MINERALOGICAL CHEMISTRY. Miner a1o gi c a 1 Chemistry. On the Constitution of Natural Silicates. By K. HAU sH o FE R (Ann. Cheni. Pharm. clxix 131-146). AN attempt to exhibit the constitution of natural silicates by means of graphic formula founded principally on known facts of pseudomorphy. Starting with enstatite SiR"03,corresponding with the neutral carbo- nates which is formulated O-Si=gIR" and regarded as a normal silicate the addition of R"O yields olivin R"I~ISiY~ZRt ;and o-O-R"' 2R"O furnishes R"Io-Si {o-R/f/o which is believed to form part of chondrodite ; whilst the introduction of (SiO,)" gives OZSi-0 1. /R",\ the general formula of petalite. OIS1-0 These formulm represent saturated compounds but they may ob- viously be so arranged as to present free atomicities.Thus SiMg,O~ (olivin) may be qISi=o-Mg-o-* O-MF?-O-X. These elementary groups may therefore combine with basGs witlh Si02 or with one another in various proportions chondrodite for instance consisting of SiRirOa + 2SiRir0,. The existence of silicates SizRfr06 points to the union of several atoms of silicon by means of oxygen. Thus steatite may be formu- lated-X -Si-O-Si-0-Si-O-Si-O-X II 11 I1 !I 00 00 00 00 ii II ii 11 Mg Mg Mg HH the two atomicities marked *being either free to combine with other radicals or else saturating one another. The primary nucleus of the garnet group is Si3(R2)ViR3r'012 repre-senting grossular and allochroite. The removal of 3SiCa03 and 2Ca0 from grossular and the introduction of HzO furnishes epidote which in the anhydrous form consists of two groups of the form of anorthite and one of euclase the latter consisting of the anorthite group, Si2Rr1'2R,"OB +R"0.Abstraction of CaO from anhydrous epidote yields smpolite. Each of these two species is known to form pseudo- morphs after the other. On the assumption that the group -Ca- Ca< -Ca -is isomorphous with -Al= idocrase may be represented as composed 1 -Al= ABSTRACTS OF CHEMICAL PAPERS. of four groups of the primary nucleus Si,AI,Ca,O,,,with one similar group in which 3Ca replaces 2A1 and the pseudomorphy of garnet and scapolite after idocrase becomes intelligible. The conversion of almandine into chlorite consists in the removal of Si2(Fez)”0~ and in- troduction of 4MgHz02 and 2H20.The further action of magnesia and water yields serpentine. Almost all these formulae are given in the graphic form in the paper. M. J. S Researches on Emeralds and Beryls-Part I. On the Colour- ing-matter of the Emerald. By C. GREVILLE WILLIAMS (Proc. Roy. SOC.,xxi 409 ; Phil. Mag. [4],xlvi 314-328). THEcolouring-matter of the emerald has been attributed to iron to chromium and to organic matter. With regard to the last the author shows that both emeralds and beryls contain carbon but he thinks that it is probably present in the form of diamond and has nothing to do with the colour of the emerald for “colourless beryls may contain as much carbon as the richest-tinted emerald.” The colour is really due to the presence of chromic oxide.Accounts are given of the author’s experiments on the effects of fusion upon (i) opaque beryls (ii) emeralds (iii) an artificial mixture of beryl ingredients. Whatever may have been the temperature at which beryls and emeralds were formed the author has convinced himself that rubies were formed at a very high temperature since the peculiar reaction between a;lumina and chromic oxide to which the red colour of the ruby is due takes place only at a heat as high as that of the oxy- hydrogen flame. G. T. A. American Minerals. By A. R. LEEDS(Amer. Jour. Sci. [3] vi 22-26>. I. Hydrated Ufiisilicate approaching Pyrosc1erite.-This mineral occurs in the Bare Hills Maryland between a wall of deweylite on one side and talc on the other.The deweylite graduates into albite and the talc is bounded by serpentine. The mineral has many of the character- istics of pyrosclerite the points in which it differs from that species beingt he following. Hardness =1.5 -2. Specific gravity = 2.558. Colour greyish inclining sometimes to bronze-yellow. Its composition is- SiO% Also,. Fe203. FeO. MgO. Na20+Li20. H20. 35.99 9.52 5.35 1.08 32.94 0.41 14°60=99*89 corresponding with the formula- Z(+(MgO) + + { (A~,0~),(FezO~),))3SiOz.4~~0 which is on the type of pyrosclerite plus one mol. H,O. 11. Pseudomorpk after Pecto1ite.-Occurs in calcite traversing the trap near Bergen Hill New Jersey. In general appearance and specific MINERALOGICAL CHEMISTRY.29 gravity it resembles talc but in hardness (= 2*5),and reaction when heated alone or in a borax-bead it corresponds with sepiolite (meer- schaum) with which it agrees in composition analysis giving Si02 MgO. A1203. FeO. MnO. CaO. HzO. 60.55 26.56 1.02 0-72 0.65 1.41 9*30=100*21 111. Leucaugite from Amity New 1Ybr7; resembles in general ap-pearance the spinel which abounds in the neighbourhood and especi- ally that of Warwick New York but its density and composition are those of leucaugite from which it differs however a little in hardness H being 5.5. Analysis gave- Si02. A1203. Fe203. MgO. CaO. HzO. 50.05 7-16 0.56 14.48 25.63 1.66=99*54 It will be seen that the relative proportion of lime and magnesia is slightly different from the proportion in leucnugite.IV. Alumino-magnesic Silicate accompanying Corzlm,durn.-The corun-dum of Chester Pennsylvania is sometimes bordered by lesleyite con- taining potash and lithia the first constituent diminishing in amount as the lesleyite recedes from the corundum. The new mineral was found accompanying the lesleyite. Analysis-SiO* Al,03. Fe203. FeO. MgO. LizO. NazO. H20. 30.62 21.73 0.42 5.01 29.69 0.11 0.14 12*26==99*98 The oxygen ratio for R,O (including the small per centage of ferric oxide due to superficial oxidation of the ferrous oxide) R,O, SiO and H,O is 4.84 3 4 3.23. The mineral resembles ripidolite in physical properties but its oxygen ratio is not that of ripidolite nor has any specimen of that mineral in this neighbourhood been found by the author to contain a trace of lithium.V. Blonstone Variety of Albite from Delaware Co. Pemwy1vania.-Composition-Si02. AlzOa. Pe203. CaO. MgO. Na20. &O. H20. 67.70 19.98 trace 1.47 0.11 8.86 1.36 0.08=99*56 It has the usual properties of albite. VI. Antholite from Xtar Rock Comord,Delaware Pennsylvania.-Has the usual properties of this amphibole. By the analytical results it will be seen that oae-third of the usual amount of ferrous oxide is replaced by other bases. Composition-Si02. Also3. FeO. MnO. CaO. MgO. E20. Na20. H20. 55.12 0.55 8.20 0.33 0.75 31.18 1.01 1.55 2*21=100*90 VII. ‘Werrheritefrom Van Arsdale’s Quarry Bucks Co. Pennsylvar~ia. -This particular variety of scapolite has usually been classed as an ekebergite hut its composition is nearer that of the least altered varieties of wernerite.Analysis gave- SiOp 47.47 Also,. 27.51 Fe203. trace MgO. 1-20 CaO. 17.59 Na20. 3.05 E20. 1.40 HsO. 1*48==99*70 73. J. G. ABSTRACTS OF CHEMICAL PAPERS. Microsommite. By GI.. TOX RATH(Jahrbuch fur Mineralogie 1873 544). THIS name was gix-en by Scacchi to a mineral which he discovered in the Vesuvian outcasts of the eruption of 1872. He describes it as occurring in very small hexagonal prisms (20 weighing about a milli- gram) resembling nephelin in form but differing from that mineral by an occasional peculiar grouping in tufts and especially by contain- ing chlorine. Vom Rath who received from Scacchi a number of the Vesuvian outcasts of the eruption of the 26th of April 1872 has made a more particular examination of the microsommite contained in them.It occurs in both kinds of blocks ejected from the great crater in the Atrio namely t'he monolithic consisting of single fragments of old porous Somma lavas and the conglomerate consisting of loosely united fragments of lava together with eisystals of augite. Both kinds are usually covered and bound together by a thin crust of new lava. In the monolithic blocks the minerals newly farmed by sublimation- leucite sodalite microsommite augite hornblende iron glance-fill the pores ; in the conglomerates they 611 the interstices between the individual lumps and crystals Microsommite belongs to the hexagonal system its form is pris- matic terminated by the dull end-face.The edges between the prism and the bases are sometimes truncated by the faces of the double six- sided pyramid. The inclination of the pyramidal to the prismatic faces P :(x P is by measurement about 111"50' hence the ratio of the axes is a c (vertical) = 2.88 1,and the calculated values of the angles are P co P = 158' 34' ; P P (lateral) = 43" 40'. The faces of the prism are vertically striated sometimes nearly rounded. The crystals are colourless and transparent about as hard as felspar ; have a sp. gr. of 2.60 (at 15") ; melt with diiliculty before the blowpipe; suffer no loss even on intense ignition. The mineral dissolves in nitric acid with separation of gelatinous silica and the solution gives with silver solution a copious precipitate of silver chloride.Analysis gave-SiO2. A1203. CaO. E20. E%O. C1. SO3. 33.0 29.0 11.2 11.5 8.7 9.1 1.7 = 1042 If we suppose t'hat the chlorine is united with sodium (9.1 C1 .t 5.9 Na the latter being equivalent to 8.0 NazO) the excess of Qhe analysis becomes reduced to 2.2 p. c. and we obtain 0.7 p. c. Na20 in addition to the 5.9 Na. The sodium given by the analysis was weighed as sulphate together with the potassium and estimated by difference after the potassium had been determined as platino-chloride. It is therefore very pro-bable that the estimation of the sodium was rather too high and that the whole of it is really combined in the mineral with chlorine. The oxygen of the silica (= 18.0) and that of the alumina (= 13.5) are nearly in the ratio of 4 3 80 that this portion of the compound = A1,03.2Si02 as in sodalite nosean and hauyne.Microsommite like hauyne contains lime and alkali in isomorphous mixtnre and MINERALOGICAL CHEMISTRY. therefore represents a hemisilicate of aluminium calcium potas- sium &c. united with sodium chloride and a small quantity of calcium sulphate. Its probable formula is- ~~~~} A1203.2Si02+ NaCl + &(Ca0.Si02) which gives by calculation- SiOz. Alz03. CaO. &O. NEL C1. SO3 33.9 28.3 10.5 10.4 6.3 9.8 1.7 = 100.0 Microsommite eonnects the sodalite group with nepheline to which latter the mineral approaches nearly in its crystalline form. In fact the most obtuse of the double hexagonal pyramids known in nephelin agrees nearly with the double pyramid of the newly formed Vesuvian prisms the formation of which must be attributed to the action of volcanic vapours laden with sodium chloride on the leucites (potash alumina) and the augites (lime) of the lava.Here therefore we meet with a new example of the action of sea-water in the formation of minerals by volcanic processes. H. W. Occurrence of Tellurium Minerals in the United States. By Dr. BURKART (Jahrbuch fur Mineralogie 1873 476-495). THEoccurrence of bismuth telluride telluric bismuth or tetradymite Bi,Te, in Virginia and Georgia has long been known. Genth also analysed a sulpho-telluride of bismuth Bi2S3.2Bi2Te3 occurring in North Carolina (Anz.J. of Sci. [el xix 16 ; xlv 317). More recently tellurium minerals have been found in California viz.silver telluride or hessite near Georgetown in Eldorado connty ; bis-muth telluride together with native gold in the Melones and Stanislaus mines in Calaveras Co. and sylvanite or graphic tellurium at the New Melones mines. Giiido Kustel (Mirtirng and ScientiJic Press qf San Francisco 20th May 1865) describes the principal tellurium ore of the Mellones mine as a telluride of siluer and gold of sp. gr. 9 to 9.4 containing 35.40 p. c. Te 40.60 Ag and 24.80 Au = 100*80. In a subsequent communication (Berg. u.. EGttenm. Zeitung 1866 128) Kustel states that he has also found in this mine silver telluride natire tellurium copper-nickel iron pyrites and native gold but no sylvanite or altaite. Tellurium ores from three other localities in Cali- fornia are mentioned by B.Silliman (Academy of Natural Science of Cul$omzia 2nd December 1867). Genth obtained from various persons a somewhat numerous collec- tion of tellurium minerals from the western declivity of the hlifornian range namely petxite hessite altaite native tellurium and three new mi&rals viz. ealaverite melonite aiid montaite. The tellurium ores of the Stanislaus mine occur in micaceous and chloritic slates together with quartz dolomite apatite a uranium ore kitanic iron iron pyrites copper pyrites and small quantities of galena ABSTRACTS OF CHEMICAL PAPERS. blende and native gold. They occur for the most part finely divided and mixed together in snch a manner that it is difficult to separate a quantity of any one of them sufficient for analysis.(a.) Petxite and Hessite.-Auriferous silver telluride or petzite is the most abundant of all the Californiam tellurium ores. The specimens from the Stanislaus and Golden Rule mines were destitute of crystal- line structure exhibited a distinctly concho'idal fracture metallic lustre steel-grey to iron-black colour and iron-black streak ; hardness = 2.51 ;sp. gr.9 to 9.4. They contained 24*80-25-70 p. c. gold 40.60- 42.36 silver and 31.94-34.16 tellurium agreeing with the formula Au2Te.3Ag,Te. They- contained therefore more gold than the pet- zite or sylvanite of Nagyag but need not on that account be regarded as a distinct species inasmuch as gold can replace silver in various proportions.Petzite is also found abundantly in the Red Cloud mine at Gold Hill Boulder county Colorado. Hessite Ag2Te containing little or no gold is found in the Stanis- laus mine but always mixed with other minerals viz. native gold altaite and quartz. (b.) A7taite PbTe occurs in the Stanislaus and Golden Rule mines mixed with hessite and petzite. It is distinguished from the other tel- lurium ores by its tin-white colour inclining to greenish-yellow. Hardness = 3. Two specimens gave the following analytical num- bers (1) after deduction of 1.03 p. C. quartz ; (2) after deduction of 1.96 P.C. :-Pb. Ag. Au. Te. (1.) 60.71 1.17 0.26 37.31 = 99.45 (2.) 47.84 11.30 3.86 37.00 == 100 These numbers show that (1) consisted of 99.25 p.c. altaite and 2.20 hessite ; (2) of 77-42 altaite and 23.11 hessite. (c.) Native TeZZuriurn occurs mixed with the preceding minerals in small greyish-white specks. (d.) MeZomite,-This is a new mineral belonging to the hexagonal system. Genth observed only one microscopic but perfectly formed hexagonal plate the mineral being for the most part indistinctly granular and laminar but exhibiting a very distinct basal cleavage. It has a metallic lustre reddish-white colour and dark grey streak. Heated in a glass tube before the blowpipe it yields a sublimate which melts to colourless drops leaving a grey residue. Heated on charcoal it burns with a bluish flame yielding a small white deposit and a greyish-green residue from which by fusion with sodium car- bonate in the inner flame a grey magnetic powder of metallic nickel is obtained.It dissolves with green colour in nitric acid the solution leaving on evaporation a white crystallhe powder of tellurous acid. Analysis gave- Silver ... 4.08 which requires 2.42 Te and therefore represents 6.50 p.c. Hemite YY JJ Lead .... 072 , 0.45 ,7 ?? 1.17 Altaite Nickel. .. 20.98 , 68.27 )) ,¶ 89.25 ,) Melonite yy Tellurium73-45 , -7) Y? 2.29 , Native L__ tellurium. 99-21 n.14 MINERALOQIOAL CHEMISTRY The nickel contains only sufficient cobalt to give a very light blue colour to a borax bead. The composition of melonite as determined by the analysis agrees approximately with the formula NiTe3 which requires 23.51 p. c. Ni and 76.49 Te.(e.) Cnlavarite AuTez.-This is also a new mineral which Genth found only once in the Stanislaus mine as'sociated with petzite. It is massive non-crystalline soft (hardness = 3) with metallic lustre bronze-yellow colour yellowish-grey streak an4 uneven fracture approaching the conchoidal. Heated on charcoal before the blowpipe it burns with bluish-green colour leaving a gold bead of a bright yellow colour. Nitric acid colours it darker and separates gold. In aqua regia it dissolves with separation of silver chloride. Analysis gave after deduction of 1.45 p. c. quartz- Gold. Silver. Tellurium. I. 40-70 3.52 55-89 = 100*11 11. 40.02 3.08 56.00 = 100 The silver is due to the admixed petzite. Deducting this the analytical numbers for calaverite agree nearly with the formula AuTe3 which re- quires44.7 p.c. gold and 55.83 tellurium. (fa> Nontade BiZO3.TeOa.HzO(or 2HzO).-This mineral formed by oxidation of tetradymite BizTe3 was first observed by Genth during the examination of the tetradymite of Highland in Montana ; subsequently the so-called tetradymite from Davidson Co. North Carolina which was in great part oxidised and gave off chlorine when treated with hydrochloric acid was found to be identical with it. The '* yellow bismuth oxide " from Whitehall in Virginia analysed by Jackson (Am. J. of Xci. [2] x 78) is also regarded by Genth as pro- bably the same mineral. Montanite is not crystallised but exhibits here and there the scaly structure of the original tetradyrnite and is in reality a pseudomorph after the latter on which it forms a coating.It is earthy soft dull to waxy in lustre yellowish to white and opaque. Before the blowpipe it exhibits the reactions of bismuth and of tellurium and gives off water when heated in a tube. Genth's analyses lead to the following results :-I. Oxygen. 11. Oxygen. 111. Oxygen. Ferric oxide .., 0.56 -1.26 -0.32 -Lead oxide.. .... 0.39 -__. -Cupric oxide.. .. -1-04! -1-08 -Bismuth oxide . . 66.78 6.85 68.78 6.29 71.90 7-37 Tellurous oxide, 26.83 7.30 25.45 7.05 23.90 6-51 Water ........ 5.74 -3.47 -2.86 - 100.50 100*00 100.06 The oxygen ratio between the bismuth oxide and the tellurous oxide is nearly 1 1,but it is as yet uncertain whether the composition of the mineral is Biz03.Te03.Hz0or IZi2O3.TeO3.2HZO IT.W. VOL XXVII. u ABSTRACTS OF CHEMICAL PAPERS. The Nickel Ores of Horbach near St. Blasien in the Black Forest. By E. KNOP (Jahrbuch f. Mineralogie 1873,521-529). THESEores are imbedded in the serpentinized gneiss of the locality in irregular nodules accompanied by copper pyrites enclosing here and there granular aggregations of iron glance. They consist of nickeli-ferous magnetic pyrites having a magnetic action a metallic aspect a pinchbeck-brown colour inclining to steel-grey darker than that of true magnetic pyrites and black streak. Hardness between 4 and 5. sp. 9:. = 4.43. Cleavage imperfect in one direction only. Analysis of material carefully freed from the accompanying copper pyrites and iron glance gave as a mean 45-87 p.c. sulphur 41.96 iron and 11.98 nickel agreeing nearly with the formula FeaNiSla which requires 45.9 S. 42.8 Fe. and 11.2 Ni. The ore for which the name h o r b a c hi t e is proposed may accordingly be regarded as an isomor- phous mixture of the trisulphides of iron and nickel- 4Fes3 + XS,. This composition has not hitherto been observed among native sul- phides all those previously known being either bisulphides monosul- phides sulphides of lower degree or compounds of these with bioxides. On the other hand the composition of the Horbach ores does not appear to be quite constant; a specimen of sp. gr. 4.7 having been found to contain 40.03 p. c. sulphur 55.96 iron and 3.86 nickel. The author here enters into a theoretical discussion of the constitu- tion of the several varieties of magnetic pyrites for which we must refer to the original paper.Horbachite subjected to lixiviation in contact with the air becomes oxidised and partly converted into the sulphates of iron and nickel which dissolve out. Knop suggests this method for the extraction of nickel from the Horbach ores instead of the treatment in the dry way hitherto adopted whereby enormous quantities of sulphurous acid are thrown off causing great injury to the neighbouring forests. H. W. Indium in American Blendes. By H. B. CORNWALL (Chem. News xxviii 28 reprinted from the American Chemist). ZINC blendes from West Ossipee and Eaton both in New Hampshire gave distinct spectroscopic indications of indium.In a specimen from Roxbury (Connecticut) the indium could be detected in the raw mineral without dissolving in acid. B. J. G. Meteoric Ikn from Shingle Springs Eldorado County Cali- fornia. By B. SILLIMAN (American Jour. Sci. [3] vi 18-22). THIS mass weighing 85 lbs. was found in 1869 and is believed to be the first discovered in California. Its cross-section was approximately a semicircle. It was very homogeneous in composition two masses of MINERALOGICAL CHEXISTRY. pyrites 3 and 5 mm. in diameter being the only accompanying mineral. For a depth of about 5 mm. from the surface the iron was harder than in any other part. Its density determined an a mass of 750 grams was 1.875 the shavings cut by the planing machine having a density of 8.024.The high density of the mass compared with ordinary meteoric iron is probably due to the large percentage of nickel which is more than twice the average amoLint. A structure resembling the Wid- mannstattian figures was developed by etching but it was visible to the microscope only. The suggestion of Berxelius that the Wid- mannstattian figures are due to the segregation of the nickel alloy in lines of the octohedron in virtue of its comparative insolubility in the etching acid seems unsupported by this result. According to this theory a high percentage of nickel would cause a clearer development than usual. The existence of a hard crust in this case would seem however to indicate that the mass had not endured the intensity and duration of heating of average meteoric iron so that crystalline structure was but feebly developed.The Cape of Good Hope iron (Rammelsberg ~~~?~e~aZc?~e~2~e, 919) has aiso no Widmannstattian figures and it contaius much nickel (15.09 per cent.). It has how- ever much more cobalt (2.56 per cent.) than the Californian iron and but five elements in its composition. The following is the composition of the Californian mass :-Fe 81-48; Ni 17.17 ; Co 0.60 ; Al 0.09 ; Cr 0.02 ; Xg,0.01;Ca 0.16 ;C 0.07 ; Si 0.03 ; P 0.31 ; S 0.01 ; I(,0.03 = 99.913. Clark analysed a meteoric iron from Tennessee which contained Ni 17.10 ; Co 2.04. Smith (Amer. Jozw. Xci. [Z] xix 155) obtained Ni,14%2; Co 0.46 with a mass from the same State. The average amount of nickel in 80 analyses referred to was 7.25 per cent.B. J. G. Product of Oxidation of Meteoric Iron; Comparison with Terrestrial Magnetites. By STANISLAX MEUNIER(Compt. rend. lxxvii 643-6443}. INformer papers the view has been developed that terrestrial rocks may be regarded as thtt crust of a globe the interior of which has a constitution similar to that of meteorites serpentine for instance forming the exterior of deeper lying masses of chantonnite. This view is now extended to magnetite which may be compared with the product of oxidation of a meteoric iron. The want of the structure which in meteoric iron gives rise to Wicimannstadt’s figures is no obstacle to this view since experiments show 6hat oxidation deprives meteoric iron of all trace of this structure.The absence of nickel from magnetite is also no insuperable objection since the author has observed that a sample of meteoric iron which had been treated with aqua regia and then washed and dried became after a time covered with ef33orescent crystals of pure nickel chloride. The tendency of iron to form insoluble peroxide might thus prevent it from being washed away while the nickel might be removed in the soluble form. This fact appears to account also for the almost invariable presence of nickel in serpentine while it is absent from chantonnite the former D2 ABSTRACTS OF CHEMICAL PAPERS. according to the preceding supposition having derived it from the washings of embedded magnetite. It is also noticeable that in reducing serpentine with carbon the iron and nickel combined forming an alloy which approached the composi- tion of meteoric iron.M. J. S. Carbon Dioxide in the Air of the Soil of Munich at Different Depths and at Different Times. By M. V. PETTENROFER (Zeitschrift f. Biologie ix 250-257). THEauthor has continued his determinations of carbon dioxide in the air of the soil (vide Chem.. Xoc. Journal [2] xi 361). The monthly means of carbon dioxide in 1,000 parts of air are as follows:- C02in COsin C02 in COz in air from air from air from air from 18’71. depth of depth of 18’72. depth of depth of 4 meters. 1Q metera. --meters. 4 meters. 1% Norember ...... 6,693 5.4’72 May. ........ 11*813 8 Y75 December ...... 6.048 4.125 June ........18 -718 11.983 18’72. July ........ 26 -110 14 547 .,.. 19 ’724 10 -308 January ........ 5.312 3.864 August February.. ...... 5.369 4.176 Sept>ember..,. 1’7*288 11 -156 March.. ........ 6-552 3.593 October .,,.. 12 ‘338 8 -227 April .......... 2.825 5.641 On comparing these numbers with those previously given it will be seen that the quantity of carbon dioxide is much larger throughout but that the variations at different seasons occur in the same order the minimum occurring in winter and the maximum in summer The mean of carbon dioxide in the air at the depth of 4 meters was 6.73 per 1,000 in 1871 and 11.81 in 1872; there is no apparent reason for this great difference. The numbers obtained by Fleck in similar experiments in the botanical garden of Dresden are given and also show an increase of carbon dioxide from the surface of the soil downwards; but here the amount is much larger being greater at a depth of 2 meters than at Munich at 4 meters and the variations with season are not so regular.E. K. The Warm Springs of Costa Rica. €37 A. V. FRANTZIUS (Jahrbuch f. Mineralogie 1873 496-510). THESEsprings more than 30 in number may be regarded as a. con- tinuation of the remarkable series of warm mineral springs discovered by Humboldt in Venezuela and extending for 150 miles from Cape Paria to Merida. The Costa Rica springs begin indeed 13 degrees of longitude farther west but are situated under nearly the same‘ parallel of longitude vie. 10” N. on a strip of land running for 30 miles from ORGANIC CHEMISTRY.east to west. Most of them occur in narrow mountain gorges on the banks of rivers or are even overflowed by the rivers so that they are visible only in the dry season. Many of them are situated near the border-line between the trachytic lavas of the volcanos (Turialba Irazu Barba Poas and Miravalfes) and the diorite and syenite moun- tain masses of Aguate Candelaria and Dota. Their temperature is higher in proportion as they are situated at a lower level. The highest observed temperature 55.6O R. (= 69.5"C. = 157.1" I?.) was exhi-bited by a spring situated about 800 Paris feet above the sea-level whilst another 4,000 feet above that level had a temperature of only 23.2" R. (=29" C. = 84.2" F.).Only one of these waters has been analysed namely that of Agua- caliente near Cartago an analysis of which was made in 1858 by Count Schaffgotsch. This water has a sp. gr. of 1.0022 and contains in 16 ounces 19.74 grams of anhydrous salts consisting of-&SO4. Na2S04. NaC1. NaaC03. CaC03. MgCO,. SiOz. Loss. 1.15 4.78 7.55 1.11 3-28 0.86 0.47 0.54 = 19.74 The water also contains free carbonic acid amounting to more than 2.35 grains but the quantity was not exactly determined. The water of most of the springs is probably similar in composition to that of Aguacaliente at least so far as regards the predominance of common salt. This is evident from the saline incrustations which form in places where the water stagnates and evaporates quickly during the dry season This circumstance gives a practical importance to t,he springs of Costa Rica inasmuch as cattle have a preference for localities where salt is deposited.They are also used as medicinal baths. The proportion of sodium chloride is however not sufficient to give them the character of true brine-springs. H. W.