ation in the absorption of light with changing temperature, for after it has been heated to 325° for many hours, cinnabar quickly regains its color on cooling. If, however, the temper- ature is carried to 445°, or perhaps to a lower point, the color remains permanently black. This was first inter preted to mean atransformation into metacinnabar, but a microscopic examina- tion failed to disclose anything but ‘cinnabar, while by grinding in a mortar the black color was found to be due merely to a thin coating. By matching the tint of a ground sample of heated cinnabar with a oround mixture of the red and black sulphides, the former was found to contain 1 per cent or less of the black sulphide. The same final product was obtained whatever form of mercuric sulphide was heated, provided the temperature was as high as 500° and no further change was found on heating to 550°. This is difficult to explain satisfac- torily. Possibly the coating is due to a condensation of the va- por to the black form. The volume of the tubes was about 05: and the weight of the sulphide taken was half a gram. On the assumption ‘that the tubes were filled with undissociated vapor | _ 116 x“ 09" at one atmosphere, they would have contained ——,—— =5:2"8 of sulphide vapor, and if this were all condensed on the sur- face of the sulphide, the mixture would have contained 1 per cent of black sulphide. Apparently little or none condensed on the glass at500°. Even if the vapor pressure were only half an atmosphere, the results would still be of the right order of magnitude. The results on the 8’-HeS, while conclusive, regard- ing the relation between it and cinnabar, do not settle the rel- ative stability of the two unstable forms. With ammonium sulphide the 8’-HgS appears to be transformed more slowly than the a’-form, though it is very difficult to detect small grains of cinnabar in a matrix of the former. Again, if we were to judge from the behavior of the 8’-HgS when heated in evacu- ated tubes we might conclude that it passed through the a’-form in its transformation to cinnabar. (See Table XI.) This would make the 6’-form the least stable which might be expected from the fact that it is not found in nature. However, the apparently large percentage of the black sulphide in the ’ -Hes during the early stages of its transtormation, may be nothing inore than a surface coating on the numerous little grains, which in grinding still remain black because they escape ‘the cr ishing action of the pestle. At the end of the transformation the B’-HeS has always changed into comparatively large cinnabar crystals in which, of course, the surface is oreatly reduced. 380 Allen and Crenshaw—Sulphides of Zine, The black coating should be thicker here if our explanation is correct, but the orinding process might well give a powder of a redder hue. We found, in accord with this supposition, that avery pure vermilion powder (unsublimed cinnabar) also gave, on heating, a product which was considerably darker after it was ground than a product formed by heating coarser cinnabar. Whether our explanation be correct or not, it remains cer- tain that metacinnabar and the new §’-HgS are both mono- tropic forms, while cinnabar is stable. Genetic conditions of the natural mercuric sulphides.— The evidence for the geologic view that cinnabar is a product of alkaline solutions is convincing. The close and constant association of the mineral with igneous rocks is signinegnt, while in two well-known localities, “Steamboat Springs, Nev.,* and Sulphur Bank, Cal., it seems to be in the process of depo- sition from alkaline waters at the present day. Posepny+ states that in the former locality pyrite is apparently forming with it. Cinnabar is thus undoubtedly a primary mineral. In some cases also it appears to be secondary, since mercurial tetrahe- drite oxidizes readily to sulphates, and the descending solution seems to be precipitated as cinnabar on other sulphides at lower levels (Lindgren). Metacinnabar, on the other hand, is regarded as a characteristic secondary mineral. It was found near the surface in the Knoxville district, Cal., and has not been found lower down. A large part of the ore in the Baker Mine and in the upper levels of the Reddington Mine were metacinnabar.t Melville§ describes an occurrence of meta- cinnabar in the New Almaden mines, Santa Clara Co., Call., where cinnabar and quartz are intimately mixed, while meta- cinnabar is crystallized on the quartz and is “certainly subse- quent to it.” The metacinnabar of Idria, according to Schrauf, || is far younger than the cinnabar which underlies it and has appar- ently been formed since the opening of the mines. It occurs here in hemispherical crystal aggregates which suggest to Schrauf that they may have formed by the action of hydrogen sulphide on the globules of metallic mercury which invariably accompany the cinnabar. Schrauf refers here to the experi- ments of Fleck, which lead him to believe that the metacinna- bar formed in the presence of sulphuric acid. It is a note- worthy fact that raetacinnabar is commonly associated with marcasite, at least in this country.4 * W. P. Blake, this Journal (2), xvii, 488, 1854. + Trans. Min. Eng., p. 228, 1893. { Becker, U. S. G. S. Monograph 13, p. 284. § This Journal, xl, 293, 1890. || Ueber Metacinnabarite von Idria und dessen Paragenesis. Jahrb. der k. k. geolog. Reichs., xli, p. 379-399, 1891. N. J. Min. 1893, I, referate 465. “| Becker, loc. cit., p. 385, Penfield, this Journal, xxix, 452, 1885. Cadmium, and Mercury. 381 Let us see how these field observations agree with the facts worked out in the laboratory. First, it was found that cinna- bar very readily forms at moderately low temperatures (80°- 100°) by the action of dilute alkali sulphide solutions on mer- euric salts. Ippen* obtained good crystals at 45° in two months. Never under any conditions have we observed meta- cinnabar as a product of alkaline solutions. Black crystalline mercuric sulphide was obtained by the slow precipitation of mercuric salts in acid solutions by the action of soluble thiosulphate. The crystals were not measurable and being opaque could not be positively identified. However there were some indications that they were regular and the identity of the two appears probable. The association of natural metacinnabar with marcasite in the deposits of the western United States is important in this connection because both are regarded by geologists as characteristic secondary minerals, 1. e., they were precipitated directly or indirectly by the action of hydrogen sulphide from solutions which had been formed by the oxidation of sulphides near the surface. Such waters would of course contain the metals in the form of sulphates, and also free sulphuric acid if pyrite or any other sulphide of that chemical type were one of the minerals oxidized. Marcasite has already been formed synthetically from such a solution. + Chemical and geological relations of deep-seated and surface qwaters. The difference in chemical character between ‘ ascending” and “descending” natural waters is well known. As previ- ously stated, our knowledge of the composition of hot springs, and the chemical behavior of common minerals with hot water, lead to the conclusion that the former class must be generally alkaline. Chlorides, bicarbonates and sulphides, especially of the alkali metals, are the characteristic constituents. On the other hand, surface waters in the vicinity of sulphides naturally contain sulphates as oxidation products and are generally acid on account of the frequency of pyrite and marcasite. Now it is a remarkable fact that the crystalline form of a number of minerals is determined by the acid or alkaline nature of the solutions from which they erystallize. We find three pairs of minerals, pyrite and marcasite (FeS,), sphalerite and wurtzite (ZnS), and cinnabar and metacinnabar (HegS), one member of which erystallizes from acid solutions only, the other member from alkaline solutions as well. Furthermore it is * Tscher. Mitt., xiv, 114, 1895. } Allen, Crenshaw and Johnston, this Journal, xxxiii, 179, 1912. 382 Allen and Crenshaw—Sulphides of Zine, always the stable form which comes out of the alkaline solu- tions, while the unstable is obtained only from the acid solutions. The alkaline solutions never give rise to any other than the stable forms, while the acid solutions may give rise to either stable or unstable or both, according to conditions. This statement is subject to only one qualification, viz., our synthetic black erystalline mercuric sulphide was not positively identified as metacinnabar though it was probably that. We find in chemical literature one well attested instance of the same rule. Schoch* finds two crystalline oxychlorides of mercury of the composition HgCl,.2HeO. One of these is red and changes readily into the other, which is black by the action of solutions of alkali carbonate or chloride. The latter form is therefore the stable one, at any rate at low temperatures. The conditions for the formation of these two bodies are essentially identical, except that for the preparation of the red form (unstable) the solution must be slightly aeéd while for the black form (stable) the solution must be slightly alkaline. The geologic evidence so far as it is at hand seems to agree remarkably well with the above facts. Sphalerite, pyrite and cinnabar are primary minerals; marecasite, metacinnabar and apparently wurtzite are characteristic secondary minerals. As explained above, pyrite, sphalerite and cinnabar may be formed from acid solutions by hydrogen sulphide under certain condi- tions, consequently we are not surprised to find that they may be secondary as well as primary. There is one fact concerning the occurrence of marcasite and metacinnabar which should be mentioned. Both are sometimes associated with calcite. Whether or not they are paragenetic is a doubtful question. (Lindgren.) If the calcite is not subsequent to the sulphides one would be led to suspect that the original solution: must have contained bicarbonates and sulphates. The synthetic work would lead us to expect pyrite from such a solution ; still the quantity of free acid required for pure mareasite is very small at low temperatures and it may be that the reaction FeSO,+H,S+S=FeS,+H,SO, would produce marcasite when the initial concentration of acid was no greater than it is in calcium bicarbonate solutions. It would probably be difficult to verify or refute this by experiment. We are usually obliged to work in the laboratory at higher temperatures and in more concentrated solutions than nature does, in order to get any erystals at all in our limited periods of time. It will be interesting to see whether other similar cases of polymorphism (or isomerism) exist where the crystal form is determined by the chemical composition of the solution. * Am. Chem. Jour., xxix, 335, 1903. ) Cadmium, and Mercury. 383 It may also well be that the acidity or alkalinity of solutions will be found to determine the composition of minerals in cer- tain cases. Thus chemical experience shows that the mineral chalcopyrite CuFeS, can be prepared synthetically out of alka- line solutions, while it is a matter of common knowledge that copper is separated from iron by hydrogen sulphide in acid solutions. Geologists regard chalcopyrite in certain occurrences as a secondary mineral. . It would be premature to enter upon a chemical discussion of this question at present; the case suf- fices to illustrate our meaning. Enough work has already been done to show that the difference in chemical character between acid and alkaline solutions, therefore in general between deep- seated and surface solutions, is of vital importance in geochem- istry. IV. Microscoric Srupy. Zine Sulphide. Amorphous.—If examined immediately after being rapidly precipitated in a pulverulent form, the particles of zine sul- phide are spherical, and have diameters of -0002 to -0005™™ (2 to 5u). If the precipitation is slow and the solution is agitated, the particles increase in size, and after standing in contact in the solution may agoregate into clusters or crusts, indicating that they are semi- HRovtal In fact, the aggregates when compressed under a cover-glass flatten and break open like a stiff jelly. Though jelly- like, the globules contain very little water. ‘On several occasions such masses after being pressed have been observed to become distinctly doubly refracting (probably erystalline). When formed very slowly in acid solutions without agitation—as in the double tube method—the precipitate is: (1) partly in the form of compact, stratified crusts having a decided double refraction with a par- allel to the surface, (2) partly in globules resembling spher- ulites, and having a parallel to the surface. These crusts and spherules are hardened to the point of being gritty and brittle, yet their refractive indices* are far lower than those of the crystalline forms of zine sulphide, sphalerite and wurtzite. The results of a large number of measurements are tabulated below. eureisphalenite: 42. ane. oi == OB a 4 reo Chil Nt eaieoms ee Sk ee ey; = 2°35 Pure wurtzite | aaa Rs Wy; = 2°33 Amorphous: SILT Aol ons Resco ee My, = 218—2°25 Doubly refracting globules __ ---- My, = 218—2°24 * Measured under the microscope in mixtures of amorphous sulphur and selenium. See this Journal, xxxiv, 42, 1912. 384 Allen and Crenshaw—Sulphides of Zine, Different layers of the hardened amorphous product may have very different refractive indices, perhaps indicating dif- ferent compositions. The double refraction shown by this ainorphous material is probably caused by. the strains induced in these nonhomogeneous masses during the process of harden- ing. At temperatures above 200° these amorphous crusts and spherules soon become partly or wholly crystalline, that is, the resulting product has the refractive index, n,;, of 2°34. Sphalerite.—The isometric crystalline form of zine sulphide, sphalerite, was produced from solutions and molten salts in the following typical habits: at about 800° from molten sodium chloride it occurred in the form of distorted dodecahedrons about 0°-1™" long; at 350° from a concentrated solution of sodium sulphide, dodecahedrons and tetrahedrons 0:01"™ in di- ameter appeared ; at 200° for 11 days in a similar solution, tetra- hedrons alone appeared. The refractive index of these crystals for lithium light determined under the microscope was found to be 2°340 + 005. Out of acid solutions definite crystals were not observed, but side by side were seen, in some preparations, spherules and crusts of amorphous zine sulphide, and similarly shaped masses having no double refraction and the high refractive index. The latter were in some cases covered with minute facets. It appears that an amorphous precipitate is at first formed and that this subsequently crystallizes, probably after having hardened. Wurtzite.—The hexagonal form, wurtzite, separated from. solutions in three extreme habits: (1) Prismatic, hemimorphie crystals strongly striated across the prism faces, and reaching 0-8™" in length, were produced in the double tube in an acid solution after 2 days at about 875°. (2) Crystals, tabular par- allel to the base and modified by a different pyramid on each end, occurred with these prisms. (3) Hardened amorphous globules from acid solutions at high temperatures become transformed into coherent masses of very small doubly refract- ing grains having the mean refractive index of wurtzite. In two cases distinctly radial-fibrous forms, having the character- istics of wurtzite elongated parallel to the prism, were seen. This structure is found in the natural schalenblendes. Exper- ience has shown that these aggregates, immersed in methylene iodide and viewed in ordinary light, may appear entirely iso- tropic, but mounted in a red mixture of sulphur and selenium having about the same refractive index as the aggregates, they are distinctly doubly refracting in artificial Jight. Wurtzite produced by sublimation may appear in slender ‘needles or small stout prisms. Such erystals made from pure zinc sulphide were used for determining the refractive indices under the microscope. The values obtained are as follows: Cadmium, and Mercury. 385 @,,= 2380, €;,;=2'350, wy, =2°356, ey, —=9'378. An independent measurement of e—@ on the prisms gave ‘019 for Li-light and 020 for Na-light. Observations on natural Sphalerite and Wurtzite. Samples from more than twenty different occurrences of natural zine sulphide minerals were studied optically. With a prism of the light amber-colored sphalerite from Sonora the refractive indices for Na-light and Tl-light were accurately determined on a spectrometer. The values obtained were 2°3688 + -0001 and 2°3990 + :0002. Although this material is the purest natural sphalerite obtainable (containing only 0°22 per cent FeS) the impurity is sufficient to raise the refractive index -0006 for Na-light and ‘0007 for Tl-light (page 386). The refractive index of pure sphalerite, probably correct within +:0002, therefore, is ny, = 23682, 7», = 2°3988. The dispersion of the Sonora sphalerite was determined from measurements on two prisms, a goniometer and monochromatic illuminator being used. The illuminator was standardized by observations on the following lines: Li, Na, Tl, Sr (blue). Wave-length Refractive Index EOS SEO a Se ae en ee oe ere ree 2-517 AS Ae ae Oe ee eee RR 2°493 ANGIE La ook esis: ieee eve eee Sey eae 2°436 COGN): as cs ee oer re 2-399 OG eee Stee Je eS rae) GSO Rebs iis eye one eie a ie as 2°353 Gitlin te no meus oe 2 BAO BAO) el Regret ete 2°320 The effect of dissolved FeS upon the refraction constants of Sphalertte and Wurtzite.—Natural blendes may contain as much as thirty per cent of FeS. Ferriferous blendes are deeply colored, but in very thin flakes they are a clear orange- brown by transmitted light. The blende from Saxony con- taming 28°2 per cent of FeS absorbs light in gradually increasing amounts from the red end of the spectrum to about 450um, where it is. practically opaque. The wurtzite formed by heating this blende is somewhat lighter in color, but is not distinctly pleochroic in any part of the spectrum. Its average refractive index is very near the refractive index of the blende —was observed to be -005 less—and its double refraction about "02. The mutual optical relations of sphalerite and wurtzite ae not sensibly altered by the presence of large amounts of FeS. The absolute values of the refractive indices are greatly increased by FeS. The following table shows the character of 386 Allen and Crenshaw—Sulphides of Zine, the change for sphalerite. The same values apply to wurtzite also, within the limits of error of the determinations of refrac- tive index—probably about =::005, due to lack of homogeneity ot the blende and to observational errors. The analyses and densities given in the table are by Allen and Crenshaw. n—1 n—1 > Peer aoe Locality | FeS d ie ae Rg en Calculated for FeS. (MnS) l Li Na Li Na Sonora 0'2 | 4:°090 | 2°34 2°37 | “328 | -335 2 Spain 8°6 | 4:023 | 2°36 2°40 | °338 | 348 "442 “488 Australia| 17° 3°98 2°38 9°43 | -846 | *858 436 “471 Saxony QSeB 739385) 122895 Wee obo memo "427 “470 | Assuming that the FeS is dissolved troilite having a density of 4°78, the average refractive index of troilite calculated on the assumption of an additive relation is 3°08 for Li-light and 2 3°25 for Na-light. However, if the formula “5 x + is used in making a similar calculation, the refractive index for Li-light is 4:7. So great a discrepancy indicates that the additive relation of the refraction constants does not hold in this case, or that it is some form of FeS other than troilite that is present in the blende. The latter alternative appears more likely, for mix-crystals of troilite and sphalerite should have intermediate densities, whereas ferriferous blende has a lower density than either of these minerals. A close agree- ment between the formulas, the observed densities and refrac- tive indices is obtained by assuming that the FeS in the blendes has a density of 3°8 and an average refractive index of 28 for Li-light. Etchjfigures and anomalous double refraction.—Triangular etch-figures on a cleavage surface of sphalerite heated in ZnCl, are sketched in fig. 7, C (p. 389). Pressure at a point develops double refraction which may become permanent. Fig. 7, D and E, show doubly refracting areas on a cleavage surface around a point at which pressure was applied. The positions of the plate are 45° apart, the lines Cl representing a cleavage direction. The tooth-like irregularities on fracture surfaces of sphalerite may be distinctly doubly refracting. Cadmium, and Mereury. 387 Optical and crystallographic relations between Sphalerite and Wurtzite. The development of double refraction is the only conclusive evidence we have of the change of sphalerite to wurtzite by heating. Inasmuch as the transformation is slow, its progress can be studied. In ferriferous blendes transformation appears to be most rapid, starting usually at a single point in a grain and progressing so that the final product has like orientation throughout. In grains of the purest sphalerites, the wurtzite usually begins developing at more than one point and in differ- ent orientations. The structure produced is an intergrowth of lamellz of wurtzite, each lamella having its principal axis parallel to one of trigonal axes of the sphalerite grain.* It is evident that the strength of the double refraction of a grain thus transformed will be conditioned by the relative develop- ment of the four possible sets of lamelle. J. Beckenkampt has considered that lamellae of wurtzite may develop parallel to the trapezohedron of sphalerite. Such lamelle would out- crop on a cleavage face parallel or normal to cleavages or bisecting the acute angle between cleavage surfaces. All of the outcropping planes would be oblique to cleavage planes. No lamellee of this sort were seen in the large number of prep- arations examined during this investigation. In material furnished by Mr. B. 8. Butler, from Beaver Co., Utah, prismatic crystals of wurtzite from a brecciated vein have fragments of sphalerite as nucleii. The traces of the prismatic cleavage of the wurtzite and of the cleavage of the sphalerite appear to be parallel. The wurtzite cleaves parallel to the second order prism. Beckenkamp has discussed in detail the very close erystallo- graphic relations of sphalerite and wurtzite. With these are now correlated the very slight changes in optical properties, volume, and energy content accompanying the inversion of these minerals. Cadmium Sulphide. Amorphous.—The flocculent precipitate of cadmium sul- -phide is yellow while moist, but it dries to an orange powder which is lumpy and nearly opaque to transmitted light. The lumps can be consolidated by pressure—as by grinding forcibly in a mortar or by compressing on a microscope slide under a cover-glass—into transparent films, some of which become erystalline during and after the compression. The progress of erystallization may be slow, several hours being required for * Hautefeuille (loc. cit.) recognized this relation. + Zs. Kryst., xliv, 248, 1908. 388 Allen and Crenshaw—Sulphides of Zine, the induced radial—or parallel—fibrous structure to attain its apparent maximum double refraction. The pulverulent precipitate is bright yellow when first formed, and may remain yellow when dry. Conditions which cause aggregation or great increase in size of the particles pro- duce orange-colored powders. The physical characters of powders of various colors are considered in a succeeding sec- tion. Amorphous cadmium sulphide has not been observed in a hard, brittle, doubly refracting form like zine sulphide. When it exhibits double refraction its refractive index approaches the refractive index of greenockite (crystalline Cd$). The tendency to crystallize is much greater in some prepara- tions of amorphous cadmium sulphide than in others. This has been observed particularly when these preparations have been embedded in the mixtures of sulphur and selenium pre- paratory to obtaining refractive indices. Large clear globules and lumps from dried flocculent precipitates have not erys- tallized under this condition, but lumpy aggregates of minute globules such as have been formed by heating the latter in strong solutions of sodium sulphide crystallize readily. How- ever, the crystals are oriented at random, and are so minute that only a very strong light reveals their double refraction. On account of the pores in these aggregates their refractive indices could not have been determined were it not for the fact that compressing and moving them about in the viscous mixture causes their surfaces to consolidate into transparent films.* Crystalline.—By whatever method produced, the erystals of cadmium sulphide as seen in the microscope were prismatic with parallel extinction, elongation c, pure yellow color, and very faint or imperceptible pleochroism. Crystals from the preparations with molten alkali polysulphide were identified as ereenockite by goniometric measurements. Three prismatic, hemimorphic crystals about 1"™ long furnished the following data: the prism angles varied between 59 and 61°; 9 angles from the prism to a pyramid varied between 26° 10’ and 28° 15’; five of these angles giving the sharpest signals were included between 27° 45’ and 28° 15’; the base terminating the end bearing this pyramid was dull; a steeper pyramid with dull faces terminated the other end. The pyramid measured corresponds to 2021 of greenockite. The observed refractive indices of these prisms for lithium light are e=2°-447, o=2-425. The crystals from cadmium sulphate, slowly precipitated by hydrogen sulphide, gave a *Tt was at first thought that these films might have taken up appreciable amounts of the constituents of the mixture, but this is not the case. Cadmium, and Mercury. 389 value for e of 2°44. Hexagonal crystals of greenockite from sublimed cadmium sulphide have a slightly higher refractive index, e=2°456. The lower refractivity of the crystals formed in contact with salts is attributed to dissolved impurities, for no other optical differences were observed. The sublimate produced by heating cadmium in hydrogen sulphide contains crystals of greenockite of four very different habits: filaments and needles, stubby prisms, twins after Fie. 7. Fic. 7. A, B, Greenockite twin. C, etch figures on sphalerite. OD, E, doubly refracting areas on sphalerite. 1011, and twins after 2023. The drawings (fig. 7), made from microscopical studies show the twins. By the elongation of twins like A, feather-like crystals are produced. The angles measured are as follows: w (1011): c(0001)=48°; m (1010): #(1011)=47°; m (1010) : 2 (2021)= 28°; m (1011): »(4041)= De: p (2028) : » (4041) = 43°; v (4041): 2 (2021)=—11°. The extinction angles measured from the trace of the twin- ning (and composition) planes are 48° and 32°. The latter angle was measured accurately more than forty times on several twins, with variations less than 1°. The value 32° + 10’ was obtained. For ereenockite the calculated angle is 31° 58’. Poorly developed twins of this type may represent the sup- posed monoclinic modification of cadmium sulphide. Optical properties of pure Greenockite. The chief optical properties of pure greenockite may be in- ferred from the upper curves of fig. 8. It is uniaxial—positive Am, Journ. Sct.—FourtH SERIES, Vou. XXXIV, No. 202.—Ocrozpmr, 1912. 26 390 Allen and Crenshaw—Sulphides of Zine, for the colors from red to blue-green and negative from blue- green to blue, the wave-length for which it is isotropic being about 523 wu. Above 500 wp (toward the violet), absorption is so strong that the double refraction has not been measured. By transmitted light, therefore, clear crystals are pure yellow. Plates 0°02" thick are opaque for all the blue and the violet light that can be obtained from a 20-ampere are through a monochromatic illuminator. The ordinary ray is more strongly absorbed in the green than the extraordinary ray, for which reason the mineral is pleochroic. In white light the pleo- chroism is not perceptible in single crystals, but in thin twin erystals it may be seen by contrast in the two parts. A closer study of the light-absorption of greenockite shows that below (toward the red) wave-length 519 wp there is very little absorption for », and from 517 to 511 it increases to nearly complete opacity. For e absorption begins near 512 and increases similarly to 506. The change of optical sign is the natural accompaniment of the more rapid increase in refractive index of than of ¢ near this region of absorption. The refractive indices for several different wave-lengths in the red and orange were measured under the microscope. The results are plotted on the diagram. The values for sodium and lithium lights with a probable error of + °003 are as follows: é€y,==2°529 5 @y,=2°5063; €;;=2°456; o,,;—=2°431. The values used for the curves in the yellow and green are extrapolated on the basis of the absorption, and of the follow- ing values for double refraction obtained by measurements on prisms -03 to ‘04"™ thick, in monochromatic light. 671 (Li) = -025 527 = 006 589 (Na) = -023 523 = :000 547 = 018 518 = —-006 535(Tl) =:013 516 = —‘016 Color of Cadmium sulphide. The color of a substance as seen by reflected light depends upon the character and relative amounts of light reflected directly from external surfaces and indirectly from internal surfaces, that is, after having passed into the substance and reflected out. The quality of the light reflected in each case depends upon the refractive and absorptive powers of the sub- stance, and upon the character (plainness, size, brightness) of the reflecting surfaces. Greenockite.—This mineral absorbs all the blue and violet and part of the green of the spectrum, and freely transmits the rest. When the greenockite grains seen in mass have diameters of 0:2 to 1:0 or more and are bounded by plane, bright faces, a comparatively large amount of blue light is Cadmium, and Mercury. 391 directly reflected and a small amount of red, orange, yellow and green are reflected after passing through the surface layer of crystals. The combined effect of all the reflected light is a lustrous, dark, yellow to yellowish-green color, or dark citrene. Similarly, plane-faced bright crystals, but having diameters of -01™™ or less reflect about the same amount of blue light directly, but they reflect much more of that which has pene- trated the surface. The resulting color is a brilliantly pure yellow. Massed crystals of the size of the last but with dull faces have a light yellow-brown or citrene color. A powder consisting of crystalline grains of which the surfaces are mostly bright but not plane, such as is formed by grinding, is in- variably a brilliant orange color. In this case there is less direct reflection, and much of the light finally reflected from the interior has penetrated deeper and thus lost more green and yellow than in a powder having plane-faced fragments. Combinations of the physical conditions described cause varia- tions between the extreme colors enumerated. The colors of dry, amorphous cadmium sulphide may be explained in much the same way. The amorphous sulphide, however, absorbs more strongly in the yellow and green than does the crystalline. By transmitted light its color is orange- yellow in films -01"™ thick, and yellow in films :001™ thick. The globules of which the puiverulent amorphous sulphide consists may have bright surfaces or surfaces dulled by wrinkles. In the former case the colors are most brilliant and pure. Powders consisting of separate globules ‘0001 to -001™™ in diameter are bright yellow with a tinge of orange ; powders having globules ‘004 to :007™™ in diameter, or compact aggre- gates of smaller globules, are bright orange-colored. (See Table XII.) The lumpy aggregates of dried flocculent precipitates have a duller orange color, owing to a less complete reflection of the light which enters the powder. The characteristics of amorphous precipitates formed in various ways are given in Table XII, and the dispersion curve for the purest material at hand is shown in fig. 8. A precipitate made by treating a 10 per cent solution of CdSO, + 20% H,SO, with H,S at boiling contained a few per cent of bright orange-yellow globules -004 to ‘007™ in diam- eter. These were cell-like, having a more highly refracting wall about -001™™ thick. Another precipitate made by heating for 2 days at 200°, 2 ¢. of amorphous CdS in a closed tube with 10 per cent HCl, was covered with a very thin film of indefinitely doubly refracting material having a much redder color than any other prepara- tion examined. 392 Allen and Crenshaw—Sulphides of Zinc, Mercurie Sulphide. Amorphous.— The tendency to aggregate into spherical | masses, which is so marked in amorphous cadmium and zine sulphides, has not appeared in amorphous mercuric sulphide. Crystalline.—At least three crystalline modifications of HeS have been prepared. Of these, cinnabar crystallizes best, in sharply-bounded nearly equi-dimensional, red, hexagonal prisms Fie. 8. Ae 700 600 500 Fie. 8. Dispersion of greenockite and of amorphous cadmium sulphide. or tables having very strong positive double refraction and very high refractive indices. The index o of crystals formed by sublimation was approximately matched under the micro- scope in Li-light with a glassy mixture of Sb,S,. and As,S,. and found to be equal within the limits of error (about +:02) to » of natural cinnabar, i.e., 2°81; €,, which is 3°14, is too high to be determined microscopically. The new form, 8’-HgS, is prismatic in habit and has not been produced in crystals exceeding ‘003™" in diameter and ‘03™™ in length. These crystals always taper toward the ends Cadmium, and Mercury. 393 TABLE XII. Size of Refractive Method of preparation particles index Color Remarks mm Ny, Sodium thiosulphate Nearly |The purest in closed tube -.---| ‘004 and less|2°38 pure orange orange | examined Do 7001—"003 /|2°32 Orange /Globules | ageregated HS cold e222 = eeUECOOOO) Meenas Glia 3 Orange- /Clustered yellow EP Srand hot b¢y HC 004 se) | bs Orange HS and hot 28% ESO eee 008 oe Weep Aggregated Boiled several hours orange with sodium thio- sulphate._ ...-..|°007—-001 |2:28+ Bright orange HS in the cold_---- "0005—-001 |_---_.--- Yellow Na,S, closed tube---| :001 2°40 Yellow Commercial powder heated dry at500°|:0006 |._..-.--- Yellow Commercial powder Amp emeteees a Compact /|2°38—2°39|Dull aggregate orange Sample: 2242-3224 Do 2°37 Do and have no distinct faces. They are commonly aggregated into stellate groups. In color they are not distinguishable from cinnabar, and like cinnabar are optically uniaxial and positive, © but their refractive indices are lower and double refraction not so strong. @ of several lots of crystals was determined for wave-length about 650 wz. Monochromatic light could not be used advantageously on account of weak illumination and dif- fraction, but for the mean wave-length of the narrow band of the spectrum transmitted by the sulphur-selenium mixture in which the crystals were imbedded, the value of » was found to be 2°60 to 2°61, the larger value being for the purest erystals. The dispersion of the crystals is not accurately known, but it appears to be about equal to that of the sulphur-selenium mix- tures used Therefore, for Li-light o = about 2°58; eo = -24 + °08 according to several determinations on prisms of meas- ured thickness. The third modification, a’, is probably identical with the mineral metacinnabar, but positive identification is not pos- sible on account of the minute size and poor development of the artificial crystals, and on account of the meager and conflict- 394 Allen and Orenshaw—Sulphides of Zine, ing data on the crystallography of metacinnabarite. Crystals of this form are black, and when typically developed have six spindle-shaped rays, apparently equal and meeting at right angles. These characteristics indicate skeletal growth par- allel to the axes of the cube. But numerous rays are present in some cases, and though the six can usually be distinguished, the exact relations of the intermediate rays are uncertain. None of the rays exceed -01™" in length. In several preparations—previously deseribed—black prisms were found which, in all but two cases, were shown to consist of transparent double salts coated with amorphous mercuric sulphide. In the two cases no satisfactory determination could be made. - Summary. 1. The two sulphides of zine are enantiotropic: $-ZnS or sphalerite is stable beiow 1020°, where it is transformed into a-ZnS or wurtzite. Sphalerite has a density of 4:090, (Pane at 25° water at 4° viz., 4087. The determinations were made on a very pure analyzed sphalerite, and the wurtzite formed by heating it to the proper temperature. The refractive indices of these forms for sodium light are: sphalerite, », = 2°3688 ; wurtzite, o, = 2°356, € = 2°378. 2. Iron sulphide in solution lowers the inversion point of sphalerite strongly and in a nearly regular manner. The inyer- sion temperatures of four analyzed ferruginous sphalerites, the highest containing 17 per cent of iron, were determined. The specific volumes of these sphalerites varied almost rectilinearly with the percentage of iron. The volume increases with the latter, although the specific volume of ferrous sulphide is only about 85 per cent as great as that of zine sulphide. The refractive indices for sodium light of both sphalerite and wurtzite are raised 00033 for each per cent of ferrons sulphide. 3. Crystals of wurtzite of considerable size were obtained by sublimation at about 1200°-1300°. Small dodecahedrons of sphalerite were obtained from molten sodium chloride at a little above 800°, while larger dodecahedrons as well as tetra- hedrons erystallized from molten potassium polysulphide at about 350°. From aqueous solutions both sphalerite and wurtzite were obtained at temperatures between 200° and 400°. Below about 200° the products were amorphous. From solu- tions of alkali sulphides (alkaline solutions), only sphalerite formed; both dodecahedrons and tetrahedrons were obtained. From acid solutions of zine salts hydrogen sulphide precipi- tates at 250° and above, both sphalerite and wurtzite. In nearly all cases (10 out of 12 experiments) so far as experiments have Wurtzite has a density very slightly less, Cadmium, and Mercury. 395 gone, temperature and acid concentration have proved the defi- nitive factors. The higher the temperature for a given acid concentration the greater is the percentage of sphalerite (the stable form) crystallized ; and the higher the acid concentration for a given temperature the greater is the percentage of wurtzite (the unstable form) crystallized. Quantitative work has shown previously that the same rule holds for the disulphide of iron. Here the stable form is pyrite and the unstable, marcasite. 4, A preliminary study was made of the precipitation of zine by hydrogen sulphide from solutions of variable acid concentration. 5. We have obtained only one sulphide of cadmium, the mineral greenockite, whatever method was used in its prepara- tion. Very pure large crystals were prepared by Lorenz’ method, viz., the action of hydrogen sulphide on cadmium mineral at ze _ was 4-820. water at 4 The refractive indices were found to be ey, = 2°529, oy, = 2°506. The various hues of different preparations of cadmium sulphide do not depend, as has been claimed, upon different allotropic forms; they depend first on whether the substance is crystallme or amorphous. The color of the amorphous products depends chiefly on the size of the grains, the yellow products consisting of more minute particles, but it is also influenced by the nature of the surface of the individual grains, and their forms. 6. Mercurie sulphide exists in three different crystalline forms, viz.: cinnabar, o-HgS, which is readily prepared by digesting any other form of mercuric sulphide with a solution of ammonium sulphide or alkali sulphide; metacinnabar, a’-HeS, which is precipitated from dilute acid solutions of mercuric salts by sodium thiosulphate; and a new erystal form, B’-HeS, which is obtained from more concentrated neutral solutions of mercuric salts in a similar way. mineral at 25° water at 4° 7 BEC: The specific gravity of a’-HgS at 25° averaged 7°60 as compared to about 7-7 for the natural mineral. For reasons stated in the text the latter figure is doubtless too high. B’-HgS has only been obtained in the form of a fine erys- talline powder, having practically the same color as vermilion. It is hexagonal. The specifie gravity at 25° averaged 7-20. ne pee of refraction for 650 mu were: o,, = 2°61, €,; 7. Cinnabar is the stable form of mercuric sulphide at all temperatures up to its sublimation point, which is about 580°. vapor. The density of these crystals, The density of cinnabar, 396 Allen and Crenshaw—Sulphides of Zine, ete. The other two forms change into it, either by heating alone or more readily in the presence of solvents like concentrated ammonium sulphide or 30 per cent suphuric acid. The absorp- tion of light by cinnabar increases markedly with rising tem- perature, “put it regains its color on cooling after long heating at 325°. Heated above 400° it becomes per manently black. This is not an inversion as some have supposed; the cinnabar contains only about 1 per cent or less of a thin coating of the black sulphide which perhaps is caused by condensation of the vapor. 8. Amorphous cadmium sulphide is so fluid, that during pre- cipitation small particles may aggregate into globules 0:005 to 0-01™" in diameter which remain permanently plastic. _Amor- phous zine sulphide aggregates similarly but the globules may harden, either without crystallizing, or by crystallizing. In the former case they simulate doubly refracting spherulites owing to the development of strains in a wholly amorphous substance; in the latter case double refraction is due to wurtzite. 9. Comparing the genetic relations of the minerals sphaler- ite, wurtzite, cinnabar and metacinnabar with the genetic rela- tions of pyrite and mareasite, we find certain remarkable reeu- larities. The stable forms, sphalerite, cinnabar and pyrite are always obtained by crystallization from alkaline solutions (solu- tions of the alkali sulphides), while the wnstable forms wurtzite, metacinnabar and marcasite are obtained from acid solutions only. Thestable forms also may be crystallized from acid under certain conditions. Of these temperature and acid concentra- tion appear to be the important ones. Certainly with pyrite and marcasite and in all probability with sphalerite and wurtzite, the higher the temperature the ereater is the percentage of the stable form obtained, while the higher the acid concentration at any temperature the greater is the percentage of the unstable form obtained. These facts appear to agree remarkably well with the field evidence which relates to the genesis of the natural minerals, while they give new significance to the general geologic distinction between deep-seated and surface waters in nature. 10. None of the sulphides of the group zinc, cadmium, mer- cury melts at atmospheric pressure. In conclusion the authors wish to express their thanks to Dr. Geo. P. Merrill of the National Museum for placing at their disposal much valuable material for study; to Mr. C. E. Sie- benthal of the U. S. Geological Survey for mineral specimens, and to Mr. B.S. Butler, and especially to Mr. Waldemar Lind- gren, also of the U. S. Geological Survey, for geological data and valuable criticism. eae ca Laboratory, Carnegie Institution of Washington, Washington, D. C., July 9, 1912. Chemistry and Physics. 397 SCIENTIFIC INTELLIGENCE. I. CHEMISTRY AND PHYSICS. 1. The Formation of Red Phosphorus.—The usual, well known, method of converting colorless phosphorus into the red modification consists in heating the substance in closed vessels for some time at about 250° C., a temperature near its boiling- point. Srock, Scuraper and Stamm have now found that when the vapor of phosphorus is highly superheated and then quickly condensed, a considerable amount of red phosphorus is found in the product. Experiments were made by heating the vapor in tubes of quartz glass to various temperatures and suddenly cool- ing by plunging the hot tubes into cold water. In this way it was found that after heating to 450° C. the condensed product was slightly colored, and that the amount of red phosphorus formed increased, as the temperature of heating was higher, up to 1175° C. and above. The duration of the heating was found to have little effect, but the rapidity of the cooling was of the greatest importance. For instance, when a tube heated to 700° C, was immediately cooled the product was very red, if the cooling was delayed two seconds after removing the tube from the furnace there was very little red phosphorus produced, after a delay of five seconds there was only a trace, and when the tube was allowed to cool in the air the condensed phosphorus was almost colorless. Under favorable conditions of high heating and very rapid cooling it was found that as much as about one- third of the total phosphorus was obtained in the red modifica- tion. It has been known for a long time that the vapor of phosphorus has a density at 300-400° C. corresponding to the molecular formula P,, while at higher temperatures the molecule becomes smaller, being intermediate between P, and P, at 1700° C. It is the opinion of the authors that the sudden cooling of the superheated vapor causes the molecules that are smaller than P, to combine either with the latter or with each other to form red phosphorus molecules. They do not believe that the molecule of red phosphorus is smaller than that of the colorless substance, since all the properties of red phosphorus indicate that the oppo- site is true.— Berichte, xlv, 1514. H. L. W. 2. Devitrification of Silica Glass—The use of apparatus blown and worked from melted quartz has become very common in chemical laboratories, on account of the resistance of the mate- rial to solvents, its high melting point, and its remarkable endur- ance of sudden changes in temperature. Several years ago it was fotind that silica glass is permeable to helium and to hydro- gen at a low red heat, and now Sir Witi1am CrooxeEs has observed that the material when kept for a long time at a high temperature becomes devitrified and opaque, and then permits 398 Scientific Intelligence. the.passage of air through the walls of an exhausted vessel. He found that a perfectly transparent bulb of quartz glass when kept at 1300° for twenty hours in an electric resistance furnace became white and translucent like frosted glass, and with several exhausted bulbs treated similarly he found a very considerable leakage of air. The leakage was found to occur not only while a bulb was heated but also at ordinary temperature after devitri- fication had taken place. A micro-photograph of .a devitrified silica bulb showed a surface cracked all over into the appearance of cells, and many of the cells showed a decided hexagonal out- line. Crookes has observed a similar appearance when a silica dish, originally clear and transparent, was used for evaporating the solution of 100™5" of pure radium bromide. Patches appeared on the bottom having a dull, roughened appearance, and upon microscopic examination they showed a structure very similar to that of the devitrified bulb. He concludes that radium at the temperature of boiling water can devitrify quartz glass, but he has not seen this effect upon the surface of glass or silica bottles in which radium salts have been kept in the cold for several years. — Chem. News, ev, 205. H. L. W. 3. The Presence of Formaldehyde in Plants.— According to Baeyer’s assimilation hypothesis, the plant first reduces carbonic acid to formaldehyde, and then condenses this to carbohydrates. In order to establish this hypothesis it is of great importance to detect the presence of formaldehyde in plants. Currivus and FRANZEN show that all the tests that have been previously used for this purpose have been unreliable on account of the interfer- ence of other aldehydes, and they conclude that formaldehyde in plants has not been detected heretofore. Therefore, these chemists have made a new investigation of the subject, and it appears that they have definitely established the presence of formaldehyde in the leaves of a certain plant, and have thus established the basis of Baeyer’s theory. 'They used the leaves of a certain kind of beech, ‘‘ Hainbuchenblatter,” in large quantity. As much as 14 tons of the leaves, altogether, were put through the process, which was an elaborate one consisting of several dis- tillations, the conversion of the aldehydes into the corresponding acids by means of silver oxide, and after further separations the final detection by several methods of the resulting formic acid. Quantitative determinations showed that 1*% of the leaves con- tained only 0:00086% of formaldehyde, or less than one part per million.— Berichte, xlv, 1715. H. L. W. 4, The Determination of Sulphur in Insoluble Sulphides.—T. Sr. Warunis, having previously worked out a method for the determination of sulphur in coals, has applied the same principle successfully to the analysis of sulphides. A portion of 0°58 of the sulphide, ground exceedingly fine, is intimately mixed in a spacious porcelain crucible with a mixture of 4 parts of dry sodium carbonate and 3 parts of copper oxide, the whole is coy- ered with a thin layer of the same mixture and is heated gently Chemistry and Physics. 399 at first, then finally for 2 hours at the full heat of a Bunsen burner. During the heating the mass must be frequently stirred with a strong platinum wire, and it is advisable to place the cru- cible in a hole in an inclined piece of asbestos board during the ignition, in order to protect its contents from the sulphur of the flame. After cooling, the mass is extracted with water, the solu- tion is filtered, the residue is boiled with sodium carbonate solu- tion and finally washed with water. The solution is acidified with hydrochloric acid, evaporated to dryness to separate silica, and the sulphur finally determined as barium sulphate in the usual manner. The author states that in this way he has obtained very good results more conveniently and quickly than by the usual methods, and he gives test analyses upon copper pyrites, iron pyrites, and mercuric sulphide which show very satisfactory results.— Berichte, xlv, 869. H. L. W. 5. Benzoie Acid as an Acidimetric Standard.—Although the employment of organic acids for this purpose is not new, it is interesting to notice that G. W. Morey has used benzoic acid with very satisfactory results. It was found best to fuse the bulky sublimed acid before weighing by heating it in a platinum dish to about 140°C. The fused material can then be broken up and kept indefinitely for use. Phenol phthalein was used as the indicator with the careful exclusion of carbonic acid. The mate- rial has the advantage of high molecular weight, so that compara- tively large quantities can be used, thus reducing the errors in weighing. It is also easily obtained in a pure condition, it is stable and not hygroscopic.— Chem. News, cvi, 66. TS be Wi 6. Hin neuer Fall von Koppelung kurz- und langwelliger Fluoreszenzbanden.—According to Stark’s theory of band spectra two kinds of absorption bands are to be distinguished. The criterion for differentiation depends on the behavior of the bands when the attempt is made tostimulate fluorescence. Theso-called long-wave bands have the property of absorbing light without appreciable fluorescence. On the other hand, absorption of light by a “ short-wave” band is, in general, accompanied by the emission of fluorescent light not only in its own interval of wave-lengths but also in the region of the “coupled” long-wave band. Furthermore, the absorption is more intense for the short-wave bands than for the associated long-wave bands. It has been shown by Steubing that mereury vapor affords a practical example of coupled absorption bands. The discovery of three other concrete cases by M. Grnpxe is obviously a matter of theoretical importance. The substances which he has shown to possess coupled absorption bands are acetone (CH,COCH,), diacetyl (CH,COCOCH,), and ethylencyanid-monoxal-ethylester (N,H,C,COCO,C,H,). These ketones were dissolved in very pure ethyl alcohol. The source of light employed was a strong spark between elec- trodes made of an alloy of iron and tungsten. ‘he condensing lens was of quartz and the spectrograph had fluorite objectives 400 Scientific Intelligence. and a single 60° prism of the last named material. The absorp- tion cell and the general assemblage of apparatus were set up in the usual manner, As already implied, the positions and inten- sities of the bands were recorded photographically. The maxima of the absorption bands of acetone are shown by the curves to have the approximate wave-lengths 275 wu and 355 wp respectively. The fluorescence bands extend from 260 pu to 320 up and from 325 wp to 460 pp. The former has aminimum of emission at 300m and the latter at 340 wp. In other words, each fluorescent band seems to be double. For the concentration used, “(1:320”, the greatest intensity of fluorescence falls at 360 pu. For diacetyl the maxima of absorption lie at 285 wy and 410 yp. The shorter wave-length fluorescent band is also double and comprises the interval 270 pp to 350 wp, the minimum being shown graphically at 320yp. The longer wave-length fluorescent band is single and extends from about 355 pp to 480 yp. The fluorescent maximum comes at 410pyp for a concentration of 1:620. For ethylencyanid-monoxal-ethylester the maxima of absorption have the wave-lengths 335 wy and about 430 py. Both fluorescent bands are single, the one extending from 290 pp to 365 pu, and the other from 395 pp to below 480 wy. For the con- centration employed, 1:11200, the intensity of emission is greatest at 435 py. The three ketones satisfy the requirements for “coupled” bands. ‘That is, when the wave-lengths of the exciting light lie within the more refrangible absorption bands, both fluorescent bands come out strongly, provided, of course, the concentration is appropriate for observation. On the other hand, fluorescence is not appreciably produced when the wave-lengths absorbed are confined to the region of the less refrangible absorption bands. Although the absorption is much greater in the short-wave than in the long-wave bands, nevertheless the more refrangible fluo- rescence is markedly less intense than the less refrangible fluo- rescence. This is doubtless due to the excessive absorption of the shorter wave-length fluorescent light by even thin layers of the solution. According to Stark’s theory, the “centers” of the absorption and fluorescent bands of the substances under consideration are the valence electrons of oxygen which are loosely bound to carbon.—Physikal. Ztschr., No. 13, July, 1912, p. 584. s H. S. U. 7. Die Ionisierung von Gasen durch Licht und das Funken- spektrum des Aluminiums im Gebiete der Schumannstrahlen.— Certain spectrograms taken by THroporE Lyman have enabled him to bring out some interesting and important facts relating to the spectra of the sources of light used by Lenard and others in the study of the ionization of gases by ultra-violet light. . The region under consideration extends from wave-length 1200 A. U. to about 1870 A. U. The first positive reproduced shows the spectrum of a spark between aluminium electrodes in an atmos- phere of hydrogen. A comparatively small number of lines Chemistry and Physies. 401 (about 20), and no continuous background can be seen on the plate. The shortest wave-length recorded photographically is about 1370 A. U. The second spectrogram was obtained when the aluminium electrodes were placed so close to the fluorite window of the vacuum grating-spectrograph as to cause the dis- charge to play along the surface of the window. The exposure lasted six minutes and the fluorite was badly damaged, How- ever, notwithstanding the latter incident, the spectrogram is very goad, the lines are much stronger than in the positive first men- tioned, additional lines appear in the neighborhood of X 1300, and a dense, continuous background fills the region from d 1650 to X 1870. In fact, most of the energy seems to be concentrated in the last named interval. (The plates do not extend to wave- lengths greater than 1870 A. U.) The electrical apparatus used by Lenard was more powerful than that employed by Lyman. From another point of view, Lyman has shown in earlier papers that a quartz plate, one or two millimeters thick, is quite trans- parent below 1600, but that it is “absolutely ” opaque for wave- lengths shorter than 1400. Furthermore, he has demonstrated experimentally that air is very opaque between 1400 and 1550, but that transparency begins again at about 1350. When all of the facts just enumerated are taken into account the conclusion must be reached that the “rays” with which Lenard was dealing consisted in the strong aluminium group at about A 1300. For, Lenard says :* “Die auf Luft wirksamen Strahlen werden also zum weitaus grossten Teile durch Quarz absorbiert, und zwar ausserordentlich viel stirker durch den Quarz als durch die gleich dicke Luftschicht, welche er ersetzte.” The third and last spectrogram illustrating Lyman’s paper per- tains to the vacuum-tube spectrum of hydrogen. In obtaining this plate no capacity was introduced in the secondary circuit, a current of about 10 milliamperes flowed, and the gas was ata pressure of about 2 millimeters of mercury. The spectrum con- sists of so many fine lines as to give a nearly continuous, strong and uniform illumination from about A 1330 to 1640. In the same interval the spark between aluminium electrodes gives only a few scattered lines. Also the hydrogen spectrum has a group of fine lines between A 1230 and A 1290. Consequently, Lyman recommends the hydrogen vacuum-tube for use in experiments on ionization. The fact that Palmer, working with a vacuum- tube, obtained nearly as great volume ionization as Lenard, although the latter had a very powerful transformer, is also accounted for by the foregoing data.—Physikal. Zischr., No. 13, July, 1912, p. 583. 8g Sh le =) 8. On the Apparent Change in Weight during Chemical Reaction.—Vhis question has been recently investigated by J. J. Mantry. Although he gives only an abstract of the original paper, nevertheless the salient points and the final result seem to be of sufficient general interest to merit attention in this place. Manley first points out that Landolt, in his classical investiga- tions, seems to bave omitted taking precautions against certain * Sitzungsber. d. Heidelb. Akad. d. Wiss., Abth. 24, 1911, p. 19. 402 Scientific Intelligence. possible sources of error, doubtless deeming them of negligible importance. These errors may arise from (1) air streams within the balance case, and (2) differences in the superficial areas of the reaction vessels and the consequent possibility of a variation in the relative masses of any aqueous films which may be formed on the surfaces. Furthermore, evidence is lacking in Landolt’s memoir to show that he paid sufficiently close attention to effects producible by very slight and varying differences in.the tempera- ture of the contents of any pair of his reaction vessels. Since Manley considered these sources of error as extremely important he accordingly introduced devices for eliminating and neutraliz- ing the errors just indicated. ‘A new and hitherto unsuspected secondary chemical reaction occurring within the reaction vessels was discovered and partially investigated. This reaction may be somewhat accelerated by heat and greatly so by the radiations from a tantalum lamp. It was shown that this phenomenon made it impossible to obtain trustworthy results with solutions of silver nitrate and ferrous sulphate. Decisive results could be obtained with solutions of barium chloride and sodium sulphate because these salts react quickly and practically perfectly. The limit of accuracy estimated by Landolt, under his working conditions, was +0°:03 mgrm., while Manley claims +0-006 mgrm., just five-fold better. Landolt found the mean apparent change in mass during chemical reaction to be not greater than 1in1 X10". In the most thoroughly investigated case, in which the reacting bodies were barium chloride and scdium sulphate, Manley found that the apparent change in mass did not exceed 1 in 1X10°.—Proc, Roy. Soc., vol. Ixxxvii, No. A 594, p. 202. H. 8. U. 9. On the Torque produced by a Beam of Light in Oblique refraction through a Glass Plate.—In the presidential address to the Physical Society in 1905 Poynting developed the idea that a beam of light must be regarded as containing a stream of momentum, and he showed that this principle may be used to solve with great ease the various cases in which a beam of light is absorbed, reflected, or refracted at a surface. The particular case in which a beam of light is refracted at a plane surface was also analyzed by Poynting and the conclusion was drawn that there will always be an outward pull along the normal, 7. e. from the more dense towards the less dense medium. Hence, when a beam of light passes obliquely through a parallel plate there is a normal pull outwards both at incidence and at emergence, and these two pulls constitute acouple. Obviously, the forces arising from the several reflections in any actual piece of apparatus must be taken into account in making numerical calculations. Prelim- inary quantitative experiments were performed by Poynting and Barlow with the object of comparing the observed torque with the value of the couple derived from the principle of luminiferous momentum. However, the apparatus used did not yield very satisfactory results. Since that time G. Bartow has devised and used an apparatus Geology. 408 which is largely free from the sources of error which vitiated the earlier work. The new system consists essentially of a cube of crown glass suitably suspended by a brass rod and a quartz fiber. The cube was made as perfect as possible and, when suspended, four of its edges were vertical and parallel to the axis of rotation. The horizontal beam of light passed diagonally through the cube and experienced two refractions and only two primary reflections. The suspended system was enclosed in a gun-metal box which was fitted with the necessary lateral tubes, windows, ete. Details of the precautions taken with respect to “ gas action” and convection currents may be omitted in this place. Also, the calculations of the torque, of the reflection corrections, and of the energy density of the beam of light afford no novelty. It is worthy of note, however, that Barlow found hydrogen far more reliable than air inside of the box enclosing the suspended system. The calculated and observed deflections agree as well as can be expected when the inherent difficulties of the experiments are given due consideration. The observed deflections are systemat- ically larger than the calculated deflections, which shows either that some source of error has been overlooked or that a funda- mental datum was not determined with sufficient accuracy. Nevertheless the following statement of the author seems justified by the final results, namely: “‘ We may therefore conclude that the oblique passage of a beam of light through a plate of refract- ing material produces on the matter of the plate a torque which has the magnitude deduced from the transfer of momentum in the beam.”—Proc. Roy. Soc., vol. Ixxxvii, No. A 592, p. 1. H, §. U. Il. Gerouoey. 1. Sub- Oceanic Physiography of the North Atlantic Ocean ; by Prof. Epwarp Huxx. Atlas with eleven folio maps in color. London (Kdward Stanford).—The maps here given show in better form than anything previously published, the borders of the continental platforms, indented by valleys and embayments, and further crossed by great river-like valleys and canyons, extending from the present continental rivers, such as the Hudson channel, as first noticed by Professor J. D. Dana; and later built up into systems by Spencer, Hull, and Nansen, in order of time. Hull’s charts show the indented borders of the Eastern Continents, like those of high plateau regions, and also many submerged river- valleys, such as those of the English Channel River, Irish Sea River, the canyons of the Adour and Tagus, of the Congo, and others in the Mediterranean; while in a condensed form, the numerous submarine valley-systems off the American coast, as developed by the reviewer, are added, thus making the subject more complete. Hull is led to conclude that these high plateaus were the courses of former land valleys, and accordingly in them are found means of measuring the late changes of land and sea. The former high elevations and continental extensions are sup- ported by the distribution of animals and plants, and late glacial 404 Scientific Intelligence. conditions. Conversely, the epoch is regarded as primarily due to elevations, now measurable, with the consequent deflection of the Gulf Stream. Even where some are indisposed to accept this conclusion as to the cause of the Glacial Period, no other expla- nation of the origin of these submarine channels has been shown to obtain, which interpretation has been given to them in this country by Dana, Chester, Lindenkohl, Davidson, Branner, Upham, and the reviewer, all who have investigated the prob- lems at first hand. : The data, shown in graphic form, are indispensable, both the physiographic as well as the oceanographic, and the whole constitutes a worthy monument of the labors of the author, which must be lasting so long as interest in the Earth’s studies lasts. These studies had been almost passed over by the earlier oceanographic writers, but they are so widespread that their study constitutes a new brand of science, which Hull most gener- ously attributes to the reviewer as the founder, by his work on the Antillean region and off the American Coast. 35. w. s. OBITUARY. James TERRY GARDINER, the civil engineer, died at North East Harbor, Maine, on September 10 in his seventy-first year. He was born at Troy, N. Y., on May 6, 1842, and was educated at the Rensselaer Polytechnic Institute and at the Sheffield Sci- entific School ; receiving the honorary degree of Ph.B. from the latter institution in 1868. His labors as an engineer were carried on in connection with the Brooklyn water works, with the con- struction of earthworks around the harbor of San Francisco, and elsewhere. His most important work was as a topographer, first with the Geological Survey of California, 1864-67 ; later with the Survey of the 40th Parallel under Clarence King, 1867-72 ; and also the U. 8. Geologe al Survey under F. H. Hayden, 1872-75. Dr. Wituiam J. McGer, the geologist and anthropologist, died at Washington on September 4 at the age of fifty-nine years ; a notice is deferred till a later number. Dr. Pavxt Caspar Freer, for ten years the able director of the Bureau of Science in the Philippine Islands, died at Baguio on April 18 in his fifty-first year. Henry Apam Weper, professor of agricultural chemistry at the Ohio State University, died on June 14 at the age of sixty- seven years. Jutes Henrt Pornearh, the distinguished French mathema- tician, died on July 17 at the age of fifty-eight years. Prorrssor Cuartes ANDRE, the French astronomer and director of the Lyons Observatory, died on June 6 at the age of seventy years. Dr. Humpnurey OwrEn Jones, the English chemist, lost his life in an Alpine accident on August 15 in his thirty-sixth year. Dr. Rupvotr Hérnes, professor of geology at Gratz, died on August 20 at the age of sixty-two years. New Circulars. 84: Eighth Mineral List: A descriptive list of new arrivals, rare and showy minerals. 85: Minerals for Sale by Weight: Price list of minerals for blowpipe and laboratory work. 86: Minerals and Rocks for Working Collections: List of common minerals and rocks for study specimens; prices from 1% cents up. Catalogue 26: Biological Supplies: New illustrated price list of material for dissection; study and display specimens; special dissections; models, etc. Sixth edition. Any or all of the above lists will be sent free on request. We are constantly acquiring new material and publishing new lists. It pays to be on our mailing list. Ward’s Natural Science Establishment 76-104 OCotneace AvE., Rocunstrr, N. Y. Warns Natura Science EstaBLisHMent A Supply-House for Scientific Material. Founded 1862. Incorporated 1890. DEPARTMENTS: Geology, including Phenomenal and Physiographic. Mineralogy, including also Rocks, Meteorites, etc. Palaeontology. Archaeology and Ethnology. Invertebrates, including Biology, Conchology, etc. Zoology, including Osteology and Taxidermy. Human Anatomy, including Craniology, Odontology, ete. Models, Plaster Casts and Wall-Charts in all departments. Circulars in any department free on request; address Ward’s Natural Science Establishment, 76-104 College Ave. Rochester, New York, U.S. A. VON TEN Ts: Pa Art, XXXJ.—Emission of .Electrons by Metals under the a Influence of Alpha Rays; by H. A. Bumsrrap and A. G. McGowan seen tes sol. 52 1s sees eee 309 XXXII.—Flashing Ares: A Volcanic Phenomenon; by F. = A. PERRED 2 See oe a ee ee eee aria ie octet 829 XX XIII.—Comparison of Two Screws; by C. Barus-_-.--.- 333 XXXIV.—Some Growth Stages in Naticopsis altonensis, McChesney ; by G. H. Girty. (With Plate I)....-.. 338 XXXV.—Sulphides of Zinc, Cadmium, and Mercury ; their Crystalline Forms and Genetic Conditions ; by E. T. ALLEN and J. L, Crensnaw. Microscopic Study by A, Mime win (2.3 eee en ae ee ore 341 SCIENTIFIC INTELLIGENCE, Chemistry and Physics—Formation of Red Phosphorus, Stock, ScHRADER and Sramm: Devitrification of Silica Glass, W. Crookers, 397.—Presence of Formaldehyde in Plants, Curtivs and Franzen: Determination of Sulphur in Insoluble Sulphides, T. St. Warunis, 398.—Benzoic Acid as an Acidimetric Standard, G. W. Morry: Hin neuer Fall von Koppelung kurz- und langwelliger Fluoreszenzbanden, M. GuLexe, 399.—Die lonisier- ung von Gasen durch Licht und das Funkenspektrum des Aluminiums im Gebiete der Schumannstrahlen, 400.—Apparent Change in Weight during Chemical Reaction, J. J. Mantuy, 401.—Torque produced by a Beam of Light in Oblique Refraction through a Glass Plate, G. BaRLow, 403. Geology—Sub-Oceanic Physiography of the North Atlantic Ocean, E. Hut, 403. Obituary—J. T. Garpiner: W. J. McGue: P. C. Frepr: HU. A. WEBER: J. H. Porncaré: C. Anpr&: H. O. Jones: R. HORNES. ibrary, Bureau of Ethnology. 40%: y 0 VOL. XXXIV. NOVEMBER, 1912. Established by BENJAMIN SILLIMAN in 1818. THE AMERICAN JOURNAL OF SCIENCE. Epiton: EDWARD S. DANA. ASSOCIATE EDITORS Prorrssors GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or Camprince, Proressors ADDISON KE. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT E. GREGORY anp HORACE 8S. UHLER, or New Haven, Proressorn HENRY S. WILLIAMS, or Iruaca, Proressorn JOSEPH S. AMES, or Bautimore, Mr. J. S. DILLER, or Wasuineron. FOURTH SERIES VOL. XXXIV—[WHOLE NUMBER, CLXXXIV]. No. 203—NOVEMBER, 1912, NEW HAVEN, CONNECTICUT. | Wil 1912. THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 123 TEMPLE STREET, NSPE ANU eos aE s aT TE TI EET IE ER AGT DE RP EE EE I IS I EE Published monthly, Six dollars per year, in advance. $6.40 to countries in the Postal Union; $6.25 to Canada. Remittances should be made either by money orders, registered letters, or bank checks (preferably on New York banks). REMOVAL NOTICE. It is with pleasure that I announce the fact that I have removed to more commodious and pleasant quarters. For years I have been handicapped by lack. of room to carry on my fast increasing business and properly display the ever- arriving new lots of goods, but the task of removal was so great, the time necessary to do this great task and yet keep my customers satisfied, nonplussed me; but I have at last “taken the bull by the horns” and made time and am now at last situated where | can properly show my vast stock as it should have been shown years ago. In going over my stock, some of which has been nuchal away for years, I was astonished myself at the many rare and beautiful minerals and other objects | came across. You will certainly miss a rare treat if you cannot come and see me, but if this is impossible, send me a list or an idea of what you want and I will ship you a box. on approval, express prepaid, and you can return me what you do not desire. In my new quarters | am going to make a specialty of the following: Minerals, Gems, Curios, Mosaics, Antiques, Jewels, Synthetic Gems, Art Objects, etc. Send for my new Catalogue, just out. It will pay you to be on my mailing list. A. H. PETEREIT, 261 West 71st St, New York City. THE +e: AMERICAN JOURNAL OF SCIENCE [FOURTH SERIES.] ——__~++>___- Arr, XX XVI.— Volcanic Vortex Rings and_the direct con- version of Lava into Ash; by Franx A. Prrrert, K.I.C. Voleanologist of the Voleanic Research Society. Tue yortex rings projected from cannon, locomotives, the mouth of a smoker, etc., are familiar to all. Sudden puffs from any smoke-filled cavity will produce, by interference at the edges of the orifice, visible annular vortices having a certain stability of form due to the persistence of the vortical move- ment. It is obvious, therefore, that the conditions necessary for the production of these rings must often obtain at voleanic craters and, in fact, such voleanoes as Vesuvius and Stromboli, during the prolonged periods of moderate activity which may be said to constitute their normal condition, project large, thin rings of vapor which frequently attain diameters of five hundred meters or more. Generally speaking, these are somewhat difficult to photograph because of their delicacy and lack of photographic contrast with the sky, and it was not until the Etna eruption of 1910 that the present writer observed volcanic vortex rings having a sufficient degree of solidity to permit of satisfactory photography. Figs. 1 and 2 show the type of rings projected at that time and of which the diameters were approximately 150 and 200 meters. It is not, however, with the purpose of discussing vortex rings, as such, that I have here referred to this phenomenon but because of the fact that these particular rings of the Etna eruption illustrate a principle of volcanic action which is of such importance as to merit attention at this time. What impressed me most in observing these rings was the fact that they were composed almost entirely of ash and yet had been projected from a crater yielding liguid lava. This was the lowermost of the seven principal groups into which the twenty or more vents of this eruption were con- Am. Jour. Sci.—Fourra Serizs, Vou. XXXIV, No. 203.—Novumpsr, 1912. 27 406 F. A. Perret— Volcanic Vortex Rings. g formed and, according to my own enumeration, which has been generally adopted, viz. from the upper end of the fissure downward, it is known as crater No. 7. The outflow of lava, which at the beginning had taken place at many points along the line of the fissure, was soon localized, as is normally the case, at the lowest principal opening, and this continued to be the source of the magnificent stream which formed so spectacular a feature of this eruption. (Fig. 3.) The gas content of the magma caused a semi-explosive emis- sion, projecting sprays and jets of incandescent liquid masses Iie, Il to a considerable height (fig. 4), the result of which was the gradual building up of a cone of the compacted scorie, best seen in the post-eruption view (fig. 5). It is important to note that neither the ejected fragments, nor the walls of the crater, nor the upper portion of the cone were “dry,” i. e., hard or brittle, and therefore capable of being broken or crushed to powder—all were in the liquid or viscous state. There was no demolition of the cone itself, which was growing by accretion, and there was present no old material whatever, yet this crater constantly emitted a quantity of ash-laden vapor having a salmon-pink tint. On the 30th of March at 6 a. m., the writer was proceeding along the line of eruptive mouths, when the rings shown in I. A. Perret— Volcanic Vortex Rings. 407 figs. 1 and 2 were projected from crater No. 7. These had the same color as the other vapors and the air was full of a fine red ash having a strongly acid reaction. The large amount of this ash present in the air may be inferred from the aspect of the sun in the lower portion of fig. 2, which appeared as a ball of burnished copper, and was photographed without irradi- ation effects. This production of ash continued along the first few hundred meters of the lava stream where the gases still escaped froin the surface in considerable quantities. The gas emission under these Fie. 2. conditions is apparently very gentle (fig. 6), producing a con- tinuous simmering sound analogous to that of water in a kettle just before ebullition, but it is probable that each tiny gas vesicle burst from the lava with an explosion which, for its size, is violent, and thus projects and carries off minute parti- cles of the exploded shell. It is this swbdivisional gaseous expansion, and not the explosion of large bubbles, which is the cause of the formation of the ash. The degree of viscosity of the lava is, in all probability, an important factor in this direct formation of ash, while secondary to the gas content as a determining cause. In the case we are considering, although the velocity of the stream at this locality 408 I, A. Perret— Volcanic Vortea Rings. was five meters per second and the mass as a whole showed all the qualities of a liquid, the viscosity was so great, especially in the outer layer, that a heavy rock thrown upon the surface rebounded as from a plate of steel, and it was only with the greatest difficulty that an iron rod conld be forced into the moving stream. In the very liquid lavas of Hawaii the gas vesicles issue almost without resistance and do not form ash in this way, but large gas bubbles seatter the lava and spin it out into olassy filaments—the well known “ Pele’s Hair.” The present writer has often observed the same copper- colored cloud of gas and ash as produced directly from glowing lava in the central crater of Etna, and also in that of Vesuvius Fic. 3. iss) and Stromboli. A cloud of white vapor over the crater will also, even in daylight, under these circumstances, appear of a cop- per tint because illuminated by the glare of the lava below, but when ash is present the color of the cloud persists to. all distances, as in the case of the Etna rings. Let us now leave the consideration of the principle as exem- plified in this superficial action, and study its function in the more fundamental processes of a great eruption. Taking the Vesuvian outbreaks of 1872 and 1906 as typical, and glancing at conditions at the outset, we find that as the result of a long period of moderate activity the cone has been built up to a considerable elevation above the crater of the preceding great eruption, and that the lava stands at a commensurately high EA. Perret— Volcanic Vortex Rings. 409 level in the central conduit. At the initiation of the eruption the cone is fissured laterally by pressure, fusion and gas tension causing a rapid and copious outflow of lava. This is accom- panied by a more or less complete demolition of the upper portion of the cone, the explosive effects at the central vent increase in violence, ejecting old and new material, and attain their maximum with the cessation of the lava outflow, to which there succeeds the truly stupendous emission of gas and ash from the main crater (fig. 7)—no more liquid lava being seen. Fie. 4. But if, in such a case, we are asked to consider the liquid outflow and the explosive ejections as constituting the only emission of co-eval lava, and the ash as being merely old cone- material and triturated rock, as is quite generally supposed, then, I believe, I may raise a point of objection and imterro- gation. Let us consider for 4 moment the condition of the magma in the deeper portions of the volcanic conduit. Its temperature will be higher and its gas tension greater than in the upper portion while it will be subjected to the pressure of the column above it. In these respects it is analogous, as I pointed out 410 I. A. Perret— Volcanic Vortex Rings. some years ago,* to the water in the lower portions of a geyser or to the water in a boiler under tension. This, by reason of the pressure upon it, is actually in liquid form, but it is potentially a vapor because of its temperature, and if the pressure is removed it will flash into the gaseous condition. Similarly, the lower zones of lava in the volcanic conduit are potentially in a condition of explosive extension and are maintained in a liquid or semi-liquid state only by the pressure from above. If now we imagine a relief of this pressure—say by a lateral outflow of lava sufficiently rapid and copious to materially Fie. 5. reduce the height of the lava column—a powerful gaseous expansion will be initiated in the magma below and this will extend progressively downward with progressive relief of pressure from above. This gaseous development will be distributed, subdivisional, intermolecular, the magmatic mass expanding to a froth and being finally ejected as a cloud of gas and ash—it is impossible that it should appear in a liquid form under these conditions. Note that this ash will have been formed under very different conditions from that which is blown off from the surface of the lava in the crater during the early part of the eruption. This latter is formed in contact with atmospheric air and is quickly cooled; it is, therefore, of a vitreous nature with surface oxidation. But the ash from the depths will have * ‘ Some Conditions affecting Volcanic Eruptions.” Science, Aug. 28, 1908. fF. A, Perret— Volcanic Vortex Rings. 411 been formed in a bath of the magmatic gases, out of contact with the atmosphere, and at a temperature which is maintained for a considerable time. The difference between the two is, however, mainly physical, and the magmatic ash, if I may so eall it, should also, by the very conditions of its formation, be virtwally identical, as material, with the co-eval lava, i. e., its ingredients should be the same, it should carry the same salts, exhibit the same crystals under the microscope, give off the same gases on being heated, ete., which could hardly be the ease if it were formed by the trituration of old and profoundly Fig. 6. metamorphosed rock. The limits of the present paper will not permit of a full discussion of this important subject, but it may be noted that Palmieri,* writing of the Vesuvian ash of 1871 and 1872, refers to its identity of composition with the co-eval. lavas, and Casoria’s analyses of the 1906 products show this to a remarkable degree; it 1s evident that some of the old tritu- rated material must be found in the ash, the present contention being that it does not constitute the bulk of it. Considering this in connection with the fact that those who have attempted to calculate the total volume of material lost by the cone have generally considered it insufficient to account for the amount of ash emitted, we may fairly state the neces- sity of admitting the emission of co-eval magmatic material during the ash phase of the eruption. *“ The Eruption of Vesuvius in 1872.” LL. Palmieri (page 119.) 412 I. A. Perret— Volcanic Vortex Pings. It will be seen that this hypothesis does not in the least degree inyalidate the division of the eruption into a ‘ Strom- bolian” and a“ Vulcanian” phase—on the contrary, it supplies a cause for the great and continued dynamism of the *“ Vul- eanian”’ phase and explains the identity of its ash with the co-eval lava. Fie. 7. It is interesting to note that the “ Vulcanian” phase, as far as actual, external eruption is concerned, is cooler than the ‘“‘Strombolian.” This is due, in part, to the downward removal of the furnace, i. e., the lava column, but also, most certainly to the enormous absorption of heat in the process of gaseous expan- sion in the deeper magina and the subdivision of the latter into minute particles which quickly assume the temperature of the gas which surrounds them. During the “ Strombolian” phase F.. A, Perret— Volcanic Vortex Rings. 418 the incandescent lava in the crater is thrown into the air in large masses by great gas bubbles rising through the liquid, without material alteration of its temperature, but the magma of the depths, if our theory be true, is atomized, so to speak, into a gaseous emulsion whose temper ature, although initially above that of the crater lava in the “ Strombolian” “phase, will be very quickly lowered by the enormous expansion—a degree of extension which can only be fully appreciated by one who, like the present writer, witnessed at close range the ereat emission of April 8th, 1906, whose volutes of vapor, even at the height of two kilometers , still expanded with an ineredible acceleration in all directions. No witness of that great, con- tinuing, trepanning blast could ever be persuaded that there remained in the throat of the volcano any broken solid materials whose trituration should furnish the ash of that and the follow- ing days, according to the ordinary conception. We must accept the absorption of heat by gaseous expansion as the cause of a lower temperature or else admit for the depths of the volcanic conduit and the magmatic pocket or fissure a temperature inferior tg, that of some imtermediate zone, as is often the case with geysers. The “ Vulcanian ”’ phase needs study. Considering this principle of direct ash production as it may obtain in the phenomena of volcanoes of different types, it would seem that the highly viscous lavas of andesitic and trachytic nature might explode subaérially, upon sudden relief of pressure, into gas and divided solid material, causing such effects as the “ Nuées Ardentes” of Mt. Pelée. At the other extreme, the ultrabasic Kilauea shows ash strata several meters in depth and ash fields many kilometers in extent, and there seems no reason to doubt that a sufficiently rapid outflow at a low level will cause even the Hawaiian subjacent lavas to froth up and be ejected as ash. In this connection it may be noted that Dana,* writing of the thread-lace scoriz of Kilauea, suggests that a further subdivi1- sion or frothing of the material might produce ash, and claimed for such a process an explanation of ash formation more reasonable than the trituration of rocks. Taking this with the many comments on the identity of ash and co-eval lava, it would seem that only a dynamic theory was required for the rounding out of a complete hypothesis. May not this conception of a subdivisional, intermolecular gaseous development and expansion in subjacent magma, upon relief of pressure from above, be accepted as a plausible explanation of the observed facts % Naples, July, 1912. * “ Characteristics of Voleanoes.” (Page 166.) 414 Pogue and Goldschmidt— Quartz from Alexander Co. Art. XXXVII.—On Quartz from Alewander County, North Carolina;* by J. E. Poaur in Washington and V. Goxp- scumipt in Heidelberg. Tur complex development of quartz crystals from Alexan- der County has been set forth in an important and thorough contribution by G. vom Rath.t This occurrence, howeve: en: shows such a wealth of forms that further investigation yields many new ones. In the United States National Museum at Washington there is a large number of crystals from this source, and the private collection of V. Goldschmidt at Heidelberg also includes an extensive series. Two crystals of peculiar interest, chosen from the two collections, form the subject of this paper. Their description may serve at the same time as an illustra- tion of the methods employed in an investigation of this kind. The measurements were made upon the two-circle goniometer in Heidelberg. Crystat 1. (U. 8S. National Museum. Cat. No. 82917.) Dimensions: 9x15"". Right-handed. Figs. 1a, 1b, 1c, show the crystal from three sides in natural development. The dis- tribution of the faces is represented in the gnomonic projec- tion (fig. 3). The forms identified appear in the following table: * Inetien some ae tee b rT p ® h K g 0 f 0 Symb. Gt. pg-- © 0 +10 —10 —#0 +20 —20 +30 —30 +40 —66 — Bravais- 1010 1011 1011 9097 2021 2021 3031 3031 4041 6061 * * Whether = == poses s L i Pe f. £. MN Symb. Gt. pq-- Le sist La cee al 21 =F —= Bravais | 112i 3253 7-6u13-7 AS fausep2) Oilele 7 onl. Of these forms those marked with an asterisk are new for quartz, ¢ = — 40(9097) is represented by three good faces. Meas- ured and calculated values are in close agreement. Measured dp = 0°05’; 58°37’ Calculated dp = 0°00’ ; 58°31’ Or 10's 58°28) 0°01'; 58°35’ The form may be accepted as certain. * Published by permission of the Secretary of the Smithsonian Institution. + Zeituchr, Kryst. vol. x, pp. 156-173, 1885. See also Hidden & Wash- ington, this Journal, vol. xxxiii, p. 507, 1887, and Miers, ibid., vol. xlvi, pp. 420-424, 1893. Etch-figures on quartz from this locality have been described by Molengraaff, Zeitschr. Kryst. vol. xvii, pp. 167-176, 1890. 416 Pogue and Goldschmidt— Quartz from Alewander Co. f= +14 (7-6:13°7) appears with 3 well developed faces. Measurement and calculations agree well. 7 Calculated ¢p =27°27'; 63°56’ 97°45's 64°08’ 27 The form may be considered as established. 7 = — 41 (4378) possesses 2 well developed faces. ured and calculated values agree closely. Meas- Measured ¢p = 25°13’ ; 68°43’ Calculated gp = 25°17' ; 68°47’ 24°56! ; 68°59’ This form may also be accepted as certain. From an attempted discussion of the new forms according to their number series,* it appeared that their relations were not clear, due to the uncertainty of many published symbols. The latter are often questionable, not only numerically, but also in respect to their + or — character. Clearness in this direction can only be attained by a critical revision of quartz, following a long series of observations on good material. Crrstat 2. (Collection of V. Goldschmidt.) Dimensions: 10x18". Left-handed. Figs. 2a, 2b, 2¢, show the crystal from three sides in natural development. The distribution of the faces appear in the gnomonic projection (fig. 4). The forms identified are shown in the following table : * Ihetteros-c soe b r p t h oO. g 7 Symb. Gt. pg oO +10 —10 +%0 +420 +40 +30 —30 — Bravais 1010 1011 1011 5053 2021 5052 3031 3031 cis * % Letters 2 32 224. is) id N: iy ¢ c C B Symb. Gt. pq +100 —190 +40 +40 —40 460 +70 —70 — Bravais 10°0°10°3 10:0°108 7072 4041 4041 6061 7071 7071 * % L@WigP 2 om cee Vv s Mn No n: It. v. a. Symb. Gt. pg —11°0 1 +31 —31 +491 —1,91 +41 —51 — Bravais 11:0°ll'1 1121 3141 3141 10°3:13°3 10°3:13°3 4151 5161 Of these forms, those designated with an asterisk are new They are well established through their measure- ments, as follows: * See Zeitschr. Kryst., vol. xxviii, pp. 1-35, 414-451, 1897. for quartz. krystal 1 Rechts QUuarx CLs 1 oe) The new form —1,0 is in accord with the series, which, however, is not the case with —230. The latter was given by Des Cloizeauxt as ers, and by Ratht as —22R. Their measure- ments, however, agree rather with —+,°0, as the following tabu- lation will show : * For explanation of such a discussion consult Zeitschr. Kryst., vol. xxviii, pp. 1-35 ; 414-451, 1897. nae sur la cryst. et la struct. int. duQuartz. Paris, 1858. Pp. 418- 432; 528. ; + Zeitschr. Kryst., vol. x, p. 159, 1885. Pogue and Goldschmidt— Quartz from Alewander Co. 419 Measured— Des Cloizeaux (p. 528); 3: et) =—10: _230= 24°49" Rath (p. 159); —R:—28R=— 10: —280=25 03 —24°55 Calculated for —10: —4,*0 = 24 55 Calculated for —10: —230 =24 46 Therefore —1,°0 appears as certain ; —280 as questionable. nu: = +11 belongs to the following zone section: enter == eer u, It: y. Symbol pg= +31 +101 +41 jo) = 0 $ 1 v eae SS 0 $ ioe) Therefore + +,°1 passes well in the series. For wn = — 4°1 we have the followirig: Weriot= == Ww ta z 1)" Symbol pg = — 31 — 41 = oh = 4 aD =O) 5 + 1 Sa ne 1 il e) ey) 3 4 1 also passes well in the series. The positive and negative rhombohedrons (+10 and —10) were determined by etching with hydrofluoric acid, and to avoid injuring the specimens the following -procedure was adopted. Crystal 1 belonged to a series of crystals of such similar development that, without danger of error, one of the others was employed for etching. Crystal 2 not permitting of such determination by analogy, the six rhombohedron faces +10 were accurately cut upon its lower, undeveloped end, by the aid of a grinding goniometer.* These artificially developed faces were then etched in the ordinary way, the eel faces being protected by a coating obtaimed by dipping the upper portion of the crystal in a mixture of molten sulphur and fluor- spar. Such a melt was used for a similar purpose by Ph. Hochschild} in a recent investigation on sphalerite. It finally remained to determine optically the right- and left- handed character of the two crystals by observing the nature of the light rotation in basal section. This was effected with- * The description of this apparatus has not yet been published + Neues Jahrb. f£. Min., ete., Beilage-Band, vol. xxvi, pp. 209- 210, 1908. 420 Pogue and Goldschmidit—Quartz from Alexander Co. out further harm to the specimens by the aid of the same erinding goniometer; whereby an oriented plane parallel to the base was prepared upon the lower end of each erystal, yielding when sliced off and polished an accurate basal section, examinable under the microscope. Crystal 1 showed itself as right-handed quartz; crystal 2 as left-handed quartz. This determination is in agreement with the rule* that the nature of light rotation in quartz may be deduced from the relative positions of the trapezohedrons. Thus, in right-handed quartz the trapezohedron faces, both + and — lie to the right of +10; and reversely, for left- -handed quartz, the trapezohedron faces are to the left of +10. These relations for each erystal are clearly brought out in the gno- monic projections (figs. 38 and 4). “ U.S. National Museum, Washington, and Heidelberg, Germany. * See Des Cloizeaux, Manuel de Minéralogie, Paris, 1862, vol. i, p. 15. NV. E. Stevens— New Jersey Palmoxyton. 421 Arr. XXXVIIIL—A Palm from the Upper Cretaceous of New Jersey; by Nem E. Stevens. Tur silicified palm stump which forms the basis of the present study was presented to the Peabody Museum of Yale University by Mr. R. W. Deforest in 1893. But it received no special examination for some ten years, when Dr. G. R. Wieland made several sections from the roots. These sections showed that the structure was unusually well conserved, and Fia. 1. Fie. 1. Palmoxylon anchorus.. Lateral view of entire specimen. -x 2/5. Photograph by G. E. Nichols. finally in the spring of 1911 the specimen was, at Dr. Wieland’s suggestion, turned over to the writer for definite study. The various additional sectious made by the writer for this study have been deposited, together with the original sections and the type, in the Paleobotanic Collections of Peabody Museum. The fossil was found by Mr. Deforest on the beach at Sea- bright, not far from Sandy Hook, and comes accordingly from near the limit of the Upper Cretaceous outcrops on the Jersey Shore. Other specimens, in less perfect preservation, were seen, though the present specimen is the only one so far recovered. The matrix appears to have been a marl, or per- haps clay with little or no lime. This specimen (fig. 1) con- sists of the much-eroded base of the trunk of a large palm with Am. Jour. Sci.—FourtTH SERIES, Vou. XXXIV, No. 203.—Novemper, 1912. 28 422 NV. EF. Stevens—New Jersey Palmoxylon. the proximal portions of the heavy and dense clump of attached roots. Whether most of the wear preceded silicification or not must of course be largely a matter of conjecture. Stem. As will be seen from figs. 1 and 2, comparatively little of the once large stem remains and, on the whole, the preservation of the stem parts is not nearly so good as that of the roots. Fic. 2. Longitudinal section of svecimen near the middle, showing erowded roots and small amount of wood. x 2/5. Photograph by G. H. Nichols. The latter are in an almost perfect state of preservation, far exceeding that of any fossil palm roots hitherto described. This circumstance, together with the fact that fossil palm roots have so rarely been found and have not been very fully de- scribed, made a rather careful study of the anatomy of this specimen seem worth while. In the following paper no attempt is made to review the literature either on fossil palms or on the anatomy of palm roots, as both subjects have been treated at length in recent monographs. Stenzel* describes sixty-two species of fossil palm * Stenzel, K. G.—Fossilen Palmenohdélzer. Beitriige zur Paldontologie and Geologie Osterreich-Ungarns und des Orients, Band XVI, Heft IV, p. 1-182, 1904, NV. E. Stevens—New Jersey Palmoxylon. 493 woods and four species of fossil palm roots, while Drabble* has studied the root anatomy in a large number of living spe- cies of palm. These papers give full citations of the literature in their respective fields. The maximum height of the specimen (ef. figs. 1 and 2) is 15™, its breadth 23, its length 28°", and only a small portion of the base of the stem remains, most of which is less well pre- served than the roots. Consequently the amount of material from which stem sections could be made was not large. More- Fie. 3. Outline showing relative size and arrangement of vascular bundles in the inner portion of the stem. x 5. The bundle shown in fig. 5 was taken from this slide. Fie. 4. Outline showing relative size and arrangement of vascular bundles in the outer portion of the stem. x 5. The bundles shown in figs. 6 and 7 were taken from this slide. over, in this basal region of the stem the course of the bundles is considerably disturbed, so that any section cuts comparatively few bundles at right angles. However, several small sections were obtained with the parenchyma of the stem and the lig- nified portions of the fibrovascular bundles in a good state of preservation. The fact that none of the bundles retain their phloem elements is perchance accounted for by the presence of numerous fungus hyphe. The parenchyma of the stem shows no unusual features, and there are no bast strands between the vascular bnndles, so that in Unger’s classification this species would belong to the second great group. Stenzei, however (p. 43), points out that this method of classification is unsatisfactory, and substitutes a sys- tem based chiefly on a comparison of the arrangement, prox- imity, structure, and size of the bundles in the outer and inner regions of the stem; and on the shape and size of the scle- renchyma portion of the fibrovascular bundles. In our specimen no marked difference could be detected between the inner and outer bundles (compare figs. 3 and 4), * Drabble, Eric—On the Anatomy of the Rootsof Palms. Transactions of the Linnean Society, Second Series—Botany, vol. vi, p. 427-487, 1905. 494 NV. £. Stevens—New Jersey Palmoxylon. so it apparently belongs in Stenzel’s class C..—the “Cocoa resembling stems.” Again, from the shape of the scleren- chyma portion of the fibrovascular bundles it should be placed in the group “ Reniformia” (p. 215), which is characterized by having the sclerenchyma portion of the fibrovascular bundle round or oval in cross section with a flat even surface or broad shallow indentation where it joins the vascular portion. This specimen does not, however, very closely resemble any one of the tive species ascribed to the “ Reniformia.’ A typical stem bundle, that is one of the “ longitudinal bundles,” is characterized (fig. 5) by having in the xylem few but rather large vessels with thick walls. As shown in the figure, the scler enchyma portion of the bundle is nearly oval in outline with a very slight indentation where it joins the vas- cular portion. The sclerenchyma cells near the phloem are considerably smaller than those farther away. No scleren- chyma, fibers are present on the axial side of the vascular bundle. The parenchyma cells adjoining the bundles are some- what smaller than those in the remainder of the stem, and the majority of those adjoining the vascular portion are somewhat elongated with the long axis perpendicular to the surface of the bundle. Besides the longitudinal bundles, a few bundles were found weet apparently belong be the classes designated by Stenzel “ Ubergangsbiindel Tor “transition bundles, ” and ‘ Kreu- maneebundele or Ean bundles.” By a “tr ansition bundle” Stenzel (p. 189) refers to the region where a bundle that goes up through the stem, that is a longitudinal bundle, turns to go out into a leaf. Jn this transition region the structure resem- bles somewhat that of a bundle going out intoa leaf. Stenzel* describes bundles of this type as follows: “The bast region is often smaller while the vascular region is larger ; the periphe- ral vessels (that is those nearest the phloem) are more widely separated and often more numerous, while in place of median vessels we find two or more lateral ones. This type of bundle is especially distinguished by the presence of numerous smaller vessels which are found chiefly toward the axial side of the vascular region.” It will be noted that the bundle shown in fig. 6 agrees very closely with Stenzel’s description of a “transition bundle,” although the sclerenchyma region is not markedly smaller than in the longitudinal bundles. This bundle, which is typical of several found in the sections examined, resembles very closely the transition bundles of Palmoxylon Aschersoni. (See Sten- zel, p. 140, fig. 234.) By the c Kreuzungsbiindel” Stenzel means those bundles which lead out toward a leaf and so are inclined at a very * Loe. cit. 1, p. 139. N. E. Stevens—New Jersey Palmoxylon. 425 slight angle. They may also be designated as “ oblique bun- dles.” The structure of oblique bundles, according to Stenzel (p. 140), differs from that of longitudinal bundles as follows : “The bast portion of oblique bundles is similar to that of the longitudinal bundles, though very often smaller. The vas- cular region is much larger and prolonged inwards. Axially to the large peripheral vessels (those nearest the phloem) and separated from them by a region or zone of parenchyma, is a Fie. 5. Fie. 6. €] IS $i O26 Oney Ie (RIE SSRIS yesece weeos : S See ) a0 S Hen CS wae 1S GBs ee Fie. 7. An “‘ oblique” bundle showing the bast region and the xyiem con- taining large peripheral vessels and a group of numerous smaller vessels separated from the large ones by a zone of parenchyma. x 60. Hie. 8. Longitudinal section of a portion of a small stem-vessel, showing tyloses. x 150. Fic. 9. Cross section of a root vessel, showing tyloses. x 1950. Fic. 10. Longitudinal section of a portion of a stem-vessel, showing fnngus hyphe. x 100. Fic. 11. Cross section of a root vessel showing fungus hyphe. x 150. ture on sapwood of Liriodendron tulipifera, develops a con- siderable mycelium extending through most of the woody tissue long before any marked effect on the lignified walls is apparent under the microscope. The fossil fungus noted here had apparently reached just this stage, having developed a considerable mycelium and destroyed much of the phloem, without affecting to any extent the more resistant tissues of the vessels and sclerenchyma. 428 N. E. Stevens—New Jersey Palmoxylon. The Root. A cross section of the root clump shows the roots closely packed together with but little space between them for some distance below their insertion. The fully developed roots, that is the large ones, are usually somewhat oval in section and about 8 or 9"™ by 5 or 6™ in diameter. Figure 12 shows a Fre. 12. Fie. 12. Polished surface cut through root clump at right angles to the roots about one inch below the region of insertion, showing marked variation in the size of the closely packed and appressed roots. x 4/5: number of these roots as they appear in section. The stele of the larger roots is about 2™™ in diameter. The numerous smaller roots vary in size down to one millimeter or less in diameter. It is, of course, impossible to determine ina section whether any particular root arises directly from the stem or is a branch of a larger root; but it is evident from the longitu- dinal sections that branching of the roots is frequent. So it seems probable that a large part of the smaller roots are branches of larger ones. Cortex.—Drabble (p. 432) divides the cortex of living palm roots into four well-differentiated regions, viz.: (1) outer limiting layer, (2) outer cortex, (3) inner cortex, and (4) endo- dermis. These four regions are described by Drabble as NV. E. Stevens—New Jersey Palmoxylon. 429 follows: The outer limiting layer is composed of cells with cuticularized, comparatively thin, walls. It does not neces- sarily form a perfectly regular sheath. The owter cortex (p. 482) is a teeumentary system consisting of several layers of elongated, lignified, and more or less pitted cells. The inner cortex (p. 434) is composed of three “zones”; an outer zone usually without air spaces, a broad middle zone with large air spaces, and an internal zone of regular cells. The middle zone Fie. 18. Fic. 18. Inner portion of cortex of large root, showing endodermis (EF), and part of the inner cortex, the inner zone of which (I) is composed of three layers of thin-walled cells. Inthe middle zone of the inner cortex are shown large intercellular spaces and the heavier lignified stone cells (A). x 10. Fic. 14. Outer portion of cortex of same root as shown in fig. 13, showing outer cortex (O shaded), outer zone of inner cortex, and a portion of the middle zone of the inner cortex. x 70. shows considerable variation in the shape, size, and number of the air spaces as well as in the number of cell layers separating these spaces. The endodermis (p. 438) consists of a single layer (sometimes locally doubled) of lignified cells from three to six times as long as they are broad. With the exception of the “limiting layer” these same regions were readily distinguishable in the present fossil specimen. Figures 13 and 14 were taken from the cortex of a tully developed root, fig. 13 being, of course, the inner and fig. 14 the outer portion. The photomicrograph, fig. 15, will give an idea of the relative size of the various zones of the inner cortex. ) 430 NV. EF. Stevens—New Jersey Palmoxylon. The outer cortex consists of from six to ten layers of elongated, lignitied cells with very thick walls and small lumen. The outer zone of the inner cortéx is also composed of lignified cells, but they are thinner-walled and shorter in proportion to their length. The cells of this region differ considerably in size, the outermost cells having about the same diameter as the adjacent cells of the outer cortex, while toward the center the cells become progressively larger in diameter. — The middle zone of the inner cortex contains numerous large air spaces. These air spaces (lacunee) are radially arranged, six to ten times as long as broad, with from one to Fie. 15. Fic. 15. Photomicrograph of cross section of palm root, showing cortex with large radial intercellular spaces in the middle zone; and stele with alternate phloem and protoxylem groups and eight internal vessels. x 8. six layers of cells separating them. The cells of this region are only about twice as long as they are broad and much thin- ner-walled than those of the outer region. In addition to the large radial air spaces, triangular spaces show plainly in the longitudinal section at the intersection of cell walls (fig. 16). The cells are apparently “ lignified parenchyma.” There occur also in this middle zone thick-walled cells with large cavity and large pits. These usually occur singly or scattered through the middle zone in groups of two or three, but are somewhat more numerous toward the inside. Seen in cross section, figs. 13 and 14, they usually appear rather round; and in fully developed roots the pits are not usually seen in cross sections. In smaller roots, however (fig. 22), the large pits are very evident. Figure 17 shows two of these NV. £. Stevens—New Jersey Palmoxylton. 431 large cells from near the origin of a root, in longitudinal sec- tion. It will be noted that these cells are from three to five times as long as they are broad and that the large pits are con- siderably elongated. It is difficult to place these cells in any of the recognized categories of lignified elements. They do not seem to correspond exactly to either the “Kentia” or “ Raphia” types of fibers described Dy Drabble (p. 435), but may perhaps be designated as “stone” cells. The internal zone of the inner cortex usually consists of three layers of rather thin-walled cells closely packed together without intercellular spaces. These cells vary considerably in size but are usually from one to three times as long as they are broad. Compare fig. 13 with fig. 16. The endodermis is almost uniformly one cell thick, the cells two to four times as long as broad; and even in the fossil the radial walls appear markedly thicker than the tangential walls. Compare fig. 13 with fie. 16. All the regions described for the fully developed root can be made out in the smaller root shown in fio. 22. In the smaller root, however, the various parts of the inner cortex are not so clearly differentiated and all the cells have much thinner walls. Nuclei in cells of inner cortex.—Three longitudinal sections showed the parenchyma cells of the inner cortex in a very remarkable state of preservation. Indeed, the majority of the cells of this region contained structure so characteristic in appearance and so constant in occurrence that, if seen in fixed material from living plant tissue, they would unhesitatingly be described as the well-stained nuclez. Drawings of such structures would naturally afford no cer- tainty as to their nature ; and the photomicrographs, figs. 18 and 19, are accordingly offered for what they may be worth as proof. While not fully convincing in themselves these figures are not wholly valueless. Practically every parenchyma cell in the field showed a nucleus in some focus ; and in the figures nuclei appear in the cells marked N as well as in some others. It will also be noted that the triangular intercellular spaces spoken of above clearly appear in these photomicrographs ; and the probability of the cells having been “fixed ” when in an actively growing condition is further denoted by the occurrence of pairs of cells which have apparently just been separated by a cross wall (y, fig. 19). In these “ daughter cells” the nuclei are still close to the dividing wall. The Stele.—The pericycle is very readily distinguished as a single layer, or sometimes locally as two layers, of rather regu- lar cells inside the endodermis. These cells tend to be some- 432 AV. EF. Stevens—New Jersey Palmoxylon. what larger than the parenchyma cells adjoining them and their radial walls are for the most part regularly perpendicular to the walls of the endodermal cells. Compare figs. 20, 21, and 23. The vascular portions show the typical root arrangement, the phloem strands alternating with the protoxylem groups. Fie. 16, Fie. 17. Fie. 16. Longitudinal section of inner region of cortex, showing endo- dermis (E), and a portion of the inner cortex with lignified parenchyma and inner zone of three layers of cells (I). x 112. Fic. 17. Two of the large ‘‘stone” cells of the inner cortex in longitu- dinal section, showing large pits. x 112. Fic. 18. Fie. 19. Fies, 18 and 19. Photomicographs of parenchyma cells of the inner cortex in longitudinal section, showing nuclei (N) and cells which by their shape seem to indicate recent division (y). Triangular intercellular spaces may be seen insome cases. x 100+. In the larger roots there are usually fifty or more protoxylem groups; in the root from which fig. 20 was taken there were fifty-five. The phloem groups and all the outer xylem ele- ments are surrounded by a continuous zone of dense scleren- chyma fibers. These sclerenchyma fibers are considerably smaller near the vascular portions than they are toward the center of the root. This dense sclerenchyma band is usually N. E. Stevens—New Jersey Palmouylon. 433 from fifteen to twenty-five cells wide, so that in the large roots the sclerenchyma region with the vascular portions it contains occupies only about half the stele. The central portion of the stele contains a variable number of “internal” vessels, frequently from four to six, which do not appear to be referable to any particular xylem group. Compare (Drabble, p. 441). Each of these internal vessels is surrounded by a region of dense sclerenchyma from six to ten cells in width. See fig. 20. The remainder of the central Fie. 20. ) oa oN OSs: O83 Fie. 20. Portions of stele of large root with endodermis, showing peri- cycle, alternate phloem (P) and protoxylem (X) groups in a zone of dense sclerenchyma. The inner region of the stele shows lignified parenchyma with large intercellular spaces, and an internal vessel (I) surrounded by a zone of sclerenchyma. x 175. region of the stele is occupied by lignitied cells with large lumen and comparatively thin walls. These cells apparently became torn apart as the root increased in size so that, in fully developed roots, this central region has large intercellular spaces. The stele of the smaller roots shows the same structure as that of the larger ones, but the number of protoxylem groups is much smaller. The root shown in fig. 21 had thirteen pro- 434 NV. E. Stevens New Jersey Palmoxylon. toxylem groups. In these smaller roots, of course, the central region is much smaller in proportion to the whole stele and contains no internal vessels or large intercellular spaces. One of the most interesting sections found was that of the very small root shown in fig. 23. While no phloem elements ‘an be distinguished in this root section, the xylem shows the arrangement of a very young tetrarch root. In this very small root, as in the larger ones, the pericycle may be readily dis- tinguished. Unfortunately the cortical portion of this root was obscured by a deposit of some foreign substance. Phloem.—As noted above, the phloem was preserved in com- paratively few cases. Not more than five of the root sections a Figs. 21-24. Fig. 21. Stele of a smaller root with endodermis, showing pericycle, alter- nate phloem and xylem groups, and a zone of sclerenchyma. x 175. Fie, 22. Cortex of same root as that shown in fig. 21, endodermis (E), stone cells (A), outer cortex (O). x 175. Fic. 23. Stele of very small root, the xylem of which has the typical tetrarch arrangement. x 175. Fic. 24. Phloem group from large root. x 265. showed well preserved phloem. Even in the root shown in fig. 20, less than half of the phloem groups remained. Fig. 24 shows, however, a group which was unusually well pre- served, and this may perhaps be taken as typical. It will be noted that the protophloem consists of small cells with thin walls while the sieve-tubes of the metaphloem are very much N. E. Stevens—New Jersey Palmoxylon. 435 larger and have thicker walls. No sieve-plates could be dis- covered in any of the sections. Neither could companion cells be definitely distinguished. It seems probable, however, that the small dark regions close to the sieve-tubes, shown in fig. 24 by the shaded areas, occupy the position of companion cells. All the preserved phloem groups were in the larger roots. Junction of bundles of root and stem.—The transition from the solid stem to the root region of variable hardness renders the making of thin sections through the root junctions difficult, the more so because the parenchyma and other tissues in this region have taken on so dark a color that sections of moderate thickness are not very useful. These circumstances prevented satisfactory study of the junction of the root and stem bundles. Description of species. Palmoxylon anchorus, sp. nov. Locality.—U pper Cretaceous, Seabright, New Jersey. Type in Peabody Museum, Yale University. Stem.—No bast strands between the fibrovascular bundles. Sclerenchyma portions of the fibrovascular bundles usually oval in cross section and but little indented next the phloem. Little difference in the size and shape of the sclerenchyma portions of the “longitudinal,” “transition,” and “ oblique ” bundles. Loots.—Roots considerably branched. Outer cortex com- posed of six to ten layers of elongated, lignified cells. Inner cortex differentiated into three zones: an outer region of thick- walled lignified cells; a middle zone of lignified parenchyma having numerous large radially arranged air spaces, six to ten times as long as broad, separated by from one to six layers of cells; and an internal zone of three layers of thin-walled closely packed cells. Hndodermis usually one cell thick. Pericycle usually one or locally two layers of cells. The larg- est roots have over fifty protoxylem groups alternating with phloem groups. Internal vessels present in the larger roots and surrounded by a dense region of sclerenchyma six to ten cells in width. * * * % * It is of much interest to append the fact that petrified stems of palms are not the rare objects that their noticeable absence from collections and general lack of mention in paleontologic texts would seem to imply. It is indeed probable that they 436 N. E. Stevens— New Jersey Palmoxylon. occur abundantly at times from the upper Cretaceous on, both on the coastal plain and in the formations of the continental interior. In addition to several Tertiary forms described by Knowlton, Hatcher is known to have secured various stems from the Laramie of Converse county, Wyoming. Wieland has col- lected from the Pierre the splendidly conserved stems arbi- trarily cailed by him Palmoxylon cheyennense.* Cannon has secured an abundance of exquisitely silicified stems from the Denver beds. And only last year Brown observed a large silicitied root clump in the “ Rattlesnake” beds on the “ big bend” of the Rio Grande in Chisos county, Texas. * This Journal, vol. xv, p. 216. E W. M. Foote—Shower of Meteorice Stones, Arizona. 487 Arr. XX XIX.—Preliminary Note on the Shower of Meteoric Stones near Lolbrook, Navajo County, Arizona, July 19th, 1912, including a Reference torthe Perseid Swarm of Meteors visible from July 11th to August 22d ; by Warren M. Foorr. History. Ir was doubtless the literary exaggerations of the 18th century and similar causes which prevented early geologists and astronomers from investigating the reports of falling sky- stones. But in the fatherland of yellow journalism we some- times find a journalistic restraint, under conditions that are worthy of remark, and which prove the labor of the news- gatherer to be of value to science. In the last week of July, the following account appeared in several Arizona papers: Friday evening about six-thirty a meteor, or some other body of a like nature, passed over Holbrook going almost due east at a rate of speed that would make a swift-moving express train seem as though it were standing dead still. The noise it created was very loud and lasted for at least a half a minute and sounded somewhat like distant thunder or the booming of a cannon in the distance. It left a large cloud of smoke in its trail and several of our citizens heard it explode a couple of times. A few saw it and nearly everyone heard the noise it made. Reports from Winslow are that several people saw the body pass over the town, and the noise was. heard at St. Joseph, Woodruff, Pinedale, and Concho. That either all or part of the body fell near the section house at Aztec, six miles east of here, there seems to be little doubt. k * * *k A few small pieces were brought in here. One piece larger than an orange fell into a tree in a yard at Aztec cutting the limb off slick and clean and falling to the ground, and when picked up was almost red-hot. Other particles of the body fell in the same vicinity and an eye-witness states that for about a mile to the east he could see little puffs of dust arising from the sand, evidently where fragments struck. *k * * k About two dozen people went to Aztec to pick up pieces of the meteor Sunday afternoon and the field is now pretty well cleaned up. The largest found weighed over 14 pounds, while several of about 5 pounds were picked up, and numerous small pieces. They are very brittle, heavy, and appear to have many small particles of iron in them. As the writer of the present article lacked the time for mak- ing the two-thousand mile journey from Philadelphia, the additional and confirmatory data were secured by correspond- Am. Jour. Sct.—FourtH SERIES, VoL. XXXIV, No. 203.—NovemBER, 1912. 29 438 W. MW. Foote—Shower of Meteoric Stones, Arizona. ence with witnesses of the fall and with finders of the stones. Following are the main facts of the fall and find, as gathered. Between 6.20 and 6.40 p.m. on July 19th, 1912, a large meteor was heard traveling in an easterly direction and pass- ing over Winslow, Holbrook, and Aztec, pomts along the Santa Fe Railroad, ‘which here parallels the Rio Puerco River. It made a very loud noise, lasting for half a minute to one minute.* This noise has been variously likened by witnesses, to the rumbling of a rapidly driven farm wagon on a rough road, to escaping steam, to distant or long continued thunder Fig, 1. oCaonen Diablo Pinedale © Fic. 1. Location of Fall. Aztec, near Holbrook, Navajo County, Arizona. 34° 57’ N., 110° 2’ W or the booming of a cannon. It was heard at Concho, St. Joseph and Woodruff and at Pinedale, some forty miles away. One large explosion was quickly followed by several small ones in “rapid succession. Charles Von Aachen and his son then saw numerous stones fall at Aztec, raising many puffs of dust for a mile or more over the dry sand of the desert, like those produced by bullets or the first drops of rain in a heavy shower. They didnot see the stones in the air. Some fell neara building, and one is said to have severed the branch of a tree. The meteor was not seen during its flight, as it was too early in the evening for its luminosity to be visible. Its speed could not be estimated, but it was “terrific” according to one account. Its *One observer states that the loud reports were followed by lesser rumblings fer four or five minutes, These were the usual echoes, W. M. Foote—Shower of Meteoric Stones, Arizona. 489 Photographs by Bond Bros., Philadelphia. Fic. 2. “‘ Holbrook peas.” 360 complete boloids, Full size, about 4 to 8mm Jong. 0:1 to 0°3 grams or 15 to 5 grains. Total weight, 70 grams. 440 W. DL. Foote—Shower of Meteoric Stones, Arizona. path was indicated to many by a train of thin smoky vapor which spread out after the meteor passed. One observer esti- mated that the explosion occurred one or two miles above the earth. The weather at the time was slightly cloudy. The stones were scattered over an ellipsoidal area roughly estimated by two finders to be about one-half mile wide and three] miles long. As frequently recorded in meteoric falls, the longest diameter of this ellipsoid was in line with the tra- jectory of the meteor, being east and west. Most of the Fig, 3. Fie. 3. Characteristic pyramidal brustseite, with two rear corners broken and later fused. x 1: diameters. smaller fragments lay on the top of the loose sandy soil; the larger pieces were about half buried, some to a depth of six inches, apparently having fallen slantinely from the west. The large and small stones, according to all answers received, were said to be indiscriminately spread over the ground, without regard to size.* In previous stone showers the small stones have been found first in the line of flight, then the medium, and finally the largest. The violent disrnptions near Hol- brook might account for the lack of such separation of the Sizes, provided an explosion occurred near the end of the flight. Just such a late disruption was evidenced by the nearly *See further, note on p. 456. W. M. Foote—Shower of Meteoric Stones, Arizona. 441 fresh fracture of many fragments. Visitors from nearby towns soon gathered the larger stones. Von Achen, who saw them fall, reported that they were too hot to pickup. Two accounts state that they became lighter in color after cooling. Except for about ten kilos sent away, all were acquired by the Foote Mineral Company of Philadelphia. There is an Aztec post-office in Yuma Oo., Arizona, but no post-office or telegraph station at Aztec, Navajo Co. Hence Fia. 4, Fic. 4. Specimen three-quarters buried in wet soil, resulting in rust and exudation of molysite. Checked surface exposed. Oblique angle of flight indicated by soil line. x1- diameters. the name of Holbrook, six miles distant, is used to designate the fall. Macroscopic Features. Externally the stones present all the commoner character- istics of aérolites. The primary crust, begun on the entrance of the meteor into our atmosphere with its high planetary veloc- ity, and prior to the first explosion, is almost universally pres- ent. It coats broadly rounded surfaces and is generally dull black, being about 0-3" thick. A checking or crackling of this crust, due to unequal expansion, is often noticeable, as shown in fig. 4. The secondary crust, formed on the fractured surfaces produced by this first disruption, is somewhat shiny 442 W. M. Foote—Shower of Meteoric Stones, Arizona. and thinner than the older crust. Moreover, the fractures it covers are hackly and irregular, and it even fails to hide occa- sional protruding chondrules, indicating that the superficial dissipation by combustion had not proceeded far enough to round off the sharper corners and smaller prominences. What may be termed a tertiary crust was begun subsequently to the second explosion when nearing the ground at reduced velocity. The genesis of this is most interestingly shown in various stages. A slight discoloration, especially a tarnishing of the metallic minerals, is sometimes seen. A mottled smoking of the surface and an incomplete incrustation of small patches is quite common and grades into a thin filmy crust (see fig. 5). Fic. 5. Fig. 5, Primary crust at left. Tertiary crust beginning as a smoky alter- ation. x 0°85 diameters. On many of the smaller stones this tertiary crust is fully devel- oped at the edges of the primary crust, being in fact a labiate overflow of the latter. The newly fractured area often shows the various degrees of fusion as its center is approached, where in some cases only a smoky alteration may be seen. Rarely the primary, secondary, and tertiary incrustations are exhib- ited in the same fragment. Several dozen individuals showed the characteristic radial flowage lines of viscous stone from the front, or brustseite, to the back. This flow is due to the backward rush of air over the molten surface. It was noted in some pieces of not over W. M. Foote—Shower of Meteoric Stones, Arizona, 443 1°" diameter. The most deeply and unusually marked brust- seite found was on a large stone of 2,400 grams, shown in figures 6 and 7. The wholly unique character of this piece suggests that it was the front end of the original large mass, or one of several large masses, which entered the earth’s atmosphere. Here the lines of flow are merged and sometimes originate in deep pits, or piezoglyphs, probably caused by a Fie. 6. Fic. 6. Brustseite with deeply marked radial fusion flow (reconstructed). x 0°64 diameters. See also Fig. 7. differential fusing or fracturing of the surface from heat and very rarely by the burning out of nodules. Unfortunately the finder of this mass treated it with scant respect, and it reached Philadelphia in three fragments scattered among thousands of other stones. The edges were much bruised from rough hand- ling, so that the reconstructed corner and cracks, shown plainly 444 W. MW. Foote—Shower of Meteoric Stones, Arizona. in the engraving, are comparatively large. The back of the stone (fig. 7) illustrates well the relatively quiet fusion of all of the similarly marked masses. In these the overflow of the molten silicates behind the rear edges is shown sometimes in a fringe-like scoriaceous “wash” or thicker crust. Otherwise Fie. 7. Fic. 7. Back of mass shown in fig. 6 (reconstructed). x 0°64 diameters. the back plainly appears as an area of lesser fusional disturb- ance than the front. Several of these specimens are superficially identical with that of Gopalpur, described by Tschermak, as having a rounded front covered with a finely striped and radiately channeled crust, with elongated pit-like depressions, becoming shallower as they recede from the radiant point. Gopalpur’s front crust overlaps the back in a well-defined and sometimes fringed border, becoming verrucose and enclosing unaltered grains of W. M. Foote—Shower of Meteoric Stones, Arizona. 445 the meteorite. One 450-gram Holbrook fragment with deeply furrowed brustseite similarly shows 1 to 2™™ fragments of unaltered stone enclosed in the back crust, the semi-fused area being 1 to 8™™ thick. A few well-marked brustseite stones (fig. 3) show fractures near the base; none at the head of the stone. This would indicate that the pressure of the air stream on the rear edge is a factor in disruption, as weJl as the expansion due to heat. In some instances the radial fusion flow is shown on the secondary crust of fragments, notably in figs. 8 and 9. Here the stone was apparently reversed in its flight Fie. 8. Fie. 9. Fie. 8. Front of stone, showing radial fusion flow on primary crust. Fre. 9. Back of same stone, showing flow on secondary crust after rever- sal of position in flight. x 0°9 diameters. after an early explosion, and a well-marked radial flow was left upon the new brustseite. In the examination of a jarge number of stones, the thick and minutely blebby character of the otherwise even crust on one face would indicate the back of the stone, whereas the fusion flow on the reverse, or front, might be but faintly marked, or even absent. The front is often brownish, the back being usually deep black. In rare cases the pits clearly result from the burning or frac- turing out of pyrrhotite nodules, as illustrated in fig. 10. At the lower end of the pit is a piece of the freshly fractured pyrrhotite. In the bottom of the pit is the smoothly altered slagey remnant of the original nodule, similar to the unaltered one shown in fig. 13. Some of these inclusions are sharply rectangular and possess a distinct parting. Qualitative tests 446 W. M. Foote—Shower of Meteoric Stones, Arizona. show the mineral to be essentially sulphide of iron. It may therefore be provisionally classed as pyrrhotite. No tendency toward any one fragmental form is observable except that the stones with well-marked fusion flow are generally of a roughly pyramidal or conical shape, the apex cor- responding with the radiant point of the fusion lines; the base is the back of the individual in flight. According to a local account, the rust noticed ona few stones was caused by rain between the falling and finding of the speci- mens. On these is observable the usual liquid exudation of moly- site (ferric chloride, FeCl,), the alteration of lawrencite (ferrous Fre. 10. ‘ Fic. 10. Piezoglyph formed by burning out of pyrrhotite nodule. x 0°77 diameters. chloride FeCl,). The remainder of the thousands of individuals examined seemed to be entirely stable and bear no signs of disintegration. On many pieces are traces of the sandy red- dish soil on which they fell. Two or three per cent showed a “soil line” (fig. 12) indicating clearly the depth of burial. A very few showed traces of the soil on all sides. This is, of course, not a reliable indication of the average depth of burial, since the amount of cleaning done by the finders cannot be deter- mined. The direction of this soil line, shown in fig. 4, would indicate that the angle of fall with the earth’s surface was about thirty to forty degrees. From the slight penetration of W. M. Foote—Shower of Meteoric Stones, Arizona. 447 the soil observed, the last stage of the flight was evidently not incomparable with the velocity of an ordinary falling body, as has been accurately calculated by investigators of previous falls. The relatively low temperature at the moment of reaching Fie. 11. Fic. 11. Flat brustseite. x 0°73 diameters. See fig. 12. Fie. 12. Fie. 12. Flat brustseite (down). Side view showing confirmatory soil line and scoriaceous overilow on back (up). x 0°76 diameters. the earth’s surface is suggested by one specimen with uncharred vegetable fiber adhering closely to the rough crust. The impact apparently pressed the fibers firmly into the minute interstices of the crust. 448 W. MW. Foote—Shower of Meteoric Stones, Arizona. Examination of a fractured surface shows a light ashy gray color and the granulated texture imperfectly reproduced in fig. 13. Irregular chondrules, visible to the naked eye, are com- mon, and here and there others become prominent by their spherical form breaking half free from the matrix. In some instances these hemispheres are quite perfect and of four or five millimeters diameter, the largest one, illustrated in tig. 14, reaching 11 millimeters. It has lost much of its definiteness in photographing. Under a lens, the chondrules breaking with the matrix are seen to be numerously distributed throughout the mass. A Fie. 138. Fic. 13. Fractured surface with pyrrhotite nodule. x 1° diameters. large nuinber of the chondrules are gray in color, others are whitish. Not infrequently a broken chondrule shows radio- foliate structure, sometimes with the radiant point at one edge. Microscopic Examination. This was made by Mr. W. Harold Tomlinson, whose report follows :— The new meteorite is an aérolite containing a very little native iron. There are three opaque minerals forming together perhaps 4 per cent of the volume. Native iron and pyrrhotite W. M. Foote—Shower of Meteoric Stones, Arizona. 449 Fic. 14. Fic. 14. Unusually large spherulitic chondrule, broken freely from the matrix. x 1° diameters. 450 W. M. Foote—Shower of Meteoric Stones, Arizona. occur in irregular but usually rounded patches and grains, in about equal amounts. The pyrrhotite oceasionally shows crystalline faces. Magnetite in small jet black or slightly bluish black grains occurs to less extent than the other metallic minerals. Fie. 15. Fie. 15. Imperfect polish showing distribution of nickel-iron. At large end a chondrule encircled by minute iron grains. x 1°‘7 diameters. The principal constituent of the stone is enstatite, which forms probably 50-60 per cent by volume. It occurs in prisms from 1™™ & *25™" in size down to minute allotriomor. phic grains. It also occurs often in chondrules with radiating structure (fig. 17), and in one section a chondrule was found with tangential structure, i.e. the fibers formed a regular polygon. Enstatite appears to have been one of the first minerals to separate and to have continued its separation until the magma cooled. In mass the enstatite has often a slightly ereenish color. It is colorless in section. Olivine and monoclinic pyroxene (diallage) make up the bal- ance of the stone. Olivine occurs usually in erystals and crystal grains set in a grayish glass. A group of olivine set in glass will be divided from the rest of the stone often by sharp demarcation as though it were an inclusion of another stone. W. M. Foote—Shower of Meteoric Stones, Arizona. 451 Olivine also occurs frequently in less perfect crystals asso- ciated with the enstatite, but segregation is rather character- istic of it in this stone. The diallage in mass is a brownish color. In section it has in some places a very slight pinkish tinge, but is usually color- less. It occurs in grains with few crystal boundaries and in large crystals. One large crystal measured approximately 2°5<2"™, Tt also occurs twinned or intergrown with the en- statite and in some of the chondrules it is twinned with the enstatite. There are some longish prisms with fibrous structur and small extinction angle that are also monoclinic pyroxene. Fic. 16. Fic. 16. Sharp octahedrons of spinel in quartz. x 30 diameters. Lower nicol in place. The most interesting feature observed was a patch of spinels set in quartz. Of ten sections examined, these minerals were found in only one. They shade in color ‘from clear and rather light ruby-red to ruby-brown. The darker are shining black by reflected light, and are probably chromite. On the light side of the patch the crystals are red by reflected light, and are therefore a ruby- or chrome-spinel (fig. 16). They occur in sharp octahedrons and are the most perfectly erys- tallized mineral in the stone. They are identified by color, erystal form, reflection, and position. The erystals are set in seml- crystalline quartz and the patch is edged with quartz that is slightly coarser. The granules of quartz are irreg- ular in shape, often interlocking, and show wavy extine- tion. No figure was obtained beyond an indistinct dot which 452 W. MM. Foote—Shower of Meteoric Stones, Arizona. had no effect on a quartz wedge. The quartz is sufliciently identified, however, by its low refraction, double-refraction showing colors up to white, and by the characteristic habit of the grains. These patches of spinel and quartz are analogous to the patches of olivine and glass previously described. The two minerals should probably be regarded as secondary minerals formed by a reaction (perhaps influenced by Cr,O,) between the olivine and anorthite or a glass near to anorthite in com- position. Chromite was found in three sections. It occurs associated with the metallic minerals. It is black by reflected light and dull red by transmitted light. Both patches and crystals were found. Fic. 17. Fie. 17. Chondrule of enstatite surrounded by native iron and pyrrhotite. x 16 diameters. Nicols crossed. Nickel-Iron Content. Mr. George C. Davis made the few chemical determinations of the aérolite which limited time permitted. The specific gravity was found to be 3:22. One hundred grams, freed from crust, were finely powdered and treated repeatedly with the electro-magnet. The metallic portion was washed with alcohol to remove the silicates, but the separation was incom- plete and the silicates finally constituted about 25 per cent of the magnetic concentrates. The material for analysis was taken from twelve individuals. W. M. Foote—Shower of Meteoric Stones, Arizona. 458 Weight magnetic portion, actual) 22355. 4°90 Weight non-magnetic portion, actual .---. - 95° Weight magnetic portion, corrected_.-. 3°675 The corrected weights give the following percentage for the aérolite : INWelceleinompy ee te te cia Scie 3°68 ilt@ a meaner mentee & eee wey Ge 96°32 100:00 Sili@aimeemierten tet n ns Oy ors peeps ede 14:12 JURGHEE AS 222) Bs Ame gk cn ao ee 60°64 aNGChce lee Oe es a i eee aE Corrected for 25 per cent non-magnetic minerals present. OTe ee ee ea ete ree Pag Soa fo 80°86 Astronomical Relationship. Astronomers have long debated the question as to whether the fall of the Mazapil (Mexico) iron* on Nov. 27th, 1885, proved it to be actually a member of the swarm of Leonids or November meteors seen numerously on the same night, or whether the occurrences were purely coincidental. The well-known August or Perseid meteors are seen from July 11th to August 22d (Denning), and reach a maximum on August 10th to 13th, while the lesser Aquarids reach their maximum on July 28th. Their nearness suggested an inquiry - as to a possible relationship between these star showers and the Holbrook meteorites. Apprized of the circumstances of the Holbrook fall, Prof. Eric Doolittle, Director of the Flower Observatory, University of Pennsylvania, wrote as follows :— October 5, 1912. I was greatly interested in your account of the extraordinary fall of meteoric material in Arizona on the evening of July 19th, last, and I take great pleasure in giving you what information I can regarding the known meteoric showers which may be expected at about this date. The nearest bright shower in point of time is undoubtedly that known as the Aquarid shower, which reaches its maximum on July *See this Journal, xxxiii, 221, 1885. Am. Jour. Sct.—Fourr Sprigs, Vou. XXXIV, No. 203.—NovemBer, 1912. 30 454 W. MW. Foote—Shower of Meteorie Stones, Arizona. 28,—but nine days later than the fall observed. Yet I hardly think it possible that the material actually collected can belong to this meteoric stream, for the reason that the region of the sky from which the Aquarids are seen to come had not. yet risen above the eastern horizon nor would this region begin to rise until about one hour later, And as at this time the constellation Aquarius would be seen in an almost due easterly direction, it seems not possible that a member of this stream could be seen coming from the west, as the Holbrook meteorite did. The next brilliant shower in point of time is the well-known Perseid shower, more commonly known as the shower of August shooting stars. Although this shower was most brilliant from August 10th-13th of the present year, the individual particles are so greatly scattered that straggling members may be seen for nearly a month before this date. The point among the stars from which the meteors of this stream are seen to come is constantly changing during this time, owing to our displacement in space caused by the motion of the earth. According to Mr. W. F. Denning,—the highest present authority upon this subject,—the occasional meteors from this shower witnessed on July 19th should apparently come from a point just without the borders of Perseus and within those of the constellation Cassiopeia. At the time under consideration this point of the heavens would be almost exactly in the north horizon, or at most but a degree or two to the west of north. ‘The greater part of the particles which reached us from this stream at this time should therefore be expected to approach our country from the north, and to at least begin their motion through our air in a path very nearly parallel to the ground. However a single, isolated member of the swarm might easily have had the direction of its motion greatly changed, either by the gravitational pull of the other members or by collision with them, so that a single such mass might be seen to enter our atmosphere from an unexpected direction.* It is therefore in my judgment not impossible that this most interesting fall might have come from the Perseid swarm, and therefore be an actual part of or an attendant to Tuttle’s Comet of 1862; but I do not think that from the data at hand we can now establish this connection. There are several other less striking showers due at about this time; notably, a second Perseid shower of faint, swiftly moving stars whose radiant is 14 degrees farther toward the south than that of the first, and also a shower coming from the constellation *That the observers’ accounts of momentary phenomena often conflict, is but natural. One witness wrote on October 18th, that on hearing the reports, he saw only a very large cloud of sand or smoke move from east to southwest, and after striking a black cloud, the first sand or smoke cloud traveled back to the east. While probably of little value, the account is recorded here because this large finder was the only one to fully indicate an ellipsoid with longest diameter east and west, in answering as to the area of the fall. W. M. Foote—Shower of Meteoric Stones, Arizona. 455 of Cygnus, which actually reaches its maximum on this very date,—July 19. But as Cygnus at the time indicated was in the northeastern part of the sky, it seems that this shower must be ex- cluded from consideration for the same reason that the Aquarid shower was dismissed. . . . —WSHric Doolittle. Quantitative Comparison. A most careful search by over one hundred persons was made under that stimulus which is usually found to be instantly effective. This search continued for two months. The discoy- eries of new stones rapidly rose and as rapidly dwindled to nothing. Following is an estimate of the entire fall. Among the 29 larger ones, three or four had an end broken off, pre- sumably by the finder; perhaps 5 per cent of the stones, counted as complete, had one face merely smoked, the balance were completely incrusted. Items 1, 2, 3, and 4 were received at Philadelphia. (1) 29 individuals over 1000 grams: 6665, 4264, 3470, 3122, 2940, 2605, 2520, 2500, 2480, 2463, 2442, 2318, 2270, 2250, 2050, 1893, 1860, 1816, 1780, 1558, 1464, 1400, 1330, 1272, 1190, 1148, 1120, 1100, 1020. Total, 64,310 grams (2) 6000 individuals of 1 gram to 1000 grams Ca Chee en ienee ee cues aS Oe A 6.000... 6 (3) 8000 individuals under 1 gram each-.--. -- 4,000 * (4) Fragments broken after finding (estimated) 4,000 “ (5) Many individuals of less than 1000 grams each, distributed as samples to institu- tions in July and carried off as curios by ISTUORSirs eras eats Sn ere aoe At ates TRO aa 10,000 <“ 14,029+ stones. Hstimated total weight of le CORES libs avid a eeeee ne cnc ee 218,310 « A record of the more notable stone showers of the 19th century was prepared for comparison. It will be observed that such showers are recorded but rarely, and that in point of number of stones, the Holbrook fall is one of the greatest in modern times. Another distinction it carries is the minute size of thousands of its individuals. ‘The smallest shown in fig. 2 average 10,000 to the kilogram, or about 4,536 to a pound avoirdupois: a single one weighs less than 0-1 gram or 14 grains, being smaller than the smallest of the Hessle stones, which were only found because they fell on the ice. These scarred and diminutive “ Holbrook peas” confirm the accepted opinion that far larger cosmic stones are usually quite consumed in the atmospheric passage. 456 W. MW. Foote—Shower of Meteoric Stones, Arizona. Principal Stone Showers of the 19th Century, The following data are gathered from Wiilfine’s “ Meteori- > Do ten”: Recorded Recorded Date Number Weight of of of Fall in Fall Locality Stones Grams April 26, 1803 L’Aigle, France 2000-3000 36,843 Dec, 14,1807 Weston, Connecticut Many 18,267 May 22,1808 Stannern, Austria 200-300 38,408 May 1, 1860 New Concord, Ohio Over 30 97,811 June 9, 1866 Knyahinya, Hungary Over 1000 423,120 Jan. 30, 1868 Pultusk, Poland About 100,000 200,932 Jan. 1, 1869 Hessle, Sweden Many 22,895 Feb. 12,1875 Homestead, Iowa Many 124,492 Feb. 3, 1882 Mocs, Hungary Over 100,000 155,632 May 2, 1890 Forest, lowa Many 122,037 July 19,1912 Holbrook, Arizona Over 14,000 218,310 No other aérolites are recorded from Arizona, the nearest meteoric fall being the siderite of Canyon Diablo, some sixty wiles distant. Philadelphia, October 7, 1912. Nore, Oct. 241h.—Of the numerous witnesses to whom a list of nineteen questions was submitted, but few replied to the following: In the area covered by the fall, were the large and small stones mixed indiscriminately or were they sorted some- what according to size? The first answered that the stones were mixed. But in a supplementary statement, Mr. Von Aachen writes on Oct. 18th that he found the small and large stones scparated. Further evidence is manifestly needed before the actual facts are established concerning a sorting of the shower. S. W. Williston— Restoration of Limnoscelis. 457 Arr. XL.— Restoration of Limnoscelis, a Cotylosaur Rep- tile from New Mexico ; by 8S. W. Wixuston. A prescription of the skeletal structure of Limnoscelis paludis Will., based upon two specimens in the Yale Museum, has been oiven by me in the papers cited below.* Since the time of my studies of these specimens they have been thoroughly and carefully divested of their conceal- ing matrix without destroying the relations of the different parts. An additional specimen of the same collections, con- sisting of pelvis, femora and tail, has also been detected and prepared. A fourth specimen of the same form, of somewhat smaller size, but in excellent condition so far as it goes, col- lected by myself in New Mexico, has also been of service. From a careful study of all this material, as now prepared, I am able to determine some additional facts in the osteology of this remarkable reptile, and to make a few corrections of my previous papers. In the restoration given herewith (fig. 32, p. 467), the skele- ton as far as the base of the tail has been drawn almost exclu- sively from the more perfect holotype specimen (No. 811, Yale University). As has been stated, the relations of the bones of this specimen had been disturbed but little in fossili- zation, except for a part of the tail. Of the tail, the parts pre- served are the proximal vertebre in position, but with the spines somewhat injured; and a nearly continuous series of twenty distal vertebre. Of specimen No. 809, Yale Uni- versity, the skull is missing, but the hind feet are in better preservation than those of the holotype; of the tail, twenty- three vertebra are preserved in articulation with the sacrum, and eighteen distal ones are preserved in several series. Speci- men No. 819 (Y. U.) comprises a fragmentary pelvis and incomplete femora, together with the nearly complete tail in natural articulation in the matrix, as shown in figs. 24-26. The caudal vertebree are in two series; the first, beginning with the fourth (fig. 25), extends to the thirty-third; the second (fig. 25) comprises twelve vertebree, with which several terminal ones were found associated in the matrix, but sepa- rated. Between the first and second of these series, the size indicates a loss of three or four vertebra, as indicated by the uniform shading of the restoration. From the tip of the tail at least eight or ten are missing, making a total of about sixty, of which forty-eight are actually present in specimen 819, * This Journal, xxxi, 380, May 1911; American Permian Vertebrates, p. 28, Oct. 1911. 458 SS. W. Williston— Restoration of Limnoscelis. forty-one in specimen 809, and twenty-six in specimen 811. Specimen No. 650 of the Chicago University collections com- prises about twenty candal vertebrae, for the most part con- nected in a basal series. Unfortunately in none of these specimens are the spines quite complete throughout. From a comparison of the four different specimens, however, most of these have been determined with certainty, as shown in the drawing. In specimen 809 three of the proximal chevrons are preserved complete in natural articulation ; in specimen 811 a number of the distal ones are also connected with the centra ; while those present in 819 are shown in fig. 25. The three specimens of the Yale collection are almost identical in size, 819 being perhaps a trifle larger than the others, though not much, the total length of the first twenty-three caudal verte- bree being about three-fourths of an inch greater than that of the same vertebrae of 809. The Chicago specimen, however, is distinctly smaller, and also presents some slight differences which might perhaps be accounted of specific value did we know what specific characters are in these old reptiles. I give herewith a number of figures of some of the best pre- served parts of these different specimens, all one-half natural size. The humeri (figs. 1, 2,31) are best preserved in the Chicago specimen ; in the Yale specimen No. 811, from which my previous figure and descriptions were made, the ectocon- dyle had evidently been crushed somewhat inward from its normal position; it really projects nearly at right angles dorsad from the distal plane of the bone, much as in Diadectes and Diasparactus. ‘The ectepicondyle is larger and better preserved than in-any of the Yale specimens of this bone, and all the processes are better defined. Perhaps the clay matrix has had something to do with the better preservation of this specimen. The radius of the Chicago specimen (figs. 4, 5) in comparison with that of 811, Y. U. (fig. 28), seems to be a little less stout. A well preserved femur and the fibule of the Chicago specimen are shown in figs. 6—9, the lower end of the femur completed from that of the opposite side. Especially characteristic of Zimnoscelis, in which it agrees with Sey- mouria, but disagrees with the Diadectida, are the large size and low position of the trochanter, and the high, thin adductor crest, as in most of the contemporary amphibians. Both the tibia and the fibula of this specimen are a little less stout, and less expanded at their ends than in the holotype. Two views of a typical posterior dorsal vertebra of this specimen are shown in figs. 15,16. In my original description of the genus I gave as a characteristic the presence of an infracentral fossa on the presacral vertebree. I am now not sure whether this character is of importance, since it seems to be absent in the S. W. Williston—Restoration of Limnoscelis. 459 Chicago specimen, and not well marked in specimen 809 of the Yale collections. In my original description I gave the probable number of presacral vertebree in Limnoscelis as twenty-five. As a fact Fies. 1-5. Fig. 1, right humerus, ventral side. C.U. 650. Fig. 2, the same, dorsal side. Fig. 3, the same, ulnar side. Fig. 4, right radius, dorsal side, C. U. Fig. 5, the same, ventral side. there are twenty-six, a. larger number than is known in any other American Ootylosaur. All these vertebree are in inti- mate matrical contact in the holotype, so that the number is 460 S. W. Welliston— Restoration of Limnoscelis. absolutely fixed. There is not a very great difference between the anterior and posterior ones. The former are a trifle smaller, the spines a very little shorter and less stout, and the diapophyses are longer, those of the notarial vertebrae (as I designate those vertebrae which support the pectoral girdle), Fre. 6-9. —— S ~S= Z Fig. 6, left femur, dorsal side, C. U. 650. Fig. 7, the same, ventral side. Fig. 8, left fibula, ventral side, C.U. Fig. 9, right fibula, dorsal side, Cau: as shown in fig. 10, projecting far beyond the zygapophyses, while posteriorly, as shown in fig. 15, they are almost sessile. I have also stated that there is but a single sacral vertebra in Limnoscelis, and functionally the statement is true. Strictly speaking, however, the ribs of the vertebra immediately suc- S. W. Williston— Restoration of Limnoscelis. 461 ceeding the true sacral vertebra touch the ilium at their tip and must, therefore, be considered as sacral ribs. In specimen Fies. 10-14. Fig. 10, fourth, fifth and sixth vertebra, with ribs, as lying in matrix, from above, Y. U. 811. Fig. 11, right sixth rib of same, inner side. Fig. 12, chevron of eleventh caudal vertebra, Y. U. 809. Fig. 13, second caudal vertebra, from the side, Y. U. 819. Fig. 14, the same, from below. 811, the one cleaned at the time of my previous studies, these ribs had been dislodged, but in specimen 809, which has 462 S. W. Williston— Restoration of Limnoscelis. been since prepared, they are in position. They are about thirty millimeters in length, tapering from the broad proximal end, and again very slightly expanded at the tip, where they touch the ilium immediately above the ilio-ischiadie noteh, and just back of the margin of the functional sacral rib. The structure here seems to be quite as in Diasparactus and per- haps as in Diadectes. On the basal caudal vertebrae the spines are elongated, as shown in figs. 24 and 25. They decrease very rapidly in Fies, 15-23. Fig. 15, posterior dorsal vertebra, from behind, ©. U. Fig. 16, the same, from the side. Fig. 17, tenth, eleventh and twelfth caudal vertebrae, from the side, C. U. Fig. 18, front end of left mandible, from without, C.U. Fig. 19, astragalus, tibial surface, C.U. Fig. 20, the same, pos- terior surface. Fig. 21, calcaneum, C. U. Figs. 22, 28, distal tarsals, C. U. length from the fourth, their height in the eleventh or twelfth (tig. 17), being less than that of its centrum. The atlas, which lies in specimen 811 nearly above the front margin of ‘the interclavicle, has a large intercentrum which bears an arch on each side. The arch is not unlike that of Ophiacodon ; it bears a rib on each side on the backwardly directed process. The presence of a proatlas cannot be deter- mined. The odontoid is only in part visible. The axis has a stout and rather broad spine, thickened behind. Ribs are present on all the vertebrae as far back as the eleventh or twelfth caudal, those of the sacral and caudal regions suturally united with body and arch. They are all holoceph- S. W. Williston— Restoration of Limnoscelis. 463 alous, the articulation continuous from head to tubercle, unless those of the atlas and axis are exceptions. The axis has a short rib, forty-two millimeters in length, with a width of eighteen at the extremity. The third rib is about sixty-five millimeters in length, and has a distal width of forty. The fourth is sev- enty-five millimeters in length, with a distal width of about sixty. It is this rib which’ was figured by me in my cited work, page 386 and plate xxxvut, as a probable hyoid. The fifth rib is scarcely longer than the preceding one, but is broader at the extremity, and the sixth rib, of the same length, is even more expanded at the extremity (fie. 11). These ribs are all thin at the distal extremity and somewhat concave on the inner side. In specimen 811 (fig. 10) the vertebree lie in position over the pectoral girdle. In settling down, after de- composition of the body had begun, these broad-ended ribs were dragged upward, especially on the right side ; all four, that is of the third, fourth, fifth and sixth vertebrae, lie on the inner side of the scapula ; the rib of the axis on the left side just underlaps the margin of the scapula. It is very evident that the function of these peculiarly expanded ribs was for the attachment and support of the scapula; in other words, they served the same purpose for the pectoral girdle that the sacral ribs did for the pelvic, and it is evident that their union was nearly or quite as firm. They may properly be called notarial ribs, and their supporting vertebre notarial vertebree. These dilated and shortened ribs are more or less characteristic of all American Cotylosauria, and, in a less degree, of Ophi- acodon, an American pelycosaur. The seventh rib has near ly the full length of those succeed- ing, but is unusually stout.and broad, especially distally. The following ribs are more slender ; they all lie in position in the holotype, and their relative lengths, with the necessary fore- shortening, I have endeavored to show faithfully in the restor- ation. The length of the seventh rib, measured on its chord, is five and three-fourths inches; of the eighth and ninth, six inches ; of the fifteenth, five and three-fourths inches; of the eighteenth, three and one-half inches. The single pair of functional sacral ribs are of the usual dia- dectid type, flattened and expanded distally and directed down- ward, lying in apposition, but not suturally united, with the inner side of the ilium above the acetabulum. The ribs of the second sacral, or sacrocaudal, vertebra are much less stout than the following ones. The first true caudal ribs are elongate, curved outward and backward, lying inside the posterior pro- longation of the ilium for the most part; they. terminate appar- ently in a pointed extremity. The second | pair of ribs are extraordinary. In specimen 819, where they are preserved 464 S. W. Williston—Restoration of Limmnoscelis. Fies. 24-26. (a (( »> “G i —=—=< F 4 f —Z ae 2 wll Fig. 24, first caudal vertebra, Y. U. 819. Fig, 25, fourth caudal vertebra and connected series. Fig. 26, second series, Y. U. 819. S. W. Williston— Restoration of Limnoscelis. 465 complete and undistorted, as shown in figs. 13 and 14, they are remarkably stout and lone. They are directed outward, down- ward, and backward, and end in a truncated extremity. They lie, for the most part, below and back of the iliae process. It seems certain that ribs of such structure and strength in this position must have served some functional use in the support of muscles, doubtless for ones controlling movements of the Figs. 27-31. Fig. 27, 28, ulna and radius, dorsal sides, Y. U. 811. Fig. 29, right tibia, ventral side, Y. U. 811. Fig. 30, left ischium, outer side, Y. U. 809. Fig. 31, right humerus, distal end, C. U. 650. leg. Back of these greatly enlarged and long ribs, the suturally united caudal ribs pragressively decrease in length and size, dis- appearing on the eleventh or twelfth. In my first paper I stated that the first chevron was attached to the third caudal vertebra ; this statement is true if the second vertebra with sacral rib attachments is considered a sacral, not otherwise. The first chevron preserved in specimen 809 is 466 S. W. Williston—Lestoration of Limnoscelis. attached to the under side of the fifth vertebra back of the func- tional sacral, articulating with the fourth. It is long and slen- der. Back of this the chevrons are more or less dilated distally and are long; on the distal part of the tail they are more slender. Large intercentra occur between the anterior caudal verte- bree; one, that articulating between the second and third true caudals, has two small processes on the under side, correspond- ing to the branches of the chevrons; and all the following chevrons have the proximal end in the form of an intercentrum ; another bit of evidence, if such be needed, that the chevrons are merely outgrowths of the intercentra. As regards the structure of the legs of Zimnoscelis I have little to add. With one of the feet of specimen 809 Yale University, two small bones are preserved attached to the calcaneum. They are evidently either the third and fourth distalia, or the centrale and a distale, probably the former. In the Chicago University specimen all these four bones are pre- served of one foot; I have figured them (figs. 19-23). The astragalus differs distinctly from that of the Diadectide in its smaller size, and more cuboidal shape. Since the discovery of these small tarsal bones in Zemnoscelis, it would seem not at all improbable that additional nodules corresponding to the first three distalia, as in Diasparactus, may yet be found. Nor have I much to add regarding the structure of the sk ull at the present time. I feel confident that I recognize both the tabulare and supratemporal, giving the full complement of cra- nial elements, with the exception of the intertemporal, known only in Seymourza. I give a figure of the anterior end of the Chicago mandible (fig. 18), showing elongate teeth correspond- ing to those of the premaxille. The teeth are implanted deeply in sockets, as in Diadectes, and show a deeply infolded dentinal structure. The characters of Zimnoscelis may be given as follows, omitting all those common to the Cotylosauria of America, as I have recently summarized them :* Lamnoscelas—Crawling, litto- ral or subaquatic reptiles, with a rather long body and long tail ; probably bare-skinned. Head elongate, narrowed anteriorly, broad behind, with elongate, premaxillary teeth, a single row of conical teeth in maxillee and dentaries; prefrontal and ~ postfrontal meeting broadly over orbits ; pineal foramen small ; tabularia and quadratojugals present, probably also supra- temporals ; occipital condyle flattened ; basipterygoid process loosely articulated with pterygoids. Twenty-six presacral ver- tebree, their spines short and not rugose at extremity ; a single functional sacral vertebra, one sacrocaudal and fifty-five or * Journal of Morphology, 1912. S. W. Williston—Lestoration of Limnoscelis. more caudals. Ribs holocephalous ; notarial ribs ot the third to sixth vertebree greatly expanded for sup- port of pectoral girdle ; seventh rib elongated, but stout ; ribs continuous to eleventh or twelfth caudal. No ventral ribs. A vestigial cleithrum. Humerus with stout ectocondyle directed dorsad ; a well-developed ectepicondyle ; entepicondylar fora- men large; ulna with olecranon ; four bones in proximal row of car- pus, and at least three in distal; un- gual phalanges short and broad ; femur with prominent trochanter near its middle, the adductor crest high ; astragalus cuboid in shape, rel- atively small, as also the calcaneum ; at least two bones in distal row of tarsus; phalanges as in front foot. Both front and hind legs short and stout. Length, seven feet. Horizon, . Permocarboniferous of El Cobre, New Mexico. As regards the habits of Limnos- celis 1 have little to add to what I have already written. The very long tail has short spines, except at the base, but the.chevrons are longer and stouter than I had expected to find them ; and they are, for the most part, more or less flattened at their extremities. The tail must have been somewhat flattened, though by no means compressed as in the croco- diles, unless it bore a carina of horny scutes above. Taking into con- sideration the very short and stout legs with their broad, flattened feet, the absence of claws, the elongate body and tail, it would seem not at all improbable that Zzmnoscelis was more or less at home in the water, though not strictly an aquatic animal. In much probability it lived in and about the marshes on the mud flats, hiding in dense vegetation, and often taking to the water for pro- tection from its enemies, of which RC A AC Ly RRR lat 3 RRR \\ > a7 a we iN S ee) | Hee J A 467 Restoration of Limnoscelis paludis Williston, one-eleventh natural size (doubtful parts are shown in uniform shading). Fig. 32. 468 S. W. Williston— Restoration of Limnoscelis. Sphenacodon may have been one, or in seeking its food. The teeth and skull remind one much of Zabidosaurus with its long prehensile teeth in the premaxillee and front end of the mandi- bles, teeth well adapted for the seizure and retention of soft and slippery invertebrates, for which it may have probed in the mud. That Lemnoscelis was piscivorous in habit seems quite improb- able. It must have been slow in locomotion, whether on land or in the water, and its prey must also have been slow-moving or stationary. That the creature could have been really burrow- ing in habit seems out of the question; with its short front legs it could not possibly have excavated burrows for the head to enter. The length of the skeleton as restored is just seventy-eight inches, a few inches less than I had estimated it to be. [It is possible, however, even probable that in my restoration I have not made allowance for the interarticular cartilages, and that the creature was somewhat longer in life, perhaps fully seven feet. It is a remarkable fact that, so far, not a trace of Limnoscelis has been discovered elsewhere than the El Cobre cation, New Mexico ; nothing that can at present be referred to the genus is known from the Puerco or its tributaries. Of the four known specimens, two are said to have come from low down in the cafion, but the specimen collected by myself was found near the top of the fossiliferous horizons, at least two hundred feet above the lowermost beds; it is possible that this differ- ence in their horizons may account for the minor differences presented by this specimen. It is also a little strange that, so far, no specimen of Sphenacodon or Ophiacodon has been found in the El Cobre cation. It was in this cation that Pro- fessor Case found a specimen of Spirifer rockymontanus, a real Carboniferous fossil, coming probably from an horizon above that of the type of this genus. From all of which it would seem probable that Limnoscelis really lived during Car- boniferous times. University of Chicago, Chicago, Ill. Gooch and Blumenthal—lodic Acid Process. 469 Arr. XLI.—The Lodic Acid Process for the Determination of Bromine in Halogen Salts; by F. A. Goocw and P. L. BLUMENTHAL. [Contributions from the Kent Chemical Laboratory of Yale Uniy.—cexxxvi. | Tue oxidation potential of a solution of chlorine in potas- sium chloride and water is, according to Bancroft,* about 0°241 volt higher than that of a solution of bromine in potassium bromide and water. lodie acid has an oxidation potential 0-064 volt higher than that of an equivalent and equally acid- ified bromine solution. The consideration of these oxidation potentials suggested to Bugarskyt+ the choice of iodie acid as an oxidizer for the liberation of bromine from mixtures con- taining a bromide and a chloride. Bugarsky’s method of separating the bromine from such a mixture consists in adding sulphuric acid with a known amount of potassium diiodate, and boiling. The bromine and the iodine which are liberated in the interaction of the free acids, according to the equation 2H1O, + 10HBr = 5Br, + I, + 6H,0, escape from the solution and may be collected in the distillate. These free halogens may be made the measure of the bromide taken, or the amount of the bromide may be calculated from the amount of the iodic acid remaining. When the liberated halogens are absorbed in potassium iodide, the amount of free iodine, as determined by titration, proves to be much less than ‘is to be expected from the equation above, and Bugarsky, con- vineed that the results are much more nearly in accord with the theory when the excess of the iodic acid remaining is made the measure of the reaction, prefers, therefore, to esti- mate the bromide by determining in an aliquot por tion of the boiled solution the iodate which remains, and to determine the chloride in another aliquot portion of the solution by Volhard’s process of titration with standard sulphocyanate. The difference in the results of these two methods of deter- mination Bugarsky attributes to loss of free bromine by the action of steam, in consequence of which hydrobromic acid is formed and oxygen set free, as in the equation 2H,0O + 2Br, = 4HBr + O,. But Andrews points out that hydrobromie acid thus formed, if returned with the condensed steam to the liquid, must be reox- * Zeitschr. phys. Chem., x, 405. { Zeitschr. anorg. Chem., x, 387. Am. Jour. Sct.—FourtH Series, Vou. XXXIV, No. 203.—Novemser, 1912. 31 470 Gooch and Blumenthal—TIodic Acid Process. idized by more iodie acid, with the result that the bromine determination by calculation from the amount of iodie acid remaining will be high rather than low. Andrews’ results for bromine were high to the extent of 1 per cent to 1°5 per cent. The differences in experience and view of these investigators have led us to further investigation of the reaction. Suitable solutions of sodium chloride and potassium bromide* were prepared and standardized gravimetrically by silver pro- cipitation. An iodine solution, approximately N/10, standard- ized against N/10 arsenite, and N/10 sodium thiosulphate, standardized against the iodine solution, were used. Finally, a solution of potassium iodatet (35670 grams per liter) was prepared, and its iodine value was determined by titrating with the thiosulphate the iodine liberated from 50™* portions by the action of potassium iodide in presence of sulphuric acid. To avoid errors of titration due to an indefinite end point, the gravimetric determination of chlorine was substituted for the Volhard process. Gaseous nitrogen trioxide, liberated from sodium nitrite and washed by nitric acid, was used instead of the sodium nitrite or potassium nitrite used by Bugarsky. The method of carrying out determinations was as follows: To 50°™* of the solution, containing known amounts of sodium chloride and potassium bromide, 50°™* of the iodide solution of known value and 10°™ of 20 per cent sulphuric acid were added and the solution was diluted to 200°, in a 500° Erlenmeyer flask. A few bits of platinum foil were added, to prevent bumping, and the solution was boiled down to about 80°, some 30 minutes being required. After cooling, the solution was transferred to a carefully calibrated 100°™* flask and made up to the mark. Half the solution was pipetted into an Erlenmeyer flask fitted with an inverted mushroom trap, through which potassium iodide was added. The liberated iodine was titrated with thio- ‘sulphate, and 1/6 of it was taken as the equivalent of the excess iodate in the portion. The remaining half of the solution was transferred to a beaker, the flask and stopper being carefully rinsed, and an excess of gaseous sulphur dioxide (vaporized from the liquid) was passed in. Gaseous nitrogen trioxide (liberated from sodium nitrite or potassium nitrite by nitric acid, and washed by the same) was passed into the solution to destroy the excess of sulphur dioxide acid and to liberate the iodine.{ The solution was boiled until free from iodine, and the chlorine of the resid- * The ‘‘analyzed” OC. P. article, free from chloride, so far as could be eee ren by qualitative tests, (this Journal (3), xl, 289). 3. {The nitrogen trioxide prepared in the manner described produced no turbidity in a solution of silver nitrate acidulated with nitric acid. Gooch and Blumenthal—TIodic Acid Process. 471 ual chloride was precipitated by silver nitrate. After standing over night, the silver chloride was filtered off on asbestos in a perforated crucible, dried at gentle heat, and weighed. The amounts of potassium bromide calculated from the iodate used up, and the amounts of sodium chloride equivalent to the silver chloride weighed, are given in the table, in com- parison with the amounts taken of these substances respec- tively. Molecular Ratio of KBr to NaCl, 1:1. KBr NaCl KBr. NaCl Error Error Final taken. taken. formed. formed. in KBr. in NaCl. volume. erm. grm. grm erm. germ. germ. em’, 0°1984 0°0974 0°2006 0°0995 +0°0022 +0°0021 ‘ 0°1984 0:0974 0°1981 0:0996 —0:0003 +0:'0022 * 0°1984 0:0974 0:1996 0'1003 +0°0012 +0°0029 x 0°1984 0°0974 0°1983 00996 —0:0001 +0:0022 = 0°1984 0:0974 0°1939 071004 —0°0045 +0:'0030 76 0°1984 0:0974 0:1966 071008 —0'0018 +0°0034 TA. 071984 0:0974 0°1990 0:0998 +0°0006 +0°0024 38 071984 0:0974 1°1945 0'1004 —0°0039 —0:'0030 74 * The final volume, 50°™’ to 75°™°, was not measured exactly. Molecular Ratio of KBr to NaCl, 2:1. 01984 0:0487 0:1940 0°0517 —0°0044 +0°0030 50 071984 0°0487 0°1904 aes —0:0080 ee 65 9°1984 0°0487 0°1926 0°1519 —0°0058 +0°0032 68 01984 0:°0487 0°1965 0:0522 —0°0019 +0:0035 65 071984 0°0487 01942 Bees —0°0042 dosvte 78 0°1984 0-0487 0°1945 0:0521 —0°0039 +0°0034 80 071984 00487 0°1949 0:0518 —0°0035 +0.0031 65 01984 0:0487 0:1942 0°0526 —0°0042 +0°0039 70 0°1984 0:0487 0°1959 0°0525 —0:'0025 +0°0038 76 Molecular Ratio of KBr; NaCl, 1:2. 01984 0:°1948 0°1964 071954 —0°0020 +0°0006 - 01984 0:1948 071952 0°1990 —0°0032 +0°0042 ze 071984 0°1948 0°1967 0°1969 —0°0017 +0°0021 ii 071984 01948 0°1978 0°1984 —0°0006 +0-°0036 = 0°1984 0°1948 0°1974 071980 —0:0010 +0°0032 ay 0'1984 071948 0°1983 071984 —0°0001 +0°0036 nS 01984 0°1948 0°1988 0°1985 +0°0004 +0°0037 HS 071984 0:1974 0:1988 0°1028 —0°0001 +0:°0054 75 0:1984 0:°1974 0°1960 0°1022 +0:0024 +0°0048 70 *The final volume, 50°? to 753, was not measured exactly. In these experiments, made with reagents carefully stand- ardized at the outset and restandardized in the course of the investigation, the errors are generally large and irregular, inde- pendent of the final concentrations within the limits shown, 472 Gooch and Blumenthal—TIodie Acid Process. and out of all proportion to any possible chloride contamination of the standard bromide. The fact that the chlorine proves to be invariably high and the bromine generally low, suggests the probability that the iodie acid reaction fails to remove all the bromine in the process of boiling. Bromine was found in the silver chloride obtained from the combined residues of several determinations by applying the chloroform test to the water extract of the sodium carbonate fusion of these residues, after acidulating with sulphuric acid and adding a drop or two of chlorine water. Furthermore, in a special experiment in which pure potassium bromide was treated according to the procedure outlined, the titration figures of the excess of iodate corresponded to only 01945 erm. of potassium bromide out of 0°1984 grm. taken; while, after treatment with sulpbur dioxide followed by nitrogen tri- oxide, and boiling until all iodine had been expelled (as shown by the starch test applied to the cool solution), silver nitrate precipitated from the solution silver bromide equivalent to 0:0036 grm. of potassium bromide. The 0°1945 orm. of the potassium bromide indicated by the determination of the iodate and the 0:0036 grm. equivalent to the silver bromide precipi- tated after reduction make up 0:1981 gr. of the 0°1984 orm. taken. In this experiment, therefore, the deficiency of bro- mine and the excess of chlorine noted were plainly due to the retention of bromine, after the iodic acid treatment, in some combination such that the process of reducing the residual iodic acid and expelling the iodine left it in condition to be precipitated as silver bromide, which was counted as silver chloride. The combination of iodine with bromine in this experiment and of iodine with bromine and chlorine in the experiments of the table, would account for the observed deficiencies in bro- mine and excess in chlorine; and a little consideration shows that the conditions of the process are favorable to the forma- tion of such combinations. Thus free bromine and free iodine, both products of the main reaction, may readily com- bine to form iodine monobromide, HIO, + 5HBr = 2Br, + IBr + 3H,0. Furthermore, Roberts* has shown that iodic acid, iodine, and aqueous hydrochloric acid react to form iodine monochloride, according to the reaction HIO, + 21, + 5HCl = 3H,0 + 5ICl, and it may reasonably be expected that in the reaction of iodic acid and hydrobromic acid in presence of free iodine (pro- * This Journal (8), xlviii, 1, 58. Gooch and Blumenthal—lodic Acid Process. 473 duced in the main reaction, previously cited) iodine mono- bromide will be similarly formed, according to the equation HIO, + 21, + 5HBr = 3H,O + 10IBr. When the acidulated solution containing an excess of iodic acid with more or Jess iodine monochloride, or monobromide, is treated with potassium iodide for the purpose of determin- ing the excess of iodic acid by means of the iodine liberated, HIO, + 5HI = 3H,0 + 31,, ; ? the iodine compounds of the halogens will contribute to the amount of the free iodine which measures the iodic acid, KI + ICl = KC) +1, KI + IBr= KBr + I, and this increase in free iodine will produce a corresponding deficiency in the bromide, which is estimated from the amount of iodate which has disappeared. When, on the other hand, the excess of iodate is reduced by sulphur dioxide preparatory to determining the chloride, the iodine monochloride is converted to hydriodic acid and hydro- chlorie acid, and the iodine monobromide to hydriodic acid and hydrobromie acid. The hydriodic acid is destroyed, and the iodine removed, in the subsequent treatment with nitrogen trioxide ; the hydrochloric acid is regularly estimated as silver chloride; but the hydrobromic acid remains to produce silver bromide, which contaminates the silver chloride and causes an error of excess in the chloride determination. These hypothetical effects are in precise accordance with the observed phenomena. It is conceivable, and even probable, that chloric acid and bromic acid may appear also as products of the hydrolysis of the halogen compounds in presence of water and iodic acid. Thus the hydrolysis of iodine mono- bromide may proceed according to the expression 15 IBr + 5HIO, + 15H,0 = 5HBrO, + 5Br, + 101, + 15H,O and that of iodine monochloride according to the equation 15 IC] + 3HIO, + 15H,O = 5HCIO, + 10HCl + 91, + 9H,0. The production of bromic acid by hydrolysis of iodine bromide, with elimination of the free halogens and substitution of bromic acid for an equivalent amount of iodic acid, will tend to reduce the tendency to negative errors in the determination of bromine; but the bromine of the bromic acid will still be 474 Gooch and Blumenthal—TIodice Acid Process. capable of contaminating the silver chloride precipitated by silver nitrate after reduction. The effect of chloric acid substituted for iodie acid will depend upon the conditions of concentration. So far as it may act as an oxidizer in the treatment of potassium iodide, it will tend to produce a positive error in the determination of bromine. So far as it escapes reduction in the treatment of sulphur dioxide and nitrogen trioxide, it will tend ‘to produce an error of deficiency in the determination of chlorine. The extent to which these various possibilities may influence results will depend upon the* conditions of action and possible balancing of errors. The experimental results given above show plainly that under the conditions defined the separation of bromine from chlorine in halogen salts by the action of iodic acid, though based upon a reaction ideal in respect to the relation of the oxidation potentials, is vitiated by secondary effects. These effects may be reasonably attributed to the action of small amounts of iodine monochloride and iodine monobromide formed in the interaction of iodic acid and free iodine with hydrochloric acid and hydrobromie acid, or to the action of chloric acid and bromic acid derived from iodine monochloride or bromine monochloride by hydrolysis. John E. Wolf—Chlorite from Northern Wyoming. 475 Arr. XLIL—A Wew Chlorite from Northern Wyoming; by Joun E. Wo rr. In December, 1910, the writer received from Mr. Albert F. Holden of Cleveland, Ohio, a tale-like mineral which he had in turn received from Mr. W. G. Swart, a mining engineer of Denver, as apparently containing a large amount of alumina. By correspondence with Mr. Swart it was ascertained that the mineral had been brought in by prospectors who were reluc- tant to disclose the actual place of the find but from whom finally the following information was obtained. To quote Mr. Swart’s letter: ‘ Location—Sheridan County, Wyoming. Occurrence—as nearly as I can make out from Mr. Heineman’s _ deseription, it occurs in a vertical vein. He describes the exposed face as being ‘as high as the office ceiling’ and about 6 to 8 feet wide. The lines of fracture run vertically in the body of the mineral. This information was obtained by Mr. Parmelee, who originally brought me the mineral.” Sheridan County lies along the northern border of the state, a little east of the center, and includes in its western half the uplift of the Big Horn Mountains, in which the pre-Cambrian erystalline rocks are exposed. In the published reports these are described as mainly granite with some bands of schist, hence the only deductions possible from this scanty informa- tion is that the mineral occurs as a vertical schist band six to eight feet wide, representing either a shear zone or an integral member of a series of crystalline schists. Description of the Mineral. Macroscopic characters.—Several pieces were received, the largest a slab 7 in. long, 3 wide, and 4 in. thick. The material is well foliated, with fine straight foliation planes like a roofing- slate, has a faint silvery green color and is decidedly translucent even in the thicker pieces, while thin splinters are colorless and almost glassy. The specimen is homogeneous and composed of minute parallel scales of the mineral. Microscopic characters.— The thin sections confirm the purity of the material, for with the exception of a few minute prisms of zircon (?) and fluid cavities it is composed of sub- parallel plates of the chlorite, sometimes interlocking with a ‘tendency to radial arrangement, with an average thickness across the base of 1/10"™ and diameter on the base of 2/10™™. No twinning is seen and no erystalline outlines, but a good basal cleavage. ; Optical properties.—Dr. H. E. Merwin, of the Geophysical Laboratory, Washington, has kindly made the following meas- urements, entirely with the microscope: a= s80 B = 1°580 to 1:°581 y= 1089 with a possible error of +:001. 2E measured on several plates 476 John E. Wolf—Chlorite from Northern Wyoming. about 1/10™™ in diameter varied from 26° to 50°, but the angle of most of the cleavage flakes is about 85°, The acute bisectrix, ‘‘Z,” is inclined 5° to 10° to ¢, On account of the indistinctness of the interference figures the dispersion of the optic axes could not be measured satisfactorily. Chemical composition.—The following analysis (I) was made by the writer, employing the usual methods and deter- mining the water directly by the Penfield method. No reaction was obtained for fluorine. To ascertain any difference in the state of the water given off above 110° the mineral after constant weight at that temperature was heated an hour at 250° to 300° without change and at 350° for half an hour with practically no change, the whole of the water going off at a white heat and hence considered water of constitution. ~ shies Mol. II SiO?__.. = 28°81. ANT. = 2 32:1 AV’O?__. = 26°43 D5 Ou es FeO? _. = 0:24 SMe aee, Seo MeO tee —asneeil GUM aely 36°7 ee ee — (1 () fy De eerie 0°6 +110 .. = 12°62 700 ; eee 0-09 Ne 5 poe NaiOnes 50135 5 ie KAO be. = )0 Old 15 | CaQze None 100-29 100°0 I. Chlorite from Wyoming. II. Leuchtenbergite, Mauléon, Delesse, Ann. - Chim. and Phys., ix, p. 396, 1843. Specific gravity (pycnometer) 2°702. Pyrognostics.—Thin fragments fuse with difficulty in the blowpipe to a snow-white enamel and give intense blue with cobalt solution. Water given. off in closed tube is neutral or very slightly alkaline. Slowly decomposed by boiling sul- phurie acid and with difficulty by hydrochloric. Conclusion.—The simplest empirical formula from the ratios of the above analysis is H,Mg,A],Si,O,,. The leuchten- bergite from Mauléon, Pyrenees, of which only the old analysis of Delesse (I1 above) is available, seems to be the only chlorite containing as little iron, although the composition differs greatly by the excess of alumina in the Wyoming mineral, which amount is in fact exceeded by few chlorites. Although it seems hardly permissible to add a new name to the forty or fifty now found under the chlorite group, yet the purity of this material, its peculiar chemical composition and the certainty that it wili be available in large quantity, perhaps justifies the name of “ Sheridanite,” from the county in which it occurs. Mineralogical Museum, Harvard University. Chemistry and Physics. ATT SCIENTIFIC INTELLIGENCE. I. OnEMISTRY AND PHYSICS. 1. The Dissociation of Hydrogen into Atoms.—It has been found by Irvine Lanemurr that at extremely high temperatures the power consumption necessary to maintain a tungsten wire at a given temperature in hydrogen gas increases with abnormal rapidity with the temperature. This rapid inerease in loss of heat could not be explained by simple convection or conduction, and the most probable explanation appeared to be that dissocia- tion of the hydrogen molecules (H,) into atoms was taking place. This dissociation in the region close to the hot wire would absorb large quantities of energy, the hydrogen atoms would then diffuse out into the gas at some distance from the wire and would then recombine and give out the heat of the reaction, thus causing an abnormally high heat conductivity. This view is strengthened by the fact that Magnanini and Malagnini have observed that the heat conductivity of nitrogen peroxide is three times as great with the partly dissociated gas as when it 1s com- pletely dissociated. Langmuir has made this explanation still more probable by calculating the energy loss from heated wires by means of simple equations and finding that the calculations agree well with the experimental results in the cases of nitrogen gas and mercury vapor up to 2500° C. (abs.), and also in the cases of carbon dioxide and air up to the melting-point of plati- num, while in the case of hydrogen there is agreement only up to about 2100° C. (abs.), and above that point the energy loss increases very rapidly until at 3300° it is four or five times as great as the calculated value. It was shown by experiments that the phenomenon is an actual dissociation which follows the law of mass action, and that the volume of the dissociation products is approximately twice the volume of the original hydrogen, as the equation H, = 2H requires. The heat of this reaction was found to be 550,000 joules, or 130,000 calories at constant vol- ume. The extent of the dissociation at atmospheric pressure was calculated to be such that the partial pressure of the atomic hydrogen at 3300° C. (abs.) is two-thirds of the total pressure.— Jour. Amer. Chem. Soc., xxxiv, 860. H. L, W. 2. The Determination of Sulphur in Pyrites.—This important determination has given much trouble to analytical chemists in the past, and it has been the subject of many investigations. ALLEN and Bisnop have now devised and have carefully tested a new method for this purpose, which gives constant and accu- rate results, and which promises to supersede the older methods. The main features of the process consist in decomposing the substance with a mixture of liquid bromine and carbon tetra- chloride with subsequent addition of strong nitric acid, the 478 Scientific Intelligence. evaporation of these reagents, the conversion into chlorides and the separation of silica by evaporation with hydrochloric acid, drying, and dissolving in the same acid, the reduction of ferric chloride ‘to ferrous chloride by the addition of aluminium powder, the precipitation of barium sulphate by the slow addition of dilute barium chloride to the cold liquid, which is diluted very largely and is not stirred during the addition of the reagent, and finally the weighing of the barium sulphate in a Gooch crucible. It was found that iron in the ferrous condition does not interfere with the precipitation, and that although a little barium chloride is occluded in the precipitate this error appears to be compen- sated by solubility, so that it is not necessary to make any correc- tion in the results.— Proceedings 8th Internat. Congress App. Chem., i, 33. H. L. Ws 3. Separation of Arsenic from Antimony and other Metals.— MosrrR and PrErsotren have devised a convenient method for removing arsenic from other metals by distilling it off with a current of air at the temperature of the water bath from a strong hydrochloric acid solution containing methyl alcohol. If the arsenic is in the higher state of oxidation, a reducing agent, pre- ferably a ferrous salt, must also be added. ‘The arsenic goes over partly as trichloride and partly as methyl arsenite, both of which are decomposed by water in a receiver, into which the air and vapors are passed. The chief advantages of the method consist in the avoidance of a stream of hydrochloric acid gas, and the rapidity and completeness of the separation. Only about 14 hours are required for the distillation. The test analyses given are very satisfactory.— Monatshefte, xxxili, 797. H. L. W. 4, Methods in Chemical Analysis ; by FRanx Austin Goocu. 8vo, pp. 536. New York, 1912 (John Wiley & Sons).—Ana- lytical chemists everywhere will welcome this summary by Professor Gooch of the methods worked out by himself and his pupils in the Kent Chemical Laboratory of Yale University, and published in this Journal during the past twenty-two years. The book in giving this condensed account of these researches is very impressive in showing not only the great number of the investigations, but also their high quality and importance. Professor Gooch is to be most highly congratulated for what he has done for the benefit of analytical chemistry, and it is to be hoped that his valuable work will continue for many years to * come. H. L. W. 5. Elementary Applied Chemistry; by Lewis B. Attyn. 12mo, pp. 127. Boston, 1912 (Ginn and Company).—At a first glance this book has a somewhat humorous aspect. The frontis- piece is a photograph of a large class of young ladies working in a laboratory where the results of ice-cream analyses are displayed upon a blackboard, and also in the preliminary “ Suggestions to Teacher and Pupil” the heating together of aniline, caustic potash solution and chloroform is recommended, “if the odor of Chemistry and Physics. 479 phenyl carbamine is unknown,” although it might be feared that the horrible odor of the isonitrile might discourage the beginner. However, upon examining the book more carefully it is evident that it presents a very interesting and useful course of laboratory work. The author says that these experiments and tests have been of personal value to hundreds of earnest students. This can be well believed, as the course deals with many matters of vital importance ; for instance, the examination of water, baking powder, milk, ice-cream, cheese, headache-powders, paint and oils, the detection of coal-tar dyes, food preservatives, arsenic, methyl alcohol, the determination of food values, etc. As the title of the book and its size indicate, the topics are treated in an elementary manner, but it seems to be very well adapted for its purpose. H. L. W. 6. A Dictionary of Applied Chemistry ; by Str Epwarp Tuorre. Revised and enlarged edition, Vol. II, 8vo, pp. 786. London, 1912 (Longmans, Green and Co.).—The second volume of this important work extends from CHI to GOV. It contains many important articles, and the matter appears to be well brought up to the present state of the science. Among the interesting longer articles are those on chlorine, coke manufacture and the recovery of by-products, disinfectants, dyeing, identifica- tion of dyestuffs on fabrics, explosives (73 pages), fermentation, fertilizers, fuel, gas, glass, etc. The volume maintains the high standard of the previous one as a valuable work of reference. HL. W. 7. Die dussere Reibung der Gase und ein neues Prinzip fur Luftpumpen : Die Molekularluftpumpe.—As early as 1875 Kundt and Warburg showed that highly rarefied gases slip along the walls of the containing vessel, and that the superficial friction of the gas decreases with the pressure. In 1909 Knudsen calculated the numerical value of the friction between the inner surface of a glass tube and the stream of gas flowing through the tube, and found good agreement between the experimental and theoretical data. Quite recently W. Gaxrpr has repeated Knudsen’s observations and found that, at pressures greater than 0-001™™ of mercury, a film or skin of gas is formed on the glass and that this layer is predisposed to throw back the molecules in the general direction of incidence. Furthermore, Gaede’s experiments indicated that the action of the gaseous skin may be fully accounted for in the following manner. Two kinds of unevennesses on the glass wall are to be distinguished, (a) “mechanical” rugosities, and (b) “molecular” inequalities. At pressures above 0°001™™ the mole- cular unevennesses are covered up by a film of gas whose density decreases towards the free interior of the gas. The gas molecules which are reflected from the solid wall must pass through the skin of gas and consequently their direction of emergence will be affected. More specifically, normal passage through the skin is the most favorable because, for oblique reflection, the length of path in the skin is greater and hence oblique emergence is more 480 Scientific Intelligence. difficult. In substance, Gaede says : “ If we compare the mechan- ical rugosities with the hills and furrows of a ploughed field, then the molecular unevennesses correspond to the grains of sand and small stones, and the gaseous skin corresponds to a thin layer of hoarfrost or to a light coating of snow upon the hills.” If a differ- ence in gas pressure exists at the ends of a comparatively highly exhausted tube, the processes favorable to equalization of pressure will be hindered by the gaseous skin and the mechanical inequal- ities. For the molecules which have a component of velocity towards the more remote end of the tube, and which also impinge against the wall, will strike the mechanical inequalities of surface which are turned towards them so that, as stated above, the gas- eous skin will, in the majority of cases, send the molecules back along their lines of incidence towards the end of the tube from which they had come. On the other hand, when the degree of exhaustion has been carried so far that no gaseous film remains on the walls, the molecules will be reflected by the molecular un- evennesses of the walls in a perfectly regular manner, that is, with- out any selective relation to the angle of incidence. Therefore, just as many molecules will be thrown back towards one end of the tube as will be sent forward towards the other end. In this case, the equalization of pressure will proceed more rapidly than when a gaseous skin adheres to the tube. At about 0:001™™ this film practically vanishes, so that for lower pressures the assumption of regular, non-selective reflection holds and the Kinetic Theory of Gases applies in its more elementary form. An obvious extension of the preceding considerations to the case of the friction between a gas and a cylinder rotating in the gas has been made by Gaede. The result is that he has designed and tested a pump which bids fair to revolutionize the production of extremely high vacua. It is so difficult and wasteful of space to clearly describe the construction of a piece of apparatus with- out the aid of diagrams that the attempt will not be made to go into full details here. Suffice it to say that the “ Molecular Air- pump ” consists essentially of a cylinder which can be revolved at very high speed inside of a coaxial cylinder or housing. ‘The housing is grooved or recessed over a certain arc and in a plane perpendicular to the common axisof the cylinders. The “leading” and “trailing” ends of the recess communicate respectively with the auxiliary pump and with the vessel to be exhausted. In prac- tice, a number of grooves and inlet and outlet tubes are placed in series or tandem. . At pressures below 0:001™™ the molecules of the gas are diffusively reflected at the metallic surfaces and fly from one wall to the other without, in general, colliding with other molecules of their kind. ‘The reflection of the molecules can be pictured as if the surface of the cylinder were covered with a great number of small cannon out of which the molecules are shot in all possible directions with a certain speed, the molecular speed.” If the surface of the cylinder moves with a linear speed greater than the molecular speed, then the molecules inside the Chemistry and Physies. 481 recess which are shot out tangentially forwards from the cylinder will have a velocity directed away from the trailing or inlet tube and towards the leading or outlet pipe. ‘Their speed will be more than twice the mean molecular speed. Theoretically, an extremely small percentage of molecules can pass in the opposite direction. From the practical standpoint, as well as from the theoretical, the molecular air-pump is almost perfect. When run at a speed of 12000 R.P.M. the pressures at the “suck-nozzle” were 0:0003, 000003, 0:000005, and 0:0000002™" when the pressures in the housing were 20, 10, 1 and 0:05"™ respectively. The commercial pumps are designed to run at 8000 R.P.M. A conception of the rapidity of working of the new pump may be formed from the following test. An X-ray tube of about one liter capacity was evacuated, in 10 secs., from about 5™™ to a pressure so low that sparks passed in a parallel gap 15°™§ long. Gaede’s well-known mercury pump required 100 secs. to bring about the same result. Another great advantage of the molecular air-pump is that it works just as well for vapors as for dry gases. In particular, therefore, the usual drying agents may be dispensed with and much time saved in this connection. Of course, the new pump alone is useless at atmospheric pressure.—Physikal. Zischr., No. 18, Sept. 1912, p. 864. H. S. U. 8. On the Emission Velocities of Photo-Hlectrons, —'The results obtained by various investigators in this field have been discordant, and hence the present paper by A. Lu. Hueues is rather important because in it he discusses the sources of error which may have vitiated, in a greater or lesser degree, the work of his predecessors, and also describes his method of experimenting, which seems to overcome most of the difficulties. The two sources of error which have not been taken into account in any previous work are: (a) the effect on the velocity-distribution curves of a magnetic field, such as, for example, that of the earth, and (b) the effect on such curves of an electron approaching the boundary obliquely. By distilling the metal to be investigated and by allowing it to deposit on a disc of nickel, which could be raised and lowered at will inside of the glass apparatus itself, Hughes avoided both the production of surface films of gas and the polar- ization of the metallic surface which arises when the method of anode sputtering is employed. The whole process of distillation and adjustment was carried out in a liquid air vacuum, so-called. A control experiment was performed by using the surface of flow- me mercury in place of the solid metal deposited on the nickel isc. A portion of the summary given by Hughes will now be quoted. ‘“(1) The maximum velocity of photo-electrons from the surfaces of a number of elements prepared by distillation in vacuo has been measured. (2) It has been shown that the energy of the fastest electrons emitted when monochromatic light falls on the surfaces is proportional to the frequency of the light. The 482 Scientific Intelligence. results are expressed in the form V= kn — V,, where V is the velocity measured in volts and 2 the frequency. LapENBURG’S law, that the velocity is proportional to the frequency, has been shown to be incorrect, (3) The values of & and V, have been found directly for the elements Ca, Mg, Cd, Zn, Pb, Sb, Bi, and As, and the values of Y, for Se and O _ indirectly. (4) The values of & and V, for elements of the same valency change regularly with the atomic volume. (5) The product of V, intoe, the charge of an electron, has been identified with the work required to sep- arate an electron from the molecule.” Conclusion (2) is especially worthy of note.——Phil. Trans. Roy. Soc., London, Vol. 212, A, p- 205. H, §. v. 9. Magnetism and Electricity ; by HE. E. Brooks and A. W. Poysrr. Pp. villi, 633; 413 figures and 414 problems. London, 1912 (Longmans, Green & Co.).—This volume has been prepared to replace Poyser’s Advanced Magnetism and Electricity, which was originally published in 1892. Nearly the whole of the subject-matter has been rewritten on modern lines, so that, for all practical purposes, the only feature of the old book which remains is the experimental form. ‘The present work is intended to afford such a range of general reading in the subject as is desirable for the majority of students, before they begin to specialize either in pure science or in the various branches of electrical engineering.” An elementary knowledge of algebra, geometry, trigonometry, and mechanics is assumed. Nevertheless, in many instances, the authors supplement the more elementary analysis by short, alternative proofs in which the processes of the differential and integral calculus are used. When the elementary demonstrations would lack rigor or would be very cumbersome, the calculus alone is used. The brevity and elegance of the latter proofs should stimulate the student who is not acquainted with this powerful branch of analysis to increase his mathematical attain- ments. The chapters entitled “Introduction to the Theory of Alternating Currents,” “ Measurement of Inductance,” “ Passage of a Discharge through Gases,” and ‘Effect of Inductance and Capacity at Starting and Stopping a Current—Hlectric Oscilla- tions—Radiation— Wireless Telegraphy and Telephony” are starred to indicate that they may be omitted because of their relatively greater degree of advancement. Of necessity, the proofs in the last of these chapters involve a knowledge of elementary differential equations. The usual symbols A and pu have been retained in the formule in order to keep the dimensions of the equations correct. Throughout the volume, 245 practical exercises are suggested in small print. Many of the qualitative experiments do not require special laboratory apparatus, so that the student can perform them at home with improvised material. The numerical problems and questions are collected in groups at the ends of the chapters, and the answers to the former are given just before the index. Most of the figures are new and all of them are clear-cut Chemistry and Physics. 483 and satisfactory. Our opinion of the book may be best expressed by the statement that we shall use it in one of our intermediate courses during the coming winter. H. §. U. 10. A Text-Book of Physics, Third Edition ; edited by A. W. Durr. Pp. xvi, 686; 595 figures and 284 problems. Philadel- phia, 1912 (P. Blakiston’s Son & Co.).—A careful comparison of the latest edition of this work with the second (see vol. xxviii, page 556) brought out the following facts. The chapter on heat, which was formerly composed by Guthe, has been entirely rewritten by Mendenhall. In the subjects of electricity and mag- netism Goodspeed has surrendered his pen to Carman. The articles on sound and mechanics have not suffered much alteration. Lewis has decreased the number of pages devoted to light from 140 to 118, while McClung has brought the numerical data in the conduction of electricity through gases and in radio-activity up to date. The sequence of subjects has been changed to the following order: (1) mechanics, (2) wave-motion, (3) heat, (4) electricity and magnetism, (5) gaseous conduction and radio- activity, (6) sound, and (7) light. Some of the figures have been replaced by new and clearer diagrams, they all appear to have been relettered, and the total number of figures has been increased by seventy. YF our-place tables of the common loga- rithms of numbers and a table of natural sines and cosines have been introduced immediately before the first index. Great care seems to have been taken to avoid the old typographical errors and to give the text a more polished and elegant appearance. Be eh ll. Gravitation ; by Franx Harris. Pp. xi, 107; 18 figures. London, 1912 (Longmans, Green & Co.).—Less than six full lines are devoted to the preface, which amounts to the confession that the volume may not be “entirely free from clerical errors” and may “involve erroneous deductions.” Furthermore, the intro- duction deals only with the question of the meaning to be attached to the word “explanation,” when dealing with natural phenomena. Hence, the author gives no explicit clue to the object which he had in writing the book or to the gap which the text is intended to fill in the literature of the subject. The text is highly mathematical and the figures are graphs of certain functions. The titles of the chapters are:—I Medium, Particle and Motion; II Spheres in Sequence; III Potential Energy ; IV Sources and Sinks; V Two Circles; VI Spheres in Parallel ; and VII The Dimensions of Space. The volume closes with a chapter on “ Atomic Forces,” which is presented as an appendix. On page 29 the author uses the term “veetal” to signify ‘“ the loss of potential energy in.an attracted particle due to its transfer from infinity to a given point.” H. S..Up 484 Scientific Intelligence. II. Gkronocy anp Mineranoey. 1. Publications of the United States Geological Survey ; GEORGE Otis Smiru, Director.—Recent publications of the U. S. Geological Survey are noted in the following list (continued from vol. xxxiii, p. 507, May, 1912) : Torocraruic Attias: Fifty-five sheets. Fouio No. 182.—Geologic Atlas of the United States Chop- tank Folio, Maryland ; by B. L. Mizruer. Surveyed in codpera- tion with the State of Maryland. Pp. 8; 2 colored maps, 3 figures, PROFESSIONAL Paprr No. 77.—Geology and Ore Deposits of the Park City District, Utah; by Joan M. Bourwe tt, with contributions by Lester H. Wootsny. Pp. 231; 44 plates, 18 figures. MinERAL Resources of the United States in 1911. Numer- ous advance chapters. Buiietins No. 471.—Contributions to Economic Geology, 1910, Part II. Ten advance chapters. No. 498. Headwater Regions of Gulkana and Susitna Rivers, Alaska, with accounts of the Valdez Creek and Chistochina Placer Districts; by Frep H. Morrir. Pp. 82; 10 plates, 9 ficures. “No. 506. Geology and Mineral Resources of the Peoria Quad- rangle, Illinois ; by J. A. Upprn. Pp. 103; 9 plates, 16 figures. No. 507. The Mining Districts of the Western United States; by Jamzes M. Hix, with a geologic Introduction by WaLpEMAR LinpGreN. Pp. 309; 16 plates, one figure. No. 509. Mineralogical Notes. Series 2; by WarprEmar T. Scatter. Pp. 115; one plate, 5 figures. Nos. 514, 516, 517, 519. Results of Spirit Leveling: R. B. Maxsuaty, Chief Geographer. No. 514, New York, pp. 139. No. 516, Florida, pp. 24, one plate. No. 517, Alabama, pp. 38. No. 519, Tennessee, pp. 45. No. 520. Mineral Resources of Alaska. Twelve advance chapters. No. 530. Contributions to Economic Geology, 1911, Part I, eleven advance chapters. Water Suprry Papers No. 279.—Water Resources of the Penobscot River Basin, Maine ; by H. K. Barrows and C. C. Bags. Prepared in codperation with the Maine State Survey Commission. Pp. 285; 19 plates, 5 figures. Nos. 285, 291. Surface Water Supply of the United States, 1910. Prepared under the direction of M. O. Leicuron. No. 285. Part. V, Hudson Bay and Upper Mississippi River ; by Rosert Fouiuansper, A. H. Horton, and G. C. Stzvrens. Pp. 318; 4 plates. No. 291. Part XI, Pacific Coast of California ; by W. B. Cuarp, F, F. Hensnaw and H. D. McGuasnan. Pp. 218; 4 plates, one figure. . Geology and Mineralogy. 485 Nos. 295, 296. Gazetteer of Surface Waters of California. Prepared under the direction of Joun C. Hoyr by B. D. Woop. Part I, Sacramento River Basin. Pp. 99. Part II, San Joaquin Basin. Pp. 102. 2. Publications of the Bureau of Mines ; Joserua A. Hotmes, Director.—The First Annual Report of the Director of the Bureau of Mines was noticed on p. 305 of this volume. There are now to be mentioned the numerous publications issued by the Bureau in the past few months (see vol. xxxiii, p. 63). These include the following Bulletins : No. 6. Coals available for the Manufacture of Illuminating Gas; by A. H. Wuire and Perry Barker, compiled and revised by Hrrperr M. Witson. Pp. 77 ; 4 plates, 12 figures. No. 10. The Use of Permissible Explosives ; ; by J. J. Rut- LEDGE and CLrareNnce Haru. Pp. 34; 5 plates, 4 figures. No. 15. Investigations of Explosives used in Coal Mines ; by CriaRENCcE Hatt, W. O. SNELLING, and 8. P. Howe x; with a chapter on the Natural Gas used at Pittsburgh by G. A. BurRELL and an Introduction by Cuartes EH. Muynor. Pp. 197; 7 plates, 5 figures. No. 16. The Uses of Peat for Fuel and other Purposes ; by Cuartzes A. Davis. Pp. 214; one plate, one figure. No. 18. The Transmission of Heat into Steam Boilers ; by Henry Kreisincer and Wattrer T. Ray. Pp. 180; 78 figures. No. 23. Steaming Tests of Coals and Related Investigations, September 1, 1904, to December 31, 1908; by L. P. Breckern- RIDGE, Henry Kretsincerand Watter T. Ray. Pp. x, 380; 94 figures. No. 25. Mining Conditions under the City of Scranton, Pa. Report and Maps, by WitiiAm Grirrira and Eri T. Conner. With a preface by Joserm A. Hoximes and a chapter by N. H. Darton. Pp. 89; also 29 maps in separate cover. No. 41. Government Coal Purchases and Specifications with Analyses for the fiscal year 1909-10 ; by Grorex 8S. Porg, with a chapter on the Fuel-Inspection Laboratory of the Bureau of Mines ; by Josepu D. Davis. Pp. 97; 3 plates. No. 44. First National Mine-safety Demonstration, Pitts- burgh, Pa., October 30 and 31, 1911; by H. M. Wizsown and A. H. Fay. Pp. 75; 8 plates, 4 figures. No. 47. Noteson Mineral Wastes ; by C. L. Parsons. Pp. 44. Fifteen Technical Papers, many of them of great practical importance particularly as dealing with the conditions in coal mining, have also been issued. 3. The Onondaga fauna of the Allegheny region ; by EK. M. Kinpiez. Bull. 508, U. 8. Geol. Surv., 144 pages, 13 plates, 1912. —The middle Devonian Onondaga limestone, which is of wide distribution and forms an important datum plane in the eastern portion of North America, is said by many geologists to be absent in the Allegheny region south of Central Pennsylvania. Kindle now shows, after a great deal of detailed work, that the Onondaga is also present in the latter region, but in the main as a calcare- Am. Jour. Sci.—FourtTH SERIES, VoL. XXXIV, No. 203.—Novemser, 1912. 32 486 Scientific Intelligence. ous shale with many characteristic Onondaga fossils. Various sections extending from New Jersey to Tennessee are described by the author and the local faunules recorded on pages 23-53, In these strata Kindle notes about 115 species, and of these 7 are described as new. Some of the Onondaga guide fossils are: Cystodictya gilberti, Stropheodonta patersoni, Spirifer acumi- natus, Anoplotheca acutiplicata, Odontocephalus selenurus, and O. cegeria. On the other hand many Marcellus forms appear in these Onondaga shales, as Strophalosia truncata, Buchiola halli, Pterochenia fragilis, Actinopteria muricata, Styliolina fissurella, Tentaculites gracilistriatus, Bactrites aciculum, and Agoniatites expansus. The greatest number of long-ranging species are, however, from the Oriskany, and if it were not that these shales occur above this sandstone the author remarks that “much paleontologic evidence could be adduced for considering the fauna to be of Oriskany age ” (54). He saysfurther; “The eastern shore line of the Onondaga sea trended southwestward across north-central New Jersey and southeastern Pennsylvania. It probably traversed the States of Maryland and Virginia near the present axis of the Blue Ridge. From southwestern Virginia this shore line appears to have trended westward not far from the Kentucky-Tennessee line as far as the valley of Tennessee River, where it resumed its south- erly trend” (65). Cc. 8. 4. Preliminary report on the geology of the coastal plain of Georgia ; by Orro Veatcu and Lroyp WitiiaM STEPHENSON. Bulletin No. 26, Geol. Surv. of Georgia, 466 pages, 30 plates, 1911.—This very valuable and detailed report is a result of the codperative work of the Geological Survey of Georgia and the U. 8. Geological Survey. The Assistant State Geologist, Mr. Veatch, describes on pages 25 to 57 the physiography of the state, and on pages 58 to 65 presents a general statement of the geology of the coastal plain, a series of sedimentaries aggregat- ing over 4500 feet in maximum thickness. The remainder of the volume, by Doctor Stephenson, presents a great mass of desirable detail regarding the physical character, thickness and fossils of the seventeen formations composing the Lower and Upper Cre- taceous and Cenozoic deposits of Georgia ; their relationship to similar strata throughout the eastern Gulf area is also given. The book is an invaluable guide to the coastal plain stratigraphy * not only of Georgia but as well of the entire eastern Gulf area. C. 8, 5. Classification of the geologic formations of the state of New York ; by C. A. Harrnacer. Handbook 19, N. Y. State Museum, 96 pages, 1912.—In this very handy booklet are briefly defined the various geologic formations of the state of New York, about 200 in number, the majority of which make up the “standard Paleozoic section” of that state. There are of Pre- cambrian 20+ terms, Paleozoic 153, Mesozoic 11, and Cenozoic 7. This is the third edition of Handbook 19 and all stratigraph- Geology and Mineralogy. 487 ers will be thankful to the New York State Survey and to Mr. Hartnagel for this up-to-date index. C. 8. 6. Central Connecticut in the geologic past; by Joseru Barrett. Proceedings and Collections, Wyoming [Pa.] His- torical and Geological Society, XII, 30 pages, 9 figs., 1912.—A popular statement, with much original matter, of the geologic history of the Connecticut valley. All teachers of historical geology will be interested in the idealized structure sections, of which ere are nine, visualizing in graphic form the chief geologic events of central Connecticut since late Paleozoic time. Cc. 8. 7. Lighth Report of the Director of the Science Division, New York State Museum, Joun M. Crarxe, Director. Museum Bulletin 158; 217 pages, many illustrations, 1912.—Besides the annual report of the Director of the New York State Museum and of the Botanist, Entomologist and Zoologist, and of the archeology section, this volume contains the following papers of a geologic nature : “Notes on the geology of the Gulf of St. Lawrence,” by Dr. Clarke, describes the demotselles of Entry Island, one of the Magdalens ; the long Silurian Black Cape section of Cascapedia Bay, a section of about 7000 feet in thickness ; and a striking unconformity between the nearly horizontal Bonaventure (Lower Mississippian) and the nearly vertical Silurian strata at Little River Hast, Gaspé. L. Hussakof describes three Upper Devonian fishes from Scaumenac Bay, Quebec. These are an almost com- plete specimen of Coccosteus canadensis ; Husthenopteron foordi, nearly 3 feet long ; and Scaumenacia curta. KR. Ruedemann has .a “Note on a specimen of Plectoceras jason.” OC. H. Smyth treats of the genesis of the pyrite deposits of St. Lawrence County. The Director also describes and illustrates a recent remarkable find of more than 400 specimens of the Middle Devo- nian starfish Palewaster eucharis, from Mt. Marion near Sauger- ties, New York. cs. 8. Bettrdge zur Kenninis der marinen Mollusken im west-euro- pdischen Plioctnbecken ; by P. Tescu. Mededeelingen van de Rijksopsporing van Delfstoffen, No. 4, 96 pages, 1 map, 1912.— The introductory sixteen pages describe the character of the Pliocene deposits of the Netherlands, and the remainder of the work is occupied by an annotated list of the fossils and their dis- tribution, recording 248 species. Cus: 9. Canada Department of Mines.—Recent publications of the Canada Department of Mines (see vol. xxxili, 289) are as follows: (1) GroLogican Survey Brancu; R. W. Brock, Director. Memoir No. 13. Southern Vancouver Island ; by Cuaruzs H. Crapp. Pp. 208 ; 18 plates, 3 figures and map. No. 21. The Geology and Ore Deposits of Phoenix, Boundary District, British Columbia; by O. EK. LeRoy. Pp. 110 ; 7 plates, 18 figures and 2 maps. No. 24-E. Preliminary Report on the Clay and Shale Deposits 488 Scientific Intelligence. of the Western Provinces; by Hertnrtcu Rigs and Joseru KreEve. Pp, 231; 41 plates, 10 figures, and 4 maps. No. 28. The Geology of Steeprock Lake, Ontario ; by ANDREW C. Lawson. Notes on Fossils from Limestone of Steeprock Lake, Ontario; by Cuartes D. Watcortr. Pp. 23; 2 plates. In the Mines Branou, EvGrne Haanet, Director. Annual Report on the Mineral Production of Canada during the Calendar Year 1910. Joun McLetsu. Pp. 328. Also Pre- liminary Report for 1911. Pp. 24. An investigation of the Coals of Canada with reference to their Economic Qualities : as conducted at McGill University, Montreal. In six volumes. By J. B. Porrzr and R. J. Durty, assisted by T. C. Denis, EpGar SransFiExp, and special assistants. Vol. I, pp. xxiii, 232 ; 45 plates, 31 figures and 5 maps. Vol. II, pp- Xlli, 189 ; 17 plates, 25 figures. ‘ Mica: Its Occurrence, Exploitation, and Uses; by Hucu S. pz Scumip. Pp. xiv, 411; 38 plates, 67 figures and 22 maps. This report is a second edition of the earlier one by Fritz Cirkel, issued in 1905 {see vol. xxi, p. 405). The mica production, which had a total value of about $1,800,000 in 1906 and 1907 for the three countries from which the world’s supply chiefly comes, fell to less than $600,000 in 1909, and $745,000 in 1910. Of this last amount, $337,000 was the production of the United States, $265,000 of India, and $153,000 by Canada. The two former countries supply chiefly muscovite, while Canada produces chiefly phlogopite, or so-called amber mica. The prominent use of mica at present is in the manufacture of electric machinery ; the possibility, however, of replacing mica plates by sheets of mica board (‘ micanite”) made of thin broken fragments threatens to further demoralize the market, since it makes possible the use of much cheap material obtained from India. The scientific interest of the present volume is largely in the minute descrip- tion of the many Canadian deposits in Quebec and Ontario, both from the geological standpoint and with respect to associated minerals. Report on the Utilization of Peat Fuel for the Production of Power; by B. F. Haaner. Pp. 145; 10 plates, 17 figures and 17 diagrams. This is a record of experiments made at the Fuel ‘Testing Station at Ottawa, 1910-1911. Bulletin No. 6. Diamond Drilling at Point Mamainse, Proy- ince of Ontario; by Atrrep C. Lanz. Introductory by Atrrep W.G. Witson. Pp. 59, vi, with 5 plates, one figure, and one map. Cataloete of Publication of the Mines Branch, 1907-1911. Pp. 135. 10. Geological Survey of New Jersey.—Henry B. Kinet, State Geologist. Bulletin 6; pp. 82, 4 plates; Bulletin 7, pp. 37. —The annual administrative reports of the State Geologist opens this bulletin 6, which also contains a report on the improvement on the Shark River Inlet. Bulletin 7 contains an account of the Geology and Mineralogy. 489 mineral industry of the state by the State Geologist. The total value in 1911 amounted to $37,700,000, with clay and allied prod- ucts first at nearly $19,000,000, zine mining at nearly $9,000,000, and the Portland cement at $3,260,000. 11. On the Origin of the Himalaya Mountains, a Considera- tion of the Geodetic Hvidence ; by Colonel 8. G. Burrarn, C.S.L, R.E., F.R.S., Surveyor General of India. Professional Paper, No. 12, Survey of India, Calcutta, 1912. Pp. 25, pls. I1.— In this paper the facts regarding the density of the subcrust beneath the several regions of northern India are concisely summed up, and the arguments throughout are clear and terse, which explains the apparent brevity of the paper. The most sig- nificant fact from the author’s standpoint is the existence of a narrow suberustal zone of remarkable deficiency of density skirt- ing the southern base of the Himalayas. Between two stations twenty-five miles apart the deflection changes 45". The change calculated from the uncompensated topographic features would be but 25”. On Hayford’s hypothesis of isostasy it would be but 15”. The actual change thus gives remarkably large and unex- plained residuals. It is shown that this cannot be explained from the lighter specific gravity of the trough of alluvium which lies in front of the mountains unless this were enormously deep. It is also shown that a horizontal displacement of compensation within a depth of seventy-six miles does not explain it. Colonel Burrard then advances the hypothesis of a great sub- terranean rift. He considers that the subcrustal shell has cracked ; the northern portion, beneath the Himalayas, has shrunk and in so doing has wrinkled the upper crust, at the same time leaving a profound crack on the south which has become filled with alluvium. In comment it may be said by the reviewer that if the crack be assumed deep enough and the alluvium retains its light surface density to an indefinite depth such a hypothesis would explain the deficiency of mass, but it involves a mode of operation which is more difficult to account for and more in unex- plained opposition with the conclusions regarding the deeper crust than is the fact of the deficiency of mass. It involves hori- zontal tension under the northern side of the Indo-Gangetic plain, enormous horizontal compression immediately north with- out adequate mechanical explanation, an absence of consolidation of sediment with depth and a volume of sediment which is unexplained. It seems to the reviewer, therefore, that although Colonel Burrard has made an admirable statement of geodetic facts, his explanation is wholly undemonstrated from the geo- logic standpoint. The possibilities do not seem to be exhausted in the several hypotheses stated and the remarkable distribution of underground densities, to the knowledge of which Colonel Burrard has contributed, offers an inviting field for investigators. a5 12. Rocks and their Origins ; by GrenvittE A. J. Corn, Pro- fessor of Geology in the Royal College of Science for Ireland. 490 Scientific Intelligence. Pp. 175, figs. 20. New York, 1912 (Cambridge University Press ; G. P. Putnam’s Sons). —This book, as stated in the preface, “4s intended for those who are not specialists i in geology, and it may perhaps be accepted as a contribution for the general reader.” Besides a description of the composition of the chief rock types, a full discussion is given of the manner of origin, the manner of weathering, the types of scenery with which the rocks are asso- ciated. The geologic meaning of the rocks is thus brought out, and those features emphasized which are of chief interest, apart from the mere classification which is a chief aim of so many texts. It should serve to stimulate a general interest in the earth and its history, presenting the subject of the description of rocks in a most attractive and significant manner. J. B. 13. The Origin of Harthquakes; by Cuartes Davison, Se.D., F.G.S. Pp. 144, figs. 26. New York, 1912 (Cam- bridge University Press ; "G. P. Putnam’s Sons). This is one of the Cambridge manuals of Science and Literature. It is clearly written, free from technicalities, and available for the general reader. The fact that the author is a well-known student of earthquakes gives it in addition an authoritative character. Earthquake phenomena are described and examples are given of different classes of earthquakes. The inferences as to origin are discussed and the part which they play in earth movements. I2Bs 14. The Identity of Parisite and Synchisite ; 3; by C. PatacuE (communicated). —In a paper published in this Journal last year, by Warren and Palache, describing the Quincy pegmatites, the parisite of that locality was described fully and it was shown that in all probability the mineral synchisite from Narsarsuk, Green- land, was identical with it. This conclusion was not accepted by the discoverer of the latter mineral, Dr. Flink, and he requested a new analysis of synchisite for which he supplied material to Dr. Warren. This analysis has not been made; but meanwhile the investigation has been completed elsewhere. EK. Quercigh* has shown that the supposed chemical differences between the two minerals are due to alteration products contained in the analyzed synchisite ; that pure synchisite has the same formula as parisite ; the same indices of refraction ; and the same specific gravity. All grounds for maintaining the distinction between the two are thus removed. The refractive indices, measured on prisms, are probably more accurate than those previously recorded. Refractive indices of Parisite, Sodium light. Muso Narsarsuk Quincy (Synchisite) paren Quercigh Warren eee Meiapelh2 1°7690 1-757 @o2 Es 1°6717 1°6730 1°676 *Sulla identita della Sinchisite con la Parisite. Rend. d. R. Accad. Lincei ; xxi, 581-588, 1912. Loology. 491 15. Introduction to the Study of Minerals ; by Austin Frnt Rogers. Pp. xx, 522; 591 figures. New York, 1912 (McGraw- Hill Book Company).—The number of text-books on mineralogy has been much increased in recent years, and among the additions made to the literature the present volume deserves commendation for the care with which the material has been selected so as to present in a single volume, capable of being carried in the pocket, all that is most essential for the student, both in his classroom work and in the field. The subject of crystallography occupies the first hundred pages, and, although brief for a rather advanced discussion covering the entire ground, will answer the needs of a student with the help of the teacher. The next hundred pages are given to the physical and chemical characters, following which we have a series of six determinative tables. Then comes the descrip- tion of two hundred prominent species, about half of these being treated in greater detail ; the rarer species are omitted entirely. The concluding pages are devoted to the subject of occurrence and association and the uses in the arts. A glossary of mineral terms is a useful feature of the work. The illustrations are mostly new and well drawn and engraved. Ill. Zoonoey. 1. Outlines of Evolutionary Biology; by Arraur DeEnpy. Pp. xiv, 454, with 188 figures. New York, 1912 (D. Appleton and Company).—In many of our secondary schools and in some of our colleges the subject of biology is taught by a study of a number of different kinds of plants and animals, or in some cases of animals alone. Certain institutions offer separate courses in zoology and botany as elementary studies, although in recent years there has been a growing tendency in some of the best uni- versities to give an introductory course embracing the general biological principles rather than the study of specific types of ani- mals and plants. The nature of the vital processes in all organisms is made the basis of the work. This study may be followed later by the general courses in zoology and botany. It is for such an introductory course that this book is planned. It contains no general descriptions of plants or animals, but con- sists rather of a discussion of general biological phenomena and the laws and theories relating thereto. With this modern conception of the subject of biology, the book is divided into five parts, treating different phases of the subject. Part I embraces such subjects as the nature of life ; the essen- tial functions of living things ; the properties of the living sub- stance ; the relationships of all organisms ; differentiation and co-operation of cells and organs; the transition from unicellular to multicellular organisms ; the general development of organ- isms ; cell structure, physiology and reproduction. 492 Scientific Intelligence. Part II includes an account of sexual phenomena in plants and animals ; the origin and development of the germ cells ; sex dif- ferentiation and inheritance ; and the evolution of sex. Part III is devoted to variation and heredity, and Part IV to the adaptation of organisms to their environment, and the evi- dences of organic evolution as revealed by a study of comparative anatomy, embryology, and paleontology. Part V, on the factors of organic evolution, discusses the more important theories of evolution, concluding with the evidence as to the ancestry of man. The book is written in an entertaining style and, with the excep- tion of the discussion of the evidence as to the inheritance of acquired characters, is free from personal bias. Ww. B. O. 2. College Zoology ; by Ropert W. Hrener. Pp. xxiv,-733, with 558 illustrations. New York, 1912 (The Macmillan Com- pany).—While this is essentially a systematic treatise on the more important groups of the animal kingdom, it differs from most text-books of zoology in emphasizing the physiological rather than the morphological aspects of the subject. Structural details are given only as far as is necessary for a clear understanding of the ‘telationships of the different groups. Of the numerous illus- trations the majority are taken from our native species. Emphasis is laid on the economic importance of the various groups and species, and, in order to increase the general usefulness of the work, more attention is paid to the vertebrates than to the other phyla. In each phylum a single species of the greatest importance is chosen as a type and described in some detail before taking up the systematic account of the various groups included. The same system is followed for the classes of vertebrates. Ww. B.C. 3. Le Zebre: Studio Zoologico popolare; by Dr. AcHittE GRIFFINI. Pp. xxvii, 298. Milan, 1913 (Ulrico Hoepli).—The Hoepli manuals, of which this little volume is one of the 1200 alr eady published, consist of treatises on all branches of science, art, liter- ature, and industries. They are all written in popular Ttalan language by authorities in the various branches of knowledge. This book contains an account of all the known species and vari- eties of the zebra, with information concerning those specimens which have been kept in various zoological gardens and the hybrids which have been secured from them. We. C, 4. Principles of Economic Zoology ; by L. 8. DavuauEerty and M. C. Daucurrry. Pp. vii, 410, with 301 illustrations. Philadelphia and London, 1912 (W. B. Saunders Company).— This is essentially a condensed Natural History of Animals, with special reference to those of economic importance. In addition to a brief description of the structure, habits, and economic rela- tions of each of the important groups of animals, there is given, in smaller type, a vast amount of information concerning the natural history of numerous individual species. The book is designed for use as a text in a course in zoology where the relations of animals to human interests are aa a BAC, Miscellaneous Intelligence. 493 5. Elementary Entomology ; by E. Dwicur SANDERSON and C. F. Jackson. Pp. vii, 372, with 496 illustrations. Boston and New York (Ginn and Company).—This book is especially adapted to the needs of students in agricultural and other colleges for a text-book to accompany a short course in the elements of entomol- ogy. The authors are practical entomologists, and this book deals with the practical rather than the theoretical side of the subject. It furnishes an excellent introduction to such courses in economic entomology as may follow, and it will also be found to supply the general reader with the essential facts regarding the structure, life- history, and classification of insects. The attractive appearance of the book and the excellence of the very numerous illustrations should arouse a wider interest in insect study. It is such a book as the school-boy interested in out-of-door life will delight in using for the identification of his “specimens” and for information as to their ways of life. As is the case with so many first editions under joint author- ship, minor discrepancies in statement occur. There are also conspicuous orthographical errors. W. R. C. TV. Misce~rnanxovus Sorentiric INTELLIGENCE. 1. Publications of the Carnegie Institution of Washington.— Recent publications of the Carnegie Institution are noted in the following list (continued from vol. xxxili, p. 385): No. 74. The Vulgate Version of the Arthurian Romances : edited from Manuscripts in the British Museum by H. Oskar Sommer. Volume V. Le Livre de Lancelot del Lac. Part III, p. 474. No: 85. Index of Economic Materials in Documents of the States of the United States. Ohio 1787-1904. Part I, A to F, pp. 1-638. Part Il, G to Z, pp. 639-1136, 4to. Prepared for the Department of Economics and Sociology of the Carnegie Institution of Washington by ApELaipE R. Hasse. 1912. No. 149, Part II. The Production of Elliptic Interferences in relation to Interferometry ; by Cart Barus. Pp. vi, 79-168 ; 29 figures. No. 150. Guide to the Manuscript Materials relating to American History in the German State Archives ; by Marion D. LEARNED. Pp. vil, 352. No. 152. Studies in Luminescence ; by Epwarp L. Nicuots - and Ernest Merritt. Pp. vi, 225; 187 figures. No. 153. The Influence of a Magnetic Field upon the Spark Spectra of Iron and Titanium ; by Arraur S. Kine. Pp. 66 ; 6 plates, 4to. Papers of the Mount Wilson Solar Observatory, Vol. Il, Pt. I. No. 164. A Physical Study of the Firefly ; by Wittram W. CosLentz. Pp. 45; one plate, 14 figures. No. 166. The Composition of the Atmosphere with special 494 Scientific Intelligence. reference to its Oxygen Content; by Franors G. Benepicr. Pp. 115; one plate. ‘This investigation confirms the conclusion as to the essential constancy in the oxygen percentage of outdoor air, The experiments extended over a period of nine months, during which this constancy was maintained through all changes of the weather as to air-pressure, temperature, humidity, and other important conditions. The average of 212 analyses of air from near the Nutrition Laboratory showed 0:031 per cent of carbon dioxide and 20°938 per cent of oxygen ; observations over the ocean gave respectively 30°936 per cent (Montreal-Liverpool, 7 anal.) and 20°932 per cent (Genoa-Boston, 36 anal.) ; while at Pike’s Peak the result was 20°927 per cent (9 anal.). The extra- ordinary rapidity with which local variations, for example, in a city, are equalized was shown by observations on street air, which indicated but a slight trace of oxygen deficit ; even in the Boston subway at 9.30 a. m., the only changes were a rise of the carbon dioxide to ‘065 per ‘cent and a fall of the oxygen to 20°897 per cent. No. 167. A Bicycle Ergometer with an Electric Brake ; by Francis G. Benrepicr and Watrpr G. Capy. Pp. 111, 44, with 16 figures. 2. British Association for the Advancement of Science.—The eighty-second meeting of the British Association was held at Dundee during the week beginning with September 4; the registration at the opening reached the large number of 2379 members. The address of the President, Professor HE. A. Schifer, on the “nature, origin and maintenance of life,” has attracted much attention. 3. Introduction to Agricultural Mycology. 1. Soil Bacteri- ology; by Dr. Avex. Kossowicz. Pp. 143, 47 illustrations. Berlin, 1912 (Gebriider Borntraeger).—This is the first of two volumes by the author on agricultural mycology. It consists of four parts: The rédle of the elements (C, O, H, N, 8, P and Fe) under the influence of micro-organisms; the mycology of the soil; the mycology of manure; and the influence of the micro- flora on the soil. ‘he book abounds in literary references, which alone should make it a valuable publication. It is a thorough compilation of important data. Special emphasis is placed on the chemical changes which take place in the soil through the agency of moulds and the lower and higher forms of bacterial life. The book should prove to be of much value to the student of soil chemistry and bacteriology. Le E.R 4. Catalogue of 9842 Stars, or all Stars very conspicuous to the naked eye, for the Epoch of 1900 ; by W. T. Backxuouse, West Hendon House Observatory. Pp. xx, 186, 4to. Sunderland, England (Hills & Co.).—This very handsome volume is pub- lished to accompany a set of 14 large stars maps on the gnomonic projection (a projection from the center of the sphere onto planes tangential to the surface). The whole work is designed for use in observations of meteors, and the greatest care has been exer- Obituary. 495 cised over every detail to secure the best results for this purpose. The arrangement is by constellations. The designation of the star is given as it appears in six different catalogues, and its mag- nitude as the average of from 2 to 12 different determinations. The maps have been engraved on copperplate with the greatest precision. The reticulation is into degree spaces and the lines of projection have been drawn directly on the plates by a machine specially designed for the purpose. The boundaries of the con- stellations and the names of stars and constellations are printed in red from a separate set of plates. No pains have been spared to make the work the most convenient possible for the purpose intended, and the result leaves nothing to be desired. Ww. B. 5. Science Manuals.—The following are the titles of several elementary volumes in science issued by the Cambridge Uni- versity Press (G. P. Putnam’s Sons, New York, 40 cents each) : Life in the Sea; by James Jounsrone. Pp. vi, 150; 4 figures. Primitive Animals; by Grorrrry Smirn. Pp. x, 1533; 25 figures. Links with the Past in the Plant World; by A. C. Szwarp. Pp. viii, 142 ; 20 figures. An Introduction to Experimental Psychology ; by Cuaruzs 8. Myers. Pp. vi, 156; 20 figures. OBITUARY. Dr. Lewis Boss.—With the death of Professor Lewis Boss, director of the Dudley Observatory, Albany, N. Y., on October 5, in his sixty-sixth year, America loses its foremost representa- tive of the old school of astronomy of precision. After render- ing strenuous practical service on the U. 8. Northern Boundary Commission, he deduced his Catalogue of Declinations of Stand- ard Stars embodying new and authoritative ideas on the Syste- matic Corrections to Star Catalogues. Appointed to the charge of the Dudley Observatory, he carried out with almost unexam- pled rapidity and precision one of the zones for the great international undertaking of the Astronomische Gesellschaft’s catalogue. Occupied mainly throughout his life with investigations of fundamental star positions, and publishing a valuable Preliminary Catalogue in 1910, he planned a comprehensive campaign which he was at last, at the age of sixty, enabled to inaugurate with the aid of the Carnegie Institution. Infusing his energy into his staff, the observational work in the Southern Hemisphere (at San Luis, Argentina) was accomplished in an incredibly short time and it is sad beyond words that he was not fated to see the final goal of his labors attained. Undoubtedly, how- ever, he has left the work in such shape that others may finish it on the plans he has carefully outlined. Ww. L. E. 496 Scientific Intelligence. W. J. McGen, the geologist and anthropologist, whose death occurred on September 4, was a self-made man of varied gifts and strong individuality ; beginning life as a worker on a farm, by his ability and force of character he made for himself a prominent place as a scientific investigator. Of formal schooling he had but little, but he carried on his private studies in leisure hours, and thus fitted himself for work in law and surveying. Geology early interested him and his first extensive investigation was a survey of Northeastern Iowa ; in 1881-82 he also made a report on the building stones of Iowa for the Tenth Census. When thirty years old, he became connected with the U.S. Geological Survey, working in the southeastern part of the United States, where he was engaged for several years in mapping an area of 300,000 square miles. He devoted himself particularly to surface and glacial geology, and the pages of this Journal between “1878 and 1892 contain numerous articles by him on these and allied subjects. He was the first president of the American Anthropo- logical Society and had much to do with its organization. In 1885 he published a geological map of the United States and later he gave a report on the Charleston earthquake of 1886. In 1893 he joined the Bureau of Ethnology, as ethnologist in charge, and for ten years his work was largely in this field: an exhaustive investigation of Tiburon island, Gulf of California, was one of his most important labors. In 1903 he took charge of the depart- ment of Anthropology of the St. Louis Exhibition of 1904 and produced remarkable results. In 1907 he was appointed a mem- ber of the Waterways Commission, and of this body he was secretary at the time of his death. Thus his activity extended over the fields of geology, anthropology, and hydrography. His personality was decided, perhaps aggressive, and on many points both within and outside the scientific field he held original views, but his energy and ability were never questioned. It could hardly be a surprise to those who knew him that in his will he left his body to a medical school for dissection, and directed also that his brain be used for study and preservation for the cause of science. Dr. Morris Lorn, Professor of Chemistry in New York Uni- versity, died on October 8 at the age of forty-nine years. He was early associated as assistant with Dr. Wolcott Gibbs of Harvard University and followed him later in his own researches on complex inorganic salts; his work also extended into other lines. He was a member of the Board of Education, and was active in philanthropic work in New York City and elsewhere ; among numerous generous gifts he presented $50,000 to Harvard University in 1911 for the Wolcott-Gibbs Library. New Circulars. 84: Eighth Mineral List: A descriptive list of new arrivals, rare and showy minerals. 85: Minerals for Sale by Weight: Price list of minerals for blowpipe and laboratory work. 86: Minerals and Rocks for Working Collections: List of common minerals and rocks for study specimens; prices from 1% cents up. Catalogue 26: Biological Supplies: New illustrated price list of material for dissection; study and display specimens; special dissections; models, etc. Sixth edition. Any or all of the above lists will be sent free on request. We are constantly acquiring new material and publishing new lists. It pays to be on our mailing list. Ward’s Natural Science Establishment 76-104 OCottecn AveE., Rocuxstrr, N. Y. Warns Naturat Science EstaBiisHMENT A Supply-House for Scientific Material. Founded 1862. Incorporated 1890, DEPARTMENTS : Geology, including Phenomenal and Physiographic. Mineralogy, including also Rocks, Meteorites, etc. Palaeontology. Archaeology and Ethnology. Invertebrates, including Biology, Conchology, ete. Zoology, including Osteology and Taxidermy. Human Anatomy, including Craniology, Odontology, etc. Models, Plaster Casts and Wall-Charts in all departments. Circulars in any department free on request; address Ward's Natural Science Establishment, 76-104 College Ave. Rochester, New York, U.S. A. CON TH NTS: Page Art. XXXVI.—Volcanie Vortex Rings and the ditect con- version of Lava into Ash ; by F. A. Purret XXX VITI.—Quartz from Alexander Cgunty, North Carolina ; by J. E. Poaur and V. GoLpscHMIpt XXXVIIL—Palm from the Upper Jersey ; by N. E, Srevens XXXIX.—Pr eliminary Note on the Shower of Meteoric Stones at Aztec, Navajo County, Arizona; by W. M. Foorw..-. 2225) 2 Se ge dk Se XL.—Restoration of Limnoscelis, a Cotylosaur Reptile from New Mexico ; by 8S. W. WILLISTON XLI.—Iodie Acid Process for the Determination of Bromine in Halogen Salts; by F. A. Goocu and P. L. Bru- MENTHAL: 30. SL Shes 3 ee XLII.—New Chlorite from Northern Wyoming ; ; by J. EB. WoLrr SCIENTIFIC INTELLIGENCE. Chemistry and Physics— Dissociation of Hydrogen into Atoms, I. LANGMUIR : Determination of Sulphur in Pyrites, ALLEN and Bisuop, 477.—Separation of Arsenic from Antimony and other Metals, Moser and PrrsorEe.: Methods in Chemical Analysis, F, A. Goocu: Elementary Applied Chem- istry, L. B. Attyw, 478.—Dictionary of Applied Chemistry, E. Tuorrx : Aussere Reibung der Gase und ein neues Prinzip fiir Luftpumpen: Die Molekularluftpumpe, 479.— Emission Velocities of Photo-Electrons, A. L. Hueurs, 481.—Magnetism and Hlectricity, KE. E. Brooxs and A. W. Poyser, 482.—A Text-Book of Physics, A. W. Durr: Gravitation, F. Harris, 483. Geology and Mineralogy—Publications of the United States Geological Survey, 484.—Publications of the Bureau of Mines: Onondaga fauna of the Allegheny region, E. M. Kinpux, 485, — Preliminary report on the geology of the coastal plain of Georgia, O. VeatcH and S. W. STEPHENSON ; Classifi- cation of the geologic formations of the state of New York, C. A. Harr- NAGEL, 486.—Central Connecticut in the geologic past, J. BaRRELL; Eighth Report of the Director of the Science Division, New York State Museum: Beitrage zur Kenntnis der marinen Mollusken im west-euro- paischen Pliocinbecken, P. Tesco: Canada Department of Mines, 487.— Geological Survey of New Jersey, 488.—Origin of the Himalaya Mountains, S. G. Burrarp: Rocks and their Origins, G. A. J. Coin, 489.—Origin of Earthquakes, C. Davison: Identity of Parisite and Synchisite, C. ParacueE, 490.—Introduction to the Study of Minerals, A. F. Roemrs, 491. Zoology—Outlines of Evolutionary Biology, A. Dmnpy, 491.—College Zool- ogy, R. W. Heener: Le Zebre, A. Grirrin: Principles of Economic Zoology, L. S. and M. C. Daucuerry, 492.—Elementary Entomology, E. D. Sanprerson and C. F. Jackson, 493. Miscellaneous Scientific Intelligence—Publications of the Carnegie Institu- tion of Washington, 493.—British Association for the Advancement of Science: Introduction to Agricultural Mycology: Catalogue of 9842 Stars, W. T. Bacxnouse, 494.—Science Manuals, 495. Obituary—L. Boss : W. J. McGre: M. Loz, 496, ~~ SS ye) y ‘ VOL, MEXIV. DECEMBER, 1912. f Established by BENJAMIN SILLIMAN in 1818. THE AMERICAN JOURNAL OF SCIENCE. Epitor; EDWARD S. DANA. ASSOCIATE EDITORS Proressors GEORGE L. GOODALE, JOHN TROWBRIDGE, W. G. FARLOW anp WM. M. DAVIS, or CamsBrince, Proressorss ADDISON E. VERRILL, HORACE L. WELLS, LOUIS V. PIRSSON, HERBERT EF. GREGORY anD HORACE S. UHLER, or New Haven, Proressor HENRY S. WILLIAMS, or Iruaca, Prorressorn JOSEPH S. AMES, or Battrimore, Me. J. S. DILLER, or Wasuinerton. FOURTH SERIES ' VOL. XXXTV—[WHOLE NUMBER, CLXXXIV}]. No. 204—DECEMBER, 1912. NEW HAVEN, CONNECTIOUTY 14 “ \ % Cy ah . \ Ay. see OF, ational Mus THE TUTTLE, MOREHOUSE & TAYLOR CO., PRINTERS, 128 TEMPLE STREET. Published monthly. Six dollars per year, in advance. $6.40 to countries in the Postal Union ; $6.25 to Canada. Remittances should be made either by money orders, registered letters, or bank checks (preferably on New York banks). OF INTEREST TO THE SCIENTISTS. During the past month we have been very busy cataloguing our stock and new additions to it and the result is a number of pamphlets a brief descrip- tion of which we give below : OUR NEW MINERAL CATALOGUE. This consists of 28 pages and contains a list of minerals with prices just to give you an iéea of the variety and immensity, as well as the large number of new minerals we now have on hand, The recent new finds will be found well represented. This catalogue will be a useful and interesting addition to your literature on Mineralogy. OUR NEW GEM CATALOGUE. : This 12-page catalogue will be found a treatise not only on all the precious and semi-precious gem stones found but also of the Synthetic Gems, describ- ing how they are made, with photo-engravings from actual examples showing the chemicals used, the gem material in raw state and the gems after being cut ready for mounting. : This catalogue describes the different collections we have prepared to show them in their various forms and colors. Not the least interesting feature are the Antique Gems and Jewelry which are worn at the present time. CALIFORNIA MINERAL CATALOGUE WITH COLORED PLATE. This gives a brief description of the immense stock of California minerals and gems, also new finds that we haveonhand, Asis well known, California has become the greatest Gem State in the Union, and to show you the beauty of some of her gems, we have prepared a plate in the natural and brilliant colors of two of the finest crystals ever found. One is in the Harvard University and the other in a well known private collection. Both were originally in my stock. This picture will be a fine addition to any studio or collection, as the colors are faithfully reproduced and the plate is not attached to the catalogue. HOW AN INSPIRATION BECAME AN ACTUALITY. This is a little sketch, illustrated with eight half tones, showing the parts of our studio and laboratory and myself and assistants at work. It was’ written by the editor of the ‘‘Guide to Nature” and is reproduced just to show how our stock of interesting specimens inspired and interested him, We trust it wil] interest you. It originally appeared in ‘‘ The Mineral Collector,” to whom we give credit. All the catalogues will be sent free of charge. In writing please mention the JOURNAL OF SCIENCE. ALFRED H. PETEREIT 261 West 71st’ St., New York City. - THE AMERICAN JOURNAL OF SCIENCE PERO RU Pre Ss hku Base ————_#e Art. XLIII.—A Buried Wall at Cuzco and Its Relation to the Question of a pre-Inca Race ( Yale Peruvian Hxpedi- tion, 1911); by Isatan Bowman. Dorine the summer of 1911 the Yale Peruvian Expedition spent some time studying the surroundings of Cuzco, and in a ravine* on the outskirts of the city came upon a group of facts in relation to a buried wall that appear to have a large signifi- cance. The wall belongs to the older type of rougher and apparently more primitive architecture which students of Peruvian archeology have for some time been inclined to regard as pre-Inca, though the belief has heretofore not depended upon secure evidence. It appears that we now have a fairly safe basis for concluding that the wall is pre-Inca, that its relations to alluvial deposits which cover it indicate its erection before the alluvial slope in which it lies buried was formed, and that it represents the earliest type of architecture at present known in the Cuzco basin. The wall extends along the border of a ravine on the south- western edge of Cuzco, BM,—BM.,,, fig. 1. The detailed topo- graphic features are brought out in “figs. 2 to 5. The lower end, BM,,, fig. 1, is now almost entirely ‘uncovered ; ; the upper and middle por tions are deeply buried in coarse but well-strat- ified gravel of prehistoric though postglacial age. ) 18000 4815 _ 5421 — 8274 - — 7627 iy 4658 4911 5525 —7887 —7922 —'7599 16 4962 5203 5664 —71703 —1774 — 75384 15000 5209 5609 5811 — 7602 —7506 —7416 14 5040 5750 5956 — 7494 —7334 —7370 13° 5730 5911 6087 —7431 — 781 —7361 12 5948 6024 6130 — 7329 — 7369 — 7285 11 6096 6060 6188 — 7285 — 7362 —7306 10000 6164 6086 6245 —7253 —7336 —7272 9 6248 6127 6341 —7379 — 7309 —7209 8 6302 6199 6455 — 7336 —7298 —7188 7 6329 6271 6044 —'7249 —7271 —T157 6 6405 6368 6643 —7198 — 7248 —7092 5000 6510 6473 6750 —7160 —7173 —7066 1 6601 6591 6853 —T131 —7123 —7024 3 6717 6694 6930 —7097 —7091 —6987 2 6825 6846 6974 —7066 —7033 — 6983 1 6969 6911 — —7045 — 6935 — 000 — = = — — — Radiation Energy Kio Exponent of Radiation A 18000 | 14844 _ 12005 4-44 17/ 16704 17488 13968 4:92 16 18497 18647 15993 5°92 15000 | 20083 19328 18060 4°36 6-82 14 21850 21544 20582 5°36 5°66 13 23883 24781 23624 3°86 5°37 4°45 12 26548 28286 27013 5°13 3°90 3°67 11 29953 32088 31143 3°87 3°70 4:03 10000 | 384502 37288 35422 4138 3°56 3°86 9 40630 42938 39717 Bai 4°25 6°28 8 45275 47822 44616 3°93 4-41 4:02 a 51345 53939 50082 4°25 4°38 502 6 57729 59582 50577 4-60 5°10 4°97 5000 64352 65980 61930 5°19 5°76 5°10 4 71739 72304 68663 3°66 4°63 3°57 3 77430 78606 74258 5°94 4°55 4:08 2 86011 86055 82802 5°24 5°64 3°94 1 94582 93656 — 6:09 5°21 — 000 4) = = =e = = rs Earth's Nonadiabatic Atmosphere. 527 6000 to the isothermal layer, and then becomes negative and probably of increasing amounts. The tendency to negative values sets in high over the equator and then diminishes in elevation towards the poles. The pressure P is higher over the middle Tropics (latitude 20°) than over the equator or over Europe, and shows the high pressure belt to the level 17,000 meters. The barometric pres- sure is B = P/gp,, where py = 13595°8 the density of mer- cury in kilograms per cubic meter, and g = 9°8060 the accel- eration of gravity in meters per second. At Lindenburg the pressure difference between computed and observed values shows that the observed values range from 0™™" to 1:5™™ larger in the strata up to the highest altitude reached ; the observa- tions in the Tropics are more nearly accordant, the observed pressures being 0:07" to 0°5™™ smaller than the computed. The pressure over Europe is less than over the Tropics in all levels, and hence there is an eastward wind movement over Europe, and a westward movement over the Tropics between the high pressure belt and the equator. Zhe density p increases from the equator to the poles up to 10,000 meters, and about that level becomes higher over the middle Tropics than to the north or south. The specific heat Cp isa variable ranging from 993 to 624 in these computations. This departure from the normal value 993 increases with the height, and by the formula, (Cp, = Cp,,) (i, ae it) = 2(9. he q ) a (Q, Te Q.), is responsible for the kinetic energy of circulation per unit mass and that of radiation. Since 4(g,*— q,’) depends upon the change of the velocity per 1000 meters, and since the velocities are small in the isothermal layer, it follows that the loss of heat (Q, — Q,) is due to the change in the specific heat at high elevations together with the low pressure and density dependent upon the observed temperatures. The complex thermodynamic adjustment involved in the formulas cannot be easily summarized in a sentence, but it will become clearer to the reader by considering the following paragraphs. The specific heat on the East African plateau was assumed to be 993 at the surface, though the elevation is 1140 meters above sea level, and so the computed Cp is not readily compared in the upper levels with the data of the Tropics and Europe. It is probably about the same in the free air at all levels in all latitudes, when the plateau effect has been eliminated. The velocity qg increases from the ground up to the iso- thermal layer, and then rather suddenly drops in amount in this layer. The velocity generally increases from the Tropics towards the poles, including latitude 52°, as is well known. 528 I. H. Bigelow—Thermodynamics of the The expended heat (Q, — Q,) per 1000 meters in elevation increases in amount by two distinct systems, one in the strata from the surface to the isothermal layer, and another in that layer. Table 2, and these are found in Table 3. TABLE 3. Take the differences in the columns for (Q, — Q,) of The variation in the loss of heat per 1000 meters. d. dQ _a yy ae Ge = aS . (Q: = Qo) = Constants. Height z Europe Tropics Victoria Nyanza 18000 —228 = —1382 17 —440 —442 — 204 16 —399 —346 —672* —878 — 265 15000 —437 —314 —190 14 — 253 —114 —141 13 — 321 —125 —118 12 —211 — 43* — 38* 1] —101 — 59% — 90 10000 + 49% — 68* —160 —143 9 — 97 — (82 —135 8 —114 —100 —140 —120 7 —127 —137 —119 —165 6 —143 —181 —131 5000 —120 —168 —145 4 —150 —135 —115 3 —139 —210 — 49 2 —164 —163 — 8 1 =i — 24 = 000 — = = Mean Ais Os) — 362 for the isothermal layer. 1000 AQ: = Qo) _ Mean OU0RGL 140 for the lower layers. x There are three facts to be noted, (1) that from the surface to about 2000 meters, in the stratum within which the diurnal convection is confined, the loss of heat diminishes less than in the strata above 2000 meters. In the lower levels there is an accession of heat which prevents as rapid cooling as in the higher levels, and this is probably due to the condensation of the aqueous vapor, which is a source of heat in addition to that Earth's Nonadiabatic Atmosphere. 529 radiated from the surface. The incoming radiation, after cer- tain depletions by scattering and absorption in the atmosphere generally, raises the surface of the ground or water to a certain . temperature which radiates with long waves. The outgoing radiation passes through three stages : (1). In the levels 000 to 2000 meters, where the heat supply is due to radiation at surface temperatures and the heat of con- densation of aqueous vapor. (2). In the levels 2000 meters to the isothermal layer, where the radiation is chiefly from the dry air, and there is no addi- tional supply of heat of importance from the condensation. The constant of accelerating loss is about —140. (3). In the isothermal layer the radiation is much more rapid, with a constant acceleration of loss of about —362, in spite of the fact that the temperature itself is nearly constant or slightly higher than at the lower boundary of the isothermal region. In Europe the isothermal temperatures are about 202° to “208°, and in the Tropics about 209°, generally, not counting the single observation at Victoria Nyanza. It should be noted, in passing, that the thermodynamic data give no support to Abbot’s hypothesis of an effective radiation layer at 5000 meters, since the only disturbance in these values is just below and on entering the isothermal layer. The vari- ous hypotheses regarding the origin of the heat found in this layer may proper ly be revised. There is apparently a con- gested or mixing region on the north side of the high pressure - belt, in the temperate latitudes at lower levels, 10, 000 to 13 ,000 meters, and over the tropics at higher levels, 13 000 to 15, 000 meters, and this may be in part due to some form of overflow during the general process of polar circulation. The theory that it is due to such by-products of the incoming radiation as ozone formation is contradicted by the fact that such formation should be able to penetrate to lower levels over the Tropics than over the temperate zones, which is contrary to the entire series of facts. In my opinion the apparent heat of the iso- thermal layer is due to the inability of the radiation to escape fast enough to reach the thermodynamic demands implied in (Cp. oe Cp,,) (2, Fr T,), although it is observed to carry off heat in the isothermal layer much more rapidly than in the lower levels. That is to say, (Cp, — Cp,,) increases so rapidly with the elevation that (I, — T,), the departure of the temperature from stratum to stratum upon which Cp,, depends, cannot proceed rapidly enough by means of radiation losses to give the temperature fall observed in the lower strata. This choking process sets in at a certain low temperature, 202° to 208°, and of course the Am, Jour, Sc1.—Fourts Srmriss, Vou. XXXIV, No. 204.—DxrcemBnr, 1912. 35 530 FE. H. Bigelow—Thermodynamies of the primary physical question becomes the natural rate of radiation of heat at certain low temperature wave lengths. In any event the marked change in the loss of heat, from —140 to —362 per 1000 meters below and within the isothermal layer, is involved in the physical conditions of the phenomena. Before crossing the boundary of the isothermal layer there is an increase in the heat supply, as at the values marked with a cross and omitted from the summation, as if there was a congestion or accumu- lation of heat, in the cirrus cloud region, before changing from the slow rate of loss —140 to the rapid loss — 362. It is appar- ent that further studies of these data are likely to lead to a differ- ential equation of radiation, which differs from the theoretical equations that have been hypothetically proposed by several authors. “ The Entropy (S,—S,) increases with the height, but more rapidly on entering the isothermal layer; it increases generally from the equator towards the poles, except in the high pressure belt, from 3000 to 8000 meters elevation. The work expended ( W,— W,) diminishes with the height, and more rapidly towards the temperate zone except in the same region of the high pressure belt. The inner energy (U,—U,) is substantially a constant, about 7200 per 1000 meters, except in the strata near the surface where it is smaller 7000, and in the isothermal layer where it is larger and increasing to 8000 or more. 0 The radiation energy K,, = eee distributed and ranges from 94,000 near the surface to 13,000 at the 18,000 meter level. The loss of radiant energy is accom- plished by means of the complex thermodynamic processes just described, and it is quite unavailing to discuss this problem by means of generalized simple equations that take no acconnt of the mechanical and thermal requirements of the atmosphere. log K, — log K, log T, —log T, 4-00 in value, and the atmosphere conforms to the radiation of a black body. In the lower levels this exponent is larger, prob- ably due to the heat contents of the condensing aqueous vapor. This can be separated from the dry air as follows : A 4:00 + (A — 4:00) 4:00 A—4:00 K, Ae wr, = sts a S-)-G) ee dry air vapor Thus if A ==5750; 4-00 1:50 is of course similarly The exponent of radiation A = is about 5381 Earth’s Nonadiabatic Atmosphere. ‘pezjIMo oLey or ‘sIojouI YNO‘ET, Suryover Ajuo ‘g Avy pue yg [lady jo suotsueose oy, pil ea, Rage a ee ae fares oe pas See ae cae 000 age CG.F 83-8 If-8 OF-9F— | P9-F 89-81 FL-9 486-96 BE = I F6- 10-9 €9-F 6G-F GL-G LS) LL-¥ 90-f 68-8 §1-8 G 80-F CL.g eg.g eL-F 6¢-P 6G.G 68.9 06-9 0S-f ¥G-L, g 1¢.8 88-9 VS-F G8- 816 PE-G Ce.G OL-6 (0-4 €9- G7 OL-g 89-4 8E-g 16-9 86-9 1@.g 81-9 Gb-F LO-P 8F-9 000¢ 16-¥ 9¢.g 80 9 CG. 7G. Os — 99.3% SI-4 LG.7 68-8 9 G0-G 69-F GL-¥ 19-6 90.¢ 8¢.F 81-¢ 96-8 PLP €6.g L G0-F 86 F L§-& 89-G L9-F 10:4 C8.f eF-€ 1g. G1-g 8 86-9 9¢.F 98-F 10-¥ 6F- GE-F CS-F 09-4 898 x1 4-0 6 98-8 €0-F 10-8 GLE aS 82-E 1G- 69-61 — C().F 16-¥ 0000T 60-7 668 60-7 81-8 80-8 69-8 $8-3 G0.89— €9.8 OF-F IT L9.€ 81-6 SL-F 10-¥ 60-F 98-€ + | II-P 91.794 17-9 + | 6681+ al CET 00-4 9F-8 60-7 66-9 + | 80-CFPI+ | 16% + 66-11 +. CEG = | ine = eT 99.¢ 96-8 16-9 98-¢ OoHe tse || GOL | Sei — 99-88 — Grill = | S2.¢ = ial! 68-9 g9- &6-P 13-4 icG = || LOLI = | iPleip = Ee ey Fe-L0L— | LT-16— | 00041 G6-¢ Bae Ise | LEaeese | iil = | OF GZi— || Soe — ake a Eloi —= || SOL 91 66-4 Becks rete gia Gal == | LO: = |) =s= es 0-87 = | 0G: + me tPF Bose eee are ORG == | ee 3 ae Ge.8. + 90:8 — | sor — | 00081 20 Aca aot em 86 oe off 06S 06g 6G 4RT 8061 9061 LOGE L061 LOGI L061 9061 6061 8061 606T qU slop wkN ‘OLA | 62 eane | ge AINE | Go SnY 6 4ydeg ce ydeg L ydeg G ydes Le Syne g Avy Z OFF, oT ae es *"I1eXkV] [BULIOW}OSE oY} SULIeJUe uO ( ) = Ul VW jO juouodxs oy], ‘Pf AAV, Vv L 5382 LF. IL. Bigelow—Earth’s Nonadiabatie Atmosphere. Further studies on this subject have been undertaken. Table 4 gives an exhibit of the values of A for ten ascensions, for the purpose of showing the turbulent nature of this exponent near the surface of the earth, and especially in the isothermal layer. Near the equator the computed values of A are approximately 400 to great altitudes, but they become very large and unsteady in the isothermal layer, in a region which increases in latitude as it descends in altitude. This is due to the fact that, ae 1. Sea feet =e 0 where T, is nearly the same as T,, that is, the temperature is nearly equal in the layer of computation. It is through such irregularities that the formulas are compelled to proceed in computing P, p, R and the other terms from the observed T in the nonadiabatic atmosphere, and this shows how impossible it is to expect to successfully apply the adiabatic formulas, for n=1 and R, Cp constant, in these meteorological problems. Transformation of the heat wnits. The K. M.S. system of units is transformed to the C. G. S. system by the factor 1000 x 100 X 100 = 10". The mechanical equivalent of heat in the K. M. 8S. system is 4185°57 kilogram-meters or 418557 x 10° C.G.8. for 1000 meters in the stratum. For example, having 660 mechanical units K. M.S. per 1000 meters; this is equivalent to 0°1577 large calories or 157-7 small calories per 1000, or 0°1577 small calory per meter of altitude per square meter of surface. In this way the mechanical units of heat Q,— Q, (K. M.S.) in Table 2 can be converted into small calories per square meter and then submitted to an integration. Conclusion. - (1) The dry air radiates like a perfect black radiator with the Stefan exponent 4:00 from the level 2000 meters to the isothermal layer. In the surface stratum it is more than 4:00, due to the added heat of condensing aqueous vapor, and in the isothermal layer it is very irregular, due to the small changes in the value of the temperature per 1000 meters. (2) The isothermal layer of relatively high temperature seems to be due to a congestion in the rate of radiation, depending upon the complex thermodynamic relations of the pressure, density, gravity, which produce a large change in the specitic heat, a small change in the temperature, but an accelerated loss of radiant heat. F. W. Very—WNote on Atmospheric Radiation. 538 Art. XLVI.—Wote on Atmospheric Radiation; by Frank W. Very. Tur heat which gives to any layer of the atmosphere its observed temperature is derived from several sources which may be designated as follows: +h,=heat transferred by convection from warmer and lower layers of air. +h, =heat transferred by molecular penetration from warmer neighboring air masses in any direction. +h, = heat derived from the condensation of aqueous vapor. + h,=heat from absorption of solar radiation by the molecules of gases and vapors, or by floating dust from which it is transferred to gaseous molecules. + h,= heat from a similar absorption of terrestrial radiation. +h, = heat produced by absorption of radiation from warmer air masses, situated either above or below, but usually below. Here each atmospheric ingredient absorbs selec- tively and solely the radiations of those special wave- lengths which it can itself emit. + h;= heat from chemical changes, such as ionization by solar ultra-violet radiation. The heat produced in this way may be included under the heat (f,) from absorption of the -sun’s rays. ‘he direct effect of ionization on temperature is probably small, but it may have important secondary consequences through the production of peculiar substances in small amount, but possessing extraordinary powers of selective radiant absorption and emission. + h, = heat imported, or abstracted by the planetary atmospheric circulation. The relative importance of these sources of thermal supply varies greatly at different elevations. In the isothermal layer (3), the heat received is Hi =h, +h, +h, + hy. Convection may not have ceased entirely, but it is very small, and has been omitted from the expression. There are no clouds here and terrestrial radiation passes through unchanged, having suffered all the absorption of which it is capable in lower layers. The heat from absorption of solar rays becomes exceptionally powerful where the solar radiation first enters the aqueous atmosphere in the higher levels of the isothermal layer. Between the isothermal layer and the air near the ground ie the great body of the cloud-producing atmosphere (2), where H,=h, th,t+th, +h t+thth th. 534. FL W. Very—Note on Atmospheric Radiation. Near the earth’s surface, the absorption of terrestrial radia- tion replaces that of the sun’s rays to a great extent, because the absorbable part of the latter has already been removed by the upper air, except as the dust of the lower air absorbs radia- tion of every wave-length. The terrestrial radiation is chiefly effective in warming a shallow layer of air (1) near the ground, and its absorbable rays are soon sifted out in passing through the lower humid layers; but it is also efficient at any level where there is cloud and where the regions of absorption in the spectrum of terrestrial radiation are much _ broader. Emphasizing this importance of heat from absorption of ter- restrial radiation by aqueous vapor, it is placed first on the list in the enumeration of heat sources for the superficial layers. WH, =A +h, +h thy th, the & hy The heat thus acquired may be removed by evaporation of precipitated water particles (—A/.), by atmospheric radiation (—h,), by convection (—/,..) and by penetration (—A,). If there is thermal equilibrium, at level (3), A, + hth, +h, —(h, +h) =0, Co (2); BSAA hee he, res es 0, CA ds (1), lets _ (A. +h. thy + hy) == (), Except in the upper air, the distance through which atmos- pheric radiation can proceed before it is reabsorbed by the air, is so small that convection and penetration might be expected to supersede radiation as available modes of thermal transfer- ence from one body of air to another. We are at last able to apply a critical test to this supposition. The study of the vertical distribution of thermal losses in the earth’s atmosphere which Professor Bigelow has made,* permits for the first time a comparison of atmospheric radia- tion, transferred from layer to layer in the free air, with that computed from laboratory measures. The radiation from a layer of pure dry air, 1 meter in depth, and having an excess of temperature of 10°C., may be taken as equivalent to a transference of 0-00006 gram-water-degree- Centigrade units of heat per second through each square cen- timeter of surface.t Bigelow divides the atmosphere into layers 1000 m. deep. ‘The difference of temperature at the top aud bottom of each layer is somewhat less than 10°, varying in fact between 5° and 7° C. throughout a considerable part of the first 10,000 meters. If radiation (R) from layer to layer is *Frank H. Bigelow—‘‘ The Thermodynamics of the Earth’s Non-adiabatic Atmosphere,” this Journal, xxxiv, 510-582, + Frank W.Very—‘‘ Atmospheric Radiation.” Bulletin G, U. S. Weather Bureau, p. 112, 1900. F. W. Very—Note on Atmospheric Radiation. 535 simply proportional to the depth of the layers and to the fourth power of the absolute temperature, we get the values in the third column of the following table. This, however, in no case represents the direct radiation of the atmosphere from layer to layer, because the selective radiations of the gaseous molecules can "proceed but a little way before they are absorbed by other molecules of the same gas. These in tum radiate again to cooler ones at a still higher level, and so on,* but the number of successive transferences of heat is very g ereat even in a layer no more than one thousand meters thick, and the process is a slow one, not to be compared for rapidity with the rate of escape of unimpeded radiation. Taking averages through the first 15,000 meters of air, we get the following mean values of energy expended as radia- tion in large calories per square meter, for comparison with Bigelow’s AQ, or “ heat expended.” Per thousand meters Elevation gurl as ce AT | R (computed) |AQ (observed) Lees ate [fae mera otk Upper layer 15000—10000™ 0°96 C. 0°057 0-408 Middle layer 10000— 5000 figs 0°429 0-237 Lower layer 5000— 0 D4 0°314 0:090 The sudden diminution of the computed radiation (from stratum to stratum) in the upper layer follows from the en- trance into the isothermal region, and from the diminution of AT; but this does not imply a cessation of atmospheric radia- tion in this nonadiabatic layer. On the contrary, the observed increase of AQ in the upper layer is presumably due very largely to direct radiation to outer space. That this outgoing radiation produces no appreciable fal] of temperature with increasing elevation, appears to be because the heat of the upper layer is continually replenished by absorption of the radiation from the lower air, but especially because the solar rays are undergoing their initial absorption, and though some of the absorbent mater ial, such as aqueous vapor, is excessively rarefied, there are certain rays which are so powerfully absorbed that they suffer loss in passing through the most ten- uous layer.+ The values of R, computed on the above assumption, are con- siderably greater than those of AQ in the lower atmosphere, * Op. cit., pp. 114-115. + F. W. Very—Atinosvheric Radiation, p. 128. 536 FF. W. Very—Note on Atmospheric Radiation. where we must presume either that atmospheric radiation is itself impeded by vaporous absorption, or that convection plays a larger part in the denser air. The actual mean direct radiation between neighboring air masses, which lies at the foundation of the successive ‘absorption and radiation of the indirect process, must be smaller than that assumed. The exten- sion of an observed air radiation for a layer 1 meter thick, at an excess of 10°, to a layer 1000" thick with the same excess, or a proportional one, is only justifiable on the supposition that in the passage of heat from molecule to molecule by radiation, the small distances traversed in successive steps compensate for the small differences of temperature and feeble radiations of the separate interchanges whose integra! finally reaches a sum total of energy equivalent to that of a direct transmission. Whatever doubt one might have as to the legitimacy of the assumption that the two processes are even moderately equiva- lent, is now set at rest by the near agreement of the exponent of T for the indirect process with the exponent (4) required by Stefan’s law. This seems to prove that, after all, the process is a radiative one, rather than convective. The potential radiation function which Bigelow derives from the internal energy of the air by the formula where U, and U, are the internal energies at the top and bot- tom of a layer, and v,—, is the change ‘of volume of unit mass of air in the same interval, at an altitude of 18,000 meters, is about 0°16 of that at the earth’s surface, which is 0°4 of that for a black body.* This ratio is not improbable considering the variation of the composition of the air in a vertical column ; but a distinction must be preserved between the passage of. * With Kurlbaum’s constant of radiation, R=5°32 x (10) x T+ joules/sq . m. sec., a black body radiates at 289° and 225° abs. C. (which are tem- peratures observed in Europe at 0 and 18,000 meters) 3,711,000 and 1,363,000 mechanical K.M.S. units, or 0°532 and 0°196 small calories per sq. cm. per min. (Ratio = 2°71:1). Bigelow obtains a ratio of potential air radiation = 6°37:1. The sum of the values of the radiant potential (K) between the surface and 18,000 meters altitude is about one-third of the difference for black radiators at the given temperatures. which may mean that the emission bands of the air spectrum cover about one-third of the entire range. On rising in the air some of the vapor bands become narrower and drop out, and thus the potential radiant function diminishes more rapidly than that of the black body for the same fall of temperature. The air radiation from laboratory measurements agrees with the average AQ, which therefore repre- sents the actual air radiation to space and increases upwards with the progressive removal of the outer obstructing layers. The surface layers of air have large K and small AQ of necessity, for it is only because these layers are not free to radiate to space in accofdance with their high potential that the lower air can maintain its relatively high temperature. EF. W. Very—Note on Atmospheric Radiation. 587 energy through the atmosphere by the modified radiant process, outlined here, and a-pure radiation. The sudden increase in the constant second differential of Q at the isother- mal level, strongly suggests an accession of immediate radiation to space above this point, and a progressive diminution of the indirect radiant process from thence outward, together with a cessation of aqueous condensation which supplies | considerable heat to the lower layers, while further loss of temperature is only permitted through the gradual removal of absorbent sub- stances which prevent the passage of radiation of great wave- length, characteristic of still lower temperatures. The methods employed by Bigelow still retain something of empiricism. Wherever a temperature gradient jess than that of the adiabatic rate prevails in the atmosphere, some heat must have been supplied by radiation, or by condensation of aqueous vapor, or in other ways; and this extraneous addition is not taken into account by the formula which connects the density with temperature and pressure. It is known from laboratory experiments that the specific heat of air at constant pressure is very nearly constant between — 39° C. and + 200°C. Wherever the values of Cp computed by Bigelow’s formula depart from the true value given by laboratory experiment, the deviation of the fictitious or apparent Cp denotes that only a part of the thermal variation with the altitude is due to simple mechanical transportation and expansion of masses of air. The deviation becomes progressively greater in the npper air where production of heat by absorption of solar radiation is large. From the intimate relation between Cp and R, the gas factor, a corresponding variation is produced in the fictitious value of R. Similarly, if the computed value of the air radia-— tion is larger than that of AQ in the lower layers of air, this must mean that the data employed include other phenomena besides those of radiation. Although the upper air is more transmissive than the lower air for radiations of wave length less than 15, the final limit to further decrease of temperature by atmospheric radiation to free space appears to result from the virtual opacity of the air to longer waves than these, coupled with the fact that a temperature has been reached at which these long-waved radiations constitute the greater part of the radiant output. This limit is only extended by the very gradual removal of the last traces of absorbent at. still oreater altitudes. These remarks are not intended as a eriti- cism of the methods, for it is difficult to see how any others can be used, but are made to point out the limitations of an analysis which is obliged to deal with data covering a variety of simultaneous processes in unknown proportions, and in the endeavor to elucidate the meaning of the results. 5388 FLW. Very—Note on Atmospheric Radiation. Although the radiation function in the lower air, where only indirect radiation exists, can be fairly well represented by a radiation formula with the exponent of T (the absolute temperature) 2a to 4, so that indirect radiation is still satistied by Stefan’s law ‘with small changes in the constant,* yet in the upper isothermal layer where free radiation ought, if anywhere, to be in control, the exponent of the radiation formula suddenly becomes negative . This discrepancy, how- ever, is more apparent than real, and appears to be due to the super position of several effects attributable to the progressive removal of absorbent vapor in the upward direction, combined with a special absorption of solar rays and the limitation of long-waved radiation. The observed value of air radiation has an immediate application to the phenomena of the nocturnal deposition of dew where the atmospheric cooling concerns an air layer only afew meters thick; but this value may require modification when applied to the conditions of the elevated isothermal region. Here the diminished radiant mass in unit volume has the outflow of its radiation of smaller wave-length than 15 proportionally less obstructed through absorption by its own substance, which tends to give constant radiation in spite of varying density, and thus there is a new freedom of radiation from the upper surface which was not possessed by the air at lower levels, while at the same time there is a strong accession of thermal energy from absorption of the sun’s rays by the first portions of aqueous vapor encountered. Tis absorbent layer immediately becomes a radiant one and its extra heat is dissi- pated by radiation to outer space. (The region of the accession of freer atmospheric radiation to outer space is about 3000 meters higher in the torrid zone than in temperate latitudes and all of the data are correspondingly modified.) This seems to be the meaning of the sudden increase in the loss of heat on entering the isothermal layer, which passes, according to Bigelow’s computation, from —140 in mechanical K. MS. units per 1000 meters for the lower layers, to — 362 per 1000 meters for the isothermal layer. Westwood Astrophysical Observatory, Westwood, Massachusetts, May, 1912. * We knew already that. within the very limited range of its selective radiation, the maximum radiant laver of a gas behaves like a black body. Linhart— Hydrolysis of Alkyl Metallic Sulphates. 539 Arr. XLVII.—On the Hydrolysis of Alkyl Metallic Sul- phates; by G. A. Linwart. [Contributions from the Kent Chemical Laboratory of Yale Univ.—cexxxvii. ] Il. Ethyl Calcium Sulphate and Ethyl Strontium Sulphate. Ir has already been shown* that for methyl-, ethyl-, and propyl barium sulphates the rate of decomposition in acid solution decreases with increase in the weight of the alkyl group. It was the purpose of this investigation to study the effect, on the rate of hydrolysis, of replacing the barium in ethyl barium sulphate by strontium or calcium. Preparation and Analysis of the Esters —The same method was used for the preparation of these esters as that described in the first two paperst of this series. In order, however, to increase the yield the mother liquor from the first crystallization was:concentrated on the steam bath and, before filtering through the hot water funnel, was treated with finely divided carbonate of calcium or strontium in slight excess to neutralize the small amount of alkyl sulphuric acid formed. In the case of the strontium ester the crystals obtained from the fourth fraction were discarded, as they con- tained some strontium sulphate, while only the first and sec- ond fractions of the calcium ester were free from calcium sulphate. After treating the crystals as described in the second papert of this series, their composition corresponds to the formulee : Sr(C,H,SO,),.H,O and Ca(C,H,SO,).2H,0. Method of Hydrolysis.—The method of hydrolysis described in previous papers§ was slightly modified. Since these esters gradually lose their water of crystallization, approximately the required amount of the crystallized substance was dissolved in distilled water and the concentration determined in the wet way. Ten cubic centimeters of this solution and 10™* of hydrochloric acid were introduced into test-tubes of about 40°™* capacity, previously constricted at about 11/2 inches from the mouth to a size just large enough to allow the stem of a carefully calibrated pipette to go through and the dis- placed air to escape. The tubes were then sealed, allowed to cool, shaken and submerged in the thermostat. The hydro- chloric acid used contained one gram of barium chloride (BaCl,.2H,O) for every 10° to transform the calcium and strontium sulphates, resulting from the hydrolysis of the cal- * This Journal, xxxiv, 292. + Ibid., xxxii, 53, and xxxiv, 292. ¢ This Journal, xxxiv, 292. § Ibid., xxxii, 53, and xxxiv, 292. 540 Linhart—Hydrolysis of Alkyl Metallic Sulphates. cium and strontium esters, to their chlorides and the equiva- lent amount of barium chloride to barium sulphate. (The addition of barium chloride slightly retards the rate of hydrol- ysis of the alkyl barium sulphates, as will be shown in a sub- sequent paper on the hydrolysis of the alkyl sulphuric acids.) In the hydrolysis of the alkyl barium sulphates no attempt was made to purify the precipitated barium sulphate-beyond washing it with hot water until the wash water-was free from chlorine, since on ignition the small amount of included alkyl sulphuric acid was destroyed and the included trace of alkyl barium sulphate was changed to barium sulphate. In the present work, however, the amount of barium chloride included is too great to be neglected, and in order to remove this included barium chloride the precipitated barium sulphate resulting from the hydrolysis was filtered through an ignited and weighed perforated platinum crucible fitted with an asbestos mat, thoroughly washed with hot water, then gently ignited until the mass completely charred. The solid cake thus formed was broken up, care being taken not to disturb the asbestos mat, and again ignited until all the carbon was burned off and the barium sulphide, resulting from the reduc- ing action of the carbon, oxidized to barium sulphate.t The ignited and cooled precipitate was again washed with hot water slightly acidulated with hydrochloric acid to remove any basic chloride which might have been formed during the igni- tion. The crucible was then gently ignited, cooled and weighed. By this treatment practically all the barium chloride was removed. Theory.—The calcium or strontium sulphate formed in this reaction, being somewhat soluble in water, is instantaneously and quantitatively removed from the reacting system by com- bining with the barium chloride present to form barium sul- phate. It is the weight of the barium sulphate which is given in grams and gram equivalents in columns two and three in Table I. The course of the reaction is similar to that of the alkyl _ barium sulphates described in previous papers,} 1. e., M(C,H,SO,), + HOH = MSO,+HC,H,SO,+C,H,OH. M = Ba, Sr or Ca. This leads to the mathematical expression * Some barium sulphide frequently escapes oxidation, but never more than a small fraction of a per cent. +It is best to prevent caking by piercing the precipitate with a platinum wire just as it begins to char. It then falls to a powder and the carbon and barium sulphide are more easily oxidized. See this Journal, xxxiv, 290. t Loe. cit. dex at Linhart—Hydrolysis of Alkyl Metallic Sulphates. = K(A—1/2 2) (B+1/2), which on integration gives K = 2% 23 og (B+ 1/22)A Tee Bites pac — 1/oie)\B Experimental Results. TABLE I. = 60° Ethyl Calcium Sulphate in presence of Barium Chloride. BaSO, t = —~ —- in hours in grams in gram equiv. K B=1:0N HCl 20°0 0°1000 0°0428 0°00733 42°5 0°2066 0:0885 0:00733 12:3 0°3340 0°1431 0°00726 - 96°0 0°4288 01837 0:00726 120°0 0°5190 0°2224 0°00729 141:0 0°5920 0°2536 0:00729 165°0 0°6656 OR 0:00726 a 0°7002 0°3000 = A B=0°5 N HCl 24°3 0:0570 0°0244 0:00673 42°5 0:0981 00420 0:00667 72:0 0°1636 0:0701 0°00667 96°0 0°2117 0:0907 0°00652 140°5 0°3020 0°1294 0:00652 191:0 0°4090 0°1752 0:00661 241°0 0°5050 0°2164 0°00667 286°5 0°5846 0°2505 0:00667 o 0°7002 0°3000 = A Ethyl Strontium Sulphate in presence of Barium Chloride. BaSO, t << ——_-_- Sil in hours in grams in gram equiy. K B=10N HCl 22°5 0°1264 00541 (000828) 48°3 02482 0°1063 0:00786 67°3 0°3260 0°1397 0°00761 76'0 0°3700 071585 0:00772 96:0 0°4592 071967 0:00786 121°0 0°5584 0°2392 0:00789 140:0 0°6248 0°2677 0:00786 o 0°7002 0:3000 = A 541 542 Linhart—Hydrolysis of Alkyl Metailic Sulphates. Tasie I (continyed). BaSO, t - aa = in hours in grams in gram equiy. K B=0'9 N HCl 29°3 0°0554 0:0237 0°00718 48°0 0°1264 0°0542 (0°00765) 66°7 0°1644 0:0704 °0:00719 95°5 0°2378 0'1012 0:00742 139°5 0'3356 0°1438 0:00730 188°5 0°4390 O'1881 0.00725 235°3 0°5280 0°2262 0:00719 286'8 0°6214 0°2662 000719 oc 0°7002 03000 = A 4 TABLE II. K for Esters in 1:0 N HCl Ba(C2H;S0,4).* S1(CoH;SO4)o Ca(C2H;SO,)2 0:00828 0:00786 0:00738 0:00828 0°00761 0:00735 0:00828 0°00772 0°00726 0°00835 0°00786 0:00729 0°00828. 0°00789 0:00729 0°00832 0:00786 0°00726 Summary.—The rate of decomposition of these esters increases with the increase in the atomic weight of the metal, while it decreases in each case with the increase in the weight of the alkyl group, as is shown in the second paper of this series. 2. The hydrolytic decomposition in acid solution of the Ba, Sr and Ca salts of methyl, ethyl and propyl sulphuric acids proceeds very slowly at 60° according to the equations: M(RSO,), + HOH —+ MSO, + ROH + HRSO, HRSO, + HOH <> SO, + ROH M(RSO,), + H,SO, —> MSO, + 2HRSO,. 3. In order to obtain an integral equation from which to calculate velocity constants the differential equation must take into account the increase in the acidity of the reacting mixture as well as the hydrolysis of alkyl sulphuric acid resulting from the hydrolysis of the metallic alkyl sulphate. *This Journal, xxxiv, 292. A. Hrdlitku— Early Man in America. 543 a Art. XLVIII. — Harly Man in America* ; by Aus HrpxioKa, Curator of the Division of Physical Anthro- pology, United States National Museum. Avr the request of the Editor of this Journal the writer here presents, in a brief form, the essential data concerning skeletal and to some extent also other remains attributed to geological ancient man on the American continent, more particularly in South America, and the conclusions reached by him and his associates, after prolonged and unprejudiced research into the subject, as to the true age and anthropological significance of these specimens. Between the years 1899 and 1907 the writer carried out a series of inquiries on the various skeletal remains which sug- gested or were attributed to ancient man in North America. The studies resulted in a number of publications,t culminating in a memoir comprehending the whole subject, which appeared as Bulletin 33 of the Bureau of American Ethnology. The results of the investigations seemed at first to lend support to the theory of considerable antiquity for some of the remains presented as evidence, as, for example, the two low skulls dis- covered at Trenton, New Jersey. Subsequent researches, how- ever, cleared up most of the uncertain points, and the entire inquiry seemed to establish the fact that thus far no human bones have come to light in North America representing other than the Indian type of man, which we have many weighty reasons to regard as relatively modern in this part of the world. The details of the above work are sufficiently well known and easily referred to, and as no further specimens for which geological antiquity is claimed have been brought forth since the publication of the above mentioned report, the territory north of Panama need not be further dealt with in this place. More interesting and much more complex conditions than those in regard to early man in North America have arisen in * Published with the permission of the Secretary of the Smithsonian. +The Crania of Trenton, N. J., and their Bearing upon the Antiquity of Man in that Region. In Bulletin of the American Museum of Natural His- a xvi, Art. 3, 23-62, 3 charts, fig. 1-4, pl. i-xxii. New York, Feb. 6, The Lansing Skeleton in American Anthropologist, N. S., V, 323-3830, 1 fig., Lancaster, Pa., June, 1903. A report on the Trenton Femur (written in 1902), published with E. Volk’s The Archeology of the Delaware Valley, Memoirs of the Peabody Musuem, v, Cambridge, Mass., 1911, pp. 244-247. Skeletal Remains Suggesting or Attributed to Harly Man in North America. Bulletin 33 of the Bureau of American Ethnology, 1-118, pl. i-xxi, fig. 1-16, Washington, 1907. 544 A, Hrdlictka—Early Man in America, South America. In-the latter continent skeletal remains of man, which came to be regarded as of geological antiquity, have been accumulating since the forties ‘of the past century, and by the end of the “last decade had multiplied to such an extent and had become so important, as to call for the closest attention on the part of anthropologists. Moreover, frequent reports were made of finds of fossil animal bones charred, striated, perforated or broken by human agency in the far past ; of burnt earth and scorize showing human activities deep in the Pampean times; and we were even told of whole ecul- tures, represented by numerous archeological objects, belong- ing to the Quaternary or even the Tertiary. Finally, the scien- tifie world was startled by an announcement of remains of not merely a number of distinct ancient species of man, but also of several human precursors, supposed to be connected, some- where in the Eocene, with the little South American primates of that period. It will be useful to introduce a condensed chronologically arranged account (pp. 546-548) of at least the skeletal mate- rial upon which rests the contentions that geologically ancient man existed in South America. It was principally on the basis of the finds shown in this chart that Professor Florentino Ameghino, the Argentinian paleon- tologist and author who was in the largest degree respon- sible for these reports, has formulated a ‘far reaching theory regarding not only the presence of early man in South America, but also man’s descent and his migrations, which, if definitely established, would greatly enlarge and modify our scope of vision. On close inspection, however, the records of the finds of the supposedly ancient remains, the descriptions of the specimens themselves and the deductions drawn from the material, as a rule were found to be unsatisfactory. They were full of defects and uncertainties which, in view of the importance of the subject, were most perplexing and owing to the remoteness of the field and other difficulties, appeared as insurmountable obstacles to the formation of a definite opinion on the merits of the evidence; indeed the whole subject threatened to become a tangle which might never be unraveled. It was under these conditions that the Smithsonian Institu- tion in 1910 sent to South America, and particularly Argentina, an expedition consisting of the writer and of Mr. Bailey Willis, a geologist of much experience with formations such as were to be met with in the course of the investigation. The objects of this expedition were to gain as far as possible a clear view of the whole problem of early man in the southern conti- nent; to examine the original specimens relating to the sub- ject; ‘to study at least the principal localities and deposits from A. Hrdlitka—Early Man in America. 545 which ancient human remains had been reported and ascertain on the spot, if still feasible, the exact circumstances of the finds; and to discover if possible and collect additional osseous, archeologic, or other specimens bearing on man’s antiquity. After a brief stop in Brazil, Argentina was reached in May, 1910, and the stay of the writer in the country lasted two months, while that of Mr. Willis was somewhat longer. The Argentina men of science received us very cordially and facili- tated our work with a great liberality. Florentino Ameghino and his brother Carlos were particularly helpful. The speci- mens which it was important to examine, even those the descriptions of which had not yet been published, were placed freely at our disposal. Ameghino and his brother accompanied us, notwithstanding the inclement season, for nearly three weeks from point to point along the coast where vestiges of ancient man or his forerunners were believed to have been discovered, and others went with us to more distant parts of the country. We found skeietal remains in the same localities and under the same conditions as some of those previously reported as ancient, and collected others that threw much needed light on some of the important points involved. Several thousands of archeological specimens bearing on the supposedly ancient cultures were discovered in undisturbed positions and collected ; and many samples of fossil bones and shells, as well as of loess, various concretionary deposits, burnt earth and scoria, supposed to be of ancient human origin, with other articles, were gathered’and brought back with us for further investigation in our laboratories. After the return of the expedition to Washington the gathered data and specimens were subjected to considerable further studies and comparison, and a large amount of recent human skeletal material from South America was examined b - the writer. The archeology was taken up by Professor W. H.- Holmes of the U. 8. National Museum; the petrology by Messrs. F. E. Wright and C. N. Fenner of the Geophysical Laboratory, Carnegie Institution ; the shell material was turned over for identification to Dr. Wm. H. Dall of the U. 8. Geo- logical Survey, and chemical examinations of numerous bones were conducted under the direction of F. W. Clarke of the same institution. The results of the several researches were without excep- tion adverse to the theory of the existence of early man and his precursors on the southern continent. Anthropol- ogy, geology, archeology, the study of the burnt earths and scorise, that of the shells which should have established the great age of some of the strata, and the chemistry of the bones, all speak independently and forcibly against the assumed existence Am. Jour. oe Cpe SERIES, VOL. XXXIV, No. 204.—DxcremsBemr, 1912. 6 Merecd. 4 an L£ A. Hrdlitka—Early Man 546 ‘eqOS}IN UueuyeT} ‘oulySeury oUuUeLIOL Ty (*N-T) | topogdépy uLoy Leep “SLV fiunu.tazong uesvyuedog jo sovduavo wory yuomeld ‘geyloomry “3 (“y) auaoong | ‘tpuny oesnyy |peuanjdn tepup) ,,ueedueg,, yoo g UvYy Sse] -MyT Ue ‘UoJOTEAG Woy “S Ivomse[enzayu0q| TQST (quawup,, (q “ru ‘Sry ‘eUL -ohpyng fo awn} (3) eed eunp purs 4aaF 9 (q (aeto[dxe) -potA «ivau ‘Loy (2) «MDD ,, (0 | eTepoesmm | —--7* STANT: 4223 GL (& STIMAS OAT, ouoLoyf ‘d “A | “IRA O1SaN O1Y | GLET sory soueng “OLY ‘TeuoloeN oesnyy foljn3 (S[ISSOJ Fo 10j09]|‘OULMIVSIOg Ivett fiunusazond) \ULSyWVUTIeI Mey) = == ~--7 7" e Jo [je ‘sseoq| doap yaoj 6 Leoutey Eebs(| -109 "4°I) WIOY 'S) ,, Oloprywes,, W | OLST “039 ; oulleMRYy “9 “aopo}ddé{H _ proydeos v GTA late Bean me jo vovdvrey , uvedureg ,, i pues viqo}IeA W| oulqsemy “7 7 = PLOT ‘ojo ‘s}ueume,dut ; Ache plo) eds S1eqjOUuR ELST | “say “eqetq pue ‘snjzeyng WOAF SBUOG Moz 10 eT ep oasnyy, ‘sn1oydo[doy , ueodueg se i 4223 6 B YFIM ‘Wox[eYS) OUuLysemy “7 nO Ns BLOT xD | (S[ISS0g finusayong | 4SOT Suyq09][00 ‘URI ‘OOIATO | S[BULTUe youT} xe IeuUTseq U “4°T) ‘OL (VV) auaooug | Oasny 0} jUeg | Jo sauoq Auvyy | ,, uvedureg,, i WO}a[ IG oulysoury “qf ‘seit ep ofomty} TST i | uMOouUyUN eovid yoerxo sjueute;dut ! yUeq IOALt emo}s ‘semoq | (TVS TOF SpISSOF | seg “anjzen eaq pue aq} FO syisodep JO SJuSTISseAz 40 10399][00) ‘S1V fanusazon® | ea104S1H,P “SUI | estoy youryxm | .,weadueg,, eyetepoy] —_—s/sesauzeyd‘qyeay, wmMseS “7 ‘RURIvOIeD OLY | FOST SUILOF Surat Auvur osye S}isodep [Isue}N euO]S T uesvyuedop ‘S[VULUIVUL STIS |10}eA PU SLIqep Yt ‘seuoq puv) (4sTjernjzeu) [izerg “eyUBS | FPRT fiupusayong | ‘tpuny oesnq | -Sofsnotemnyy jeAvo JO oANAXI) [eMeyeuMUy |s[[Nyssnorewny) puny “M ‘dq | BoseyT ‘seavrp | cesT | NIGHAUNGSaUg | NOlLVIOOSSY | Hida GLY aoag t Lisodaqy SNIVNGY NVWOF UTUAA OOS, ALIIVOO'T ‘VYOIUGNVY HLOOS NI NVW ATUVA OL GHLOAGIYLLY SNIVNGY TVLYTHNS 547 A. Hrdlicka—Early Man in America. spreysjod ‘diy ‘oraleaCd ,, weedmeg TIA Setlog euros opuoy out | 4 S06 fiwnusaj0ne FD A 4 OOBURLIY) Ioredneg ,, 4a0y © ynoqy pur s[[nys % SeLIVO op ‘W | pus aang oz | 906T STRUTT eS qoulyxe , ueedueg ,, (;, omoygoad Oulysouly auaoory ‘ i SuOl snoreun yy Areiy.10 J, (3) -B14oJ, ,,) INE sopre9 OSOTIIOFT 9}UOW| -06T (eouRjsIp 4e ; 4 7 ({eAey ‘Va reqsiy Aysys | ,, BOSOT, ,, (,,oumoyyordi , ,) JSIMOT) auav0yg |‘[euoloeN oesnyy| UopoyddéyH) — | aepun ‘pueg |() yo03 OP ynoqy deo-][nyx¢ S10.10qe] Selly soueng | 968T (eq UI Op oesnyy jo seA4o;dume) “Is y ‘1100049 (i) fitpusaqone ; 3 (i) « Uvedueg ,, IeAVyT OeJING W0}9TOS equInitey “7 ‘ohoray S881 (eqeT (N-T) ®] ep oosnyy fiupusaqzong eyed (i) “049 lTeug (,Sneedueg *,,)| Jo eeXoydure) (‘y) auaoong €] ep oesnyy |‘UINIIeyj0pijeog | ,, uredueg ,, (3) 149 esoury VW ‘Sry ‘reurediyy | 88st SoIly souaeng ‘svijaT (OVN oesnTy & eygosoity o7VIopoyy, 10jye1ede1d) ‘B1YV () fipusozond ep peqlnorT i) « Ubedmieg ,, () ets JO] MSO “ |‘Sesloorty op o1y| SgeT qoun7, a ‘onbrutyoe4 « Ueodueg hanya, -AJod 900g ) Arerpeurtezuy ,, (1) 9993 g = __ EOFETOAS WON S ‘Sry ‘orepereg | L88T (,,snevs : -00U OUI0F ,, eyelq CN-T) f.00290,, eed (, oumloyjoad eT op oosnyy ‘S1V (YW) auasoipy ®] ep oasnyy (2) ,ernjeu { sse0ry (3) “BIZ9T, 5») SeIFW | Jo vakojdm@ |‘osomsey oyuoyy) Bgl 9.198 (oeNT Oesny, (N-1) uledg ey} jo yueq 4ST[eIngeu ‘SLY finusaqong “erouaye A (4) Ul ,, uvadureg SulpeAety) ‘Tog uLOTOg URS ((Y) auav0ng ye Unesny{ | UNIeyJOptjeog| oulgsnoerT ,, | (2) a0F F ogy moyalayS SOL’ ep “| ep of01Ly S881 A. Hrdlicka—EKarly Man in America. 548 (pooytoqysieu aq} UI yisodep (,.804 VE sci aules UL punoyz SSOO] -eulfoIjndevo owl oulyseury “S1y aua00u"d ‘[eMoleN Oesny_| e200 uopozdAT4)) , Ueedueg ,, (4) 3207 $e -OFf,,) W0zaTeXS | SOTIVO ‘oSsvig oforry | OI6T 5 uojejeys - | (couep.zes) ee SN SNES Sys SESS 3 5 rake] VovzJING ® FO syleg evIpoleg “T Bayoooe NT OI6L 3 gouRysip aTOS 42 “040 (,,0} mou auonabacreptoyeys rake -9UIS OULOF{,,) BIBER IL oa 5 » _|7[29g “uopozd ATH 40 54 aoBJINS UL VSSO,T| SUOJITPONS | | IOpIVs W eunsey OI6I Teqjour (N-T) jo SyUeUISRIT ~ fiunusayon?) a Tp |e hoes aT 45 or ReU 2) | adeFING WIA ‘TnyAS a ‘5 7 3 ( uvedurg ,, seuog JO sjueUL (1eTe pres) OI6L (7y) auao0nd 5 Ap peas pepnueg ToAR] GOVFANG | -Svay pwe [[NAS rporeg “7 “say ‘eeyoo0eN | LOGI « Ueodurg ; Seuoq Fo syred “ oe 3 Re Esa | aomedng,, eovyIns wean |{uUezUL FO [[NyS)} ,, 5 09409 Aqieou TINLIeyyesoy seuoq pue ” ” » Jo Inula UTUNLAUTT VY 499F 9 FHOGV |TIMAS JO worytod) ,, ” epeury eT | souo0q JO syUeuIsvaz | ‘19qjOUR JO forays | ea » () qe2F 0G WAG Jxed B pue TINA,» ” O1LET890S | ‘Vv ‘a ,uvedmeg joey ¢ uvyy ssoT somo0q jo fiumusayong \‘{euoroeN oesnyq| 7 _ donedng ,, (3) syed pur [[NyS| septwO ep q | orefeaQ eaoqy ao1aag | NI aaauasaug | NOLVIOOSSY Lisodaq Hida SNIVNGY NVWO| WLAAAOOSIG ALITVOOT aLV( ‘panuyuog—VOINANV HINOS NI NVW ATYVEA OL GHLOGIYLLV SNIVNGY TVLATHS A. Hrdlicka—Early Man in America. 549 of ancient human or of any prehuman forms in South America. The evidence obtained attests nothing more than the presence in the south, as similar evidence has formerly shown in the north, of the already differentiated and relatively modern American Indian. The bibliography of the subject, the historic data, the details which led to the above conclusions, will be found in a volume recently published by the Bureau of American Kthnology* and need not be here repeated. The only question which requires to be dealt with in this place is that of the causes which have led to the remarkable conclusions concerning the antiquity of man in South America reached by Ameghino and other South American men of science who have occupied themselves with the problem. How has it come about that a number of inves- tigators, including Ameghino, the foremost exponent of South American paleontology, have arrived at, maintained and even strenuously defended conclusions, which. after a serious and allsided research into the subject, cannot be accepted and must in fact be entirely subverted by other students. The causes may never be fully analyzed but comprise, in the main, defective collection, imperfect criteria of comparison, a lack of experience in anthropology, and finally, in at least “some eases, the allure of the new and wonderful. As to defective collection, it may be said that with the sole exception of the Lagoa Santa material, not one of the speci- mens advanced as representing early man in South America _ was gathered in a way to satisfy the requirements of science. Let us turn to the records: The “ Rio Carcaraia” bones were brought to Buenos Aires by F. Seguin, a collector and dealer in fossils. No written report was ever made regarding the circumstances of the find by Seguin, his oral information was very deficient in details, and the stratum from which the bones came, their association, and even the place where the discovery was made are uncertain. The first ‘“ Arroyo de Frias” find was made about 1871 by F. Ameghino, at that time less than 18 years old, acting as a a subpreceptor ” at a nearby school and beginning to interest himself, while searching to regain lost health, in fossil bones. The “ Saladero” skeleton was found in 1876 by Santiago Roth, at that time a young collector of fossils, and was not even men- _tioned in literature until twelve years later. At the time of its discovery the bones were thought nothing of and were given to * Harly man in South America, by Ale’ Hrdlitka, with the collaboration of W. H. Holmes, Bailey Willis, Fred. Eugene Wright and Clarence N. Fenner. Bull. 52, Bureau American Ethnology, Smithsonian Institution, Washington, D. C., 1912, 8vo, pp. i-xy, 1-405, with 68 plates and 51 figures, 7 See ‘‘ Dr. Florentino Ameghino,” por Juan B. Ambrosetti, Anales de Museo Nacional, etc., Buenos Aires, xxii, page xii. 550 A. Hrdlicka—Karly Man in America. a companion. About one year later Roth happened to see in the garden of his companion some fragments of “fossil” bones and on asking where they came from he was informed that they were the remnants of the skeleton dug out near Saladero. And these fragments constitute the sole evidence of the Sala- dero representative of the ancient man of Argentina. The “Arrecifes”’ skull was found by a preparator attached to the Museo Nacional of Buenos Aires “in terrane belonging to the Pampean formation which was left exposed by water.” This is absolutely all that has ever been recorded in regard to the circumstances of this discovery. The “ Samborombén”’ skeleton was found in 1882 by a traveling naturalist of the Museum at Buenos Aires, and the first meager details regard- ing the specimen were not given until 1889. The skeleton itself is lost without ever having been studied, but neverthe- less poses as a representative of “fossil” man in Argentina, and has even given rise to the coining, by Kobelt in Germany, of a new human species, the ‘ Homo pliocenicus.” The ‘*Chocori” skeleton was found about the year 1888 by an employee of the La Plata Museum. The first notice of it was not published until nineteen years later. The remains lay “abandoned on the surface of the ground, partly covered by indurated sand.” The “Ovejero” bones were collected at different times by one of the traveling naturalists of the Museum at Buenos Aires. They were found at different levels in wind-blown loess in the proximity of a fair-sized river, and in association with the bones of several recent forms of animals. The “ Baradero” skeleton lay with most of the bones in their natural relations in eolian loess, about 3 feet deep below the surface. The “Arroyo del Moro” skeleton, which gave ' rise to the new species of Homo sinemento, was discovered by a sailor and his wife, and later excavated at the initiative of a local physician by the sailor, his boy, and a gardener. The “ Ta Tigra” skull, which resulted in the establishment of Homo pampaeus, attributed to the Tertiary, was found in 1888 by an unscientific employee of the Museo La Plata near the Arroyo La Tigra. ‘The employee was charged with collecting fossils for the Museum. He went to a point at which fossil animal bones were previously discovered, excavated in the neighborhood, and among other things discovered a human skull. This is all we know of the circumstances of the discovery of this specimen, which has been given such importance. Another small lot of bones relating to the Lomo pampaeus were discovered on and near the surface of some partly denuded ground near Necochea by the above mentioned gardener. The “ Diprothomo” (or nearest but one precursor of man) fragment was found by com- mon laborers and for thirteen years lay unnoticed in the A. Hrdlicka—Early Man in America. 551 Buenos Aires collections. As to the “ Tetraprothomo” (the fourth precursor of man, counting backward from the latter), finally, the atlas was brought by an employee of the Museo de la Plata from a fossil collecting trip to Monte Hermoso, and no details are known of this discover , even the year of the find being uncertain ; while the femur was found, with other fossil bones, some time during the early years of the present century, the exact year being also uncertain, by Carlos Ameghino. The atlas, which is human, after being brought to the Museum was forgotten and lay for many years unnoticed. The first pub lished notice of it appeared about twenty years after its discov- ery. The femur belongs to an ancient small-sized carnivore. The above notes could be extended; however, the subject may be briefly summarized by the statement that not one of the osseous specimens which represented the “ ancient” man in South America and particularly in Argentina has been dis- covered or exhumed by an experienced anthropologist or arche- ologist, or by a person well trained in or employing the methods recognized to-day as requisite in dealing with objects of such importance. And this applies equally to the other objects than human bones which represent the “ early” man in Argentina. A more meager and defective record could searcely be imagined. _ Following unscientific collection of the specimens came defec- tive judgment in adjusting their age and, in the case of numer- ous specimens other than human bones, of recognition of their true character. Such defects of judgment were, as has appeared from the observation of Mr. Willis, an imperfect and in some instances decidedly faulty identification of the deposits in which the human remains were discovered; a general but wholly unwarranted conclusion that the human bones were contemporaneous with the deposit in which they lay and with the bones of various animals found at the same levels; the noxious opinion that the mineralization of a human bone meant generally and of necessity a great antiquity of the speci- men; the assumption that certain refuse and by-products of the manufacture of stone implements were sufficient to estab- lish ancient and otherwise unknown primitive cultures; the failure to recognize or admit the accidental nature of numerous markings on the bones of ancient animals; and the attributing of anthropic significance to baked earth and scoriae which are in all probability secondary volcanic products having noth- ing to do with man’s existence. A lack of experience in anthropology, with a dearth of material for comparison, resulted in such sad occurrences as the giving of wholly faulty positions to more or less incom- plete human crania and ascribing the apparent differences 552 A. Hrdlictka—Eurly Man in America. which they presented in consequence, to morphological inferi- ority ; in giving undue weight to various actual features which, with a more extensive view, would have been seen to be well within the limits of variation of the same parts in man and particularly in the Indian of the present day ; and above all, in the failure in numerous cases to recognize an artificial defor- mation of the skull, with the mistake of taking the results of such deformation, particularly the lowered forehead, for marks of anthropological inferiority of the specimen, and even regard- ing them as species characters. ‘In conclusion it cannot be denied that in at least some cases the part of a proper critical analysis has been taken by the more enticing and less dithicult task of enthusiastic theorization, one of the most striking examples of this being Ameghino’ S hypotheses about the various “ Precursors.” All the above points are dealt with in detail in the main report of the writer and his collaborators on this subject and therefore a further discussion of them in this place would be quite superfluous ; ; but it may not be amiss to give here just a few concrete instances illustrating the conditions. The Argentinian writers did not hitherto clearly distinguish the holocene and the Quaternary in the pampas deposits ; most often everything beneath the vegetal layer was regarded asa part of the Pampean formation and as of definite geological age ; yet the upper and sometimes large portions of the deposit are evidently of a very recent origin, the paleontological remains which they hold being of secondary inclusion. Many uncertainties exist also in the recognition of the Tertiary as dis- tinguished from the Quaternary Pampean deposits. _ One of the most important strata in relation to ancient man, the so-called ‘‘ Interensenadean” of Ameghino, could not be traced at all by the geologist of the Smithsonian Expedition, and what was pointed to by Ameghino himself as representing this layer, proved to be a modern sea shore agglomerate of shell detritus and sand, containing remnants of molluses of liv- ing species only. And at Ovejero, what was represented as a Quaternary Superior Pampean bed was found to be nothing but a wind-blown deposit of no great age. In a number of instances, particularly at ‘‘ Necochea” and Arroyo del Moro, the human remains ‘recovered represented clearly burials and, hence, introductions into the earth; yet they were described as contemporaneous with the deposits which they barely penetrated. Mineralization of the human bones was taken invariably as proof of the great age of the specimen, notwithstanding the well-established fact that it depends far more upon the environ- ment than upon time. Actually not a single perceptibly min- eralized human bone was seen by the Expedition i in Argentina >. A. Hrdlicka—Karly Man in America. 553 which was not regarded by one or more of the local scientists as geologically ancient. Yet along the coast and in other places, on or near the surface of the ground, lay many bones of domestic and other animals of living species showing plainly various phases of “ fossilization.” Really no bones from bur- ials or inclusions in the Pampean deposits were met with that were not more or less mineralized, which is easily explainable:by the high percentage of calcareous and other mineral contents favoring “fossilization” that are held by the formation. The results of lack of experience in anthropological matters manifested itself especially and painfully in the case of the “Diprothomo.” This form is represented by a frontal bone with a portion of the parietals. The fragment was described, not in the position which it would occupy in a normally poised skull, but in that which it assumed when laid on the table. This error was responsible for the creation of a genus of human ancestors. The descriptions of various specimens extended to and made much use of minute details, which are known to be of little or no biological significance. In the cases of the Diprothomo and Tetraprothomo, as published by Ameghino, there are page upon page of minutiz through which even a trained anatomist wades with difficulty and which only obscure the true char- acter of the specimens. The Monte Hermoso atlas, which in every respect is well within the range of variation of the same bone in the prehistoric and pr obably even in the historic American Indian, was at the same time and largely by such minutiz being made a part of the Tetraprothomo by Ameghino, and given as a representative of a Tertiary species of Ameri- can man by Lehmann—Nitsche. As to the part that theory played in this connection, it is sufficient to point to the system of human descent and migra- tion constructed on the basis of the various reports assumed to indicate the presence of human man in South America by Ameghino. He not only derived the whole human family from certain little primitive forms in South America and peopled that continent with hitherto unsuspected species of man and genera ot precursors, but he also considered that all these species and genera became extinct before South America was peopled by the Indians. The latter, he assumed, came from the emigrants who originally left the southern continent and spread over Africa, Asia, and North America, finally reaching again the southern part of the continent. The above examples could be much enlarged upon, but this is hardly necessary in view of the facility with which the detailed report on “ Early Man in South America” can be con- sulted. Here it is only. just to the other South American 554 A. Hrdlitka—Early Man in America. authors who are involved in this subject to say, that the majority of the failures referred to were those of the prin- cipal exponent of ancient man in Argentina, Florentino Ameghino. And lest the above examples may seem partial and unjust the interested reader is strongly urged to peruse the detailed accounts of these matters. The conclusions which the members of the Smithsonian Expedition inevitably reached and hold in regard to early man in South America were as follows: An unbiased study of all the available facts has shown that the whole structure erected in support of the theory of geolog- ically ancient man on that continent rests on very imper- fect and incorrectly interpreted data and in many instances on false premises, and as a consequence of these -weak- nesses must completely collapse when subjected to search- ing criticism. It fails to establish the claim that in South America there have been brought forth thus far tangible traces of either geologically ancient man himself or of any “precursors of the human race.* The position is maintained, and should be maintained, it seems, by all students, that the final acceptance of the evidence on this subject cannot be justified until there shali have accumulated a mass of strictly scientific observations adequate in kind and volume to establish a proposition of so great importance. *In the opinion of the writer, based on the published data as well as a personal examination of the specimens, the recent ‘‘ ancient” or ‘‘ prehis- toric” man found in Peru and reported on by Professor Bingham, offers noth- ing which would necessitate a recasting of these conclusions. Washington— Constitution of Some Salic Silicates. 555 Art. XLIX.—The Constitution of Some Salic Silicates ; by Hunry 8S. Wasuineron. Introduction.—In this paper is suggested an explanation of the chemical constitution of the feldspars, lenads, scapolites, and zeolites, which accounts for the intimate relationships that they show. In brief, the hypothesis is that they are alumino- silicates, possess a chemical feature in common, the presence of certain atomic groups, and that the formulas of the feldspars, lenads,* and zeolites, are of identical type, while that of the seapolites is different but closely analogous. The difficulties presented by the silicate minerals in the study of their molecular constitution are so great, especially as compared with organic compounds, that we are still only on the threshold of knowledge adequate for discussion, and any suggestions put forth at the present time must be regarded as tentative and speculative. Still, certain facts are known, and these, with recent applications of physical chemistry to miner- alogy, give justification for dealing with the subject. The literature of the mineral groups mentioned is so exten- sive that it cannot be gone into exhaustively here, and only a few of the more recent papers will be mentioned. Lelationships.—The minerals of the four groups under con- sideration are alike chemically in being silicates of aluminum and of potassium, sodium, or calcium, any two or all three of the last being often present together. Some feldspars contain barium in place of calcium, and this element and strontium are present in a few. zeolites, while lithium is essential in eucryptite. All the minerals containing these elements, however, are of very rare occurrence. The function of the invariably present aluminum is unknown, though it is generally assumed to be present as a base. The zeolites and analcite are hydrated, containing either water of crystallization, acid hydrogen, or constitutional hydroxyl, in either case very loosely combined. None of these minerals ever contains magnesium, ferrous or ferric iron, manganese, titanium, boron, or fluorine, as essential constituents, in this differing radically from the large and important groups of the pyroxenes, amphiboles, micas, garnets, vesuvianites, epidotes, tourmalines, and olivines. The seapo- lites alone may contain small amounts of Cl and SO,. These are also present in the minerals of the sodalite group, which show some affinities with the lenads, but which are now usually *The name lenad has recently (C, I. P. W., Quant. Class. Ign. Rocks, Chicago, 1903, p. 182) been proposed to replace the term feldspathoid, including leucite, nephelite and their congeners, and this will be used throughout this paper. 556 Washington—Constitution of Some Salic Silicates. regarded as allied constitutionally to the garnets. The heavy metals are never present in nature, though artificial feldspars and lenads have been prepared which contain lead, zine, silver, or thallium replacing calcium and the alkali metals, and ferric iron replacing aluminum. In the feldspars, lenads, and zeolites the ratio of K,O+ . Na,O+Ca0O to Al,O, is constant, 1:1, while in the scapolites the ratio does not differ much from this. The relation of silica to these constituents varies from orthosilicate to trisili- cate ratios, and the various minerals are commonly regarded as either orthosilicates, metasilicates, trisilicates, or mixed inter- mediate compounds. That is, in spite of their chemical resemblances and other evidences of close relationship, pres- ently to be noted, they are usually regarded as salts of at least three distinet silicic acids. The intimate chemical relations between these minerals and mineral groups are also evident from their alterations in nature and from experiments which have been made in their artiticial reproduction. The feldspars are quite readily alterable, chang- ing into scapolites and zeolites, as well as to muscovite, zoisite, and so forth. Feldspars, nephelite, analcite, and other zeolites are common natural alteration products of leucite, and leucite and analcite are readily transformable artificially, the reaction being reversible. Pseudo-leucite will be mentioned later. Nephelite readily alters, especially to analcite and zeolites, and analcite has been formed artificially from some feldspars, as andesine. The scapolites are often derived from the sodicalcic feldspars and are easily alterable themselves, forming a varied number of products, among which the zeo- lites do not often occur. The zeolites are always secondary, with the exception of analcite in some igneous rocks, and are generally derived from feldspars and lenads, of which most of them are commonly regarded as hydrated torms. Many of them yield feldspars on fusion. Kaolin is the most usual end product of the alteration of the feldspars, lenads, scapo- lites, and also the zeolites. The crystallographic similarities and relations between the minerals of these groups, as the feldspars and leucite, and the feldspars, lenads, and zeolites, are commented on by several authorities, as Dana and Hintze, and only need mention here. It also seems scarcely necessary to dwell on the intimate relationship shown by the feldspars and lenads in igneous rocks, as their petrological importance and mutual relations are too well known to need more than mention. Lehavior with Acids.—Ot apparently minor importance, but with bearing on their constitution, as will be seen later, is the action of hydrochloric and other acids on the various Washington— Constitution of Some Satie Silicates. 557 minerals under discussion. A brief recapitulation will suftice here. Some of the feldspars and scapolites, as orthoclase, albite, microcline, and marialite, are entirely unattacked. Leu- cite is soluble, with production of pulverulent silica, while nephelite, anorthite, and many zeolites are easily soluble, giving gelatinous silica. In general, resistance to the attack of acid increases with the silica content, but the bases seem to _have little influence, leucite, nephelite, and anorthite being about equally soluble. Polymorphism.—aA very marked characteristic common to the feldspars, lenads, and zeolites is the frequent dimorphism or polymorphism of their molecules, though certain cases (the monoclinic feldspars) may rather be attributed to polysym- metry. Examples are: orthoclase and microcline; albite and barbierite,* the recently discovered monoclinic soda-orthoclase ; nephelite and the triclinic carnégieite ; anorthite, celsian (the monoclinic barium salt), and the presumably hexagonal form present in calcic nephelite; kaliophilite (phacolite) and the artificial triclinic, isometric and tetragonal (?) forms of the same molecule ;+ dimorphous leucite, which is enantiomorph- ously transformable at about 500°; heulandite and epistilbite, and possibly lanmontite and levynite ; and dimorphous analcite. The existence of pseudo-leucite, an intimate mixture of orthoclase and nephelite with the crystal form of leucite, is of interest and importance in this connection. There are good erounds for rejecting the usual view that it is a pseudomorph in the ordinary sense, by which soda has partially replaced the potash of an original leucite, and for considering it as repre- senting an original definite and distinct, sodi-potassic, leucite- like mineral, which is only stable at high temperatures and which breaks down on cooling into a microscopic mixture of (K,Na)AISi,O, and Na,Al,Si,O,, crystallizing as a sodic ortho- clase and nephelite. This view has been previously suggested. The composition of pseudo-leucite from various localities varies somewhat, but the amount of nephelite is always subordinate to that of sodic orthoclase. The most recently described§ cal- culates out about 52°9 orthoclase, 27°8 albite, and 19°3 nephe- lite, neglecting the small amounts of biotite and scapolite which are present. This is approximately (% or $ ab),ne,, which cor- responds to the formula KNaAl],Si,O,,, or a molecule each of potassium and sodium leucite with one extra of silica. *Barbier and Prost, Bull. Soc. Chim.. vol. iii, 1908, p. 894. Gonnard, Bull. Soe. Geol. Fr., vol. xxxi, 1908, p. 303. W. T. Schaller, this Journal, vol. xxx, 1910, p. 358. + Z. Weyberg, Centralblatt Min., 1908, p. 395. $C. W. Knight, this Journal, vol. xxi, 1906, p. 71. H.S. Washington, Jour. Geol., vol. xv, 1907, p. 387. §C. W. Knight, op. cit., p. 292. 558 Washington—Constitution of Some Salic Silicates. A closely analogous substance is “ sigterite,” originally de- scribed by Rammelsberg* as a new plagioclase, but which was later shown by Tenne} to be an intimate mixture of albite and nephelite, though with well-detined cleavage and polysyn- thetic twinning lamellae. Rammelsberg’s an aly ysis yields the composition 26 orthoclase, 13 albite, 60 vepltehiel or (% or $ab),ne,. He assigns to the supposed feldspar the formula (Na, K), ‘Al 1,0..; and it is noteworthy that this is closely like the formula which represents the composition of the Linosa © anemousite,t (Na,,Ca)AI,Si,O,,, which has been shown to be a solid solution or mixed crystal of labradorite (Ab,An,) and car- negieite, the triclinic form of Na,A1,Si,O,. The fact that the stable mixed er ystals of anemousite contain only 5°6 per cent of the carnegieite molecule, while sigterite, with about 60 per cent of the hexagonal form of the same molecule, evidently represents a mixed crystal which is unstable at ordinary temperatures, is in harmony with the suggestion made in the paper cited as to the limited miscibility of the iso- dimorphous Na,A1,Si,O, and CaAl,Si,O,; and serves to confirm our views as to the comparative instability of the triclinic form of the former. Taking thesé facts and the observed cleavage and twinning into consideration, there is little doubt that sigter- ite repr esents an original mineral, with triclinic plagioclase- -like form, to which the name sigterite would be applicable, but which is unstable at ordinary “temper atures. A somewhat similar case is the alteration product of spodu- mene, the so-called @-spodumene, described by Brush and Dana.§ This is an intimate mixture of 67:6 albite and 32-1 eucryptite, and may be expressed by the formula NaLiA1,8i,0.,. This does not seem to be a case of metastability, as with the preceding, but rather due to replacement through weathering. It will be discussed in a later paper on the constitution of the pyroxenes. LIsomorphism.—Isomorphism is the relation between two or more erystalline compounds of identical type of chemical formula and essential agreement in crystal form; the con- stituent atoms being replaceable by elements or atomic groups of the same valency and similar chemical character and fune- * Rammelsberg, Neues Jahrb., vol. i, 1890, p. 71. + Tenne, Neues Jahrb., vol. ii, 1897, p. 206. ¢ This Journal, vol. xxiv, 1910, p.57. Professor F. Zirkel (letter dated. 10, III, 1910) kindly called my attention to the fact that the composition of some of the feldspars of the Monte Amiata toscanites, described and analyzed by J. F. Williams (Neues Jahrb. Beil. Bd., vol. v, 1887, p. 419-421), is closely like that of the Linosa feldspar. The analyses are not satisfactory, having been made on small amounts of material and the silica determined by differ- ence, but the general resemblance is striking, and they all calculate out with notable amounts, up to 13 per cent, of the Na, Al,Si.0; molecule, presumably present as carnegieite. § Cf. Dana, Syst. Min., pp. 368 and 426, 1892. Washington— Constitution of Some Salie Silicates. 559 tion; this functional substitution (isomorphous replacement) involving only slight (morphotropic) change in the erystal form, and the isomorphous compounds (mutually either wholly or partially miscible) forming homogeneous mixtures whose properties (chemical and physical) are continuous functions of their composition. The replacement* may be complete, as is illustrated by barite and celestite, or apatite and mimetite, and an element can be re- placed by an atomic group (radical) of like valency, as K or Na is replaced by NH,, or F by OH. Isomorphously replaceable ele- ments or radicals may partially replace each other, their total molecular amounts being equal to that needed to satisfy the for- mula. This is of very rare occurrence or quite unknown among organic compounds, but is so frequent among minerals that examples need not be given. Finally there is a third type of isomorphous replacement, namely, that between atomic groups of the same total valency. Thus a bivalent atom or radical replaces or is replaced by two univalent ones, two bivalent atoms by a univalent and a trivalent one, or two trivalent atoms by three bivalent ones, and vice versa. This is illus- trated by the mutual isomorphous replacement of Ca and Na, in many silicates, of Pb and Cu, in sulpho-salts, of CaMg and NaFe’” in the pyroxenes, and so on. These last mentioned atomic groups differ from radicals in the ordinary sense. In radicals the atoms are linked together, and the valency of the group is that of the free bond or bonds, and may therefore be spoken of as restdual. In the Jast case the atoms are not linked together between themselves, but are in stoichiometric ratios and together replace a number of other atoms the sum of whose valencies is equal to theirs. While apparently not joined together enter se, they act together, and the valency of such a group may be spoken of as ¢otal. Our knowledge of this sort of replacement is comparatively recent and is still not well understood, but so many examples are known that it is now generally accepted.t As this type of atomic group is of importance in mineralogy and the study of isomorphism, it will be well to distinguish it from the ordinary radical by a special name. For this the word “ by Fo dh. Witson, s 22-725 seat oe eee ee XLV. — Thermodynamics of the LEarth’s Nonadiabatic Atmospheres by F.y Bichiow<., 2292 cn eo eee eee ee 51S XLVI.—Note on Atmospheric Radiation ; by F. W. Very 538 XLVII.—Uydrolysis of Alkyl Metallic Sulphates ; by G. A. TANDART™ So 0s ons foes Rees ao Saee es eee 539 XLVIII.—Early Man in America ; by A. HepiiéKa.._.-.- 548 XLIX.—Constitution of Some Salic Silicates; by H. S. W ABHINGTON™ oie: 6. a USS) a el ae ee a SCIENTIFIC INTELLIGENCE. Chemistry and Physics—Preparation of Perchlorie Acid, H..H. WiLuarp; Centenary Celebration of the First Commercial Gas Company: The Analyst’s Laboratory Companion, A. E. Jonwson, 572.—A College Text- Book on Quantitative Analysis, H. R. Moopy: Per-acids and their Salts, T. S. Price: Lehrbuch der Chemischen Technologie und Metallurgie, B. Neumann, 573.—Riumliche Intensitiétsverteilung der X-Strahlen, die von einer Platinantikathode ausgehen, W. FrimepRicu, 574.—Experimen- telle Bestimmung des Verhiiltnisses der spezifischen Warmen c,/c, bei Kalium und Natrinmdampfen und daraus sich ergebende Schiussfolge- . rungen, M. Rosirzca, 575.—Untersuchungen tiber magnetische Zerlegung feiner Spektrallinien im Vakuumlichtbogen, Cu, Waxi-Monammap: Absorption Spectrum of Tellurium Vapor and the Effect of High Tem- perature upon it, H. J. Evans, 576. Geology—New York Potsdam-Hoyt Fauna and Group Terms for the Lower and Upper Cambrian Series of Formations, C. D. Waxcott, 578.— Volcanic vortex rings and the direct conversion of lava into ash, F. A. Perret, 579, Miscellaneous Scientific Intelligence—National Academy of Sciences, 579. Obituary—J. W. Marunr, 579. InDEX TO VOLUME XXXIV, 580. ere a SMITHSONIAN INSTITUTION LIBRARIES “TONIC 3 9088 01298 5859