IMAGE EVALUATION TEST TARGET (MT-3) 1.0 1.1 £ Itt |Z0 6" ^^V Hictographic Qr*ionrioc Corparadon a»Mn«MMSTtin \Mntm,N.Y. I4SM ** * ^ CIHM/ICMH Series. CIHM/ICMH Collection de microfiches. Canadian Instituta for Historical IMicroraproductions / Inttitut Canadian &m microraproductions hifttoriquas ■^ M ■^1^ T«chnical and Bibliographie Notaa/Notaa tachniquat at bibllooraphiquaa Tha Inttituta has attamptad to obtain tha baat original copy avaiiabia for fiiming. Faaturaa of thia copy which may ba bibliographicaliy uniqua, which may altar any of tha imagaa in tha raproduction, or which may aigniflcantly ehanga tha usual mathod of fiiming. ara chaclcad baSow. □ Colourad covars/ Couvortura da coulaur I I Covars damagad/ Couvartura andommagAa □ Covars rastorad and/or lamlnatad/ Couvortura rastaunia at/ou pailfcul4a □ Cover title missing/ Le titre 6» couvartura manque I I Coloured maps/ D D D Caites gAographiquas en couleur Coloured ink (i.e. other than blue or blacic)/ Encre de couleur (i.e. autre que bieue ou noire) □ Coloured plates and/or illustrations/ Planches et/ou illustrations en couleur □ Bound with other material/ ReliA avac d'autres documents D Tight binding may causa shadows or distortion along interior margin/ La reiiure serr^e pout causer do I'ombre ou de la distortion la long de la marge intirieure Blank leaves added during restoration may appear within the text. Whenever poaaibia, these have been omitted from filming/ (! se peut que certaines pages blanchaa ajoutiaa ics d'une restauration apparaiaaant dans la texte, mais, lorsque cela Atait possible, cas pages n'ont pas M filmtes. Additional comments:/ Commentaires suppSAmentaires; L'Inatitut a microfilm* la maillaur axamplalra qu'il lui a *t4 posaibia da aa procurer. Lea details da cat axamplalra qui aont paut-4lra uniques du point da vua bibllographiqua. qui pauvant modifier una image raproduita, ou qui pauvent axiger uno modification dans la mAthoda normala de filmage •!ont indiquAs ci-dessous. n~| Coloured pagea/ Pagaa da couleur Pagea damaged/ Pagaa andommagAaa □ Pagaa restored and/or laminated/ Pages restaurtes at/ou pellicuMes r~1/Pagas discoloured, stained or foxed/ LJd Pages dAcolortes, tachatAes ou piquAes □ Pages detached/ Pagea d^tachias nn^ Showthrough/ L_J Tranaparance □ Quality of print varies/ Quali*i inAgala de I'impression □ Includes supplementary material/ Comprand du matirlel suppltfmentaire □ Only edition available/ Seuie Mition disponible D Pagaa wholly or partially obacured by errata slips, tiasuaa, etc., have been refllmed to enaura tha baat poaaibia image/ Lea pages totaiamant ou partialiement obscurcies par un fauillet d'ertata, una pelure. etc., ont At* filmies A nouveau de fa^on A obtanir la mailleure image poaaibia. This item is filmed at the reduction ratio checked below/ Ce document est filmi au taux da reduction indiqu4 c^dasaoua. 10X 14X 18X 22X MX 30X x 12X 16X aox a4x 28X 32X Th« copy filiiMd h«r* has b««n r«produc«d thanks to tho gonoroslty of: Library of tha Public ArchivM of Canada L'axamplaira film* fut raproduit grica i la g4n4roaiti da: La bibliothAqua da« Archivas publiquas du Canada Tha imagaa appaaring hara ara tha bast quality poasibia oonsMaring tha condition and laglblllty of tha original copy and in Icaaping with tha filming contract apacif icationa. Laa imagaa auivantaa ont 4t4 raproduitaa avac la plua grand aoin, sompta tanu da la condition at da la nattat* da l'axamplaira film*, at an conformity avac las conditiona du contrat da filmaga. Original coplaa in printad papar covara ara filmad baginning with tha front covar and anding on tha last paga with a printad or illuatratad !mpraa> alon, or tha back covar whan approprlata. All othar original copkM ara filmad baginning on tha firat paga with a printad or Illuatratad impraa* alon, and anding on tha iaat paga with a printad or illuatratad Impraasion. Tha last racordad frama on aach microficha ahaii contain tha aymbd — ^^ (moaning "CON- TINUED"), ir tha aymbol ▼ (moaning "END"), whichavar appllaa. Laa axampbiiras originaux dont la couvartura an paplar aat ImprimAa aont fiimAs an commandant Pf r la pramlar plat at an tarmlnant soli par la darnlAra pagu qui comporta una amprainta d'Impraaaion ou d'iiluatration, aoit par la aacond plat, aalon la cas. Tous las autraa axamplairas originaux sont fllmte an commandant par la pramlAra paga qui comporta una amprainta d'imprassion ou d'iilustration at an tarmlnant par la damlAra paga qui comporta una taila amprainta. Un daa aymbolas suivants apparattra sur la darnlAra Imago da chrqua microficha, salon la cas: la aymbola — »^ aignifia "A SUIVRE", la aymbolo ▼ aignifia "FIN". Mapa, plataa. charta, ate, may ba filmad at diffarant raduction ratio*. Thoaa too iarga to ba antiraly included In ona axposura ara filmad baginning in tha uppar laft hand comor, loft to right and top to bottom, aa many framaa aa raquirad. Tha following diagrams illuatrata tha mathod: Las cartaa, planchas, tablaaux, ate, pauvant Atra fllmto A das taux da rMuctlon dlff«rants. Lorsquo la document aat trap grand pour Atra raproiduit an un aaul clichA, 11 aat filmA A partir da I'angia aupAriaur gaucha, da gauclia A droita. at da haut an baa, an pranant la nombra d'imagaa nAcaaaaira. iiaa diagrammas suivants lliustrant la mAthoda. 1 2 3 1 2 3 4 S 6 CONTRIBUTIONS TO THE CHEMISTRY OF NATURAL WATERS. BY T. STERRY HUNT, LL.D., F.R.S. Reprinted frou Silliuan's Ahbrioan Journal or Science for March, July, and Seftsubbr, 1865. MONTREAL 1865. ^Bmi^mmmmmmmmR CHEMISTRY OF NATLRAL WATERS. I. GENERAL PRINCIPLES AI action of the ancient sea-waters, holding a largo anion lit of chlorid of cnlcium, upon the hydrated and half- decomposed foKlspars which constituted the clays of the period, may have given rise to those double silicates which formed the lime- soda feldspars so abundant in the Ltibrndor scries. 1^ 8. The reaclions just described assume an importance in the case of wators impregnated with soluble matters from vegetable de- cay ; and in this event, another and not less important class of phe- nomena intervenes, which are due to the deoxydizing power of the dissolved organic matter. By the action of this upon the insoluble peroxyd of iron set free from the decomposition of ferruginous minerals and disseminated in the sediments, protozyd of iron is formed, which is soluble both in carbonic acid, and in ihe excess of the organic (acid) mutter. By this means not only are great (juantities of iron dissolved, but masses oi' sediments are sometimes entirely deprived of iron-oxyd, and thus beds of white clay and sand are formed. The waters thus charged with proto-salts of iron absorb oxygen when exposed to the air, and then deposit the metal as hydratcd peroxyd, which when the organic matter is in excess, carries down a greater or loss proportion of it in combina- tion. Such organic matters are rarely absent from liraonite, and fh some specimens of ochre amount to as much as fifteen per cent.* The conditions under which hydrous peroxyd of manganese is often found are very similar to those of hydrous peroxyd of iron with which it is so frequently associated ; and there is little doubt that oxyd of manganese may be dissolved by a process like that just pointed out. A portion of manganese has been observed in the soluble matters from decaying peat-moss ; and it seems to be generally present in small quantities with iron in surface-waters. § 9. There is reason to believe that alumina i.s also, under cer- tain conditions, dissolved by waters holding organic acids. The existence of pigotite, a native compound of alumina with an organic acid, and the occasional association of gibbsite with lim- onite, point to such a reaction. That it is not more abundant in * Geology of Canada, p. 512. 6 :i solution, is due to the fact, that, unlike most other metallic ozyds, alumina, insteaa of being separated in a free state by the slow dtioomposlvion of its silicious compounds, remains in combination with silica. The formation of bauxite, a mixture of hydrate of alumina with variable proportions of hydrous peroxyd of iron, which forms extensive beds in the tertiary sediments of the great Medi- terranean basin, indicates a solution of alumina on a grand scale, and perhaps owes its origin to the decomposition of solutions of native alum by alkaline or earthy carbonates. Emery, a crystallino anhydrous form of alumina, has doubtless been formed in a sioiilir manner. Silliman's Journal [2] xxxii, 287. The existence in many localities of an insoluble s^s of the chalk in England. On the absence of free hydrated alumina from soils, see Miiller, cited in Silliman's Journal [2] xxxv, 292. § 10. The organic matter dissolved by the surface-waters serves to reduce to the condition of sulphurets the various soluble sulphates which it takes up at the same time or meets with in its course. These sulphurets, decomposed by carbonic acid, which is in part derived from the atmosphere, and in part from the oxyda- tion of the carbon of the organic matter, give rise to alkaline and earthy carbonates on the one haml, and to sulphuretted hydrogen on the other. In this way, under the influence of a somewhat elevated temperature, are generated sulphurous waters, whether of subterranean springs, or of tropical sea-marshes and lagoons. The reaction between the sulphurets thus formed and the salts or oxyds of iron, copper, and similar metals which may be present, gives rise to metallic sulphuieis. The decomposition of sulphur- etted hydrogen by the oxygen of the air, produces native sul- phur ; with which are generally found associate i sulphates of lime and strontia. By virtue of these reactions, soluble sulphates of lime and magnesia may be completely eliminated from waters, the bases as insoluble carbonates, and the sulphur as sulphuretted hy- drogen, free sulphur, or a metallic sulphuret. Moreover, as Foroh- hammer has pointed out in the paper already cited, sulphuret of potassium in the presence of ferruginous clays is also completely separated from solution, the sulphur as sulphuret of iron, and the alkali as a double aluminous silicate. § 11. We hive thus far considered the composition of surface- traters as modified by the decay of vegetation, or by the reactions between the matters derived from this source and the permeated sediments. Not less important however than the elements thus removed by substitution from sedimentary strata are those which arc liberated by the slow decomposition of the minerals composing these sediments. It has long been known that in the transformation of a feldspar into kaolin, the double silicate of alumina and alk)*li takes up a portion of water, and is resolved into a hydrous silicate of alumina ; while the alkali, together with a definite portion of silica, is separ- ated in a soluble state. The feldspar, an anhydrous double salt formed at an elevated temperature, has a tendency under certain con- ditions to ('.ombine at a lower temperature with a portion of water, and break up into two simplei silicates. Paubr^e has moreover shown that when kaolin is exposed to a heat of 400° C. in presence of a soluble silicate of potash, the two silicates unite and regenerate feldspar. These reactions are completely analogous to those presented by very many other double salts, ethers, amides, and similar compounds. The preliminary conditions of this con- version of feldspar into kaolin and a soluble alkaline silicate, how- ever, still require investigation. It is known that while some feldspathic rocks appear almost unalterable, others containing the same species of feldspar are found converted to a depth of many feet from the surface into kaolin. This chemical alteration, ac- cording to Fournet, is always preceded by a mechanical change of the feldspar, which first becomes opaque and friable, and is thus rendered permeable to water. He conceives this alteration to be molecular, and to be connected with the passage of the silicate into a dimorphous or allotropic condition.* § 12. The researches of Ebelman on the alterations of various rocks and minerals have thrown considerable light on the relations of sediments and natura! waters. I* From the analyses of basaltic and similar rocks, which include silicates of lime, magnesia, iron, and manganese in the forms of pyroxene, hornblende, and olivine, and which undergo a slow and superficial decomposition under atmos- pheric influences, it appears that durin»» the process of decay the greater part of the lime and magnesia is removed, together with a large proportion of silica. It was found moreover that in the case * Aanales dc Ghimie [2] Ir. 225. t Ebeiman, Recueil dca Travuux, ii, 1-79. 8 of a rock apparently composed of labradorite and pyroxene, the removal of the lime and magnesia from the decomposed portion was much more complete than that of the alkalies ; showing thus the comparatively greater stability of the feldspathio element. The decomposition of the feldspar in these mixed rocks is however at length eflected, and the final result approximates to a hydrous silicate of alumina, or clay. This slow decomposition of silicates of protoxyd-bases appears to be dre to the action of carbonic acid, which removing the lime and magnesia as carbonates, liberates the silica in a soluble form ; while the iron and manganese passing to a state of higher oxydation, remain behind, unless the action of organic matters intervenes to give them solubility. § 13. It is to be remarked that apart from the peculiar and complete decomposition resulting in the production of kaolin, to which orthoclase, oligoclase, and some other feldspathides, asleucite, beryl, and perhaps also the scapolites and albi*^e, are occasionally subject, orthoclase is loss liable to change than the soda-feldsprrs, albite, oligoclase, and labradorite. Weathered surfaces of these become covered with a thin, soft, white, and opaque crust from decomposition, while the surfaces of orthoclase under similar conditions still preserve their liardness and transluccncy. The de- composition of feldspathides, and other aluminous double silicates, whether rapid and complete, or slow and partial, apparently yields the same results. A gradual process of this kind is constantly going on in the feldspathic matters which form a large proportion of the mechanical sediments of all formations ; and in deeply buried strata is not improbably accelerated by the elevation of temperature. The soluble alkaline silicate resulting from this process is in most cases decomposed by carbonates of lime and magnesia in the sediments, giving rise to silicates of these bases (which are for the greater part separated in an insoluble state), and to carbonate of soda. Only in rare cases docs potash appear in large proportion among the soluble salts thus liberated from sediments, partly because soda-feldspars are more subject to change, and partly from the fact that potash-salts would be separated from the percolating waters in virtue of the reactions mentioned in § 5. Hence it happens that apart fiom the neutral soda- palts of extraneous origin, waters permeating sediments containing alkaliferous silicates, generally bring to the surface little more than soda combined with carbonic and sometimes with boric acid, and carbonates of lime and magnesia with small portions of silica. ni ft § 14. This ezplauation of the decomposition of alkaliferous sili- cates P.nd of the origin of carbonate of soda is opposed to the view of Bisohof, who conceives that carbonic acid is the chief agent in decomposing feldspathic minerals.* The solvent action of waters charged with carbonic acid is undoubted, as shown by various experimenters, especially by the M.^srs. Rogers, f but this acid is not always present in the quantities required. The proportion of it in atmospheric waters is so inadequate that it becomes necessary to suppose some subterranean source of the gas, which is by no means a constant accompaniment of natron-spring!*. A copious evolution of carbonic acid is observed in the vicinity of the lake of Laach, where the alkaline waters studied by Bischof occur. | The same thing is met with in many other localities of such springs, among which may be mentioned the region around Saratoga, where saline waters containing carbonate of soda, and highly charged with carbonic acid, rise in abundance from the Lower Silurian strata ; but further northward, along the valleys of Lake Cham- plain and the St. Lawrence, similar alkaline-saline waters, which .abound in the continuation of the same geological formations, are not at all acidulous. From this the conclusion seems justifiable that the production of carbonate of soda is a process, in some cases at least, independent of the presence of free carbonic acid. In this conneotion, it is well to recall the solvent power of pure water on alkaliferous silicates, as shown more especially by Bunsen, and also by Damour, who found that distilled water at temperatures much below 212° takes up from silicatos like palagonite and calci'icd mesotype, comparatively large amounts both of silica and alkalies. (Damour, Ann. Chim. et Phi/s. [3] xix, 481 .) § 15. Another and an important source of mineral impregnation to waters exists in the soluble salts enclosed in sedimentary strata, both in the solid state and in aqueous solution, and for the most part of marine origin. In order to form some conception of the amount of saline matters which may be contained in a dissolved state in the rocky strata of the earth, we have made numerous experiments to determine the porosity of various rocks; some, few of the results of which may here be noticed. Fragments of the rocks were dried at a heat of 150° to 200° F., in a current of • Bischof, Chem. Geol. ii, 181. t Silliman's Journal [2] v, 401. + Bischof, Lehrbuch, i, 357-363. \l \U H\ [I w dry air until they ceased to lose weight. They were then soaked in distilled water, and kept under it for many hours beneath an exhausted receiver. When thus saturated, they were wiped from adhering water, and weighed ; first in air to determine the aug- mentation of weight from absorption, and secondly, in water to give, by the loss in weight, the volume of the specimens. These data furnish the means of determining the volume of water ab- sorbed, which is given below for 100.00 parts of different rocks from the paleozoic strata of the St. Lawrence basin. Potsdam formation, (sandstoue) 3 specimens.. 2.26— 2. 7L " " " 3 " 6.94—9.35 Calciferous " (crys. dolomite) 4 " 1.89—2.53 " «' «' « 2 " 5.90—7.22 Chazy " (argil, limestone) 4 " 6.45-13.55 Trenton " (grey crys. ") 4 " 1.18—1.70 " " (black impalp. " ) 2 " 0.30—0.32 Utica " (black shale) 3 '• 0.75—2.10 Hudson River " (arenac. " ) — 7.94 Medina " (argil, sandstone) 2 specimens. .8.37-10.06 Ouelph " (crys. dolomite) 3 " .... 9.34-10.60 Niagara " (impalp. " ) 2 " .... 9.69-10.92 The above data might be much more extended, but sufficient have been given to show the porosity of the principal paleozoic rocks of the basin.* § 16. If we take for the Potsdam sandstone the mean of the first three trials, giving 2-5 per cent for the volume of water which it is capable of holding in its pores, we find that a thickness of 100 feflt of it would contain in every square mile, in round numbers, 70,000,000 cubic feet of water; an amount which would supply a cubic foot (over seven gallons) a minute for more than thirteen years. The observed thickness of the Potsdam sandstone in the district of Montreal, varies from 200 to 700 feet, and the mean of 500 feet may be taken. To this are to bo added 300 feet for the Calciferous formation, whose capacity for water may be taken, like the Potsdam sandstone, at 2*5 per cent. We have thus in each square mile of these formations, wherever they lie below the water-level, a volume of 490,000,000 cubic feet of water, equal to a supply of a cubic foot per minute for 106 years. • A great many similar determinations will be found in a Report on Building Stones to the British House of Commons in 1839, by Barry, Delabeche, and Smith. See also Delesse, Bui. Soc. G6ol. [2] xix, 64. tl The capacity of the 800 feet of Chazy and Trenton limestones which succeed these lower formations, may be fairly taken at one half that of those just named. But it is unnecessary to multiply such calculations: enough has been said to show that these sedimentary strata include in their pores great quartities of water, which was originally that of the ocean of the paleozoic age. These strata throughout the great Silurian basin of the St. Lawrence, are now for the greater part beneath the sea-level ; nor is there any good reason for supposing them to have ever been elevated much above their present horizon. Wells and borings sunk in various places in these rocks show them to be still filled with bitter saline waters ; but in regions where these rocks are inclined and dislocated, surface-waters gradually replace these saline waters, whifh in a mixed ar,t\ diluted state appear as mineral springs. These saline solutic i, other things being equal, will be better preserved in limestones or argillaceous rocks than in the more porous and permeable sandstones. § 17. But besides the saline matters thus disseminated in a dis- solved state in ordinary sedimentary rocks, there are great volumes of saliferous strata, properly so called, charged with the results of the evaporation of ancient sea-basins. These strata enclose not only gyp- sum and rock-salt, but in some regions large quantities of the double chlorid of potassium and magnesium, carnallite; and in others sulphate of soda, sulphate of magnesia, and complex sulphates like blodite and polyhallite. Besides these crystalline salts, the mother liquors containing the more soluble and uncrystallizable compounds, may also be supposed to impregnate, in some cases, the sediments of these saliferous formations. The conditions under which these various salts are deposited from sea-water, and their relations to the composition of the ocean in earlier geological periods, are reserved for consideration in § 22. Infiltrating waters remove from these saliferous strata their soluble ingredients- which, together with the ancient sea-waters of other sedimentary rocks, give rise to the various neutral saline waters ; while the mingling of these in various proportions with the alkaline waters whose origin has been described in § 13, produces intermediate classes of waters of much interest. § 18. I have elsewhere described the results of a series of experiments on the mutual action of the waters of these two classes.* When a dilute solution of bicarbonate ot soda is gradu- 64. Silliroan's Journal [2] xxviii, 170. r 12 ally added to a solution which, like soa-watcr, coiituins besides thlorid of sodium, the chlorids and sulphates of calcium and mag- neaium, the greater part of the lime separates as carbonate, carry- ing down with it only from one to three liuiidredths of carbo- nate of magnesia ; a portion of lime however remaining in solution as biearbonate. When the chlorid of calcium is wholly decom- posed, the magnesian salt is attacked in its turn, and there finally results a solution in whicli the whole of the earthy chlorids are replaced by chlorid of sodium. A farther addition of the solution of carbonate of soda gives them the character of alkaline-saline waters ; which moreover contain abundance of earthy carbonates. The substitution of neutral carbonate for bicarbonate of soda in the above experiment does not affect the result, except in causing a somewhat larger proportion of mngnciiia to bo thrown down with the carbonate of lime. The resulting liquid still retains large quantities of earthy carbonates in solution. ''- § 10. In the saline waters just considered, chlorids generally predominate, the sulphates being small in amount, and often altogether wanting. Some exceptions to this are however met with; for a )art from waters impregnated with gypsum, wljose origin is readily understood, there are others in which sulphate of soda or sulphate of magnesia enter largely. The soda-salt may sometimes be formed by the reaction be^een solution of gypsum and natriferous silicates referred to in § 7, or by the decompo- sition of gypsum by solution of carbonate of soda ; while in other cases its origin will probably bo found in the natural deposits of sulphates, such as glauberite, thenardite, and glauber-salt, which occur in saliferous rocks. A similar origin is probable for many of those springs in which sulphate of magnesia predominates. This salt also effloresces abundantly in a nearly pure form upon certain limestones, and is in some cases due to the action of sulphates from decomposing pyrites upon magnesian carbonate or silicate. In by far the greater number of cases, however, its appearance is unconnected with any such process ; and is, according to Mits- cherlich, due tn a reaction between dolomite and dissolved gypspai. § 20. In support of this view, it was found by the chemist just named that when a .solution of sulphate of lime was made to filter for some time through pulverized magnesian limestone, it was de- composed with the formation of carbonate of lime and sulphate of ^ff Geol. Survey of Canada, Report 1853-56, p. 4G8. ¥ 13 ajagncttia. This reaction I have been unable to verify. A solu- tion of gypsum in distilled water was made to percolate slowly through u column of several inches of finely powdered dolomite, and after ton filtrations, occupying as many days, no perceptible amount of sulphate of magnesia had been formed. Solutions of gypsum were then digested for many months with pulverized dolomiic, and also with crystalline carbonate of magnesia, but with similar negative results ; nor did the substitution of a solution of chlorid of calcium lead to the formation of any soluble magnesian salt. Solutions of gypsum were then impregnated with oarbonio acid, and allowed to remain in contact with pulverized dolomite and with magnesitc, as before, during six months of the warm season, when only inappreciable traces of magnesia were taken into solution. These experiments show that no decomposition of dissolved gypsum is effected by native carbonate of magnesia, or by the double carbonate of lime and ma^^nesia, at ordinary tem- perature. §21.1 find however that hydrated carbonate of magnesia readily and completely decomposes a solution of gypsum when agitated with it, with formation of carbonate of lime and sulphate of magnesia ; ard the same result is produced with the native hydrate of mag- nesia when mingled with a solution of gypsum in presence of oar- bonio acid. Now there maybe dolomites which contain an admix- ture of hydro-carbonate of magnesia, as there oertainly are others which like predazzite, are penetrated with hydrate of magnesia The reaction between solutions of gypsum and such magnesian limestones, (with the intervention, in the case of predazzite, of atmospheric carbonic acid,) would suffice to explain the results obtained by Mitscherlich, and the appearance in certain cases of sulphate of magnesia as an efflorescence on dolomites. In the experiments above described, the nearly pure crystalline dolomites from the Guelph and Niagara formations were made use of. § 22. When sea-water is exposed to spontaneous evaporation, the lime which it contains separates in the form of sulphate, gypsum being but sparingly soluble in a concentrated brine, and the greater portion of the chlorid of sodium crystallises out in a nearly pure state. The mother-liquor of specific gravity 1.24, having lost about four fifths of its chlorid of sodium, still contains dissolved a large proportion of sulphate of magnesia. If the evaporation is continued at the ordinary temperature, till a density of 1.32 is attained, about one half of the maguesian sulphate separates, mixed 14 \vith common salt ; and by reducing the temperature to 6** C, a large portion of pure sulphate of magnesia now orystalliies out. The farther evaporation of the remaining liquor by the heat of summer cause8 the potassium-salt to separate in the form of a hydrous doublo chlorid of potassium and magnesium, an artificial carnallite.''' "By varying somewhat the conditions of temperature, thesulphatc of magnesia and the chlorid of sodium of the mother-liquor undergo mutual decomposition, with the production of sulphate of soda and chlorid of magnesium. Hydrated sulphate of soda crystallizes out from such a mixed solution at 0° C, and by reducing the temperature to — 18" C. the greater part of the sulphates may be separated in this form from the mother-liquor of 1 .24, previously diluted with one tenth of water; without which addition a mix- ture "f hydrutcd chlorid of sodium would separate at the same time. If, on the other hand, the temperature of the mixed solu- tion be raised above SO'' C, the sulphate of soda crystallizes out in the anhydrous form, as thenardite. By the spontaneous evapora- tion during the heats of summer of the mother-liquors of density 1.35, u double sulphate of potassium and magnesium separates. These reactions are taken advantage of on a great scale in Balard's process, as modified by Merle,t for extracting salts from sea-water. § 23. The results of the evaporation of sea-water would however be widely different if an excess of lime-salt were present. In this case the whole of the sulphates present would be deposited in the form of gypsum at an early stage of the evaporation, and the mother-liquor, after the separation of the greater part of the common salt, would contain little else than the chlorids of sodium, potassium, calcium, and magnesium. * The bydroas double chlorid of potassium and magnesium (carnal- lite of H. Rose) occurs in large quantities in a stratum of clay overlying a great bed of rock-salt 100 feet thick, at Stassfurth in Prussia. It is associated with considerable quantities of sulphate of magnesia. According to Glemm, this sulphate of magnesia, to which the name of kieserile has been given, and which occurs also in Anhalt, contains but one equivalent of water, (MgO,S03-l-HO). It is not more soluble than gypsum, and unlike the ordinary sulphate of magnesia, loses the whole of its acid at a red heat in a current of steam, the acid passing off undecomposed. This salt is found in such large quantities as to be of economic importance. (Bull, Soc. Ghim. de Paris, 1864, p. 297.) t See my paper in Silliman's Journal [2] zxv. 361 ; also Report of the Jaries of the Exhibition of 1862, class ii, p. 48. n 15 § 24. A coneidoration of the conditions of the ocean in earlier geological periods m\\ show that it must have contained a much larger quantity of lime-salts than at present. The alkaline car- bonates, whose origin has been described in § 13, and which from the earliest times have been flowing into the sea, have gradually t modified the composition of its waters, separating the lime as car- cj bonate, and thus replacing the chlorid of calcium by chlorid of I sodium, as I have lor^g since pointed out.^ This reaction has § doubtless been the sovirco of all the carbonate of lime ' i the rth's crust, if we except that derived from the decomposition of ^: calcareous silicates. (§12). In this decomposition by carbonace of soda, as already described in § 18, it results from the incompati- l bility of chlorid of calcium with hydrous carbon^ie of magnesia, '- that the lime is first preoipitatad, with a little aJhcring carbonate of magnesia ; and it is only when the chlorid of calcium is all decomposed that the magnesian chlorid is transformed into car- bonate of magnesia. This latter reaction can consequently take place only in limited basins, or in portions cut off from the oceanic circulation. § 25. It follows from what has been said that the lime-salt may be eliminated from sea-water cither as sulphate or as carbonate. In the latter ease no concentration is required ; while in the former the conditions are two, — a sufficient proportion of sulphates to convert the whole of the lime into gypsum, and such a degree of concentration of the water as to lender this insoluble. These conditions meet in the evaporation of modern sea-water ; but the M evaporated sea-water of earlier periods, with its great predomi- v~ nance of lime-salts, would still contain large amounts of chlorid of calcium ; the insolubility of gypsum in this case serving to eliminate all the sulphates from the mother-liquor. Evaporation alone would not suffice to remove the whole of the lime-salts from waters in which the calcium present was more than equiv- alent to the sulphuric acid ; but the intervention of carbonate of soda would be required. § 26. In concentrated and evaporating waters freed from lime- salts by either of the reactions just mentioned, but still holding sulphate of magnesia, another process, which I have elsewhere I described, may intervene.f The addition of a solution of bicarbon- * Canadian Jonrnal for 1868, p. 202 ; Silliman's Journal [2] zxv, 102, and Comptes Rtndut, June 9, 1862, p. 1191. t StUiman's Jonrnal [2*] zzviii, 174. 16 ate oFIimo to such a solution gives rise, by double dcoompositioD, to sulphate of lime and bicarbonate of magnesia. The former being; much the \em soluble salt, especially in a strongly saline liquid, is deposited as gypsum ; and subsequently the magnesian carbonate is precipitated in a hydrous form. The effect of this reaction is to eliminate from the sea-vater both the sulphuric acid and the magnesia, without the pcrmanont addition to it of any foreign element. § 27. Gypsum may thus be separated from sea-water by two distinct processes, — the one a reaction between sulphate of magnesia and chlorid of calcium, and the other between the same sulphate and carbonate of lime. The latter, involving a separation of bicarbonate of magnesia, can, as we have seen, only take place when the whole of the chlorid of calcium has been eliminated ; and if we suppose the ancient ocean, unlike the present, to have con- tained more than an equivalent of lime for each equivalent of sulphuric acid, it is evident that a lake or basin of sea-water free from lime-salts could only have been produced by the intervention of carbonate of soda. The action of this must have eliminated the whole of the lime as carbonate, or at least have so far reduced the amount of this base that the sulphates present would be sufficient to separate the remainder by evaporation in the form of gypsum, and still leave in the mother-liquor a quantity of sul- phate of magnesia for reaction with bicarbonate of lime. The source of the magnesian carbonate, whose union, under certain conditions, with the carbonate of lime, gives rise to dolo- mite,^ may thus be due cither to the reaction just described be- tween bicarbonate of lime and solutions holding sulphate of mag- nesia, or to the direct action of carbonate of soda upon waters containing magnesian salts ; but in either ease the previous elimi- nation of the incompatible chlorid of calcium must be con- sidered an indispensable preliminary to the production of the magnesian carbonate. § 28. To the three principal sources of mineral matters in mineral waters already enumerated, viz., decaying organic matters, decomposing silicates, and the soluble saline matters in rocks, a few other minor ones must be added. One of these is the oxydation of metallic sulphurets, chiefly iron pyrites, giving rise to sulphate of * SiUiman's Journal [2] xxviii, 180-186 ; and further, Geol. Surrer of Canada, Report for 1859, 214-218. 4 17 *l|ji; 1 iron, and more rarely to sulphates of copper, cino, cobalt, and nickel ; and by secondary reactions to sulphates of alumina, lime, mag- nesia, and alkalies. This process of ozydation is necessarily super. Uoial and local, but the soluble sulphates thus formed have probably played a not unimportant part. (§ 9.) § 29. Besides these last, which contain chiefly neutral and acid salts, there is another class of waters characterized by the presence of free sulphuric or hydrochloric aciu, or both together. These acid waters sometimes occur as products of volcanic action ; during which both hydrochloric acid and sulphur are often evolved in large quan- tities. This latter element generally comes to the surface us sul- phuretted hydrogen, which by the ozydation of the hydrogen may deposit its sulphur in craters and fissures. In other cases, as shown by Dumas, the sulphur and hydrogen may be slowly and simultaneously oxydized at a low temperature, giving rise directly to sulphuric acid. Not less frequent, however, is probably the direct conversion, by combustion, of the sulphuretted hydrogen into water and sulphurous acid, which afterwards absorbing oxygen from the air is converted into sulphuric acid. § 30. The source of the hydrochloric acid and the sulphur of vol- canoes is probably the decomposition of chlorids and sulphates at high temperatures. It is known that for the decomposition of earthy chlorids, water and an elevated temperature are sufficient ; and at a higher temperature, chlorid of sodium is readily decom- posed in presence of silicious and aluminous minerals, with the intervention of water. Another agency which probably comes into play in volcanic phenomena is that of organic matters, which, reducing the sulphates to sulphurets, enable the sulphur to be subsequently disengaged as sulphuretted hydrogen by the operation of water, either with or without the intervention of carbonic acid or of silicious and argillaceous matters. Even in cases where this reducing action is excluded, the ignition of sulphates in contact with earthy matters must liberate the sulphuric acid as a mixture of sulphurous acid and oxygen ; and these uniting in their distillation upward through the strata, may give rise to springs of sulphuric acid.* To reactions similar to those just noticed, involving borates like stassfurtbite and hayesine, or boric silicates like tourmaline, etc., are to be ascribed the large amounts of boric acid which are sublimed in some volcanoes, or volatilized with the watery vapor of the Tuscan suffioni. * See the note to § 22, oa kieserite. B I II t'^ lll'i 18 § 31. Tho action of oubtorranoan heat upon buried strata con- taining sulphates and chlorids is then sufficient toexpluin thcappoar- anco of hydrochloric and sulphurous acids and sulphur, oven without the intervention of organic matters, wliich are, however, seldom or never wanting; whether as coal, lignite, bitumen, and pyro8ohists,or in a more divided condition. The presence of hydrogen and of marsh- gas, as observed by Deville among volcanic products, is an evidence of this. Tho generation of marsh-gas is, however, in most cases clearly unconnected with volcanic action or subterranean heat. To the decomposition of carbonates in buried strata by silioious matters, with the aid of heat, is to be ascribed the great amounts of carbonic acid gas which are in many places evolved from the earth, and, impregnating the infiltrating waters, give rise to acidu- lous springs. Tho principal sources of this gas in Europe are in regions adjoining volcanoes, either active or recently extinct ; but their occurrence in the paleozoic strata of the United States, far remote from any evidence of volcanic phenomena other than slightly thermal snrings, shows that an action too gentle or too deeply-seated to manifest itself in igneous eruptions, may evolve carbonic acid abundantly. Tho sulphuric acid springs of western New York and Canada, to be described further on, are not less remarkable illustrations of the same fact. § 32. The frequent presence of ammoniaoal suits in volcanic exhalations is here worthy of notice, especially when considered in connection with the rarity of niirio and ammoniacal com- pounds in natural waters, except in some local conditions, as in the wells of cities, etc., where they arc sometimes observed in comparatively large amounts. The explanation of this is evident ; for although nitrates themselves arc not directly removed from the water, they are, by the reducing action of organic matters, converted into ammonia, which is retained by the soil. In con- sequence of this affinity, the argillaceous strata, whether of the present period or of older formations, hold in a very fixed form a considerable quantity of nitrogen. This, from the slowness with which it is eliminated in the form of ammonia under the influence of alkaline solutions, probably exists as an ammoniaoal silicate. (§ 6.) The action of acids, however, as well as alkalies, may bo supposed to liberate it from its combination; and thus generate the ammoniacal salts which are such frequent accompaniments of volcanic phenomena. The numerous experiments of Delesse show that ammonia, or at least nitrogen capable of being evolved by 19 hoat and alkalieH in the form of ammonia, is prosopi in tho lime- ntones, marls, argil litea, and sandsronea of former goologioal periods, in quantities soarooly inferior to those in similar deposits of modern times, amounting, for most of tho ancient sedimentary strata, to from one to fivo thousands of nitrogen ;*" iVom which it will bo seen that tho amount of this element thus retained in tho rooky strata of tho earth's crust i.* very greut.f § 33. If wc attempt a chemical clussitioution of natural waters in accordance with tho principles laid down in tho preceding see- tions, they may bo considered under the following houds : A. Atmospheric waters. B. Waters impregnated with the soluble products of vegetable decay. C. Waters impregnated with tho salts from decomposing feld- spathic rooks, and holvling a portion of carbonate of suda as a characteristic ingredient. D. Waters holding neutral salts of sodium, calcium, or magne- sium from strata where they existed as solid salts, or as impregnating brines. E. Waters holding chiefly sulphates from decomposing pyrites; copperas and i^lum waters. F. Waters holding free sulphuric or hydroohlorio acid. § 3i. The name of mineral waters is popularly applied only to such as contain sufficient foreign matters to give them a decided taste ; and hence the waters of the divisions A and B, and many of the feebler ones of C and D, are excluded. Those of E and F have peculiar local sources ; but those of and D are often asso- ciated in adjacent geological formations, and their commingling in various proportions gives rise to mineral waters intermediate in composition. In accordance with these considerations, a classifica- tion of mineral waters for technical purposes was adopted by me in the Geology of Canada, p. 531, including only those of C, D, and F, which were arranged in six classes. I. Saline waters containing chlorid of sodium, often with large portions of chlorids of calcium and magnesium, with or * ^nn. de$ Minet [5] xviii, 151-523. t For an exposition of the views pat forward in the four preceding sections, see my paper in the Canadian Journal for 1858, p. 20tf. m n 11/ 1''^ without salphates. The carbonates of lime and magnesia are either wanting, or present only in small quantities. These waters are generally bitter to the taste, and may be designated as brines or bitterns. II. Saline waters which differ from the last in containing, besides the chlorids just mentioned, considerable quantities of carbonates of lime and magnesia. These waters generally contain much smaller proportions of earthy chlorids than the first class, and are hence less bitter to the taste. III. Saline waters which contain, besides chlorid of sodium and the carbonates of lime and magnesia, a portion of car- bonate of soda. I V . Waters which differ from the last in containing but a small proportion of chlorid of sodium, and in which the carbon- ate of soda predominates. The water.9 of this class gener- ally contain much less solid matter than the three previous classes, and have not a very marked taste until evaporated to a small volume, when they will be found, like I'.e last, to be strongly alkaline. Of these four classes, I corresponds to the division D, and IV to C, while II and III are regarded as resulting from the admixture of these in varying proportions. Sulphates are sometimes present in these waters, but never predominate ; in their absence, salts of barium and strontium are often met with. The chlorids are generally, if not always, associated with bromids and iodids. Small quantities of potassium-salts are also present, while borates, phosphates, silicates, and small portions of iron, manganese, and alumina, are generally present. These various waters are occasionally sulphurous, and those of the last three classes may be impregnated with carbonic acid. Y. The fifth class includes acid waters remarkable for containing a large proportion of free sulphuric acid, with sulphates of lime, magnesia, portions of iron, and alumina. These waters, which are characterized by their sour and styptic taste, generally contdn some sulphuretted hydrogen. VI. The sixth class includes some neutral saline waters, in which the sulphates of lime, magnesia, and the alkalies predomi- nate ; chlorids being present only in small quantities. These waters, like the last, are often impregnated with sulphuretted hydrogen. t 21 Tilt above classification, although adopted originally for the convenient description of the mineral waters of Canada, vtiVi, it is thought, be found to embrace all known classes of natural waters, with the exception of those included under E, and of some waters from volcanic sources holding muriatic acid. These may consti- tute two additional classes. In the first three of the classes above described, chlorids predominate ; in the fourth, carbonates ; and in the fifth and sixth, sulphates. The waters of the first, second, and sixth classes are neutral ; thdse of the third and fourth, alka- line ; and those of the fifth, acid. II. ast, to Analyses op Various Natural Waters. CoNTKNTS OF SiCTioNS. — 35, modo of analysis, date of collection ; 37, waters of the first class ; 37, their probable origin ; the eliminntion of sulphates ; 38, separation of lime-salts from waters : 39, earthy chlorids in saliferous formations ; brides of New York, Michigar , and England ; foot-note on errors in water-analyses ; 4Cj brines of western Pennsylvania ; waters in which chlorid of calcium predominates ; 41, origin of such waters ; separation of magnesia as an insoluble sil- icate ; 42, waters of the second class ; 43, w%*ers of the third class ; 44, waters of the fourth class ; Ghambly ; 45, other waters of the same class ; Ottawa River ; 46, waters of Highgate and Alhurg ; 47, changes in the Caledonia waters ; comparative analyses ; 48, waters of the fifth class ; sulphuric-acid springs of New York and Canada ; 49, changes in the composition of these waters ; their action on calcareous strata ; 50, waters of the sixth class, their various sources ; 51, examples of neutral sulphated waters ; sulphate of magnesia waters. § 35. The analyses of the various mineral waters to be given in the second part of the present paper, were made according to the modes laid down in the treatise of Fresenius on Quantitative Analy- sis. The carbonate of soda in the alkaline waters was determined by the exuess of the alkaline bases over the chlorine and sulphuric acid present. This was generally controlled by the amount of the carbonate of baryta thrown down from a solution of chlorid of bariui:* by a solution of the soluble salts obtained by the evapo- ration of the mineral water ; and in some cases, to be specified farther on, this latter process was relied on as the only means • ■, ' -I' 22 l!ii; m ' il of determining the amount of carbonate of soda. For some remarks on the earthy carbonates of the waters, and on their relation to the results of analysis, see part III of this paper. The date at which the various waters were collected for analysis is in each case appended to the notice of the spring. This is of the greater importance, inasmuch as it will be shown that in the course of years, some of those springs here described have suffered con- siderable changes in their composition. § 36. In the following table are given the analyses of several waters belonging to the first class, as defined in § Si.''' 1. — This water is from a well thirty feet in depth, near the village of Ancaster, on the western shore of Lake Ontario. It is sunk in the Niagara formation ; but like the other waters of this class, probably has its source in the Lower Silurian limestones. The water rises nearly to the surface, but there is no perceptible dis- charge. Its temperature was found to be 48" F. when collected for analysis in September 1847. 2. This water is from a copious spring which issues from the limestones of the Trenton group at Whitby, on the north shore of Lake Ontario. It contained small portions of baryta and strontia, and was collected in October 1853. 3, 4. Several wells have been sunk in the Trenton limestone in the township of Uallowell, on the Bay of Quints, Lake On- tario, in search of brine for salt-making, and have yielded bitter saline waters, of which the two here noticed are examples. No. 3 was obtained from a well twenty-seven feet deep, in October 1853. No. 4 was taken in the summer of 1854 from a well a mile or two distant from the last. Neither of these waters was examined for baryta or strontia. 5, 6. At St. Catherines, near Niagara Falls, a boring of five inches in diameter was carried to a depth of about 500 feet, and after traversing the Medina formation, is said to have penetrated fifty or sixty feet into the Hudson River shales. It yields about twenty gallons a minute of a saline water, whose analy- sis by Professor Croft of the University of Toronto, a few years since, afforded the results given under 5. This water, which was u * Of the thirty-seven analyses of waters here given, ten have already appeared in Silliman's Journal [2] viii, ix, xi, bat for the purposes of comparison it is thought well to reproduce them in the present con- nection. Of tht others, the greater part have appeared in the Geology of Canada, but several are now for the first time in print. -m CO < n H fa O ta H I I— I 00 7.227 undet. 2.102 1.763 undet. 2 2.388 .400 Ok CO • CO 1 rt ^ *j 2 <^ *: •t © 1» « « P? OT ii lO 1- 00 lA cJ t- 29.8 und 12.4 7. und 0> CO 11 1 19.94 ndet. 6.49 1.95 ndet. • ■ . •A m © p • CO 8034 3555 8544 3977 adet. e . § O i-t o> ■* CO s^* e eq o 50 ^H ■* ® *- OS o> o> ■* . f4 . . . <= 00 ♦' in e^ 00 CO »H f1 «0 00 m lO 1>. M 00 ii ^- . X M 00 © rH M q> CO ^ O CO •<* T»i o 'I' CQ tS o 00 -^ t- OS © o © a CD r-t '(f © O *- t- 00 i- pH 00 e>i es» CO t- CC V , Oi C *- l-l 1-i u a o 00 c 00 o •-• »• OS • ■ 5 CO ** p «fl • CO • eS • a * • • m • •a c3 a E '5 00 CS e '5 i . 1 E ' 1 - J i o to 5 5 t ' -a O 13 •♦J c3 1 a © s S 1 2 o •g 1 o o o I-t a 24 m I first sought for the manufacture of salt, is now much used for medicinal purposes. Its strength seems subject to some variation, since a specimen from the same well in December 1861 gave me, by a partial analysis, chlo^'id of sodium 23.00, chlorid of cal- cium 9.66, chlorid of magnesium 2.40, sulphate of lime 1.75; = 36.81 parts in 1000. No. 6, examined at the same time, is from a second well sunk in 1861, not far from the last. 7, 8. — These are analyses of the waters from two borings in the Trenton limestone at Morton's distillery in Kingston. The analyses are by Dr. Williamson of Queen's College in that city, and were made probably ten or twelve years since. They have been rocalculated bo as to represent the whole of the sulphuric acid as combined with calcium. The first of these waters gave to Dr. Williamson both bromine and iodine, and the second was found to be sulphurous. These waters differ from the preceding in containing considerable amounts of earthy carbonates, and in this refipect are related to those of the second class, while they still show a large predominance of earthy chlorids. . § 37. The waters of the above table contain, besides chlorid of •odium and a little chlorid of potassium, 'large quantities of the chlorids of calcium and magnesium, amounting together, in several cases, to more than one half the solid contents of the water. Sul- phates are either absent, or occur only in small quantities, and the same is true of earthy carbonates. Salts of baryta and strontia are sometimes present, while the proportions of bromids and iodids, though variable, are often considerable. In the large amount of magnesian chlorid which they contain, these waters resemble the bittern or mother-liquor which remains after the greater part of the chlorid of sodium has been removed from sea-water by evaporation. The bitterns from modern seas, however differ in the presence of sulphates, and in containing, when sufficiently concentrated, only traces of lime. The reason of this, as already pointed out in § 22, is to be found in the fact that in the waters of the present ocean the sulphates are much more than equivalent to the lime, so that this base separates during evaporation as gypsum.^ But as shown in § 23 and § 24, the waters of the ancient seas, which held in the form of chlorid of calcium the greater part of the lime since deposited as carbonate, must have yielded by evaporation bitterns containing a large proportion of chlorid of calcium. Such is the nature of the * See farther on this point, Bischof, Gbem. Geology, i, 413. 25 brines whose analyses arc given in the above table, and such we suppose to have been their origin. The complete absence of sul- phates from many of these waters points to the separation of large quantities of earthy sulphates in the Lower Silurian strata from which these saline springs issue ; and the presence in many of the dolomitic beds of the Calciferous sand-rock of abundantly dissemina- ted small masses of gypsum, is an evidence of the elimination of the sulphates by evaporation. The frequent occurrence of crys- talline masses of sulphate of strontian in the Chazy and 'Black River limestones of this region, is also to be noted as another means by which the sulphates were separated from the waters of the Lower Silurian seas. From the proportions of chlorid of sodium, varying from about one third to more than two thirds of the solid contents of the above waters, it is apparent that in most cases the process of evaporation had gor^ so far as to separate a part of the common salt ; and thus successive strata oi' this ancient saliferous formation must be impregnated with solid or dissolved salts of unlike composition. The mingling of these in varying proportions aflFords the only apparent explanation of the diiFerences in the relative amounts of the several chlorids in waters from the same region, and even from adjacent sources. These dififerences are seen on comparing the waters from the different wells of St. Catherines, Hallowell and Kingston, with each- other. § 38. The great solubility of chlorid of calcium renders it diffi- cult to suppose its separation from the inother-liquors so as to be deposited in a solid state in the strata. The same remark applies to chlorid of magnesium. It is however to be remarked that the double chlorid of potassium and magnesium (carnallite) is decom- posed by deliquesence into solid chlorid of potassium and a solu- tion of chlorid of magnesium ; and thus strata like those which at Stassfurtb contain large quantities of carnallite (§ 22), might give rise to solutions of magnesian chlorid. This however would require the presence of a large amount of chlorid of potassium in the early seas. It will be observed by referring to the analyses above given, that the chlorid of magnesium sometimes surpasses in amount the chlorid of calcium ; and sometimes, on the contrary, is equal to only one half or one fourth of the latter salt. While it is not impossible that the predominance of the magnesian chlorid in some waters may be traced to the decomposition of carnallite, it is undoubtedly in most cases connected with the action of solutions of carbonate of soda ; the effect of which, as already pointed out, is to first 20 r3 I Hi t'M m .,. IS- separate the soluble liiiio salt an carbonate, leavin<^ to a subsequent stage the magnesian chlorid (^ 18.) As this reaction replaces the calcium-salt by chlorid of sodium, it might be expected that there would be an increase in the amount of the latter salt in the water wherever the magnesian chlorid predominates, did we not remember that evaporation separates it from the water in u solid form ; and that the two processes, one of which replaces the chlorid of calcium by chlorid of sodium, while the other eliminateii the latter salt from the solution, might have been going on simul- taneously or alternately. As the nature of the waters now under consideration shows that the process of evaporation had been carried so far as to separate the sulphate in the form of gypsum, and pro- bably also a portion of the chlorid of sodium in a solid state, it is evident that we have not yet the data necessary for determining the composition of the water of the Lower Silurian ocean, as re- gards the proportions of the sodium, calcium, and magnesium which it held in solution ; and we can only conclude from these mother- liquors, that the amount of the earthy bases was relatively very large. § 39. As already remarked in § 22, the mother-liquor from modern sea-water contains no chlorid of calcium, but, on the con- trary, large quantities of sulphate of magnesia ; the lime in the mcdcrn ocean being less than one-half that required to combine with th 3. sulphate present. If however we examine the numerous analyses of rock-salt and of brines from various saliferous forma- tions, we shall find that chlorid of calcium is very frequently present in both of them ; thus supporting the conclusions already announced in § 24 with regard to the composition of the seas of former geological periods. The oldest saliferous formation which has been hitherto investigated is the Onondaga Salt-group of the New York geologists, which belongs to the upper part of the Silu- rian series, and supplies the almost saturated brines of Syracuse and Salina in New York. These, notwithstanding their great purity, contain small proportions of chlorids of calcium and magne- sium, as shown by the analyses of Beck, and the recent and careful examinations of Goessmann. In the brines of that region the solid matt«rs are equal to from 14.3 to 16.7 per cent., and contain on an average, according to the latter chemist, 1.54 of sulphate of lime, 0.93 of chlorid of calcium, and 0.88 of chlorid of magnesium in 100.00; the remainder being chlorid of sodium.* * Goessmann. Report on the Brines of Onondaga : Syracuse, 1862 and 1864. Also Report on the Onondaga Salt Go. : Syracuse, 1862. m The nearly saturated brines from the Saginaw valley in Michi- gan, which have their source at the base of the Carboniferous series, contain, according to my calculation from an analysis by Prof Dubois, in 100.00 parts of solid matters: chloridof calcium 9.81, obloridof magnesium 7.61, sulphate of lime 2.20, the remain- der being chiefly ohlorid of sodium. Another brine in the same vicinity gave to Chilton an amount of chlorid of calcium equal to 3.76 per cent.* In a specimen of salt manufactured in this region, Ooessmann found 1.09 of chlorid of calcium ; and in two specimens of Ohio salt, 0.6^ and 1.43 per cent of the same chlorid. The rock-salt from the Lias of Cheshire, according to Niool, contains small cavities, partly filled with air, and partly with a concentrated solution of chlorid of magnesium, with some chlorid of calcium. f * Winchell ; SillimaD's Journal [2] xxxiv, 311. t Cited by Bischof, LeLrbuch, ii, 1671. The results of the analyses by Mr. Northcote of the brines of Droitwich and Stoke in the same regioa (L. E. & D. Philos. Mag. [4J ix, 32), as calculated by him, shaw^ no earthy chlorids whatever, and no carbonate of lime, but carbonates of soda and magnesia, and sulphates of soda and lime. He regarded the whole of the lime present in the water as being in the form of sulphate. If however we replace in calculating these analyses, the carbonate of soda and sulphate of lime by sulphate of soda and carbonate of lime, we shall have for the contents of these brines, chlorid of sodium, with notable quantities of sulphate of soda, some sulphate of lime, and car- bonates both of lime and magnesia ; a composition which is more in accordance with the admitted laws of chemical combinations. From these results, it would appear that the earthy chlorids, which according to Nicbol are present in the rock-salt of this formation, are decomposed by sulphates in the waters which, by dissolving it, give rise to the brines. It is to be regretted that in many water-analyses by chemists of note, the results are so calculated as to represent the co-existence of incompa- tible salts. Of the association of carbonates of soda and magnesia with sulphate of lime, as in the analysis just noted, it might be said that I have shown tliat it may occur in the presence of an excess of carbonic acid. (Silliman's Jour. [2] xxviii, 174). By evaporation, however, such solutions regenerate carbonate of lime and sulphates of soda and mag- nesia ; and by the consent of the best chemists these elements are to be represented as thus combined. But what shall be said when chlorid of magnesium, carbonate of soda, and silicate of soda are given as the con- stituents of a water whose recent analysis may be found in a late number of the Chemical News ; or when bi-carbonatcs of soda, magnesia, and lime are represented as co-existing in a water with sulphates and chlorids of magnesium and aluminum ? These errors probably arise from 28 Iri W § 40. Tho brines from the valley of the Alleghany River, obtained i'roni borings in the .Coal formation, are remarkable for contain- ing large proportions of chlorids of oaloium and magnesium ; though the sum of these, according to the analyses of Lenny, is never equal to more than about one fourth of the chlorid of sodium. Tho presence of salts of barium and strontium in these brines, and tho consequent absence of sulphates, is, according to Lenny, a constant character in this region over an area of two thousand square miles. (See Bischof, Cbem. Gcol., i, 377.) A later analysis of another one of these waters from the same region, by Steiner, is cited by Will and Kopp, Jahresbericht, 1 861 , p. 1 1 1 2. His results agree closely with those of Lenny. See also the analysis of a bittern from this region by Boy^ (Silliman's Journal [2] vii, 74). These remarkable waters approach in character to those of Whitby and Hallo (Vell ; but in these the chlorid of sodium forms only about one half the solid contents, and the proportion of the chlorid of magnesium to the chlorid of calcium is relatively much greater than in the waters from western Pennsylvania, where the magnesian chlorid is equal only to from one third to one fifth of the chlorid of calcium ; the proportions of the two being subject in both regions to considerable variations. In this connection may be cited a water from Bras d'Or, in the island of Cape Breton, lately analyzed by Prof. How, which con- tains in 1000 parts, chlorid of sodium 4*901, chlorid of potassium 0'650, chlorid of calcium 4'413, and chlorid of magnesium only 0-638, besides sulphate of lime 0*134, carbonates of lime and mag- nesia 0*085, with traces of iron-oxyd and phosphates; = 10*821. (Canadian Naturalist, viii, 370.) The analysas of European waters furnish comparatively few examples of the predominance of earthy chlorids.* determining in the recent water, or in water not suflBciently boiled, the lime and magnesia which would by prolonged ebullition be separated as carbonates, together with portions of alumina, silica, etc. In the subsequent calculation of the analyses, these dissolved earthy bases being regarded as sulphates or chlorids, instead of carbonates, there remains an excess of soda, which is wrongly represented as carbonate, instead of chlorid, or sulphate of sodium. * Lersch, Hydro-Chemie, Zweite Auflage : Berlin, 1864 ; vide p. 207. This excellent work, which is a treatise on the chemistry of natural waters, in one volume 8vo. of 700 pages, was unknown to me when I prepared the first part of this essay. the § 41. We have already shown in § 38 how the action of carbonate of Boda upon sea-water or bittern will destroy the normal propor- tion between the two ohlorids of magnesium and calcium by con- verting the latter into an insoluble carbonate, and leaving at last only salts of sodium and magnesium in solution. A process the reverse of this has evidently intervened for the production of waters like that from Cape Breton, and some others noticed by Lersoh, in which chlorid of calcium abounds, with little or no sulphate or chlorid of magnesium. This process is probably one connected with the formation of a silicate of magnesia. Bischof has already , inttsted upon the sparing solubility of this silicate ; and he observed that silicates of alumina, both artificial and natural, whea digested with a solution of magnesian chlorid, exchange a portion of their base for magnesia, thus giving rise to solutions of alumina ; which, being decomposed by carbonates, may have been the source of many of the aluminous deposits referred to in § 9. He also observed a similar decomposition between a solution of an artificial silicate of lime and soluble magnesian salts, (Bischof, Ohem. Geology, i, 13, also chap, zxiv.) In xepeating and extending his experiments, I have confirmed his observation that a solution of silicate of lime precipitates silicate of magnesia from the sulphate and the chlorid of magnesium ; and have moreover found that by digestion at ordinary temperatures with an excess of freshly precipitated silicate of lime, chlorid of magnesium is com- pletely decomposed ; an insoluble silicate of magnesia being formed, while nothing but chlorid of calcium remains in solution. It is clear that the greater insolubility of the magnesian silicate, as com- pared with silicate of lime, determines a result the very reverse of that produced by carbonates with solutions of the two earthy b ises. In the one case the lime is separated as carbonate, the magnesia remaining in solution ; while in the other by the action of silicate of soda (or of lime), the magnesia is removed and the lime remains. Hence carbonate of lime and silicate of magnesia are everywhere found in nature ; while carbonate of magnesia and silicate of lime are produced only under local and exceptional conditions. The detailed results of some experiments on 4his subject are reserved for another place. It is evident that the production from the waters of the early seas of beds of sepiolite, talc, serpentine, and other rocks in which a magnesian silicate abounds, must, in closed basins, have given rise to waters in which chlorid of calcium would predominate. ... 30 I . I m § 42. Of the waters of tho sooond olass whose analytics are hero given, the first three occur, with many others of similar character, on the south side of the Ottawa river, below the city of that name, The remaining four are on the north side of the Si. Lawrence, between Montreal and Quebec, where also similar waters abound. All of these springs rise from the Lower Silurian limestones of the region. 1,2. These two waters arc from the township of Plantagenet. The first is known as Larocque's, and the second as the Georgian spring. These waters were examined in 1849 and 1851. Two other springs have been observed in tho same vicinity, one resembling Larocque's spring and containing borates, with a notable pro- portion of strontia, while the other is an alkaline-saline water of the third class. 3. Caledonia Intermittent Spring. This spring owes its name to the intermitting dischufL^o of oarburetted hydrogen which takes place from its waters. It ia in the township of Caledonia, not far from Plantagenet, and near three other waters from the same township, to be mentioned in the next class. The water was collected in September 1847. 4. Lanoraie. This is from the seigniory of Lanoraie. It contains both baryta and strontia, and evolves an abundance of carburetted hydrogen. The water was collected in March 1851. 5. Is from a copious spring in the seigniory of Berthier, and was collected in July 1853. 6. Is from the township of Caxton, and yields six or eight gal- lons of water a minute, besides a great abundance of inflammable gas. The carbonic acid was found to equal 1.126 parts, of which .651, or more than one half is required for the neutral carbonates present. The water was taken from the spring in October 1848. 7. Is from the seigniory of St. L^on, and is a copious spring which, like the last, disengages iuflummable gas. The carbonic acid was equal to 1.224 parts, of which .651, or not quite one half is required for the neutral carbonates found by analysis. The water was collected in October 1848. 8. 9. These are from two springs in the parish of Ste. Genevieve on the Batiscan River, and are remarkable for the largo proportion of iodids which they contain. The first is known as Trudel's spring, and the second is at the ferry opposite to the church. The waters were collected in August 1853. Several other saline springs occur in the same neighborhood. m in ■< u o o u n H O eg H 1 S! *i • V * CO n ■* 00 e 00 s S § 0> o "O w o> C'l ^H "^l ff »o \\ ««• 1 m a ■ M 00 o o © "^ a ^14 o : ' ■ • ' ♦^ a CO ; ■^ 9l 00 ^4 v>4 CO t» M e (0 m a h- ' t- © n CI 00 CO M © Ci 00 1 00 co *•• • o « •o -^ f-t in ^ a M M CD © © o © t- ?! o> *- * * • M * * • • ♦rf a o 1 ^^ c< ! s M e> « 00 (0 vrf CO • CO 00 m !• m m CO CO — 1— » ^H M A ■«»< • ot 00 •f CO •"f d 1 *- Oi 00 o © »- CO o © • f CO ^^ 00 ^H CO • ^ -« o © © CO © © • fO a © © 9 * p-l 1— 1 • CO 9 ■^ PM p^ o © « m w 0> © CO ^ 9i 9 CO CO 1 ■A © © ^ «»• M • CO (A lO «- in CO (0 »- 00 lO t-i CO o • •-« ■n o f c •n f o o • n • © • o • *? o • © 9 CO 9 i-H w^ • p-H CO 9 f^ " i-l ■"C • , CO CO « o t ^ >n V * CO in V fl t- « ^ \n •^ »o ■* QO •o ■>*< CO CO o a o o a © 00 9 00 3 ; *rf a> o © M in © o CO M CO t- o CI n e^i © CI o CO © 00 N A 00 in o CO M e>» m CO «»< ■^ ■^ ■^ CO r^ ■»f r- M © PM F-t m CO m -a a p 00 1-1 .-1 O o M M 9 © o o ■* »*< ^ 9 00 a> • • • • • • • ■ 9 ^^ C4 9 1 1-1 fm »^ o i« © 00 00 ^^ 1" CI no in m CO a o o l- m CO ej to CO CI 0, A CO in M 00 CO N © c» CO 5 c» CO 9 e^ o M © © o 00 © ? CO • • 9 e» • fH •* l-H i-i PM o © n N o> t- Ci © A m in : •* 00 o n" Ttl •^ e>» •-H (M 00 e © o Ci 01 CO ?! CI T" ) oc • •«*< o •* o o p-l CI CO 9 1 ^ 00 • • ■ • • • 4^ • c: ' o 9 9 1— < o o • I* M © CI • o •* CO 9 J 00 Oi CO ^ • • CO lO ou in • • • CO o OS O i t~ CO i-H CO o ■ m •^ o o • ■ • CO 0> o t- i « CO •—1 • • l-H N o © ■ • • © 00 o 9 ' •H 0> • ■ • • • ■ • ■ • • • • • • • 9 ^A • • • • • « 9 ^H PM 1-4 • • ^ i i • • • • • 08 to 1 .2 S 1 1 .2 a 2 a s m 0) a a - = 4> s d 2 CO 1 a a d 2 t' O O OS 03 f^ Cm ' !«• Ol P< .a 10 u tn . . o :'i o « '." ^» 'X' -»• 11 • • T Tt o rt >o 41 •t '« ft M o § : . M m^ © f lO ,; ifS on -r o o . M O o M M ?. © lO cc -K • • O 'fi o « M 00 f) Ifl ri M M Ol 11 • o ~» »»< -* CO -f (U 'D f-' CI iM F^ O . • I'- •M ^-* i"5 »c o -f ■M "* F-< © o • • © o CO l.'^ © a © t- «•' O CO in o t- IM -»• ••O '» o 1^ '0 o o CO ci ■■o ^» M • M at) ij © o 01 -r fo r- CO m^ .~ • *^ OfJ o 10 (-) CO •^ lO r-4 o o • © i.O t- ? o •— < CO <0 00 © © © r/l fO 00 © o 'ft f:> © r- CO CO o a? o o 1< a (Tl ■^ <-> -f iM I..I rt r-^ (^ ..^ « /^ 00 •^ 00 © o © — t" • M CI ','*. o © OJ CO • •* o «> 05 -t* ro IM >o •M m . »fj <-< o a> (O ^H "i< '■O r- l- ?! n »- 'f i?» fM o © t~ f-4 .^ T1 »t< -f -J< © O o © f— • f-^ m i*i a © CO o -^ a J- © © © © © CO o o in o CO CS O «0 CO © lO © irs CO -< o © © o © © O O a a 3 o a o o 00 © © M ro r»t W '■* Ji ■* "^ T1 S © -< 'o 2 © ■rj O o '« ^ o • .2 m a s i 2 ;= a .is w »::i o OS 2 -5. o .a fl a. o r/l .O O >.< to 0^ u a i 2 i U3 COi CO I; 34 11 m ll!) hi The water was collected in September 1852. Several other mineral springs occur in this vicinity, one of them belonging to this class, and others to the second and fourth classes. 8. This water, from the seigniory of Belocil, was collected in 1851. § 44. We shall now proceed to the springs wiiieh, in ^ 34, have been referred to the fourth class — and begin with three analyses of a mineral water from Chambly. Here, on a plateau, over an area of about two acres, the clayey soil is destitute of vegetation and impregnated with alkaline waters ; which in the dry season give rise to a saline effloresence on the partially-dried up and fissured surface. A well sunk her«j to a depth of eight or ten feet in the clay, which overlies the Hudson River formation, affords at all times an abundant supply of water, which generally flows in a small stream from the top of the well. Small bubbles of carbu retted hydrogen are sometimes seen to escape from the water. The temperature at the bottom of the well was found in October ISfil to be 53° F., and in August 18G5to b(! nearly 51° F. The mean temperature of Chambly can differ but little from that of Mon- treal, which is 44''.6 F., so that this is a thermal water. Another alkaline and saline spring in the same parish has also a temper- ature of 53° F. The water of the spring here described has a sweetish saline taste, and is much relished by the cattle of the neighborhood. Three analyses have been made of its waters, the results of which are here given side by side. The first was collected in October 1851 ; the second in October 1852 ; and the third in August 18G4, during a very dry season. I. II. 111. Chlor id of potassium, undet. -0324 -0182 " sodium, -8689 -8387 -8846 Carbonate " 1.0295 1.0604 -9820 " lime, -0540 -0380 0253 " magnesia -0908 .0765 -0650 " strontia, undet. -0045 undet. " iron, " 0024 " Alumina and phosphate, " 0003 " Silica, -1220 -0730 -OIGO Borates, iodids and bromids, . . undet. undet. undet. In 1000 parts 2.1652 2.1322 1.9917 A portion of barium is included with the strontium salt. The water contains mcreover a portion of an organic acid, which causes it to assume a bright brown color when reduced by evapo- ration. Acetic ucid gave no precipitate with the concentrated and filtered water ; but the subsequent addition of acetate of cop- per yielded a brown precipitate of what was regarded as apocrenate of copper. The organic matter of this and of many other mineral springs has probably a superficial origin. The carbonic acid was determined in the third analysis, and was equal in two trials to .903 and .fl05. The neutral car))onates in this water require .452 parts of carbonic acid. § 45. In the following table are given the analyses of several other waters which belong like the last to the fourth class. TABLE IV.— WATERS OF THE FOURTH CLASS. 1 2 3 4 5 Chlorid sodium " potassium... Sul^i.iate soda " potassium.. Carbonate soda " lime " magnesia. Iron, alumina, pbos.. Silica .0207 .0490 • • • • .0081 .1340 .1740 .1287 traces. .0161 .0347 .0076 traces. • • • • .1952 .0710 .0278 .0110 .3818 .0067 .0215 • • • • .2301 .0020 .0257 .0245 .3920 .0318 traces. 1.1353 undet. II II II .0169 .0188 .0122 .0410 .2480 .0690 traces. .2060 In 1000 parts i iln 10,000 parts I .5311 .3473 .'523 1.5591 ■ • • • .... ... .6116 i 1 . This spring was met with some years since in constructing a lock on the Richelieu River at St. Ours, and was enclosed in such a way that it is only accessible through a pump ; so that it is impossible to determine the amount of water furnished by the spring, or its freedom from admixture. The water was obtained in November 1852, and is remarkable for the large pro- portion of potassium salts. 1000 parts of the water gave of alkalies determined as chlorids, 0.2250; of which 0.0565 parts, or 25.11 per cent, were chlorid of potassium. Another trial gave 24.52 per cent, ; while a portion of the water taken from the spring three m I 'I n i weeks carlic- yavc a larjicr proportion of alkalies, o(jual to 0.3400 ofchlorids; of whicli 0.0r»0<">. or 17.r).'; per foiil, wore clilorid of potassium. 2. Tills spriiic; occurs on tlio bank (>rtlic Jacques Cartler River, a little above Quebec. It is strongly inijiregnatcd with sulphuretted hydrogen, and appears to contain a considerable proportion of borates. It was collected for analyses in the summer of 1852. o. This water is from a sprinu' in the township of Joly, on the opposite side of the 8t. Lawrence, a few miles south from the last, and lik ' it is sulphurous, and affords a strong reaction of boric acid. It was collected for analysis in July 1853. "^ 4. A small area of marshy ground in the seigniory of Nicolet. near the line of St. Grcgoire, is, like the similar tract in Chambly. so impregnated with mineral water as to be destitute of vegeta- tion. The water collected in a small pit dug in this locality in the autumn of 1853, was yellowish colored, and alkaline to the taste, and gave by analysis the above results. Several other alkaline springs occur in this vicinitj'. All of the preceding waters, with the exception of No. 2, which comes from out the Utica slates, rise, like that of Chambly, from the Hudson River formation. 5. This water, unlike the preceding, is that of a large river, the Ottawa, which drains a region occupied chiefly by ancient crystalline rocks, covered by extensive forests and marshes. The soluble matters whirh it contains arc therefore derived in part from the superficial decomposition of these rocks, and in part from the decaying vegetation. The water which was taken at the head of the St. Anne's rapids, on the 9th of March 1854, before the melting of the winter's snows had begun, had a pale amber-yellow color from dissolved organic matter, which gave a dark brown hue to the residue after evaporation. The weight of this residue from 10,000 parts, dried at 300° F., was .6975, which after ignition was reduced to .5340 parts. As seen in the above table, one half of the solid matters in this water were earthy carbonates, and more than one third was silica, so that the whole amount of salts of alkaline bases was .088 (of which nearly one half is carbonate of soda) ; while the St. Ours water, which resembles that of the Ottawa in its alkaline salts, contains in the same quantity 4.248, or more than forty-eight times as much. The alkalies of the Ottawa water equalled as chlorids, -0900, of which .0293, or 32.5 per cent, were ohlorid of potassium. The results of some obser- 37 vations on the silica and organic matters of this river water will be given in part HI. It will Ix; observed that in the above table the figures given for the iirst five waters arc for 1000 p;trt.s. while those of the Ottawa are for 10,000 parts. i^ 4(5. In this connection may be given the aualy.se.> of two similar springs from Vermont, the llighgate and Alburg springs. The waters were sent me in 0(!tober and November 18G1, and the results have already appeared in the (Jcoloiji/ of Vcmiont, ii, 92t). Both of these waters, when examined, were slightly sulphurous, and yielded tlic reactions of boric acid. The amount of carbonate of soda was estimated from the carbonate of baryta obtained by the process already mentioned in § 35. lligligate. Chlorid of sodium 402 , Sulphate of soda 0412 . rarbonato of soda 235 . •' lime 024 . 010 I'Dtagh and borates iindet. magnesia, .71.". Alburg. .140 .024 .230 .03C .022 undet. •452 In 1000 parts § 47. On the 5th January 1H65, after a lapse of more than seventeen years, I again visited the three springs of Caledonia whose analyses have been given in the table § 43, and collected. their waters for a second examination. The results of my recent analyses show that considerable changes have occurred in the composition of each of these springs, and tend to confirm in an unexpected manner, the theory which I long since put forward, — that the r 'cr.-^ of the second and third classes owe their origin to the mingii'v u!' saline waters of the first class with alkaline waters t>f liic fo' rth class. It will be observed that the three Caledonia wa ;is in 1847 were all alkaline, though the propor- tions of the carbonate of soda were unlike. Sulphates were also present in all of them, though most abundant in the Sulphur spring, which, although holding the smallest amount of solid mat- ters, was the most alkaline. In January 1865, however, the first and second of these waters had ceased to be alkaline, and con- tained, instead of carbonate of soda, small quantities of earthy chloril causing them to enter into the second class. They no longer < . ntained any sulphates, but, on the contrary, portions of baryta und strontia. Only the Sulphur spring, which in 1847 contained the largest proportion of carbonate of soda and of sul- 38 lit I it P': phatcs, still retained those elements, though in diminished amount.s, and was feebly impregnated with sulphuretted hydrogen. If we sui)posc these wfitors to arise from the commingling of saline waters like those of Whithy and Lanoraie, containing earthy chlorids and salts of baryta and strontia, with waters of the fourth class hold- ing carbonate and sulphate of soda, it is evident that a sufficient quantity of the latter water would decompose the earthy chlorids and precipitate the salts of baryta and strontia present ; while an excess would give use to alkaline-saline waters containing sul- phate and carbonate of soda, such as were the three springs of Caledonia in 1847. A falUng-off in the supply of the sulphated alkaline water has however taken place, and the result is seen in the appearance of chloi, ; ' -r ?gnesiura and of baryta and strontia in two of the springs, and . Jiminishcd proportion of carbonate of soda in the Sulphur spriiij^. These later analyses being directed chiefly to the determination of these changes, no attempt was made to determine the potassium, iodine, bromine. For the purposes of comparison, the two series of analyses are here put in juxtaposition ; the elements just men- tioned being included with the chlorid of sodium, and the figures reduced to three places of decimals. The precipitate by ? solution of gypsum from the concentrated and acidulated water was re- garded as sulphate of strontia, and calculated as such, but was in part sulphate of baryta. TABLE V.—SHOWING THE CHANGES IN THK CALEDOxVIA SPRINGS. Chlor. sodium... " magnesium Sulph. potash .... ijCarb. soda j " lime i " magnesia.. 1 " strontia.... ' Silica 1. Gas Spriug. 3. Saline Spring. 8. Sulphur Spring. 1847, 1865. 1847. 1865. 1846. 1865. ' 7.014 .005 .048 .148 .526 .021 7.772 6.570 .024 .096 . 45r> .009 .020 6.488 • • • • .005 .176 .117 ! .517 • • • ■ .042 7.345 6.930 .026 1 .095 .469 j .012 .015 3.876 .018 .456 .210 ; .294 .084 3.685 1 .021 .091 .077 j .228 ( • • • • .021 In 1000 parts.... 7.174 7.547 4.938 4.123 i 39 .021 .091 .077 .228 • ■ • • .021 In the recent analyses of these waters, the carbonic acid in the Gas spring was found to etjual for 1,000 parts, 671 ; of which .278 were required for the neutral carbonates. The Saline spring con- tained .664 of carbonic acid ; of which .290 go to make up the neutral carbonates. The Sulphur spring, in like manner, gave of carbonic acid .573 ; while the neutral carbonates oftho water require only .191 . All of these waters, in January 1805, thus contained an excess of carbonic aeid above that required to form bicarbonates with the carbonated ba.ses present ; while the analy.ses of the same springs in 1847, showed, as we have seen in § 43, a quantity of car- bonic acid insufficient for the formation of bicarbonates. The questions of this deflciency, and of the variation in the amount of carbonic acid in these and other watcn-s, will be considered in the third part of this paper. § 48, The waters of our fifth and sixth eliissus, tu dolincd in j^ 34, are distinguished by the prcjsunee of sulphattis; the former being aeid, and the latter being neutral waters. In the fifth class the principal element is sulphuric acid, associated with variable and accidental amounts of sulphates of alkalies, lime, magnesia, alumina, and iron. Apart from the springs of this kind which occur in re- gions where volcanic agencies are evidently active, tlie only ones hitherto studied are those of New York and western Canada ; which issue from unaltered, and almost horizontal Upper Silurian rocks. (§ 31.) The first account of these remarkable waters was jiivon in Silliman's Journal in 1829 (vol. xv, p. 238), by the late Prof- Eaton, who described two acid springs in Byron, Genessee Co. N. Y. ; one yielding a stream of distinctly acid water sufficient to turn a mill-wheel, and the other affi)rding in smaller quantities a much more acid water. The latter was afterwards examined by Dr. Lewis Beck (Mineralogy of New York, p. 150). He found it to be colorless, transparent,and intensely acid, with a specific gravity of 1.113; which corresponds to a solution holding seventeen per cent of oil of vitriol. Xo chlorids, and only traces of lime and iron, were found in this water, which was nearly pure dilute sulphuric acid. Prof. Hall (Geology of New York, 4th District, p. 134) has noticed, in addition to these, several other springs and wells of acid water in the adjacent town of Bergen. Farther westward, in the town of Alabama, is a similar water, whose analysis by Erni and Craw will be found in Silliman's Jounial [2] ix, 450. It contained in 1000 parts about 2.5 of sulphuric acid, and 4.G parts of sul- phates, chiefly of lime, magnesia, iron anO alumina. In this, as in I'M 40 the succeeding aualyscK, hydratod sulphuric acid, SO^HO, if meant. The earliest quantitative analyses of any of ...zaa waters were those by Croft anl myself of a spring at Tuscarora, in 1345 and 1847, of which the detailed results appear in Sillitnan's Journal [2] viii, 3G4. This, at tlic time of my analysis in Sep- tember 1847, contained in lOOO parts, 4.29 of sulphuric acid, and only 1.87 of sulphtitcs; while the previous analysis by Prof. Croft gave approxlmatively 3.00 of neutral sulphates, and only about 1.37 of sulphuric acid. Similar acid waters occur on Grand Island above Niagara Falls, and at Chippawa. All of these springs, along a line of more than 100 miles from east to west, rise from the outcrop of the Onondaga salt-group; but in the township of Niagara, not far from Queenston, are two similar waters which issue from the Medina sandstone. One of these is iu the southwest part of the township, and fills a small basin in yellow clay, which, at a depth of three or lour fi'ot, is underlaid by red and green sandstones. The water, which, liki; those of Tuscarora and Chippawa, is slightly impregnated with sul- phuretted hydrogen, is kept iu constant agitation from the escape of inflammable gas. It contained in 1000 parts about two parts of free sulphuric acid, and less than one part of neutral sulphates. This water was collected in October 1810, and at that time another half-dried-up pool in the vicinity contained a still more acid water. Another similar spring oocurs n(!ar St. I)avia1t9 of lime and magnesia, relations of chlorids and carbonates; 55, solubility of earthy carbonates ; 5G, snper-saturatcd solutions of carbonates of lime and magnesia ; 67, salts of barium and strontium, solution of their sulphates ; 68, iron, manganese, alumina and phos- phates ; 59, bromids andiodids ; the small portiun of bromine and tho excess of iodine in saline springs as compared with the modern ocean ; 60, probable relation of iodids to sediments; 61, sulphates, their elimi- nation from waters ; 62, water holding a soluble sulphuret ; 03, borates, their detection and determination ; 64, analysis of a borax water from California ; 65, carbonates, their amount in the Caledonia waters ; 66, intervention of neutral carbonate of soda ; 67, dclicicncy of carboniv acid in waters ; 68, reactions of various waters ; 69, silica, its source and its proportion ; 70, its conditions ; formation of silicate; , 71, organic matters ; 72, geological position of the waters here described ; 73, succession of paleozoic strata ; lithological relations of successive formation ; 74, Quebec group, its waters; 75, sources of various classes of waters; 76, their relation to the forma- tions ; 77, association of unlike waters, changes iii constitution ; 78, temperature of springs, thermal waters ; 79, geological interest of the above analyses ; possible results of the evaporation of these springs ; 80, relations of mineral springs to folding and to mctamorphism of strata; 81, on the supposed origin of the primeval ocean and the earliest sediments ; 82, on the theory of metalliferous deposits. § 52. Salts of the Alkaline Metals. — These salts abound in most saline waters, and except in the few cases in which sulphate of magnesia prevails, form a large part of the soluble matters present. The salts of sodium are by far the most abundant, and the proportion of potassium salt is generally small. The chlorid of potassium in modern sea-water constitutes three or four hundredths of the alkaline chlorids, while in the brines from old rocks, and in saline waters of the first two classes alike from Germany, England, the United States, and Canada, its proportion is much less, sometimes amounting to traces only. In the waters of classes III and IV, where alkaline carbonates appear, and even predominate, the proportion of potassium salt becomes greater. Thus of the waters of the latter class (§ 45), the alkalies of the )rtion and litivc sea ; irbonatea ; )IutionB of strontium, aud plios- ne and tho Rrn ocean ; Iheirelimi- huret ; G3, of H borax Calodoniu , deficiency katers ; GO, rmation of tlie waters itliologieal naters ; 75, the forma- tution ; 78, terestof tbc sc springs ; orphism of an and tbc posits. Its abound iu which lie soluble the most ally small. cs three or rines from alike from proportion the waters ', and even 33 greater, lies of the Nicolct spring calculated as chlorids contain 1.89 per cent of chlorid of potuasiuin, and those of tho Jacques-Cartier 2.95 ; while for the St. Ours spring tho chlorid of potassium is equal to not less than 25.0 per cent. There does not however appear to be any relation between the proportion of alkaline carbonate and that of potassium, since the salts from tho waters first named are more alkaline than those of St. Ours ; while those of the alkaline water of Joly contain less than one per cent of potassic chlorid. The amount of this suit obtained from the water of the Ottawa River is worthy of notice, being equal to not less than 32.0 per cent of the alkaline chlorids, while in the waters of the St. Lawrence it amounts to IG.O per cent.* A large proportion of potassium relatively to the sodium has already been observed in the cose of many ordinary river and spring waters, and this is readily explained when we consider the extent to which potash is set free by the decomposition of both vegetal and mineral matters at tlic earth's surface. The process by which this base is eliminated in filtering through soils has already been explained in § 5. The occasional presence of considerable amounts of potash in sulphated mineral waters (Lersch, Hydro-chemie, p. 346) is explained by the power of solutions of gypsum to set free this alkali from soils (§ 7), and also probably in some cases by the dissolution of double potassic salts like polyhallite. Strata holding glauconite, which occurs alike in paleozoic and more recent formations,! may also be conceived to yield potash salts to infiltrating waters. § 53. It will be seen that the waters above noticed, in which the proportion of the potash to the soda is large, are but feebly saline, so that the real amount of potassium is in no case great. I have however recently examined the water of a borax lake from California, which contains in 1000 parts 17.250 of solid matters, of which 1.818 is carbonate of potash, the remainder being soda- salts, carbonate, borate, chlorid, and a little silicate, with no sulphate • T. S. Hunt, L. E. and D., Phil. Mag. (4) xiii. 239 j and Geology of Canada, page 565. fFor a notice, with analyses by the author, of a green by drated silicate of alamina, iron and potash, allied to glauconite, from the paleozoic rocl^s of Canada and of the Mississippi valley, see the Geology of Canada,, pages 487, 488 ; where also will be found an analysis by the author of the glanconite from the Cretaceous formation of New Jersey. See also Silliman's Journal, [2]. xxxiii. 277. 4e I m m I; ! (§ 64). ThiH amouut, if represented as ohlorid of potasBium, in equal to LOG!', or to 11.46 per cent of the alkulicB calculated an chlorids. The amount of potassium salt in this water is consequently about forty times greater than in that of St. Ours. The fact of Hpecial importance as regardH the alkaline metals in the waters whose analyses wo have given in this paper is the very small amount of potassium in the strongly saline muriated waters of the first three classes, which we conceive to be more or loss directly derived from the waters of the ancient ocean. To this primeval sea, almost destitute of potassium, the process of mineral decay has been for ages adding potash suits, and despite the partial elimination of these by vegetation (§ 5), and by the formation of glauconite, wo find a notable proportion of potash in the waters of the modern ocean. In the analyses of the saline waters hero given lithia was sought fur in a few instances, and was detected in the waters of Vuronnes. Most of these analyses were made before the discovery of the new metals caosium and rubidium. § 54. Salts Ok' Calcium and Magnesium. — Wo have to consider under this head the relations both of tho chlorids and the carbonates of these bases. Tho bitter saline waters of the first class, although containing large (quantities of chlorids of calcium and magnesium, are, as we have seen, generally destitute of earthy carbonates. These latter, however, arc found in small (quantities in the alkaline waters of the fourth class, and in some- what larger amounts in those intermediate waters which fornj classes II and III, and are apparently formed by admixtures of the two classes previously mentioned. Besides the carbonates of lime and magnesia which the waters of the fourth class hold in solution, tho carbonate of soda which they contain gives rise, by its rc-action with the chlorids of calcium and magnesium, to additional quantities of the carbonates of these bases. In the Waters of Kingston, (§ 3G), a large amount of chlorid of calcium is associated with earthy carbonates, and these waters thus o£fer a passage from the first to the second class. In most of the waters of the second class, as will be seen from the table § 42, there appears but a small amount of chlorid of calcium ; and even this depends upon the manner in which tho analysis has been conducted. We may suppose in the recent water such a partition of bases between the chlorine and the carbonic acid that chlorid of calcium, chlorid of magnesium, bicarbonate of lime 47 isium, M luted an iquontly 2 metals cr is the iiiuriatcd more or To thitt niinornl ic partial nation of 10 watcns ns sought ^urcnncH. ry of the have to orids and srs of the hlorids of J destitute d in small d in some- hich form ixtures of bonates of 9S hold in es rise, by lesium, to i. In the of calcium thus offer seen from ehlorid of which the 3cent water rbonic acid Ate of lime and bicarbonate of magnesia coexist. When suuli a solution is submitted to evaporation at ordinary tomperaturos, provided there is present u sufficient amount of ehlorid of calcium, carbonate of lime alone is deposited, and ehlorid of magnesium remains in solution. In ease the ehlorid of calcium isiiisufHcient, the lime is still first deposited as carbonate, and the more soluble magnesian carbonate is precipitated by further evaporation. When however such a water is boiled, a reverse process takes ])laee; the carbonate of lime slowly decomposes the magnesian ehlorid, and carbonate of mag- nesia /s deposited, while ehlorid of calcium remains in solution. Hence if the amount of ehlorid of magnesium bo great enough, and the ebullition sufficiently prolonged, the precipitate will at length contain only carbonate of magnesia ; while an equivalent of ehlorid of calcium, now found in the solution, represents the carbonate of lime which the analysis of the precipitate at an earlier stage of the ebullition would have furnished. As an example of this may be cited the analysis of the water of te. Genevieve ( §42, No. 8), where the precipitate after a few lutcs boiling contained carbonates of lime and magnesia in the proportion 12 : 750. When howc'cr another portion wos boiled down to one sixth, the precipitate was found to be pure carbonate of magnesia. Again, the Plantagcnet water gives, by ebul- lition, the results set forth in §42, No. 1 ; showing chiefly carbonate of magnesia, together with a portionof ehlorid of calcium. When however this water is left to spontaneous evaporation, the whole of the lime separates as carbonate, and the liquid remains for a time charged with carbonate of magnesia, probably as sesqui-carbonate This solution is however after a time spontaneously decomposed even in closed vessels, with deposition of a portion of crystalline hydrated carbonate of magnesia; another portion remains in solu- tion, together with ehlorid of magnesium, but is precipitated by ebullition. (Silliman's Journal [2]. xxvii. 173) § 55. Bicarbonate of magnesia and ehlorid of calcium, when brought together in solution, undergo mutual decomposition with separation of carbonate of lime if the solutions are not too dilute. At the ordinary temperature and pressure, water saturated with carbonic acid will not hold in more than about one gram of carbonate of lime to the litre (1 : 1000) ; equal to only 0.88 grams of carbonate of magnesia. (The solubility of carbonate of lime in pure water is well known to be much less, and is, according to 48 1 h' Bineau, equal to 1 : 30,000 or 1 : 50,000.) Wo should not there- fore expect to find that water holding chlorld of calcium in solution would yield, by boiling, more than the latter amount of magnesian carbonate; so ntuch might evidently be formed by the action of dissolved carbonate of lime which the water might hold as bicarbonate. I have elsewhere described a series of experiments on the solubility of bicarbonate of lime both in pure water and in saline solution.s, and have shown that the presence of salts of soda, lime and magnesia docs not increase the amount of bicarbonate of lime which water is capable of holding jjermancntli/ in solution. In view of those facts it seems at first sight difficult to explain how a mineral water like that of Kingston (§ 30, No. 7), holdin^- .\ large (juantity of chlorid of calcium, could yield, as appears from Dr. Williamson's analysis, 1.287 grams of carbonate of magnesia, equal to 1 .462 of carbonate of lime to the litre. Recent experiments have however shown me that supersaturated solutions of a certain stability may be obtained, in which comparatively large quantities of neutral carbonates of lime and magnesia exist in the presence of sulphates and ehlorids of calcium and magnesium. Reserving for another occasion a description of the details of these investiga- tions, I shall briefly state the results obtained. § 5G. In a memoir on the salts of lime and magnosia published in 1859 (Silliman's Journal [2] xxviii. 171), it was shown that by the additionof bicarbonate of soda to asolutionholdingchloridsof sodium calcium and magnesium, with or without sulphate of soda, iind satu- rated with carbonic acid, it was possible to obtain transparent solutions holding from 3.40 to 4.16 grams of carbonate of lime to the litre ; of which however the greater part was deposited after twenty-four hours ; wlien the solutions were found to contain some- what less than 1.0 gram in the form of bicarbonate. Boutron and Eoudet had previously shown that by saturating lime-water with carbonic acid, solutions were obtained holding in a litre 2.3 grams of carbonate of lime ; of which one half was soon deposited, even when the solution was kept under' a pressure of several atmospheres. It would thus seem that saline liquids favor this temporary solu- bility of the carbonate of lime. In all of the above experiments, an excess of carbonic acid was present, but this I have since found is not essential, since super-saturated solutions may be obtained holding as much as 1.2 grams of carbonate of lime, together with sulphate of mag- nesium and chlorid of calcium, in a litre of water, without any 49 excess of carbonic acid. The power of alkaline clilorids ami of chlorid of calcium to prevent the precipitation of chlorid of calcium by carbonate of soda has already been observed by Storcr, (Dictionary of Solubilities, p. 110). I have found that the precipi- tate produced by the admixture of solutions of these two salts is readily dissolved, when recent, by a solution of chlorid of calcium, or of sulphate of magnesia ; and thus liquids may be prepared liolding at the same time from 1.0 to 1.2 grams of neutral carbonate of lime, and 1.00 of neutral carbonate of magnesia, in presence of sulphate of magnesia. Thesesolutionsof carbonate of lime, which are strongly alkaline, may be kept for twelve hours or more without perceptible change at ordinary temperature.", but after a time deposit crystals of hydrated carbonate of lime. The addition of alcohol immediately throws down the whole of the carbonate of lime in an amorphous condition. The carbonate of magnesia is still more soluble than the carbo- nate of lime under similar conditions, and it is possible to obtain 5.0 grams of neutral carbonate of magnesia dissolved in a litre of water holding seven per cent of hydrated sxxlphate of magnesia. without any carbonic acid. These solutions, which are strongly alkaline to test-papers, yield a precipitate by heat, which vc-dis- solves on cooling. It is evident that the mingling of saline and alkaline waters may give rise to solutions like those just described, and thus explain apparent anomalies in composition like that of the Kingston water. See also in this connection the observations of Bineau, and my own on the properties of solutions of sesqui-carbonate of magnesia. (Silliman's Journal [2] xxvii. 173.) § 57. Salts op Barium and Strontium. — As will be seen from the preceding tables, the salts of thes two bases are found in very many of the saline and alkaline waters of Canada. Their carbon- ates probably sustain to the magnesian chlorid a similar relation with that of calcium, and hence these bases appear in some of the analyses partly as carbonates, and partly as chlorids of barium and strontium. The precipitate formed in the concentrated and acidulated water by dilute sulphuric acid was, whenever submitted to analysis, found to contain both barium and strontium. For the separation of these the mixed sulphates were first converted into chlorids ; the barium was then thrown down as silico-fluorid, and the strontium subsequently precipitated by a solution of gypsum. The insolubility of its sulphate must have excluded baryta from 50 i ,! ! I •J .^ the waters ot the primeval sea, and when set free, as we may suppose by the decomposition of its silicatcd compounds existing in the primitive crust, (§ 81) its soluble bicarbonate carried down to the sea would there be precipitated by the sulphates present. A similar process must still go on with all the dissolved barytic salts which tind their way to the ocean. The sulphate of baryta thud accumulated in sedimentary strata, may be partially decomposed by infiltrating solutions of alkaline carbonates, and thus be rendered capable of being sub- sequently dissolved as carbonate ; but the most probable mode of its solution, is, we conceive, through its previous reduction by organic matters to the form of a soluble sulphuret (§ 10), ready to be converted into carbonate or chlorid of barium. In this way we may explain the frequent occurrence of baryta salts in the saline waters of the first three classes, and the consequent absence of sulphates, which will be further considered in § 61. From the similarity of its chemical re-actions, the preceding remarks apply to strontia as well as baryta. § 58. Iron, Manganese, Alumina and Phosphates, — None of the waters of the four classes here described contain any notable (juantity of iron, yet this element is never wanting in those waters which contain earthy carbonates. Whenever a portion of one of tliese waters, or better the earthy precipitate separated from it by boiling, is evaporated to dryness with an excess of hydrochloric acid, the residue treated with acidulated water yields a portion of silica, and tlic solution v/ill then be found to yield with ammonia a precipitate. This, which is partially soluble in caustic alkalies, is often colorless, and will be found to consist of alumina and peroxyd of iron, with phosphoric acid and a trace of manganescj which latter metal is seldom or never absent. The small quantity of alumina which these waters contain appears not to be derived Irom suspended argillaceous matters, but to be held in a state of solution. The phosphates are generally present only in very small quantities in these waters, for the reason pointed out in ^ 5. The largest amount which I have met with was in an alkaline water from Fitzroy {§ 43, No. 4); where it is equal to 0124 of tribasic phosphate of soda in 1000 parts of water. § 59. Bromids and Iodids. — The chlorids in these ancient mineral waters are always accompanied by bromids and iodids> but the proportion of the bromids to the chlorids appears to be 51 much less than in the waters of the modern seas. According to Usiglio, lOr parts of the salts from the Mediterranean contain 1-48 of bromid of sodium ; while ten analyses by Von Bibraof the waters of different oceans, give from 086 to 146, affording for 100 parts of salts, a mean of 116 of bromid of sodium, equal to 1*04 parts of bromid of magnesium. The waters of Whitby and Ilallowell, on the contrary, which arc the richest in bromids of those described in this paper, contain only 0-54 and 69 parts of bromid of sodium in 100 parts of solid matters; while few of the saline springs of the second class contain more than one-half of this proportion, and some of them very much less. With regard to the iodids in many of these waters, however, the case is very different. The waters of the modern ocean, as is well known, contain but traces of iodine, and in some strongly saline springs of the first class, like that of Whitby, it is only in the alcoholic extract of the salts from this water that iodine can bo detected. The Ilallowell water (§ 36, No. 3), which closely resembles this in its general composition, and in the proportion of bromids, is however so rich in iodine that its presence can readily be discovered without previous evaporation. It is sufficient to add to the recent water acidulated by hydrochloric acid, a little Bolution of starch, and a few drops of nitrite of potash to produce an intense blue color. The iodid of sodium in the first-named water was found equal to 0-0017 parts of the solid matters, and that of the second to O'OID or nearly twelve times as mucl' The unconcentrated saline waters of Ste. Genevieve, of the secoi. i flaR*. ulso give a strong re-action for iodine, and when acidulated with hydrochloric acid, without previous evaporation, yield with a salt (»f palladium an insoluble precipitate of iodid of palladium after a few hours. The salts from these two springs of Ste. Genevieve, though poorer in bromids, are much richer in iodids than the waters of Hallowell ; the spring No. 8, containing in 100 parts of salts no less than 0-138 of iodine, so that there appears to be no constant proportion between the chlorids, bromids, and iodids of these saline waters. § 60. The relations of bromids and iodids to argillaceous sediments have yet to be determined. It would appear from the facts just cited that bromine has in the course of ages been slowly eliminated from insoluble combinations, and like potassium, has accumulated in the waters of the ocean ; while the facts in the history of iodine seem to point to a process the reverse of this ; 62 h ^ ^ ;l T Pi K in other words, to a grailuul elimination of iodine from the sea- waters, and its fixation in the earth's crust. The observations of numerous chemists unite to show the frequent occurence of small portions of iodine in some unknown combination, in sedimentary rocks of various kinds; from which we may conjecture that it was in former times abstracted from the sea, either directly or through the intervention of organic bodies (as in the case of potash, which is separated and fixed by means of algoe, § 5). Experiments after the manner of those of Way and Voelcker may throw liglit upon this interesting question. We are aware that insoluble combinations of soluble chlorids with silicates of alumina are found under certain conditions, as appears in the generation of sodalite, eudyalite, and the chloriferous micas, and it is not improbable that the soluble iodids may give rise to similar compounds. By such a process might be explained the rarity of this element in modern seas, while the occasional re-solution of the iodine from these insoluble compounds by infiltrating waters, would help to explain the variable and often large proportions in which this element is met with in some of the waters noticed above. § 61. Sulphates. — In the preceding sections we have already discussed the principal fa-^ts in the history of those neutral waters in which sulphates predominate, or prevail to the exclusion of chlo- rids (§ 50, 51 .) The history and the probable origin of those curious springs which contain free sulphuric acid has also been considered (§ 31, 48, 49) ; and it now remains to notice the relation of sulph- ates to the muriated waters. The first fact that excites our attention is that of the total absence of sulphates from numerous springs of the first, second and third classes ; as shown in the preceding analyses, and also in the observations of Lenny and others on the saline waters over a great area in western Pennsylvania (§ 40). The elimination of sulphate in the form of gypsum from evapo- rating waters containing an excess of chlorid of calcium, has already been discussed in § 37; but the b itterns resulting from such a pro- cess still retain small portions of sulphates ; while it is to be re- marked that the saline waters under consideration contain no traces of sulphates, and in many instances hold portions of baryta and strontia, bases incompatible with the presence of sulphates. The modes in which this complete elimination of sulphates may be eJOfected are two in number. The first has already been suggested in § 10^ and depends upon the deoxydizing power of organic matters, which, reduce the sulphates to sulphurets. These in their turn may be 7. 5;3 converted into carbonates, the sulphur being separated either as sul- phuretted hydrogen (giving rise by oxydation to free sulphur), or as insoluble metallic sulphurets. This reducing action not only de- composes the soluble sulphates of soda, lime and magnesia, but also) as has been pointed out in § 57 may extend to sulphate of baryta, and thus sulphuret or carbonate of baryta be formed. It is the action of these soluble baryta salts which constitutes the second snodc of desulphatizing waters ; and this, if we may judge from the frequence with which baryta salts occur in the saline waters in question, appears to have been the most general process. It is a fact worthy of notice that a saline spring at Sabrevois near Lake Champlain, which holds both baryta and strontia in solu- tion, is at the same time slightly impregnated with sulphuretted hydrogen. Another saline and sulphurous spring, which rises within ton feet of this, contains however a portion of sulphates. (Geology of Canada, page 542.) § 62. I am indebted to Prof Croft of Toronto, for some notes of a recent examination by himself of a saline of the first class, which contains at the same time a soluble sulphuret. This water, from a boring sunk to a depth of several hundred feet through the Devonian limestone at Chatham, Canada West, had a specific gravity of 1039.3, and yielded for a thousand parts about 51.0 of solid matters. It contained large portions of chlorids of calcium and magnesium, with very little sulphate, traces of carbonate, and no free carbonic acid. The water, which gave an alkaline reaction with turmeric, was greenish in color, very sulphurous to the taste, and yielded a purple color with ni ?o-prussid of sodium, and a black j»rccipitatc of sulphuret with a solution of sulphate of iron. A current of carbonic acid rendered the recent water opalescent, and by exposure to the air it deposited sulphur. A quantitative anal- ysis of this water is to be desired. § 63. Borates. — The reddening of the yellow color of turmeric paper in presence of free hydrochloric acid, affords, with certain precautions, the ordinary means for detecting small portions of boric acid. Most of the waters of the third and fourth classes, and some of those of the second have been tested in this way, and have never failed when reduced to a small volume, and acidulated with hy- drochloric acid, to give this reaction ; which was however most marktd with the waters of the fourth class. The determination of the amount of boric acid in saline waters presents no small dif- ficulty. In the case of the alkaline water of Joly (§ 45, No. 3) the 54 Is'' Wl'l' J' following process was however attempted. The salts left by its evaporation were treated with carbonate of ammonia to separate a portion of silica, and then with recently precipitated carbonate of silver, by which the alkaline chlorids were converted into carbonates. The solution now retained in some undetermined form a portion of silver, winch was separated by fusing the evaporated saline resi- due in a silver crucible. By a second evaporation and fusion there was obtained a mixture of soda and potash, combined only with carbonic, sulphuric and boric acids. By directly determining the other ingredients the boric acid was estimated from the loss, and was found equal to 0*028 parts in 1000 of water, which contained 0-752 of solid matters. The conversion into carbonate of the sul- phates in the mixed salts, by the aid of bicarbonate of baryta, would simplify this process. In § 35 it has been explained that the amount of carbonate of soda in the waters of the third and fourth classes was generally calculated from the excess of the alkaline bases, and controlled by the amount of carbonate of baryta preci- pitated from chlorid of barium by the alkaline salt. It was found, however, that this last method always presented a certain deficit, due to the borate of soda, whose quantity in many of the waters, is too large to be disregarded. The precipitate of carbonate of baryta contained a portion of sparingly soluble borate of baryta, which was not completely removed by long and continued washing. § 64. I have recently had an opportunity of examining from California the waters of a borax lake, which contains, beside borate and carbonate of soda, a portion of chlorid, and a little silicate, traces only of phosphate, and no sulphate. It held in solu- tion very small quantities of earthy carbonate, and was remarkable for a large proportion of potash, already referred to in § 53. The evaporated and fused saline residue was treated by the ordinary methods for the determination of the chlorine, carbonic acid and silica ; while the bases were obtained in the form of sul- phates by the aid of sulphuric and hydrofluoric acids, and after- wards separated as chlorids by the aid of chlorid of platinum. From the data thus obtained the following ingredients were found by calculation for 1,000 parts of the water : Carbonate of soda 9.476 Biborate of soda 4.395 Chloride of sodium 1.702 Carbonate of potash 1. 818 Silica 0.129 17.520 The potassium, as above determined, equals 11,46 per cent, of the bases weighed as chlorids ; another trial gave 11.41. Although for convenience we have represented the potassium as carbonate, it will be seen that the amount of chlorine is such that it might, for the greater part, have been rervresented as chlorid of potassium, with an equivalent portion additional of carbonate of soda. § 65. Carbonates. — In describing in § 43 the alkaline-saline waters of Caledonia it has been shown that these contained a quantity of carbonic acid insufficient to form bicarbonates with the carbonated bases present. It was partly with this fact in view that, after an interval of more than seventeen years, I undertook the new analyses of these waters, which in § 47 are given side by side with the earlier results. In these recent analyses, as there remarked, a slight excess of carbonic acid was met with. In the interval the springs had undergone changes in composition, and while the third one still retained in a slight degree its alkaline character, the other two had become waters of the second class. holding instead of carbonate and sulphate of soda, chlorid of mag- nesium, and baryta salts. The amount of carbonic acid had how" ever undergone but little change ; and as will be seen by compar- ing the figures below with those in the table in § 47, the slight diminution in the first and third corresponds very closely with the falling off in the amount of solid matters between 1847 and 18(15 ; while, on the contrary, the augmentation in the amount of carbonic acid in the second is accompanied with a corresponding increase in the amount of fixed matters present. CARBONIC ACiD IN ONE LITRE OF THE CALEDONIA WATERS. 1847. 18C.-). Gas spring 705 gram. .G71 ^rr.ain. Saline spring 648" .664'' Sulphur spring 590" .573' While the amounts of fixed matters and of carbonic acid in the several waters have undergone but little change, we find, however, that there has been a great diminution in the proportion of car- bonated bases. Thus in the Gas spring in 1847 the carbonic acid required for the neutral carbonates found in the analysis was .356, while for the same water in 1865 only .278 of carbonic acid was required. In the Sulphur spring, in like manner, the neutral carbonates required .449*. or more than three-fourths of * By mistake this is printed .349 in § 43. 50 " ii I tlio carbonic acid present; while the falling off in the anioant of cMibonatcs in 1805 is such that only .191 of carbonic acid, or just about one-third of the carbonic acid present, is required for the neutral carbonate. Nor is this change due entirely to a less amount of carbonate of soda ; the carbonates of lime and magnesia in IS 17 required .24(5, and in 1865 only .15Ji of carbonic acid. T ho changed conditions which we here meet with may bo explained liy suppoijing that the carbonated bases are due to the mingling in (lilVoront proportions of neutral carbonate of soda (generated by the reaction indicated in ^ 13,) with an earthy saline water holding a constant amount of free carbonic acid ; which, in some cases, is more than is required to form bicarbonates, but in others, as wc have seen above, shows a deficiency. ^ 00. If we admit, aw I have already assumed, that the waters of the second and third classes have been generated by the ming- ling of solutions of carbonate of scjja with waters of the first class, it can readily be shown that these solutions contained chiefly or exclusively the neutral carbonate. If we add a solution of bicarbonate of soda to earthy saline waters of the first class it is easy to obtain solutions of holding twenty grams or more of bicavuonate of macucsia to the litre ; ' while in none of the natural waters of the second class do our analyses show the existence of much overone gram to the litre. Again, if we sup- pose any considerable amount of chlorid of calcium to be decom- posed by bicarbonate of soda, the separation of the lime in the form of neutral carbonate, and the liberation of the second equivalent of carbonic acid, Avould yield waters holding an excess of carbonic acid above that required to form the bicarbonates of the solution. From the absence of such an excess, as appears in the case of the waters of Caledonia, Varennes, St. Leon and Caxton, and from the small amount of bicarbonate of magnesia in these waters, it may bo concluded that the alkaline salt whose addition has changed their character was the neutral carbonate of soda. § 07. Examples are not wanting of waters in which, as in those of Caledonia in 1847, the carbonic acid is insuflScient to form bicarbonates, or even neutral carbonates, with the bases uncombined with sulphuric acid or chlorine. Thus, according to Pagenstecher and Miiller, the spring and well-waters of Berne do not contain sufficient carbonic acid for the lime present, a part of which they suppose to be held in solution as a silicate. See Bischof, Chem. Geology, i. 5 ; who remarks that Lowig seems to have observed the cnme foot in the thermal spring of Pfiiffers. For further examples of this kind see Lersch, Hydro-Chemie, page 333. The carbonic acid in the water of Toplitz is according to him not sufficient to form bicarbonates unless the silica present be supposed to be com. lined with a portion of bases ; while in the alkaline thermal sprin*^' of Bertrich, according to the analysis of Mohr, a similar deficiency of carbonic acid exists ; leading to the conclusion that a part of the earthy bases present is in combination with silica and organic matters. The existence of solutions holding comparatively large amount of neutral carbonates of lime and magnesia, as described in ^ 56, is not without interest in this connection ; since it at once afibrds an explanation of the nature and origin of all such alkaline waters, and waters deficient in carbonic acid, as contain earthy sulphates and ehlorids. § 68. It wos found that the waters of Chambly in 1864, and of the Sulphur spring of Caledonia in 1865, gave with lime-water :i precipitate which was soluble in an excess of these mineral waters, but to a much less extent than in the acidulous saline water from the High- Rock spring of Saratoga. The latter, which contains bicarbonate of soda, and is highly charged with carbonic acid, turns to a wine-red the blue color of litmus tincture, which is not changed by the Chambly or the Caledonia water. The Saratoga water, after some time, gives a feeble alkaline reaction with dahlia paper ; this is more distinctly but slowly changed by the Caledonia water, and almost immediately turned to green by that of Chambly. This latter water readily browns yellow turmeric paper, which is scarcely affected by the water of Caledonia. § 69. Silica. The silica which exists in solution in cold saline springs is generally very small in amount, as might be expected from the insolubility of earthy silicates, which is such that super- ficial drainage waters in filtering through the soil lose the silica which they held in solution (§ 5). We have further shown that as :i result of this tendency to the formation of insoluble silicates, the silicate of soda liberated in the sediments by the decomposition of feldspar, generally appears at the surface as carbonate of sodo, having been decomposed by earthy carbonates (§ 13). In two cases, however, considerable quantities of silica are found dissolved in natural waters. The first is met with where the rapid solvent and decomposing action of heated waters is exerted upon alkaliferous silicious minerals (§ 14), as seen in springs like the Geysers. The second case is that of those rivers and streams E I *l r)8 : .1 Mi which (Iruin tturiiuTS t'ovorcd with (U'cnying vo^'ctation nnd ilccoin* posing silicates, fnim botli of whic'i tlicy ilcrivo dissolved silica. ISueh waters contain but small amounts of solid matters, but the proportion of silica is relatively considerable, amounting, as we have seen in the water of the Ottawa River, (which contains in 10,(100 parts. 0-01 IT. of solid matters). t(» 02060, or thirty-two per cent. ; while in the St. Lawrence, which contains for the flame amount of water, 1 -0050, the silica ecpials -IJTOO, or twenty-four per cent, of the solid ingredients. The aiialysi.s by H. Seville of the river-waters of France .show, in like manner, large amounts of silica, which seem to have been hitherto overlooked in the analyses of most chemists. (Ann. de ('him. et IMiys., [II] xxiii, '>i2.) It will be seen by a reference to tlie tables of analyses given in the second part of this pajter, that in the waters of the second cla^s the amount of silica is ecjual to from 01 5 to 000 parts for lOO'Od of solid matter. In the alkaline waters of the third and fourth cla.sses its proportion is greater, and up to a certain point appears to increase with that of the carbonate of soda. In the following table the proportions of carbonate of soda and silica for 1000 parts of solid matters are given for certain springs, who.'re Ouelle and Ste. Anne, are bitter waters belonging to the first class ; while a sulphurous spring at the latter place, and another at Quebec are alkaline waters of the fourth class. § 75. Of the waters of tlv; western basin, which alone are no- ticed in this paper, many have been qualitatively analyzed which are not here described. Including two from Vermont, twenty-one al- kaline waters of the third and fourth classes have been examined. Of these, as already stated, the waters of Caledonia rise from tlife Trenton group, and that of Fitzroy from the Chazy or Cal- ciferous, while two others, at Ste. Martine and Itawdon, appear to have their source in the Potsdam. All the other waters of these two classes issue from the Hudson-River shales, with the exception of those Varennes and Jacques Cartier, which seem to rise from the Utica formation. Of the waters of the second class, of which about thirty have been examined from the western basin, some five or six i ssue from the shale formatio'^is Nos. 5 and 6, but all the others are from the underlying limestones. The bitter salines of the first cla.ss flow from the limestones of the Trenton group, with the exception t^f that of Ancaster, which is from a well sunk in the Niagara formation, and that of St. Catherines, from a boring carried through the Me- dina down into the Hudson-lliver shales. The source of both of these is probably, like that of the other very siniilar w.itcrs, the Lower Silurian limestones. § 76, From this distribution of the waters of the first four classo.'* it would appear that the source of the neutral salts, which consist of alkaline and earthy chlorids, is in the limestones and other strata from the Potsdam to the Trenton inclusive, while the alkaline c:ir- bonates are derived from the argillaceous sediments which make up the Utica and Hudson-River formation. These sediments are never deficient in alkaline silicates, whose slow decomposition yields to infiltrating waters (§13) the alkaline carbonates which charac- terize the mineral springs of the fourth class. These, mingling in various proportions with the brines which rise from the limestones beneath, produce the waters of the second and third classes in the manner already explained. The appearance of several springs of the third class, as those of Caledonia and Fitzroy, from the Lower 64 Silurian limestones is not suprising, when it is considered that the Chazy formation in the Ottawa valley includes a considerable thick- ness of shales, sandstones and argilaceous limestones, approaching in composition to the sediments of the Hudson-River formation. § 77. As an evidence that the different classes of waters have their origin in different strata may be cited the fact that springs very unlike in composition are often found in close proximity, and apparently rising from a common fissure or dislocation. Thus in the seigniories of Nicolet and Labaie du Febvre, I have examined six springs, all of which rise through the Utica formation along a line, in a distance of about eight miles. Of these springs two be- long to the second, two to the third, and two to the fourth class ; these last being probably derived entirely from the shales, while the others have their source in the underlying limestones, and are more or less modified in their ascent. Again, at Sabrevois, within a few feet are two springs of the second class, of which one contains salts of baryta and strontia, and the other soluble sulphates. In like manner at Ste. Anne, in the Quebec group, a spring of the second class and one of the fourth are found not far apart. The springs of Caledonia offer another and not less remarkable example. In 1847 there were to be seen, not far from a spring of the second class, three others of the third class very near together, one of them sulphurous, but all sulpbated, and differing in the proportions of carbonate of soda present. In 1865, while one of these still re- tained its character of a sulphurous sulpbated water of the third class, the others were changed to waters of the second class, and held salts of baryta in solution. These relations, which we have already) pointed out (§ 47), not only show waters holding incom- patible salts issuing from different strata along the same fissure, but mingling in such varying proportions as to produce from time to time changes in the constitution of the resulting springs. § 78. The temperature of none of the springs which we have here described exceeds 53°, which has been observed for two springs at Chambly, about twelve miles from Montreal. Inasmuch as the mean temperature of this city, as deduced from the observations of twenty-seven years, is 44°.67, the Chambly waters are to bo regarded as slightly thermal. No other springs in Canada are known to present so high temperature, unless possibly the acid watera of the fifth class, for which we have pointed out the importance of farther observations, (§ 48). The St. L^on spring was found to ,bc 46°, and that of Caxton, 49° F. 65 ^ 79. The extendc'l series of analyses which we have given in the preceding pages presents many points of interest. Nowhere else, it is believed, has such a complete systematic examination of the waters of a region, and of a great geological series been made. Additional importance is given to these results by the fact that the waters are all derived fVom paleozoic strata, and we are thus enabled to compare these saline materials of an ancient period with those which issue from, and in may cases owe their saline impregnation to strata of comparatively modern origin. Compari- sons of this kind, such as I have already instituted between brines of different geological epochs in § 39, possess great geological interest. It is a consideration not without interest, that the valley of the St. Lawrence under different meteorological conditions might become a region abounding with saline lakes, affording sea-salt, na- tron and borax, the results of the evaporation of the numerous saline and alkaline springs which have just been described. § 80. A few considerations are here suggested by the fact already mentioned of the apparent absence of mineral springs from the altered paleozoic strata of the Quebec group. Metamorphisni and disturbance or displacement of strata are generally concomi- tants, not, as I conceive, because the process of alteration is in any way connected with the disturbance of the rocks, but because a great accumulation of superincumbent strata, a necessary preliminary of metamorphism, is the efficient cause of the fold- ing of the deeply buried and subsiding rocks, in a way which I have already elsewhere pointed out.* The subsequent continertul uplifting of the altered, plicated, and more or less fissured strnta, and their irregular erosion, give rise to the broken surfaces of mt ra- morphic regions, and at the same time permit the saline solutli us impregnating the strata to flow out ; while solid soluble salts, unless enclosed by impermeable strata, are removed by lixiviation. Hence we shall rarely find muriated waters issuing from crystalline and dis- turbed strata. Those saline products which result from the decom- position of feldspathic minerals, and the«eparation of alkaline car- bonates ; or from the decomposition by these, or other agents, of the irypsum which is often present in metamorphic strata, may, however, Silliman's Journal [2] xxxi: 412 66 readily give rise to waters of the fourth and sixth classes ; so that wo arc not surprised to find alkaline and sulphated waters issuing i'rom crystalline strata. § 81. I have in a previous section (§ 57) alluded to the con- dition of the primeval ocean, and in this connection it may be well to refer to a hypothesis which I some years since advanced, to explain the origin of its salts and the primeval sediments- starting from the notion " of a cooling globe, such as the igneous theory supposes our earth to have been at an early period, and con- sidering only the crust with which geology makes us acquainted, iind the liquid and gaseous elements which now surround it, I have ondeavored to show that we may attain to some notion of the chemical conditions of the cooling mass by conceiving these materials to again ro-act upon each other under the influence of an intense heat. The quartz, which is present in such a great proportion in many roc'vs, would decompose the carbonates and sulphates, and aided 1' tlic presence of water, the chlorids both of the rocky strata and <'i the sea; while the organic matters and the fossil carbon would Ije burned by tlie atmospheric oxygen. From these re-actions wuuld result a fused mass of silicates of alumina, alkalies, linio. magnesia, iron-oxyd, etc. ; while all the carbon, sulphur ami chlorine, in the form of acid gases, mixed with watery vapor, nitrogen, and a probable excess of oxygen, would form an exceedingly dense atmosphere. When the cooling permitted con- densation, an acid rain would fall upon the heated surface of the earth, decomposing the silicates, and giving rise to chlorids and sulphates of the various bases, while the separated silica might take the form of crystalline quartz. In the next stage of the process, the portions of the primitive crust not covered by the ocean would undergo a decomposition under the influence of a hot moist atmosphere charged with carbonic acid, and the feldspathic silicates become converted into clay, with separation of the alkali. This, absorbing carbonic acid from Ihe atmosphere, would find its way to the sea, where, having first precipitated from its highly heated waters various metallic bases then held in solution, it would decompose the chlorid of calcium, giving rise to chlorid of sodium on the one hand, and to carbonate of lime on the other. In this way we obtain a notion of the processes by which from a primitive fused mass may be generated the silicious, calcareous and argillaceous m Ik m 67 rocks which make up the greater part of the earth's crust ; and wo also understand the source of the salts of the ocean. "^• § 82. A further development of this view would lead us too iar for the scope of this paper. It will however be seen that the first precipitates from the ocean would contain most of the metals, and that in the subsequent re-solution and deposition of these pre- cipitates is to be found an explanation of the origin of metalliferous deposits, and of their distribution in various formations; cither as integral parts of the strata, or as deposits in veins, the former channels of mineral springs. In an essay on American Geology, published in SilUman's Journal in 1861, [2] xxxi, 405, I have nlready sketched the outlines of what I conceive to bo the true theory of metalliferous deposits, a subject to which it is proposed soon to return. Montreal, July 4, 1865. * Canadian Journal, May, 1859, 201, and Silliman's Journal, [2]xsv-, 102 ; also Comptes Rendus, Juue Otli, 1862, and Can. Naturalist, vii, 202. JOHN LOVELL, PKINTKH, ST. NICHOLAS STREET, MONTREAL.