merce, fumes strongly in the moist air, so great is its avidity for water. Thrown upon several times its own weight of ice, the two liquefy and the temperature stands above 0°. The strongest commercial solution of chloride of aluminium, however, when at the temperature of the air or at 0° or at 100° C., will reduce the temperature to -13° when poured upon three or four times its own weight of ice. I suppose the anhydrous chloride may be viewed as separating the atoms of the water-molecule, as is supposed to be the case with the chlorides of phosphorus.

$52. Hydrochloric Acid as a Cryogen with Ice.-Ordinary commercial hydrochloric acid when poured upon ice may produce as much as 22° of cold. The amount of HCl in the hydrochloric acid, however, is here to be as much considered as it would have to be if we dealt with ice and a solution of NaCl. In the experiments which are given in the succeeding paragraphs as far as $54, the hydrochloric acid employed was formed by saturating water with the gas, while the liquid was kept after absorption had ceased for half an hour at 0° while the acid passed through. The barometric pressure was 770 millims. This may be considered a normal acid.

§ 53. The acid at 0° was poured in various proportions upon ice at 0°. Fifty grms. of ice were used in each experiment. The Table shows within what small limits of ratio the minimum temperature is reached. In the Table the weight of the ice is taken as unity.

We are therefore, when dealing with a cryogen one of whose constituents is a liquid, much more limited in the range of ratio which we may employ to procure the maximum cold than is the case when both are solid.

§ 54. Accordingly, in examining the effect of the alteration in the temperature of one or both of the constituents, it is necessary always to use the same ratio. In the Table below, the ratio used was that which gave the maximum cold when both were at 0° C., namely 1 of ice to 0.4 of the saturated solution of HCl. It must be understood that in those cases where the hydrochloric solution was used below 0° C., it is not meant that the saturation with HCl took place at that lower temperature, but that, having been saturated at 0°, the solution was subsequently cooled to the lower temperature.

Phil. Mag. S. 4. Vol. 49. No. 324. March 1875.

P

All these numbers are sufficiently near to -26°, the number got when both ingredients are at 0°, to justify the conclusion that the temperature of the ice has something, but little, to do with the temperature of the freezing-mixture. And it appears that the temperature of the HCl solution is also without effect, provided that the saturation has been effected at the same temperature and pressure. The water acts, indeed, here much as the water of crystallization acts in the case of solid salts.

$55. Table of Freezing-mixtures.-In the following Table the lowest attainable temperature is given. This, as we have seen, is independent of the temperature of the salt and its degree of crystalhydration. The temperatures are got on mixing the salt with three to six times its weight of ice in lumps of the size of a pea downwards. The salts are arranged according to the degree of cold they furnished.

13.0 Hg Cl

130 | NH CO,

We shall study this Table in connexion with the temperature

and molecular ratios of the cryohydrates of the same salt.

CRYOHYDRATES.

Cryohydrates of the Halogen Alkalies.

§ 56. As was anticipated in § 31, the study of the nine salts

resulting from the union of the halogens with the alkali metals has brought to light many points of interest. The two members of this family which I had previously examined were NaCl, solidifying at -23° with 10-5 molecules of water, and NH, Cl, solidifying at -16° with 12 molecules of water. I have reexamined the cryohydrate of chloride of ammonium; and the mean of several analyses gives 19-27 per cent of chloride, which brings the ratio to NH, CI+12·4 H2 O.

There is nothing particular to remark in the analysis of this group. All, excepting the ammonium salts, bear heating to 300° or 400° C. without decomposition or volatilization. The liquid cryohydrates were weighed in covered basins of Bohemian glass, evaporated to dryness, strongly heated, covered, and quickly cooled. The ammonium salts were heated on the water-bath till they ceased to lose weight.

§ 57. Iodide of Potassium.—This salt, which at the ordinary temperature is so abundantly soluble in water, yields a large proportion of anhydrous salt when cooled. An ice-salt cryogen fails to solidify it, but just brings it to the verge of solidification, removing all the free salt in the same manner as it does with NaCl solution, excepting that there does not appear to be any intermediate cryohydrate corresponding with NaCl +2H, O. The KI solution appears to be homogeneous at -8° or -12°. When subjected to solid CO, and ether, it at once begins to solidify at -22° to -23°, and retains this temperature to dryness, Of the four parts into which the substance was divided as it solidified, two were analyzed—namely, the second crop of crystals, and the finally solidified and remelted mother-liquor. Of the first, 5.8240 grms. gave 30120 anhydrous KI, or 31.72 per cent.; of the second, 6·3960 grms. gave 3.331, or 52.07 per cent. The latter corresponds to the molecular ratio

KI +8.5 H, O.

§ 58. Bromide of Potassium.-This forms a cryohydrate of great beauty. At -13° the characteristic crystals are seen to be produced. They have the fern-like shape of ice-crystals. The fronds are studded with a fructification of opaque crystals, whose opacity seems to spread through the at first transparent fronds till the whole resembles frosted silver. The two portions which were taken for analysis were the final part, or that which solidified last, and the immediately preceding crop of crystals. Of the former, 7·8285 grms. gave 2·5170 of KBr; of the latter, 9.4560 grms. gave 3.0070 dry salt. These correspond respectively with 32.15 and 31.80 per cent. The first points to the molecular relationship

KBr+13.94 H2O.

$59. Chloride of Potassium.-Since KI is thus seen to combine with 8.5 of water and KBr with 14 of water, the one a fraction and the other a whole number, I spared no care in the preparation and analysis of the chloride of potassium, to see whether the fractional relationship did not arise from experimental error. Perfectly pure K, CO, was dissolved in a slight excess of HCl, evaporated to dryness, and ignited in a platinum basin. The KC was recrystallized four times. Six ounces of the saturated solution were cooled; solid matter separated down to -10°. At this temperature the separated salt appeared as a cloudiness in the liquid, a little heavier than the latter, and of a granular but translucent appearance. The liquid may now be cooled to -15°; but it is now a supersaturated solution of the true cryohydrate-a solution, it is to be observed, of which the solid crystals present are unable to determine the solidification. At a little below -15°, long ice-like crystals shoot out and the temperature rises to -11°4, which is the crystallizing-point of the cryohydrate. From this behaviour I am disposed to admit that there may be an intermediate hydrate similar to the bihydrate of NaCl.

Of the last portion which solidified, 6.588 grms. gave 1.320 of anhydrous salt. Of the previous crop of crystals 7.869 grms. gave 1-579 of salt. These show the respective percentages 20-03 and 20.07, both of which point to the relationship

KCl+16.5 H2O.

§ 60. The halogen salts of potassium arrange themselves according to the cold required for the solidification of their cryohydrates, and according to the molecular ratio of water, as follows:

:

We find here the general rule confirmed, that among like salts the lower the temperature of solidification of the cryohydrate, the smaller is the number of molecules of water which it contains.

§ 61. Iodide of Ammonium.-This salt gave me considerable trouble. It is difficult to obtain in perfectly colourless crystals; and both the solid and its solution are disposed to change colour, becoming brown when kept, even in the dark. Heating in a water-bath with constant stirring restores the dry altered

salt to a light-grey colour. Analysis showed that this coloration did not sensibly affect the percentage composition of the salt; but still its power of assuming water might be seriously affected. A saturated solution was exposed for some hours to an ice-salt cryogen at -22°; it was then exposed to the carbonic-acid-and-ether cyanogen. It solidified wholly at from-27° to 28°. While solidification is taking place, the cryohydrate is nearly perfectly white. When dry and over-cooled, it assumed a pinkish grey colour, resembling chloride of silver which is beginning to be affected by light.

The analysis of these portions is given, namely:-(1) the last to solidify or remelted mother-liquor; (2) the immediately preceding crop of crystals; and (3) the crop before (2). As the results are by no means in good accord, I give them all:

NHI. per cent. 2.9375, or 55.49 3.2120, "

4.8675,,, 57.56

(1) (2)

58.42

The portion (1), which, being the last, is in one sense likely to be the most homogeneous, corresponds to the relationship

NH2I+6·44H2 0.

The portions (2) and (3) would indicate as much as 0.7 molecule less of water.

$62. Bromide of Ammonium.-This salt separates as a cryobydrate from a saturated solution at -17° C. Its formation is well marked, and its analysis was satisfactory. Of the final portion, 4-2591 grms. gave 1.3680 of NH, Br. Of the immediately preceding crop of crystals, 5.5990 grms. gave 1.8010 of NH Br. The percentages are 32-12 and 32-17 respectively these numbers correspond to the relationship

§ 63. Taking the value of the cryohydrate of chloride of ammonium from 56 as NH, C1+12·4H2O, we have for the halogen salts of ammonium :—

4

whence it is gathered that the relative faculties of chlorine, bromine, and iodine, in depressing the temperature at which the cryohydrates are formed, are similar with ammonium to their