MISERALOGICAL CHENISTRT. M i n e r a l 0 g i c a l Chemistry. 21 Asphalt in Utsh and Colorado. By G. H. STOXE (Amel.. J. Sci., 42,148--159) .-In western Colorado and north-eastern Utah, asphalt deposits of four classes are represented : (1) asphaltic sand-rock, known also as sand-asphalt and bituminous rock, the most abundaiit of all the asphaltic deposits; (2) bituminous shales or mark; ( 3 ) bituminous limestones ; and (4) outflow or overflow asphalt,, in- cluding all forms of asphalt that have oozed out of the rock that originally contained them. These deposits are described in detail by the author, and the various theories t h a t have been propounded to explain the origin of petroleum and asphalt are fully discussed. T h author intends t o complete a map of the asphalt exposures, and to publish a more complete account of them.Gmelinite from Nova Scotia. 3 y L. V. PIRSSON (Amel.. J. Sci., 42, 57-&3).--The zeolites of Nova Scotia have long been noted for the size and perfection of their crystals, and arriongst them gmeliriite has held a prominent place. The author has consequently made a careful investigation of the crystalline form and physical properties of this mineral. The material employed was collected at Pinnacle Island, Nova Scotin. I n order to control the crystallographic work, two analyses were made. In I the outer shell WAS aualysed, anti in €I tlie inner nucleus, the results being as follows :- B. H. B. SiO,. A1203. Fe20,. CaO. K20. N%O. H20. Total. I. 509.5 18-33 0.26 1-01 0.15 9.76 20.23 100.d9 IJ. W 6 7 18-50 0.15 1.05 0 16 9.88 20.15 10@*56 In considering the bearing of these results on the identity of this mineral with chabazite, there is an ampparent discordance. The re- sults of the crystallographic work point to a difference in axial ratio:, and there is also a different habit and cleavage.On the other hand, t,he twinning and the chemical coiistitution indicate the identity of the species. These appaxent discrepancies trhe author explains by the hypothesis that the effect of sodit is to lengthen tliu vertical axis; the analyses of chabazite and gmelinite showing that soda arid lime may22 ABSTRACTS OF CHEJIICAL PAPERS. replace each other t o any extent. According to this view, gmelinite would bear much the same relation to chabazite as enstatite does t o hypersthene. Whether i t should be considered a distinct Apecies would be largely a matter of choice or convenience.Newtonite and Rectorite, two new Minerals of the Kaolin Group. B y R N. BRACKETT and J. F. WILLIAMS (Amer. J. &i., 42, 11-21 j .-The authors describe two hydrous silicates of aluminium which they believe have not before been obsewed. The mineral, which they term newtode, is found on Sneed’s Creek, Newton Co., Arkansas: in R dark-grey clay, of lower carboniferous age. It is a plire white, soft, compact, liomogeneous snbstance, having a sp. gr. of 2-37. B. H. B. Analysis gave the f’ollowing results :- SiO,. AbO, Loss on ignition. Fe,O,,. CaO. K,O. Pu’a,O. Total. 40.28 35.27 22.89 0.21 0.54 0.99 0-73 100.85 These results ane in accord with the formiila AI2O3,2SiO2,4H2O. Under the microscope, the rriineral is seen to be entirely made up of minute rhombs.The second hydrous silicake of aluminium, rectorite, which is also to be regarded as new, is found in the Blue Mountain mining district, Arkansas. When pure, i t is a soft, white mineral occurring in large leaves. Analysis gave the followiug results :-- SiO?. A1.10,. Loss on ignition. Total. 54.32 37-69 7-99 100~00 I n this analysis, small pi*oportione, of ferric oxide, lime, magnesi:], and alkalis have been disregarded. The formula of the mineral is AI2O3,2SiO2,H20 + Aq. From the authors’ reseamhes it is probable that three members of the kaolin series out of the possible four are known ; and the present status of the series may be concisely stated as follows :- 1.Rectorite.. . . . . . . . 2. Kaoiin and members A1203,2Si02.2 t1,O Monoclinic or 0. of the kaolin group Ai2O3,2SiO2,2H20 + Aq 0. 3. - A120,,2Si0,,3H,O - 4. Newtonite . . . . . . . . Al2O3,2SiO2,H20 + Aq Monoclinic (?). A120,,2Si0,,4H20 + Aq Rhombohedral. B. H. B. New Analyses of Astrophyllite and Tscheffkinite. By L. G. EAK~KS (Amer. J . Sn’., 42, 34-373.--1. Ast.l.oPhi/ZZile.-Piaom a dis- cussion of the xnaljses published by Hiickstrom m d by Konig, Brogger deduces the general fortnula R”,R’,Ti( Si04)r for this mineral. The new analysis made by the author closely confirms this formula, agreeing with it better than those from which it was derived. Cal- culating the small amount, of ferric oxide present in with the R” group, the molecular ratios of the author’s analysis give the following elementary proportions :- Si,,0,9T i (Zr) 15qR’t33GR t231H,a.If ISERALOGICA L CHENISTRY.23 which reduces to Si,O,,.,Ti 1.Ji''3.65( RH)4.7. 2. !lk&efkin;fe -A fragment of this rare mineral from Bedford Co., Virgiiiia, analysed by the author, showed a distinctly banded structure of lustrous black and dull-black material. The analyses-, however, show that tliesc two bands are practically identical, the dull being somewhat more hydrated. The molecular ratios seem to lead to no definite or satisfactory formula, a result quite i n accordance with the eyidence furnished 1))- the microscopic examination of sections. B. H. B. Minerals in Hollow Spherulites of Rhyolite. By J. P. IDDINGS and S. L. PENFIELD (rln~er. .J. Lvc?., 42, 39--46).--The o c ~ ~ r r e n c e of fayalite with other mirierals in the litbophysae and hollm spherulites a t Obsidian Clig has been described by the authors (Abstr., 1891,26), and they have also called attention to the occurrence of faj'tlite in obsidian a t Lipari and Vulcano (Abstr., 1891, 1.58).In the prment paper, they contribute further to the knowledge of these aqneo-igneous products in siliceous lsvas by describing a somewhat different develop- ment of hollow spherulites in rhyolites at Glade Creek, Wyoming. In this rhyolite, as in the obsidian of Ohsidian Cliff, fayillite occurs i n atssoc*i;i tiort witli abundant quartz, B S the result of the mineralising action of vapours in the cooling acid lavas. The qiiartz in both localities has an unusual, simple, and pedect development, and is accompanied by an uncommon form of smidine and by tridymite.11 oreoi-w, in certain hollow spherulites the fayalite is replaced by himbleride arid biotite. R. H, B. Siliceous Sand of Monte Soratte. By G. GIORGIS (Gnazetta, 21, 514--516).-At St. Orest,e, near Monte Soratte, there is a deposit of qunrtzose sand of the pliocetie age lying unconfoi*mably on the jurrlwic limestone of the district. Microscopically it is seen to consist of quartz, orthoclasic felspar, and small quantities of mica. Its corn- position is as follows :-- Water at 100" = 0.20 per ceiit. SiO,. AI20,{. FePO,. CaO. MgO. Alkalis. Dry residue.. . . . . 93.50 3.6'2 traces traces - 2.88 Ditto aftel. lcviga- tion.. . . . . . . . . . 94.41 2.99 0.11 - - 2.42 S.B. A. A . Reproduction of Acid Rocks. By H. IJE CBATETJEII (Cowpf. rpnc?., 113.370- SCS).-'l'he distinctness, purity, and uniform devPlop- rnent of felspxr crystals show undoubtedly that they have heen formed within a f l u i d mass and not from nt solid mass under the influerice of mineralising agents;, or from the devitrification of a glass The grent pressure under which crystallismtion takes place has prevented ihe felspnr from becoining vitreous or amorphous, as it does wlim hented under atmospheric pressure. Kxperiments made nnder a pressure of24 ABSTRACTS OF CHEMICAL PAPERS. 5000 stmos. gave ouly glasses, probably because cooling took place much too rapidly. C. H. B. Kctmacite, Taenite, and Plessite, from the Welland Meteoric Iron. By J. M. DAVlSON (Amer.J. Sci., 42, G&ti6).-The WeI- land siderolite was described by E. E. Howell (PTOC. finchester Acnd. Sci., 1890, 86--87). Its analysis gave 91.17 per cent. of iron, and 8-54 per cent. of nickel. I t is singularly free from troilite and schreibersite, and thus offered a good opportunity for the analysis of its separated nickel-iron alloys. Between the decomposed ou hide and the compact centre there was a zone in which the oxidatinn was super- ficial, and confined to the planes of contactof the diffewnt alloys that form the Widmanstatten figures. It thus became possible to separate the kamacihe and tamite in quantities sufficient for analysis. It was intended to analjse the plessite as a whole ; bnt, on examination, its fine layers were SO suggestive of kamacite and of tcenite that an attempt was made to analyse them separately. The analytical iTesults were as follows, the analyses of knmacite and taenite being givcn, each next to its correspondtiig part of- the plessite :- Plessite.rI.-h----- 7 li nmwi tc. haiiiscit e-like part. 'I'uenite-like part. Tsenite. Pe.. . . 93-09 92-81 72.98 75. i 8 'Ni ... 6.69 6.9 7 25.87 24 32 Co.. . . 0.25 0.19 0.83 0.33 c * . .. 0.02 0.1'3 0-91 0.S) 100.03 100.16 100*39 The physical and chemical correspondences appear to justify the concliision that in the Welland meteoric iron there are but two dis- tinct nickel-iron alloys-kamacite and taenite, and that the so-called plessite consists merely of thin, alternating lamella3 of these two alloys. It is unsafe to generaiise on a single analysis; but an ex- aminatioii of the markings of other meteoric irons suggests the thought that in them also there may be but two distinct alloys. A Gold-bearing Hot-spring Deposit.By W. H. WEISD (Amei.. J: Xci , 42,166-169).--The author has examined a series of speci- mens from the Moant Morgan gold mine, Queensland, with a view to cornpace ttieir~ with the siliceous sintcrs from the hot-spring region of the Yellow,tone Park. These specimens possess unustial interest, inasmuch as the observations of the Government Geologist of Queens- land show that the Mount Morgan mine, which paid a dividend of S1,200,000 in 1889, works a depcsit of a hot spring, the ore being a siliceous sinter impregnated with auriferous hematite (compare Abstr., 1886, 21). The Steamboat Springs of Nerada are surrounded by deposits of sinter, in the fissnres of which ore dcposition is now taking place, L hmall amourit of gold being found in these contem- poraneous mineral reins.The Mount Morgan deposit is, however, the only hot-spring deposit known that has been found to contain B. H.B.ORQANZC CHEMISTRY. 25 gold in workable qiiantities. A careful search for such deposits has been inade for the past eight years in the Yellowstone Park, the most remarkable hot-spring district of the world, without bringing to light a single caseof the sort. Hot-spring waters and deposits have been most carefully analysed, without indicating the presence of even a trace of the precious metals. Note by Abstractw.-In discussing the origin of the Mount Morgan deposit, the author accepts Mr.H. L. Jack's theory, that the deposit is that of a geyser, without pointing out that the origin of this ore- deposit has been the therue of much controversy. According to some, the deposit is thought to be an auriferous zone traversed by a series of quartz veins of auriferous qundic. Others think that it is the decomposed cap of a large pyrites lode. The last contribntion to the discussion i s afforded by a paper read before the American Iristitute of Mining Engineers, in June, 1891, by Mr. T. A. H,icknrd, who adro- cates the theory of metamorphosis and replacement. The ore-deposit, he thinks, represents a n altered portion of shattered country rock, which, by reason of i t s crushed condition, was readily acted on by miiieral solutions, and t h a t these solntions replaced the basic and felspathic with acid and quartzose material, which was also anriferous.I t is its quartzose and pei*meable c4iamcter which bas saved from dis- integrntiou the mass t h n s affected, and has preserved it as an ore- body on the summit of the hill. Analysis of a Hot Mineral Spring at Sclafani. By E. PATERK~ (Gazzetta, 21, ii, 4~J--.51).--This water issues at, a temperature of 32-9", and has a sp. gr. of 1.0074 compared \kith water a t Uo. It con- tains 0.1982 gram per litre of free carbonic anhydride, 0.0171 gram of free liydrogen sulphide, and 16.9 C.C. of nitrogen per litre. The total weight of carboiiic anhydride per litre is 0.3527 gram, and of hydrogen sulphide 0.0185 gram. The residue from a litre, when dried at 110", weighs 12.510 grams.B. H. B. One litre of the water contains in grnms : SiO,. so,. c1. R 1. I. CaO. Sr0. 0.0746 0-0790 G 6900 0.0148 0.0062 0.4720 0.1145 MgO. NaJ3. K,O. Fe,O, and A1,03. 0.3550 5-5512 0.0170 0.001 5 togebher with 0*0003 gram of organic matter and traces of phosphoric acid, lithiurn. barium, arid manganese. The results obtained differ considerably from those published by Cappa in 1847. W. J. P.MISERALOGICAL CHENISTRT.M i n e r a l 0 g i c a l Chemistry.21Asphalt in Utsh and Colorado. By G. H. STOXE (Amel.. J. Sci.,42,148--159) .-In western Colorado and north-eastern Utah, asphaltdeposits of four classes are represented : (1) asphaltic sand-rock,known also as sand-asphalt and bituminous rock, the most abundaiitof all the asphaltic deposits; (2) bituminous shales or mark;( 3 ) bituminous limestones ; and (4) outflow or overflow asphalt,, in-cluding all forms of asphalt that have oozed out of the rock thatoriginally contained them.These deposits are described in detail bythe author, and the various theories t h a t have been propounded toexplain the origin of petroleum and asphalt are fully discussed. T hauthor intends t o complete a map of the asphalt exposures, and topublish a more complete account of them.Gmelinite from Nova Scotia. 3 y L. V. PIRSSON (Amel.. J. Sci., 42,57-&3).--The zeolites of Nova Scotia have long been noted for thesize and perfection of their crystals, and arriongst them gmeliriite hasheld a prominent place. The author has consequently made a carefulinvestigation of the crystalline form and physical properties of thismineral.The material employed was collected at Pinnacle Island,Nova Scotin. I n order to control the crystallographic work, twoanalyses were made. In I the outer shell WAS aualysed, anti in €Itlie inner nucleus, the results being as follows :-B. H. B.SiO,. A1203. Fe20,. CaO. K20. N%O. H20. Total.I. 509.5 18-33 0.26 1-01 0.15 9.76 20.23 100.d9IJ. W 6 7 18-50 0.15 1.05 0 16 9.88 20.15 10@*56In considering the bearing of these results on the identity of thismineral with chabazite, there is an ampparent discordance. The re-sults of the crystallographic work point to a difference in axial ratio:,and there is also a different habit and cleavage.On the other hand,t,he twinning and the chemical coiistitution indicate the identity ofthe species. These appaxent discrepancies trhe author explains by thehypothesis that the effect of sodit is to lengthen tliu vertical axis; theanalyses of chabazite and gmelinite showing that soda arid lime ma22 ABSTRACTS OF CHEJIICAL PAPERS.replace each other t o any extent. According to this view, gmelinitewould bear much the same relation to chabazite as enstatite does t ohypersthene. Whether i t should be considered a distinct Apecieswould be largely a matter of choice or convenience.Newtonite and Rectorite, two new Minerals of the KaolinGroup. B y R N. BRACKETT and J. F. WILLIAMS (Amer. J. &i., 42,11-21 j .-The authors describe two hydrous silicates of aluminiumwhich they believe have not before been obsewed.The mineral,which they term newtode, is found on Sneed’s Creek, Newton Co.,Arkansas: in R dark-grey clay, of lower carboniferous age. It is aplire white, soft, compact, liomogeneous snbstance, having a sp. gr. of2-37.B. H. B.Analysis gave the f’ollowing results :-SiO,. AbO, Loss on ignition. Fe,O,,. CaO. K,O. Pu’a,O. Total.40.28 35.27 22.89 0.21 0.54 0.99 0-73 100.85These results ane in accord with the formiila AI2O3,2SiO2,4H2O.Under the microscope, the rriineral is seen to be entirely made up ofminute rhombs.The second hydrous silicake of aluminium, rectorite, which is also tobe regarded as new, is found in the Blue Mountain mining district,Arkansas. When pure, i t is a soft, white mineral occurring in largeleaves.Analysis gave the followiug results :--SiO?. A1.10,. Loss on ignition. Total.54.32 37-69 7-99 100~00I n this analysis, small pi*oportione, of ferric oxide, lime, magnesi:],and alkalis have been disregarded. The formula of the mineral isAI2O3,2SiO2,H20 + Aq.From the authors’ reseamhes it is probable that three members ofthe kaolin series out of the possible four are known ; and the presentstatus of the series may be concisely stated as follows :-1. Rectorite.. . . . . . . .2. Kaoiin and members A1203,2Si02.2 t1,O Monoclinic or 0.of the kaolin group Ai2O3,2SiO2,2H20 + Aq 0.3. - A120,,2Si0,,3H,O -4. Newtonite . . . . . . . .Al2O3,2SiO2,H20 + Aq Monoclinic (?).A120,,2Si0,,4H20 + Aq Rhombohedral.B.H. B.New Analyses of Astrophyllite and Tscheffkinite. By L. G.EAK~KS (Amer. J . Sn’., 42, 34-373.--1. Ast.l.oPhi/ZZile.-Piaom a dis-cussion of the xnaljses published by Hiickstrom m d by Konig,Brogger deduces the general fortnula R”,R’,Ti( Si04)r for this mineral.The new analysis made by the author closely confirms this formula,agreeing with it better than those from which it was derived. Cal-culating the small amount, of ferric oxide present in with the R” group,the molecular ratios of the author’s analysis give the followingelementary proportions :-Si,,0,9T i (Zr) 15qR’t33GR t231H,aIf ISERALOGICA L CHENISTRY. 23which reduces toSi,O,,.,Ti 1.Ji''3.65( RH)4.7.2. !lk&efkin;fe -A fragment of this rare mineral from BedfordCo., Virgiiiia, analysed by the author, showed a distinctly bandedstructure of lustrous black and dull-black material.The analyses-,however, show that tliesc two bands are practically identical, the dullbeing somewhat more hydrated. The molecular ratios seem to leadto no definite or satisfactory formula, a result quite i n accordancewith the eyidence furnished 1))- the microscopic examination of sections.B. H. B.Minerals in Hollow Spherulites of Rhyolite. By J. P. IDDINGSand S. L. PENFIELD (rln~er. .J. Lvc?., 42, 39--46).--The o c ~ ~ r r e n c e offayalite with other mirierals in the litbophysae and hollm spherulitesa t Obsidian Clig has been described by the authors (Abstr., 1891,26),and they have also called attention to the occurrence of faj'tlite inobsidian a t Lipari and Vulcano (Abstr., 1891, 1.58).In the prmentpaper, they contribute further to the knowledge of these aqneo-igneousproducts in siliceous lsvas by describing a somewhat different develop-ment of hollow spherulites in rhyolites at Glade Creek, Wyoming.In this rhyolite, as in the obsidian of Ohsidian Cliff, fayillite occurs i natssoc*i;i tiort witli abundant quartz, B S the result of the mineralisingaction of vapours in the cooling acid lavas. The qiiartz in bothlocalities has an unusual, simple, and pedect development, and isaccompanied by an uncommon form of smidine and by tridymite.11 oreoi-w, in certain hollow spherulites the fayalite is replaced byhimbleride arid biotite. R. H, B.Siliceous Sand of Monte Soratte. By G.GIORGIS (Gnazetta, 21,514--516).-At St. Orest,e, near Monte Soratte, there is a deposit ofqunrtzose sand of the pliocetie age lying unconfoi*mably on the jurrlwiclimestone of the district. Microscopically it is seen to consist ofquartz, orthoclasic felspar, and small quantities of mica. Its corn-position is as follows :--Water at 100" = 0.20 per ceiit.SiO,. AI20,{. FePO,. CaO. MgO. Alkalis.Dry residue.. . . . . 93.50 3.6'2 traces traces - 2.88Ditto aftel. lcviga-tion.. . . . . . . . . . 94.41 2.99 0.11 - - 2.42S. B. A. A .Reproduction of Acid Rocks. By H. IJE CBATETJEII (Cowpf.rpnc?., 113.370- SCS).-'l'he distinctness, purity, and uniform devPlop-rnent of felspxr crystals show undoubtedly that they have heen formedwithin a f l u i d mass and not from nt solid mass under the influerice ofmineralising agents;, or from the devitrification of a glass The grentpressure under which crystallismtion takes place has prevented ihefelspnr from becoining vitreous or amorphous, as it does wlim hentedunder atmospheric pressure.Kxperiments made nnder a pressure o24 ABSTRACTS OF CHEMICAL PAPERS.5000 stmos. gave ouly glasses, probably because cooling took placemuch too rapidly. C. H. B.Kctmacite, Taenite, and Plessite, from the Welland MeteoricIron. By J. M. DAVlSON (Amer. J. Sci., 42, G&ti6).-The WeI-land siderolite was described by E. E. Howell (PTOC. finchester Acnd.Sci., 1890, 86--87). Its analysis gave 91.17 per cent. of iron, and8-54 per cent. of nickel. I t is singularly free from troilite andschreibersite, and thus offered a good opportunity for the analysis ofits separated nickel-iron alloys.Between the decomposed ou hide andthe compact centre there was a zone in which the oxidatinn was super-ficial, and confined to the planes of contactof the diffewnt alloys thatform the Widmanstatten figures. It thus became possible to separatethe kamacihe and tamite in quantities sufficient for analysis. It wasintended to analjse the plessite as a whole ; bnt, on examination, itsfine layers were SO suggestive of kamacite and of tcenite that anattempt was made to analyse them separately. The analytical iTesultswere as follows, the analyses of knmacite and taenite being givcn,each next to its correspondtiig part of- the plessite :-Plessite.rI.-h----- 7 li nmwi tc.haiiiscit e-like part. 'I'uenite-like part. Tsenite.Pe.. . . 93-09 92-81 72.98 75. i 8'Ni ... 6.69 6.9 7 25.87 24 32Co.. . . 0.25 0.19 0.83 0.33 c * . .. 0.02 0.1'3 0-91 0.S)100.03 100.16 100*39The physical and chemical correspondences appear to justify theconcliision that in the Welland meteoric iron there are but two dis-tinct nickel-iron alloys-kamacite and taenite, and that the so-calledplessite consists merely of thin, alternating lamella3 of these twoalloys. It is unsafe to generaiise on a single analysis; but an ex-aminatioii of the markings of other meteoric irons suggests thethought that in them also there may be but two distinct alloys.A Gold-bearing Hot-spring Deposit.By W. H. WEISD (Amei..J: Xci , 42,166-169).--The author has examined a series of speci-mens from the Moant Morgan gold mine, Queensland, with a view tocornpace ttieir~ with the siliceous sintcrs from the hot-spring regionof the Yellow,tone Park. These specimens possess unustial interest,inasmuch as the observations of the Government Geologist of Queens-land show that the Mount Morgan mine, which paid a dividend ofS1,200,000 in 1889, works a depcsit of a hot spring, the ore being asiliceous sinter impregnated with auriferous hematite (compareAbstr., 1886, 21). The Steamboat Springs of Nerada are surroundedby deposits of sinter, in the fissnres of which ore dcposition is nowtaking place, L hmall amourit of gold being found in these contem-poraneous mineral reins.The Mount Morgan deposit is, however,the only hot-spring deposit known that has been found to containB. H.BORQANZC CHEMISTRY. 25gold in workable qiiantities. A careful search for such deposits hasbeen inade for the past eight years in the Yellowstone Park, themost remarkable hot-spring district of the world, without bringing tolight a single caseof the sort. Hot-spring waters and deposits havebeen most carefully analysed, without indicating the presence of evena trace of the precious metals.Note by Abstractw.-In discussing the origin of the Mount Morgandeposit, the author accepts Mr. H. L. Jack's theory, that the depositis that of a geyser, without pointing out that the origin of this ore-deposit has been the therue of much controversy.According to some,the deposit is thought to be an auriferous zone traversed by a seriesof quartz veins of auriferous qundic. Others think that it is thedecomposed cap of a large pyrites lode. The last contribntion to thediscussion i s afforded by a paper read before the American Iristituteof Mining Engineers, in June, 1891, by Mr. T. A. H,icknrd, who adro-cates the theory of metamorphosis and replacement. The ore-deposit,he thinks, represents a n altered portion of shattered country rock,which, by reason of i t s crushed condition, was readily acted on bymiiieral solutions, and t h a t these solntions replaced the basic andfelspathic with acid and quartzose material, which was also anriferous.I t is its quartzose and pei*meable c4iamcter which bas saved from dis-integrntiou the mass t h n s affected, and has preserved it as an ore-body on the summit of the hill.Analysis of a Hot Mineral Spring at Sclafani. By E. PATERK~(Gazzetta, 21, ii, 4~J--.51).--This water issues at, a temperature of32-9", and has a sp. gr. of 1.0074 compared \kith water a t Uo. It con-tains 0.1982 gram per litre of free carbonic anhydride, 0.0171 gramof free liydrogen sulphide, and 16.9 C.C. of nitrogen per litre. Thetotal weight of carboiiic anhydride per litre is 0.3527 gram, and ofhydrogen sulphide 0.0185 gram. The residue from a litre, whendried at 110", weighs 12.510 grams.B. H. B.One litre of the water contains in grnms :SiO,. so,. c1. R 1. I. CaO. Sr0.0.0746 0-0790 G 6900 0.0148 0.0062 0.4720 0.1145MgO. NaJ3. K,O. Fe,O, and A1,03.0.3550 5-5512 0.0170 0.001 5togebher with 0*0003 gram of organic matter and traces of phosphoricacid, lithiurn. barium, arid manganese. The results obtained differconsiderably from those published by Cappa in 1847. W. J. P