# Transactions of the Geological Society, 1st series, vol. 1/On an Aluminous Chalybeate Spring in the Isle of Wight

IX. A Chemical Account of an Aluminous Chalybeate Spring in the Isle of Wight.
By Alexander Marcet, M.D. F.R.S.
One of the Physicians to Guy's Hospital, and Member of the Geological Society.

The accurate analysis of a mineral water, although attended with considerable difficulty and labour, must be allowed, in a general point of view, to be an object of so little importance, that unless there be some interesting medical question to investigate, or some new analytical methods to point out in the course of the inquiry, it may be questioned whether researches of this kind are worth the time and attention which they require, or deserve to be placed amongst the records of natural science.

Having thought it necessary, in the present essay, to confine myself to the natural and chemical history of the spring in question, without any digression upon its medicinal qualities, and being well aware that chemical details are considered by geologists merely as collateral objects, some apology may be required for the length of this communication. But if the relation which the history of mineral waters bears to geological and mineralogical inquiries, and the peculiarities of composition for which this spring is remarkable, entitle the subject to the attention of this Society, I hope that the general views and investigations which I have occasionally introduced respecting the analysis of mineral waters, and the composition of several salts connected with this inquiry, will be deemed a sufficient excuse for having thus expanded an account from which they were almost inseparable.

It is about two years since my attention was directed to this chalybeate spring by Dr. Saunders, to whom in consequence of his valuable treatise on mineral waters, inquiries of this kind are frequently referred. Having been requested by him, and soon afterwards by the discoverer of the spring, Mr. Waterworth, Surgeon, of Newport, to examine this water, I soon perceived by a few preliminary experiments, that its principal ingredients were sulphat of iron and sulphat of alumine, and that it possessed a degree of strength far more considerable than any mineral water of the same kind that had ever come to my knowledge.

This last circumstance, and the probability that this spring might some day attract public notice from its medicinal properties, induced me to undertake the present analysis, which, after many interruptions, I have at length brought to a conclusion.

§ 1. Situation and Natural History of the Spring.

This spring is situated on the south-west coast of the Isle of Wight, about two miles to the westward of Niton,[1] in one of those romantic spots for which that coast is so remarkable.

In its present state it may be said to be of difficult access, for there is no carriage road, nor even any regular foot-path along the cliff leading to it, and the walk would appear somewhat arduous to those unaccustomed to pedestrian excursions. But it would be practicable, and probably not very expensive, to render this path equally easy and agreeable. It was in walking along the shore, a few years ago, that Mr. Waterworth's attention was accidentally directed to this spring, which he traced to its present source, by observing black stains formed by rivulets flowing from that spot.

With regard to the mineralogical history of that district, I have been favoured through the kindness of my friend Dr. Berger, who visited the spot very lately, with so much more accurate an account of it than I should, from my own observation, have been able to offer, that I shall make no apology for transcribing it in his own words.

“ The albuminous chalybeate spring,” says Dr. Berger, “ issues from the cliff on. the S.S.W. coast of the Isle of Wight, below St. Catherine's Sea Mark, in the parish of Chale. The bearing of the needles from the spot is N.W. while that of Rockenend, not far distant, is S.E. by S.”

“ The elevation of the spot, as far as I could ascertain it by the barometer, is one hundred and thirty feet above the level of the sea. Its distance from the shore may be about one hundred and fifty yards. ”

“ The water is received into a bason formed in the rock for that purpose, and flows, as I was informed, at the rate of two or three hogsheads in a day. Its temperature I found to be 51°, that of the atmosphere being 48°; and it may be worth while to observe that this temperature corresponds with that of several springs of pure water which I have met with in the island.”

“ The lower part of the cliff is rather incumbered with masses of rock, or portions of soil, which have fallen from the upper strata. Immediately above these, the spring issues from a bed of loose quartzose sandstone containing oxyd of iron. This sand, in which vestiges of vegetable matter are discoverable,[2] alternates with a purplish argillaceous slate of a fine grain, disposed in this layers with a few speck; of silvery mica interspersed through the mass. Black stains, or impressions of vegetables, are seen on the natural joints of this rock. Above this, lies a stratum of several fathoms in thickness, of a bluish calcareous marl, with specks of mica, which has an earthy and friable texture, and contains imbedded nodules or kidneys of sulphuret of iron. Many of these nodules have undergone a partial decomposition, to which, no doubt, the existence of the principal ingredients of the spring is to be ascribed. The upper strata of the cliff are composed of a calcareous free stone, alternating with a coarse shelly limestone, accompanied by nodules or layers of chert or flint.”

“ As the same arrangement of rocks here observed prevails in several other parts of the Isle of Wight, and even along the coast of Hampshire, it is not improbable that other springs of a similar nature might be discovered. May not Alum Bay, which lied to the north of the Needles, have derived its name from a circumstance of this kind?”

“ On the road from Shorwell to Chale, the soil consists of a ferruginous sandstone, and chalybeate iridescent waters are to be seen in several places; To the east of Fresh-water Bay, not far from the place where the cliffs of chalk begin to make their appearance, there is a rivulet, the taste of which strongly indicates the presence of iron. At Blackgang Chine, a little to the N.W. of the aluminous chalybeate, is another ferruginous. stream running to the sea. The rock there, is a sort of decomposed iron-stone under the form of balls. The sound compact iron-stone, having the appearance of flat pebbles worn by the rolling of the sea, occurs not infrequently along the shore.”

§ II. General Qualities and specific Gravity of the Water

a. The water issues from the sand rock above described, perfectly transparent, and it continues so for any length of time, provided it be collected immediately, and preserved in perfectly closed vessels; but if allowed to remain in contact with the air, or even if corked up after a temporary exposure to it, reddish flakes are soon deposited, which partly subside, and partly adhere to the inside of the vessel.

b. It has no smell, except that which is common to all chalybeates, and this it possesses but in a very slight degree.

c'. Its taste is intensely chalybeate, and, besides a considerable degree of astringency and harshness, it has the peculiar kind of sweetness which sulphat of iron and sulphat of alumina are known to possess.

d. Its specific gravity somewhat varies in different specimens. In three different trials I obtained the following results:

 1st specimen 1008,3 2d specimen 1007,2 3d specimen 1006,9 ───── 3022,4 which gives a mean specific gravity of 1007,5

§ III. Preliminary Experiments an the ejects of Reagents.

A. Paper stained with litmus was distinctly reddened by the water.

B. Paper stained with Brazil-wood was changed to a deep purple.

C. When agitated in contact with the air, or repeatedly poured from one vessel into another, the water became turbid, and on standing deposited reddish Hakes.

D. On applying heat to a portion of the water just uncorked, and boiling it quickly, till it was reduced to one half or even one third of its original bulk, no precipitation whatever took place; but on continuing the evaporation, a white feathery crystalline substance appeared on the surface of the fluid, and on pushing the process still further, a saline matter of a pale yellowish green colour appeared, which continued to increase till the whole was reduced to a dry yellowish mass. These were the phenomena observed with water recently uncorked; but when, previous to the evaporation, it had been for some time exposed to the air, or when the evaporation was conducted very slowly, an appearance of reddish flakes was the first circumstance observed.

E. The mineral acids produced no obvious change in the water.

F. Oxalic acid produced a slight yellowish tinge; but no immediate precipitation or turbidness.

G. Oxalat of ammonia, in small quantity, likewise produced a yellow colour, without precipitate: but on adding more of this test a white precipitate appeared.

H. Prussiat of potash and infusion of galls produced abundant precipitates, the one blue, and the other black or dark purple; and the colour of these precipitates was much paler when the water had not previously been exposed to the atmosphere.

I. Alkaline solutions produced copious greenish flocculent precipitates, which became darker on standing in the air.

K. Nitrat of silver occasioned a dense, white, but not considerable precipitate.

L. Both muriat and nitrat of barytes occasioned copious white precipitates.

M. A piece of marble being boiled for some time in a few ounces of the water, the marble was found to have undergone no sensible loss of weight by that operation; but its surface had acquired a faint yellowish tinge.

N. A quantity of the water being evaporated to dryness, and a considerable degree of heat applied to the dry residue, a solution of this in water had the same effect of reddening litmus as before.

§ IV. Inferences arising from those Effects.

1. From experiment A, connected with experiments C, H, I, M and N, and from the circumstance of taste, and other general properties, it appeared highly probable that the water contained sulphat of iron, and perhaps also sulphat of alumine, without any uncombined acid.[3]

2. From experiments C and D, it appeared evident that iron and lime were contained in the water, and that their solvent was not carbonic acid.[4]

3. The experiments D and E concurred to show that the water did not contain any sensible quantity of carbonates.

4. The experiments F and G afforded additional evidence of the presence of iron, and whilst they shewed the existence of lime in the water, seemed to indicate that the quantity of this earth was not considerable.

5. It appeared probable from experiment K, that the water contained a small quantity of muriatic acid.

6. The change produced in experiment B, on the infusion of Brazil-wood, appeared at first ambiguous; it could not be owing to the prevalence of an alkali or carbonated earth, since the water turned litmus red, and since the presence of carbonated earths had been disproved by other results. But having found by comparative trials, that solutions of sulphat of iron changed paper stained with infusions of Brazil-wood to a black, or at least intensely dark violet colour, and that solutions of alum turned it crimson; and observing that a mixture of these solutions produced a dark purple hue, the appearance in question was easily explained.

7. The result of experiment L indicated the presence of sulphuric acid.

8. Upon the whole, and from a review of the foregoing experiments, the substances which, at this early stage of the analysis, the water appeared most likely to contain, were sulphat of iron, sulphat of alumine, sulphat of lime, and a small quantity of muriatic salts. Some sulphat of magnesia, and some alkaline sulphats, might possibly be contained in the water, though their presence could not be satisfactorily ascertained by these preliminary experiments.

§ V. Gaseous Contents of the Water.

A quantity of the water measuring ten cubic inches, being boiled briskly over mercury, the gas given out, together with the air contained in the apparatus, was received in a graduated tube; on admitting caustic alkali into the tube, one-tenth of a cubic inch of gas was absorbed. It appears therefore that one hundred cubic inches of the water contain one cubic inch of carbonic acid gas, which is equivalent to about three-tenth of a cubic inch to each pint. The water was uncorked at the moment of being examined, but I had not an opportunity oft ascertaining the quantity of gas.

§ VI. Evaporation of the Water, and Estimation of the Quantity of solid Ingredients.

1. Sixteen ounces of the water by measure, being evaporated down to a soft mass over a lamp, and afterwards desiccated in a drying apparatus at the heat of 180°,[5] the solid mass weighed eighty-six grains. During the evaporation the same appearances were observed as have been already related (in § III., D,) and the dry saline mass assumed a pale greenish colour. On standing in the air, it slightly deliquesced, and its colour became somewhat darker. This saline mass, though slowly evaporated, never assumed a distinct crystalline appearance.

2. I have stated before (§ II. d.) that some difference prevailed in the specific gravity of the several specimens of the water which were examined. A similar want of uniformity was observed in regard to the quantity of solid ingredients, as will appear from the following statement :[6]

 grains The 1st specimen yielded 86. ${\displaystyle \left.{\begin{matrix}\ \\\ \\\ \\\ \\\ \\\ \\\ \\\ \\\ \end{matrix}}\right\}}$In the pint of sixteen ounces. 2d 92. 3d 63.6 4th 80.4 5th 82.8 6th 77.2 7th 84. [7] 8th 78. ───── 644

These eight results therefore give 80,5 grs. dried at 180°, as the average quantity of solid ingredients in each pint of the water.

In the first of these trials, a whole pint was evaporated; but in the

§ VII. Of the different Methods of Analysis applicable to the present Inquiry.

In analysing a mineral water, two modes of proceeding occur from the very first. We may either evaporate the water first, and apply our reagents to the solid residue; or operate at once upon the water itself. The former plan is in general found expedient when the quantity of solid contents of the water is small; but when, as in the present instance, the impregnation is considerable, it may be more convenient to adopt the latter method. But at all events, as the re-dissolution of the solid residue, when the first mode of proceeding is resorted to, generally requires the introduction of an acid, which may modify or complicate the process, it is always desirable that both methods should be tried in succession, in order to obtain comparative results.

We may also, if necessary, precipitate from the same portion of the water the several ingredients which it contains, by applying to it in succession their respective reagents; or if our supply be considerable, we may use a fresh portion of it for each successive operation, a mode of proceeding which is generally preferable. No difficulty being experienced during the present inquiry in regard to the supply of water, a variety of methods was tried, with the details of which I shall not trouble the Society: but in order to convey a general idea of them, and in hopes that a summary review of this kind may afford some assistance to chemical inquirers not yet accustomed to researches of this nature, I shall briefly enumerate the different plans which presented themselves at this period of the analysis, and it will be seen afterwards how these plans were gradually modified.

1st method. To precipitate in succession from a known quantity of the water, the iron by prussiat of potash—the lime by oxalat of ammonia—the alumine and magnesia by caustic potash, which, by boiling, re-dissolves the alumine and leaves the magnesia untouched.

2d method. To precipitate the iron and earths by subcarbonat of ammonia. To evaporate the remaining clear solution to dryness, and apply a red heat. To re-dissolve this saline residue, and evaporate the solution slowly, in order to discover any fixed alkaline sulphat or muriat which may exist in the water. To boil in caustic potash the precipitate containing the iron and earths, in order to separate the alumine and silica. To dissolve the remaining mass (supposed to contain iron, lime, and magnesia.) in nitric acid, evaporate to dryness, and apply a red heat, in order to render the peroxyd of iron thus formed insoluble in acid. To add to the mass minutely pulverized, nitric or acetic acid, as either of these acids will only dissolve the lime and magnesia, which may be separately obtained by their respective reagents. And lastly, to ascertain the quantity of oxyd of iron, supposed to have been left untouched by the acid.

3d method. To precipitate from another portion of water, the iron, lime, alumine and silica, by a solution of neutral carbonat of ammonia, which reagent retains the magnesia in solution. To boil the precipitate in caustic potash, which takes up the alumine and silica. To re-dissolve in muriatic acid the residue not taken up by potash, which consists of lime and iron—separate the iron by pure ammonia, and the lime by oxalat of ammonia.[8] Precipitate the magnesia[9] from the clear solution by an alkaline phosphat.

4th method. To evaporate to dryness a known quantity of the water and to boil the residue in caustic potash, which will dissolve the alumina and silica, both of which may be precipitated again by muriat of ammonia.[10] Treat the residue, insoluble in potash and supposed to contain iron, lime and magnesia, in the manner pointed out in the 2d method.

5th method. After having obtained by the preceding methods a knowledge of the proportions of iron and earthy substances, and formed an estimate of the nature and quantities of acids with which they are united, to ascertain in a direct manner the quantities of acids by their respective reagents, with a view to obtain a confirmation of the preceding results.

6th method. To boil a known quantity of the water in succinat of ammonia, till all the iron and alumine are precipitated—edulcorate, precipitate and separate the alumine from the iron by boiling in caustic potash. From the clear concentrated fluid, to separate the lime by oxalat of ammonia, and the magnesia by pure ammonia; to evaporate the remaining clear fluid to dryness, and to apply a red heat, in order to burn or volatilize any remaining portions of the tests used in the processes above described. To re-dissolve the residue in order to ascertain by subsequent evaporation the presence and quantity of sulphat and muriat of soda.[11]

7th method. To boil a known quantity of residue of the water, in alcohol, in order to ascertain what salts it may contain which are soluble in that menstruum.

Although I found it expedient, before advancing farther in the examination of the water, and in order to regulate my steps in the progress of the inquiry, thus to trace the various plans which seemed adapted to the purpose, yet I apprehend it would be superfluous to detail here in regular succession all the trials which arose from these different methods. I shall therefore confine myself to such as belong more immediately to my object; and in relating them, shall consider singly and under separate heads, the various ingredients of the water, stating, as I proceed, the proportions in which they were ultimately obtained.

§ VIII. Sulphat of Iron.

The presence of iron, in the state of sulphat, having been abundantly proved by the preliminary experiments, the next step was, to ascertain the proportion of this salt in a given quantity of the water. The first reagent which I tried for this purpose, was prussiat of potash; but after many trials which afforded uncertain and discordant results, I convinced myself that this test, however useful for detecting the presence of iron, is quite inappropriate when our object is to ascertain the quantity of that substance.[12]

1. Fifty grains of residue[13] dried at the temperature of between 170 and 180, (as described in § VI.) and therefore equal to ten ounces of the water, were boiled in successive solutions of potash, so as to saturate all the acid contained in that residue, and to dissolve the alumine. The remaining solid residue (which had passed first to a dark green, and some hours afterwards to a dark brown or nearly black colour) was dissolved in nitric acid and the solution evaporated to dryness, after which a red heat was applied in order to bring the iron to the state of peroxyd, and thus render it insoluble in the same acid. The mass being now treated with nitric acid, in order to separate the lime and magnesia supposed to be mixed with the oxyd of iron, and the whole being thrown into a filter, the clear solution was found still to contain a good deal of iron. This last solution was, like the former, evaporated to dryness, and to the residue, again heated to redness, acetic,[14] instead of nitric acid, was this time added, and the solution filtered. The filtered fluid still contained a quantity of iron, which however, from subsequent examination, appeared very inconsiderable. The oxyd of iron left in the filter being roasted with wax and heated to redness, in order to bring it to an uniform state of oxidation, weighed 6,8 grs.[15]

2. With a view to repeat and vary the last experiment, another portion of residue, also weighing 50 grs. was thrown into a solution of neutral carbonat of ammonia, the quantity of the latter being more than sufficient to saturate any acid present, and to dissolve the magnesia suspected to exist in that residue. A considerable effervescence took place. The mixture, after this, was gently heated and filtered. The residue left in the filter was of a pale yellowish-brown colour. The dear solution deposited on standing a small quantity of precipitate similar to the residue left in the filter, to which residue this precipitate was added. The contents of the filter were then treated with potash, in the manner before described (§ VIII. 1), in order to separate the alumine, after which the residue (now supposed to contain nothing but carbonat of lime and iron) was treated with dilute muriatic acid, which dissolved it with effervescence. From this solution, the lime was precipitated by oxalat of ammonia, and the remaining liquor, now containing nothing but muriat of iron, was treated with carbonat of ammonia, so as to precipitate the whole of the iron, which, in subsiding, assumed a pale reddish colour. The clear fluid being decanted off, and the precipitate carefully washed, dried, and ultimately heated to redness with a little wax in a platina crucible, weighed 7,2 grs.

3. It will be observed that between this and the former result there was a difference of 0,4 grs. in the quantity of oxyd of iron contained in 50 grs. of residue. But when it is considered that in the first of these analyses, a small quantity of iron was positively detected in the acetic solution, which, from the best estimate I could make, would have brought the quantity of iron very near that obtained in the second process, it will readily be admitted that the coincidence was such as to authorize me to consider the last result as sufficiently accurate.[16]

4. If therefore we consider 7,2 grs. of peroxyd of iron, as the quantity of this metal contained in 50 grs. of the residue, which corresponds to 11,59 grs. of the oxyd for 80,5 grs. of residue (that is for each pint of the water, according to the average before established, § VI. 2), we shall be able to infer the quantity of sulphat of iron contained in the water.

5. In order to do this, however, it was necessary to ascertain by a comparative experiment the proportion of oxyd which a known quantity of sulphat of iron yields by a process similar to that which I have just described. For this purpose, 50 grs. of transparent crystallized green sulphat of iron were dissolved in water, and treated with carbonat of ammonia as long as any precipitate appeared. This precipitate, after being carefully separated, edulcorated, dried, and ultimately heated to redness with wax in a platina crucible, weighed exactly 14 grs. It appeared in the form of a red-brown magnetic powder.[17]

6. Since therefore 50 grs. of crystallized green sulphat of iron gave 14 grs. of this oxyd, the 7,2 grs. of oxyd obtained from 50 grs. of residue, would represent 25,7 grs. of green sulphat of iron; and 11,59 grs. of oxyd (which is the quantity contained in an English pint of the water), would represent 41,4 grs. of that salt.

§ IX. Sulphat of Alumine.

1. Fifty grains of residue [18] were boiled in two successive lixivia of caustic potash (as in § VIII. 1), so as to take up all the alumine present ; the residue was separated and well washed, and the washings were added to the alkaline solution. The clear liquor had a brownish colour, and on being tried with muriatic acid and prussiat of potash, a blue tinge was produced, which appeared to have arisen from a few particles of oxyd of iron which were suspended in the lixivium rather than actually dissolved; for the solution being left at rest for some time, these particles subsided.

2. To the clear alkaline solution, muriat of ammonia was added, till no further precipitate took place; the precipitate was edulcorated and collected in a filter. It was white and gelatinous. Caustic potash being added to the clear fluid, ammonia was disengaged, showing that it contained an excess of muriat of ammonia; and acetic acid being added to another portion of the same liquor, no turbidness appeared, both circumstances showing that all the alumine was precipitated. This precipitate being dissolved in muriatic acid, in order to separate a minute portion of silica, which it contained[19], and being again precipitated by succinat of ammonia with excess of ammonia, formed a gelatinous mass, which being edulcorated, dried, and ultimately heated to redness, weighed 2,4 grains.

3. Another portion of residue, weighing thirty grains, being treated in a manner exactly similar to that just described, with this exception, that the redissolution of the alumine in muriatic acid and its subsequent precipitation by succinat of ammonia, were omitted, the gelatinous precipitate, heated to redness, weighed 1,4 grains[20], which afforded as close a coincidence with the former result as may be well expected: in processes of this kind.

4. Having never been able to obtain, by the mere evaporation of the water, any appearance of crystals resembling alum, I was desirous, for the sake of obtaining further evidence on the subject, to bring the sulphat of alumine to a crystallized state, by artificially supplying what I conceived to be wanting for the completion of that process. For this purpose, having dissolved about thirty grains of residue in distilled water, I added to the filtered solution two or three drops of a solution of carbonat of potash, and evaporated it very slowly; crystals were thus obtained, dispersed in the saline mass, which, though of a size scarcely exceeding that of a pin's head, had a distinct octahedral form, and when separated and chemically examined, had all the properties of alum.

5. With regard to the proportion of sulphat of alumina, contained in the water, it will be seen, that by connecting together the results of the experiments just related (1, 2, 3), eighty grains of residue, or a pint of the water, yield 3,8 grains of alumine heated to redness, which, according to the proportion of twelve parts of ignited alumine, in one hundred parts of crystallized alum[21], would be equivalent to 31,6 grains of alum in each pint of the water[22].

§ X. Sulphat of Lime.

1. Some of the former experiments (§ III. d and g) had shown, beyond all doubt, the presence of selenite, and indeed; from the general composition of the water, lime could scarcely be supposed to exist in it in any other form of combination.

To ascertain the quantity of this substance, a variety of methods was used, the principal results of which I shall cursorily relate.

2. It would have been in vain, in this instance, to have applied, without any previous step, oxalat of ammonia, the usual test of limes in order to obtain an accurate estimate of the quantity of lime present in the water; for as oxalic acid also acts upon iron, some ambiguity would necessarily have occurred. Indeed that oxalat of ammonia did not, in this case, re-act upon the lime in the manner that it usually does, had been noticed (§ III f, g) in some of the preliminary experiments[23]

3. It was therefore necessary to separate the iron previous to the precipitation of the lime. This was done in one instance by prussiat of potash, and in another by succinat of ammonia. I shall not trouble the society with the detail of these operations. It will be sufficient to state, that the two most unexceptionable experiments indicated the one 8 grains, and the other 8,3 grains of oxalat of lime, dried at 160°, for each pint of the water, making an average of 8,15 grains of oxalat of lime, or 10,17 grains of sulphat of lime dried at 160°; or 7,94 grains of the same salt dried at a red heat[24].

§ XI. Inference: obtained from the application od Alcohol.

1. Having ascertained (§ III. k), that a small quantity of muriatic acid was present in the water, it became desirable, before proceeding any farther, to discover, by the agency of alcohol, which has the well known property of dissolving the earthy muriats, with what bases this acid was combined. With this view, 20 grains of residue were digested in successive quantities of alcohol of great purity, and the solution filtered. The residue, by this operation, acquired a lighter colour and a more pulverulent appearance. Part of this residue being treated with muriatic acid and oxalat of ammonia, oxalat of lime was precipitated ; and another portion being treated with neutral carbonat of ammonia and phosphat of soda, some magnesia was precipitated in the form of triple phosphat, circumstances which confirmed the presence of lime in the form of selenite, and that of magnesia, in the form of sulphat or Epsom salt.

2. The alcoholic solution being evaporated to dryness, a yellowish deliquescent residue was obtained, which, being dried at 160° weighed 0,9 grains. Water being added to this residue, a small portion of it remained undissolved. The filtered watery solution was yellowish, though perfectly transparent, and being examined by the usual reagents, appeared to contain iron, sulphuric acid and muriatic acid, with imponderable vestiges of lime and magnesia, without any trace of alumine.

3. From these circumstances, it was inferred that the only deliquescent salts yielded by the residue, in ascertainable quantities, were sulphat of iron, and muriat of iron, both of which had probably been formed in consequence of some new orders of attraction taking place during the process of evaporation to which the water had been subjected.[25]

§ XII Sulphat of Magnesia.

1. The presence of magnesia[26] was ascertained beyond all doubt, in the following manner:

50 grains of residue minutely pulverized, were boiled in a solution of neutral carbonat of ammonia, so as to decompose all the sulphat of iron and earthy salts, and dissolve all the magnesia which might be present.[27] This process was, of course, attended with considerable effervescence, and when this had subsided, the liquor was filtered. The clear solution deposited on standing a brownish sediment, which was separated and proved to be oxyd of iron. The residue left in the filter had passed from a greenish-yellow to a pale brown colour.

2. Phosphat of ammonia being added to the clear solution, a precipitate appeared, having all the characters of the ammoniaco-magnesian phosphat, and in particular, that of forming white stripes on the inside of the vessel when scratched with a pointed instrument. This precipitate dried at a temperature of about 120°, weighed 1,9 grains,[28] and being made red hot in a platina crucible, was reduced to exactly 1 gr. = 0,385 grains of pure magnesia = 2,26 grains of crystallized sulphat of magnesia in 50 grains of residue, or 3,63 grains in a pint of the water[29] The magnesia phosphat became slightly brownish during the calcination, owing to the presence of a few particles of iron, the quantity of which was too minute to be ascertained.

§ XIII. Precipitation of the sulphuric and muriatic Acids, with a view to ascertain their Quantity.

Before drawing any ultimate conclusion respecting the contents of the water and the proportions of its ingredients, I found it necessary to ascertain the quantities of sulphuric and muriatic acids which it contained, in order to enable me to try how far these quantities might coincide with the conclusions obtained by the separation of the basis, and also to assist me, as will be seen hereafter, in forming certain inferences with regard to the alkaline salts. For this purpose I made the following experiments.

1. To four ounces of the water was added nitrat of barytes till the whole of the sulphuric acid was precipitated; the sulphat of barytes thus obtained being carefully edulcorated and heated to redness in a platina crucible, weighed 18,5 grains, which correspond to 74 grains of sulphat of barytes from a pint of the water.

2. Four ounces of the water were treated with nitrat of silver as long as any precipitate appeared, and the muriat of silver thus obtained, being well edulcorated, and afterwards brought to a state of incipient fusion by the heat of an Argand lamp, weighed 2,05, which is equivalent to 8,2 grains of luna cornes, or four grains of muriat of soda,[30] in each pint of the water.[31]

§ XIV. Sulphat and Muriat of Soda.

1. The mode in which I first attempted to ascertain the presence of alkaline salts in the water, was that alluded to in a former part of this paper, which consisted in precipitating the iron and the earths by subcarbonat of ammonia, evaporating the clear solution to dryness, heating the dry mass to redness, with a view to drive of the sulphats and muriats of ammonia, redissolving the residue in water, and evaporating again very slowly in order to obtain crystals. But the saline mass yielded by this process did not crystallize regularly, and on being examined by reagents, was found to contain only sulphat of soda, with minute quantities of sulphats of alumine and magnesia, which had escaped the action of the carbonat of ammonia.

2. In hopes of obtaining more satisfactory results I had recourse to the following process: five ounces of the water were boiled with a solution of succinat of ammonia till the whole of the iron and alumine were precipitated.[32] The lime was precipitated by oxalat of ammonia, and the magnesia by ammonia. The solution was then concentrated over a lamp, and gradually evaporated to dryness in a platina crucible. A white pungent smell arose, and on raising the heat to redness, these fumes took fire and burnt with a blue flame, till the whole was fused and reduced to a fixed saline mass mixed with a black coaly matter. Distilled water was poured upon this. mass and the solution filtered. This clear solution being now evaporated and dried at a gentle heat, so as to obtain the salts in a crystallized state, the mass weighed 6,3 grains,[33] which would give 20 grains of alkaline salts in a pint of the water. The centre of this mass exhibited no distinct crystallization, though from its appearance and disposition to effloresce, it evidently contained sulphat of soda; but the circumference was strewed with numerous and perfectly regular crystals of muriat of soda.[34]

3. This saline mass being dissolved in water the solution had the following properties:

a. It was neither acid nor alkaline.

b. Its most obvious taste was that of muriat of soda.

c. It formed copious precipitates with nitrat of barytes, nitrat of silver, and nitrat of lime.

d. Oxymuriat of platina, oxalat of ammonia, and prussiat of potash, produced no precipitate whatever.

Therefore the only salts contained in this solution were sulphat of soda, and muriat of soda.

4. As to the proportions of those two salts, it would have been easy to ascertain them by precipitating their acids. But it occurred to me that the sulphat of ammonia formed in the solution by the ammoniacal salts which had been introduced for the precipitation of the earths, had probably reacted upon the muriat of soda when urged by heat, so as to decompose it partially, and form the sulphat of soda obtained by the process just described; so that muriat of soda might perhaps in fact be the only alkaline salt contained in the water.

5. In order to ascertain this, another portion of the chalybeate having been treated in the way just described with succinat of ammonia, the residue was gradually desiccated, and then heated to redness in a platina crucible, which was at first kept closed, in order to retard the escape of the sulphat of ammonia, and thus promote its action on the muriat of soda. The remaining mass being dissolved and very slowly crystallized, assumed the form of clusters of regular prismatic efflorescent crystals of sulphat of soda, amongst which scarcely any vestige of muriat of soda could be discovered.

6. The decomposition of muriat of soda by the above process being thus well established, it became necessary to determine the proportions of sulphat and muriat of soda by some less direct method; and the expedient which appeared the most appropriate was that of inferring the point in question from a reference to the quantities of acids as estimated in the preceding section. Thus as it was obvious that, whatever the case might be with regard to sulphat of soda, the presence of muriat of soda in the water was unquestionable; and as the whole quantity of muriatic acid discovered in the water (§ XIII. 2), corresponded to a quantity of muriat of soda which fell far short of the sum total of alkaline salts, I naturally inferred that the whole of the muriatic acid was united with soda, and that the water must also contain a quantity of sulphat of soda sufficient to complete the 20 grains of alkaline salts which the experiments just related had shewn to exist in each pint of the water.

7. Since therefore, the whole of the muriat of soda, as was before computed (§ XIII. 2), amounted only to 4 grains in a pint, the quantity of crystallized sulphat of soda contained in each pint of the water will be 16 grains.

§ XV. Comparison of the quantities of Acid actually obtained from the water by precipitation, with the quantities inferred from the precipitation of the bases.

1. It appears evident, from. all that precedes, that the only acids contained in the water are the sulphuric and muriatic. The whole of the muriatic acid having been shewn to exist in the form of muriat of soda, nothing further remains to be said on this head. But it will be curious to examine how far the total amount of sulphuric acid, obtained from a portion of the water, would coincide with that which might be inferred from the quantities of bases with which it was combined. This inquiry will give rise to the statement of certain results respecting the proportions of acid and base in some of the salts concerned, and the precipitates obtained from their decomposition, which, from their general import in chemical analysis, appear to deserve some attention.

2. It was ascertained by a direct experiment (§ XIII. 1), that the whole of the sulphuric acid contained in a pint of the water, formed, when precipitated by a barytic salt, a quantity of sulphat of barytes, which, after being ignited, weighed 74 grains.

I shall now recapitulate the several sulphats discovered in the water, and from the quantities of each, compute the quantities of barytic sulphat which would result from their decomposition.

Sulphate contained in a pint of the Water.
 sulph. of baryt,ignited Sulphat of iron (§ VIII. 9) 41,4 grs. crystallized = 31,8 grs. [35] Sulphat of alumine (§ IX. 5) 3,8 grs. ignited alumine = 17,7 ditto[36] Sulphat of lime (§ X. 4) 10,17 grs. dried at 160° = 13,9 ditto[37] Sulphat of magnesia (§ XII. 2) 3,63 grs. crystallized = 4,0 ditto[38] Sulphat of soda (§ XIV. 7) 16,0 grs. crystallized = 11,6 ditto[39] ─── Total amount of the sulphat of barytes 79,0 grs.

3. It appears therefore that the aggregate of the analytical results would indicate 79 grs. of ignited sulphat of barytes, instead of the 74 grs. obtained by a single direct operation. This difference I apprehend to be in a great degree owing to my estimate of the proportion of acid in sulphat of alumine being over-rated, from the circumstance of not having been able to obtain a neutral sulphat of alumine in the experiment just relaxed from which that estimate was deduced.

§ XVI. Silica.

I. During the various solutions of the residue in acid, I had repeatedly observed that, besides the selenite, (the solution of which was attended with some difficulty, and required a considerable quantity of water) there always remained a small proportion of earthy matter, which resisted all solvents, caustic potash excepted. This insoluble matter, I had thought from some of the first trials, amounted to about 1 gr. in 100 of the residue; but from some subsequent experiments in which the silica was separated by caustic potash, there appeared to be reason to suppose that this estimate was rather over-rated. I shall relate the process, to which, after various trials, I gave the preference.

2. 50 grains of residue being boiled with very dilute muriatic acid, a white flocculent substance remained undissolved, upon which neither acid nor water could make any impression. This substance being separated and boiled in a solution of caustic potash, readily redissolved with the exception of a few particles of highly oxydated iron which subsided. Muriat of ammonia[40] being added to the clear alkaline solution in sufficient quantity to saturate the whole of the potash with muriatic acid, the white flocculent substance reappeared, which, after being well washed and heated to redness, weighed between 0,3 and 0,4 gr. This substance when heated with alkali ran into a vitreous globule, and muriatic acid being poured upon this, the alkali was dissolved, and the earthy matter remained untouched. It was therefore silica, the quantity of which may be estimated at 0,7 gr. in a pint of water.[41]

§ XVII. Conclusion.

On reviewing and connecting together all the foregoing results, it appears that each pint, or sixteen-ounce measure of the aluminous chalybeate, contains the following ingredients:

 Of carbonic acid gas three-tenths of a cubic inch. granns Sulphat of iron, in the state of crystallized green sulphat 41,4 Sulphat of alumine, a quantity which if brought to the state of crystallized alum, would amount to 31,6 Sulphat of lime, dried at 160° 10,1 Sulphat of magnesia, or Epsom salt, crystallized 3,6 Sulphat of soda, or Glauber salt, crystallized 16,0 Muriat of soda, or common salt, crystallized 4,0 ─── 107,4

I am not acquainted with any chalybeate or albuminous spring, in the chemical history of mineral waters, which can be compared, in regard to strength, with that just described. The Hartfell water, and that of the Horley-green spaw near Halifax, both of which appear to be analogous to this in their chemical composition, and were considered as the strongest impregnations of the kind, are stated by Dr. Garnett to contain, the one only about 14 grs. and the other 40 grs. of saline matter in each pint.

No doubt therefore can be entertained that the water which is the subject of this essay, will be found to possess in a very eminent degree the medical properties which are known to belong to the saline substances it contains. Indeed there appears to be in that spring rather a redundance than a deficiency of power, and it is probable that in many instances it will be found expedient to drink the water in a diluted state; whilst in others, when it may be desirable to take in a small compass large doses of these saline substances, it will be preferred in its native undiminished strength.

INDEX.
 § I. Situation and Natural History of the Spring § II. General qualities and specific gravity of the Water § III. Preliminary experiments on the effects of Reagents § IV. Inferences arising from those effects § V. Gaseous contents of the water § VI. Evaporation of the water, and estimation of the quantity of solid ingredients § VII. Of the different methods of analysis applicable to the present inquiry § VIII. Sulphat of Iron § IX. Sulphat of Alumine § X. Sulphat of Lime § XI. Inferences obtained from the application of Alcohol § XII. Sulphat of Magnesia § XIII. Precipitation of the Sulphuric and Muriatic Acids, with a view to ascertain their quantity § XIV. Sulphat and Muriat of Soda § XV. Comparison of the quantities of Acid actually obtained from the water by precipitation, with the quantities obtained from the precipitation of the bases § XVI. Silica § XVII. Conclusion

1. On an estate belonging to Michael Hoy, Esq.
2. On being sprinkled on heated shovel, this sand scintillates as if undergoing a partial combustion. When submitted to chemical analysis, it yields a quantity of iron, but no lime, nor alumina, nor any other earthy matter soluble in an acid. Close to the spring, this sand contains some traces of sulphuric acid, but not at a distance from its it is evident therefore that the sand-rock is not the medium through which the spring is impregnated.
3. Solutions of sulphat of iron and sulphat of alanine, though made from there salts in their crystallized state, have, like acids, the power of imparting s red colour to litmus.
4. The reddish flakes mentioned in C & D, and in § II, a, are uniformly found to be sub-sulphat of iron.
5. This is the heat I have usually employed for desiccation, because it is that which is obtained by the water-bath which I use, and can scarcely be raised higher by that apparatus. By a heat of 180° however, I generally mean some intermediate point between 170° and 180° for it is impossible to regulate the temperature with perfect accuracy.
6. subsequent ones, the quantity of water was diminished to eight, six, and sometimes only four ounces, all of which, for the sake of uniformity, I have reduced in the table to the common standard of the pint.
7. This specimen I brought myself from the spring; the others were sent me in sealed bottles from the Isle of Wight.
8. It is necessary to precipitate the iron before the lime, whenever any considerable quantity of sulphat or muriat of iron is present. For oxalat of ammonia acts upon solutions of iron, as will be fully explained under the head of sulphat of lime.
9. The magnesia might be equally, and perhaps more conveniently separated, by boiling a known quantity of the solid residue in the neutral carbonate of ammonia, Instead of applying this reagent to the water itself.
10. The mode in which the silica, my be separated from the alumina, will be detailed in a subsequent part of this paper.
11. This process is liable to an objection which will be hereafter fully stated, namely, that muriat of soda is decomposed by sulphat of ammonia at a high temperatur.
12. Prussiat of potash, as a precipitant of iron, is liable to the following objections.

1st. It is apt, although apparently well prepared and crystallized, to precipitate certain earthy substances, and in particular alumine; this I found distinctly to happen in two experiments in which the mixture was heated.

2dly. If the solutions be used cold, and if the metal be not highly oxydated, some of the Prussian blue unavoidably passes through the filters; or if no filters be used, it subsides but slowly and imperfectly.

3dly. If the solutions be heated, the prussiat of potash is itself decomposed, and yields quantity of oxyd of iron which vitiates the results.

13. By the word residue, thus generally used, is always meant the residue of the water under examination, dried at the temperature of between 170° and 180°. And in comparing a quantity of residue with a corresponding portion of the water, the average proportion of 80,5 grs. for each pint (§ VI. 2) is always assumed as the standard of comparison.
14. The acetic acid, as well as the nitric, is said to be incapable of dissolving any Iron, which has been peroxydated by the process just described. In this instance a few particles of oxyd were taken up by the acid; but it is probable that if, instead of heating the residue to redness only for a few minutes, the oxyd had continued exposed to a red heat for half an hour or more, the whole of it would have become insoluble.
15. It may be asked in what state of oxidation the iron is after that operation? It has generally been supposed to be reduced to the state of protoxyd in consequence of the affinity of the combustible matter for oxygen; but in an experiment which I made some years ago to ascertain that point, (the particulars of which my be seen in my account of the Brighton chalybeate) this process appeared to bring the iron to the state of peroxyd; for 100 parts of iron gave 147,6 parts of oxyd, proportions which are now considered as constituting the red oxyd of iron. And as a confirmation of this, I observe that Dr. Thomson in his valuable paper on the oxyds of iron, published in the twenty-seventh volume of Nicholson's Journal, states (page 379) that some of the red oxyd being mixed with oil and heated to redness till it became black and magnetic, no diminution of weight took place. Indeed I have always obtained by this process, not a black, but a brown oxyd, which in cooling passes to a red-brown colour, somewhat varying in shade, but mostly resembling powdered cinnamon, and being more or less magnetic.
16. In one experiment in which the iron was precipitated from a similar quantity of residue, by prussiat of potash, and the prussiat of iron roasted with wax, the quantity of oxyd obtained amounted to 11 grs. from which I infer either that a portion of the oxyd of iron, always contained in prussiat of potash, must have been precipitated with the Prussian blue, or that the prussiat of iron was not completely decomposed in the process in question, or that some earthy substance was precipitated along with the iron.
17. This result which was obtained in two different trials, with the variation of only 0,1 gr. corresponds exactly with the proportions given by Mr. Kirwan in his treatise on mineral waters (table iv.), in which 28 grs. is the quantity of oxyd stated to exist in 100 grs. of green sulphat. But in order to establish the perfect coincidence of these results, it would be necessary to know the process which Mr. Kirwan followed. The iron in his experiment, is stated to have been obtained in the state of black oxyd.
18. These fifty grains had been previously boiled in neutral carbonat of ammonia, in order to separate the magnesia as will be detailed hereafter. The previous intervention of a carbonated alkali renders the subsequent application of caustic potash for the separation of the alumine, more unexceptionable, as a solution of caustic potash might redissolve a small portion of the lime, if it were not previously carbonated.
19. The particulars of the manner in which the silica is separated, by the intervention of muriatic acid, will be detailed under the head Silica, in another part of this paper.
20. The real weight was 1,6 grains, but 0,2 grains were deducted, on account of the quantity of silica known, by other experiments, to have been present, as will be seen under the head Silica. It may be proper to mention, that the gelatinous precipitate, during its gradual desiccation, shrunk into small fragments resembling coarsely pulverised glue, an appearance which is well known to characterize alumine.
21. These are the proportions stated by Mr. Kirwan, and which I obtained myself on a former occasion (See the Analysis of the Brighton Chalybeate.)
22. It is scarcely necessary again to observe, that the sulphat of alumine contained it the water does not appear to exist there in the state of alum ; but it is perhaps better to express the quantity of alumine by the quantity of alum which it would form, as the crystallised state of a salt affords a much more precise standard of comparison.
23. By adding a considerable quantity of oxalat of ammonia, and concentrating the solutions by heat, the whole of the lime appeared to be precipitated, together with a portion of iron; but in order to obtain the oxalat of lime pure, it was necessary to calcine the precipitate so as to drive of the oxalic acid, to redissolve the residue in muriatic acid, and to precipitate the lime again by oxalat of ammonia. The small quantity of iron present did not than interfere, and this process, however circuitous, proved tolerably accurate.

I was drawn by this part of the subject into an experimental inquiry respecting the action of oxalat of ammonia on solutions of iron, and the unfitness of that test for the precipitation of lime when iron is present, the principal results of which I shall state summarily.

1. If to a strong solution of sulphat of iron, a small quantity of sulphat of lime be added, and then a little oxalat of ammonia, no precipitate or cloudiness appears ; whilst the same quantities of sulphat of lime and oxalat of ammonia added to a bulk of water equal to that of the solution of iron, instantly form a precipitate.

2. If oxalat of ammonia be added to a solution of sulphat of iron, a bright yellow colour is produced, and presently after this a copious white precipitate appears, which, in subsiding, assumes a pale lemon colour. If at the moment the cloud is forming, the vessel be scratched with any pointed instrument, white lines appear, as in the precipitation of magnesia from carbonic acid by phosphoric acid.

3. This precipitate being washed, and gently heated over a lamp, assumes a bright cinnamon colour, and becomes magnetic, in consequence, no doubt, of the carbonization of the oxalic acid, and these changes take place ata heat much inferior to ignition.

4. If a solution of potash be added to the washed precipitate, previous to the application of heat, a strong smell of ammonia arises, and the oxyd passes to a dark greyish colour, showing that the precipitate is a triple salt of oxalic acid, iron, and ammonia.

24. I avail myself, in forming these various estimates, of the proportions given by Dr. Henry, in his valuable 'Analysis of several varieties of Sea Salt' (published in the Philosophical Transactions for 1810, page 114), where he states that 100 grains of ignited sulphat of lime (which he finds to be equal to 128 grs. dried at 160°), give 102,5 grs. of oxalat of lime dried at 160°; so that 100 grs of oxalat of lime dried at 160°, correspond to 124 grs. of sulphat of lime dried at the same temperature.
25. Namely, the red sulphat from the hyper-oxygenation of the iron, and the muriat from the decomposition of muriat of soda, as will be explained hereafter.
26. The presence of this earth in the form of sulphat had already been proved by the application of alcohol, (§ XI. 1).
27. It is scarcely necessary again to state here the well known fact, that carbonat of ammonia, when fully saturated with carbonic acid, has the power of dissolving magnesia.
28. In a subsequent experiment in which the water itself, instead of the residue, was treated in the same manner with neutral carbonat of ammonia, the quantity of magnesia appeared somewhat greater; but the difference did not amount to more than one-tenth of a grain.
29. It will be necessary here to state the grounds of this computation, which will afford me an opportunity of relating some general results concerning the proportions in which magnesia and phosphoric acid combine.

By dissolving 11,82 grains of the purest magnesia (perfectly free from carbonic acid and water) in muriatic acid, and precipitating it by a mixture of phosphat of ammonia and neutral carbonat of ammonia, I obtained 65,8 grains of the triple phosphat dried by exposure for near forty-eight hours to a temperature which never exceeded 120°, a degree of heat under which this salt appears to retain the whole of its ammonia These 65,8 grains of triple salt being exposed for half an hour to a strong red heat in a platina crucible, were reduced to 30,8 grains. The salt appeared then in the form of a friable cake or loose aggregate, a fragment of which, on being urged by the blow-pipe, run into a white opaque vitreous globule, without any further diminution of weight. In its friable state it was readily dissolved by muriatic acid; in its vitrified form it required heat and trituration. This salt was perfectly tasteless and shewed no attraction for water. With regard to the proportions of acid and base to be inferred from this experiment, it is obvious that if 80,8 grains of phosphat of magnesia contain 11,82 grs. of earth, the remainder, viz. 18,98 grs. represents the proportion of phosphoric acid; which is equivalent to 38,37 grs. of magnesia, in 100 of phosphat. In another experiment conducted in a similar manner, the magnesia amounted to 38,7 grains, so that by taking the mean between these two very nearly similar results, we have the following proportions, viz.

 Magnesia 38,5 ${\displaystyle \left.{\begin{matrix}\ \\\ \end{matrix}}\right\}}$ in 100 grams of ignitedphosphat of magnesia. Phosphoric acid 61,5

We may infer therefore that one grain of phosphat of magnesia, the quantity yielded by the twenty grains of residue, indicated 0,385 of pure magnesia; and if, according to the statements of Kirwan and Wenzel (which very nearly agree) one hundred grains of crystallized sulphat of magnesia contain seventeen grains of magnesia, 2,26 grains of that salt will be the quantity corresponding to 0,385 grains of magnesia. And I have the satisfaction of observing that the proportions obtained by Dr. Henry, of one hundred grains of ammoniaco-magnesian phosphat dried at 90°, for one hundred and eleven grains of crystallized sulphat of magnesia, would have led to a very similar result. (See Dr. Henry's Analysis of several varieties of Salt, in Philos. Trans. for 1810, page 113.)

30. I have found by direct experiments that one hundred grains of pure muriat of soda heated to redness, and decomposed by nitrat of silver, yield 241,6 grains of luna cornea heated to fusion.
31. The same experiment was tried three times upon different specimens of the water, and I here give the average. The smallest quantity of luna cornea obtained was two grains, and the largest 2,5 grains, a difference too great to arise from mere inaccuracy. From this and several other circumstances I have reason to suspect that the water is subject to occasional variations in the proportions, as well as in the aggregate quantity of its solid contents.
32. This is a long operation, because the iron does not combine with the succinic acid at low degree of oxygenation, so that the mixture must be long digested with access of air, or repeatedly boiled and allowed to stand in the air for some hours during the intervals, before the process can be completely effected. This operation necessarily requires one or two days, but is remarkably accurate as to the precipitation of both the iron and alumine.
33. This was the combined result of two separate experiments tried on three and two ounces of the water, the first of which yield 3,5 grains, and the other 2,8 grains of alkaline salts.
34. This result shews the compatibility of muriat of soda with sulphat of iron, the latter being in excess, which has been questioned by some chemists. Being desirous of obtaining a confirmation of this by a direct experiment, I mixed together solutions of two parts of sulphat of iron and one part of murat of soda. The mixture became yellowish, and on applying heat reddish Hakes subsided. On separating these by filtration, and repeating this process two or three times, I nevertheless obtained by evaporation distinct crystals of muriat of soda, partly cubic, partly octohedral, deposited in the centre of a saline yellowish mass, without any appearance of efflorescence or of any thing resembling sulphat of soda. Therefore muriat of soda is compatible with sulphat of iron, although these two salts evidently exert some degree of action on each other, as appeared from the change of colour and the formation of reddish Hakes, which I suppose to be sub-sulphat of iron. I may take this opportunity of mentioning that by an analogous experiment on sulphat of iron and muriat of alumine, and by the assistance of alcohol, I satisfied myself that those two salts could not exist together.
35. These proportions were deduced from the following experiment : 50 grains of crystallized green sulphat of iron were dissolved in water, and nitrat of barytes was added as long as any precipitate took place. The sulphat of barytes after being carefully edulcorated and heated to redness in a platina crucible, weighed 38,5 grs. Therefore 50 : 38,5 :: 41,4 : 31.
36. It may be recollected that 3,8 grs. of ignited alumine, would, according to the proportion before stated (§ IX. 5,) correspond to 31,6 of crystallized alum. I found by a direct experiment that 100 grs. of regular octohedral crystals of alum formed by gradual deposition from a saturated solution of common alum, being dissolved in water and precipitated by muriat of barytes, produced 88,2 grs. of ignited sulphat of barytes; so that the 31,6 grs. of alum would correspond to 27,8 grs. of the barytic sulphat. This, however, could not be an accurate estimate of the real quantity of sulphuric acid, since the sulphat of alumine does not exist in the water in the state of alum.

With a view to learn the proportions of acid and base in pore sulphat of alumine, I made the following attempt. A quantity of alumine (which had been prepared by precipitation from alum, re dissolution in muriatic acid, and second precipitation by carbonat of ammonia, and appeared to contain no impurity except a vestige of muriatic acid), was dissolve din sulphuric acid, and the solution evaporated to siccity. When reduced to the consistence o fa thick syrup, and allowed to cool, the saline mass congealed into a hard whitish deliquescent cake, capable of being pulverised. This was redisolved and re-evaporated four successive times, and the last time was made red-hot, in order to expel the excess of sulphuric mid which always appeared to prevail. By this last operation a portion of the salt was decomposed and rendered insoluble in water, in spite of which the remainder still exhibited signs of acidity. The clear solution of this mass being divided into two equal portions, one of which was precipitated by succinat of ammonia, and the other by nitrat of basytes, yielded 4,5 gra. of ignited alumine for 21 grs. of ignited sulphat of barytes. From which it may be inferred, that the 8,8 grs. of ignited alumina found in a pint of the water, were combined with a quantity of acid equal to 17,7 grs. of ignited sulphat of barytes. But it is assumed in this computation that the artificial sulphat of alumina subjected to analysis, was in the same state of combination as that which exists in the water, a supposition which may not be strictly accurate.

37. The quantity of sulphat of barytes produced by the precipitation of a given quanttity of sulphat of lime, was ascertained in the following manner: some pulverized crystals of native selenite, apparently perfectly pure, were dissolved in water and afterwards slowly precipitated by evaporation. The object of this previous operation was to obtain the sulphat of lime in a state more lit for subsequent re-dissolution. Fifteen grains of this selenitic residue were dissolved in water, slightly acidulated by muriatic acid, in order to supersede the necessity of using a large quantity of water; and the solution, after being neutralized by pure ammonia, was precipitated by muriat of barytes. The sulphat of barytes thus obtained, weighed, after careful edulcoration and ignition in a platina crucible, 26,75 grs. which is equivalent to 175,6 grs of harytes for 100 grs. of ignited sulphat of lime.
38. According to Dr. Henry 100 grs. of crystallized sulphat of magnesia give 111 grs. of ignited sulphat of barytes. See Philos. Trans. 1810, p. 114.
39. These proportions were deduced from the following experiment: 40 grs. of crystallized sulphat of soda, being dissolved in water and precipitated by nitrat of barytes, the sulphat of barytes, well edulcorated and ignited, weighed 29,1 grs.
40. This precipitant, which was, I believe, first proposed by Mr. Chenevix, is much more appropriate than acids, because if an excess of acid be incautiously added, the precipitate is re-dissolvsd; whilst with muriat of ammonia, an excess of the test is attended with no inconveniences.
41. The presence of silica was also shewn, and its quantity attempted to be ascertained by the following process. A portion of residue was boiled in caustic potash: this dissolved not only the silica, but also the alumine; both these earths were precipitated from the alkaline solutions by muriat of ammonia, and separated; muriatic acid being now added, both the silica and alumine were re-dissolved (for silica, just precipitated from its solution, and not desiccated, is soluble in acid); and this solution being evaporated to dryness in a water-bath, by which means the silica parts with its acid and becomes insoluble, the muriat of alumine was washed of by distilled water, and the silica remained undissolved. This method, though affording a very useful means of discrimination, must obviously be liable to inaccuracy as to proportions, when very minute portions of silica are to be separated from considerable quantities of alumine. This however was the process to which I trusted on a previous occasion (IX. 2), to free the alumina from the silica which was mixed with it.