Page images

§ 30. The above are arranged according to the molecular ratio, NaCl having the least number of molecules of water. At once an important law declares itself: those cryohydrates which have the lowest solidifying temperatures have the fewest molecules of water. This law holds true with all which have been examined, with the exception of the bichromate of potassium, which, if placed according to its temperature of solidification, would be above the sulphate of sodium. Is this due to its abnormal composition? Such questions suggest themselves by the score on contemplating this Table VIII.

§ 31. I have found the solidifying-points, and have hermetically sealed specimens, of the cryohydrates of a few other salts, and am collecting more previous to analyzing them. I give their solidifying-points in the order of the degree of cold; and it will be of great interest to see whether their molecular ratios fall in accord with the above rule.

[blocks in formation]

Perhaps the most promising direction of inquiry for the establishment of such a uniformity of result as may be used for the prediction of untried experiments will consist in a careful study of the nine salts between K, Na, and NH on the one hand, and Cl, Br, and I on the other.

Some Experiments with Sea-water.


§32. Freezing sea-water.-The sea-water with which the following experiments were performed was procured from Dover. After filtration, it was found to have at 760 millims. the boilingpoint 100°-6, while the temperature of its vapour was 100°.2. This sea-water began to freeze at -2° C. On evaporation on a waterbath and keeping at 100° C. for two hours, the percentage of solid residue was 6.5786. A large beaker of this sea-water was cooled to 0° C. A tin vessel was supported inside the beaker so that its bottom just touched the surface of the water; and a freezingmixture was placed in the tin vessel. When .about of the whole had solidified, the solid was removed and divided into two parts: one was allowed to melt, and its percentage of solid matter was determined as above; the other was broken up and frequently pressed between linen and flannel in a screw press, being allowed to melt as little as possible. The percentage of Phil. Mag. S. 4. Vol. 49. No. 322. Jan. 1875.


solid matter in this also was determined. The following numbers show the result of this examination :

[merged small][ocr errors][merged small][merged small][ocr errors][merged small][ocr errors][merged small][merged small]

It appears, then, that under these conditions the freezing of sea-water is little more than the freezing of ice, and that the almost undiminished saltness of the unpressed ice is due, as suggested by Dr. Rae, to the entanglement amidst the icecrystals of a brine richer in solid constituents than the original water itself. Such brine, which is here squeezed out in the press, drains in nature down from the upper surface of the icefloe by gravitation, and also is replaced by osmic action by new sea-water which again yields up fresh ice; so that while new floes are porous and salt, old ones are more compact and much fresher, as the traveller observed.

§ 33. But, bearing in mind the existence of the cryobydrates, certainly of sulphate of magnesium and doubtless also of chloride of calcium at temperatures not far below 0° C., a rapid fall of temperature may be accompanied by more complex phenomena of gelation; for if the ice be quickly removed from a large mass of water by freezing, the resulting brine may easily be so enriched as to throw out one or more cryohydrates, which thus perpetuate in situ a definite amount of saline matter. How far such cryohydrates are soluble in the chief cryohydrate, namely that of NaCl, which by itself resists the cold the longest, is an important matter for future research. But there can even, viewed in the light of the experiments given above, be little doubt that the degree of saltness of a floe depends not only upon its age, but also upon the rapidity with which it was at first formed, and upon the lowest temperature to which it has subsequently been exposed.

§ 34. Since sea-water has no maximum density below its freezing-point, when a mass of sea-water is uniformly cooled to -2° C. ice will be formed at any point, whether at the bottom or at the surface, which loses more heat. Even the middle of a mass of sea-water may lose heat by radiation, and crystals of ice be thus formed in the mass. Or the bottom of the sea may radiate sufficient heat through the icecold layers above to freeze the water in contact with it, and thus generate large masses of ice which break off and carry with them parts of the sea-bottom. But I suppose the ice of the sea is mainly formed at or near the surface by radiation from the surface into space and by contact with the colder air. To imitate as nearly as I could the condition which I suppose to exist, I

cooled the sea-water to -2° C. in a beaker, which I enveloped thickly with flannel. I tried in vain to freeze the surface by blowing over it dry air which had passed immediately before through a long pewter worm immersed in a freezing-mixture. But I succeeded in getting a sheet of ice when I hung a freezingmixture contained in a blackened tin pan within about inch of the surface of the water, the whole being plentifully enveloped in flannel. Perhaps here the actual conditions which obtain when sea-water freezes were reproduced. I found that the pressed ice contained only 0.4052 of solid residue at 100° C.

§ 35. The question suggested itself to me whether, when one part of a solution of a salt is cooled, there may not be an accumulation of salt in the cooler part, although not accompanied by any solid separation. I accordingly cooled a saturated solution of nitre to -1° C. and decanted from the separated nitre. I then warmed the solution in a tall beakerglass to 60° C. and placed the bottom of the beaker in melting ice. In an hour's time a thermometer at the bottom stood at 10° C., at the top at 33°. A specimen from the bottom contained 11.3 per cent. of nitre; one from the top contained 11.7 per cent. of nitre-showing that there was no sensible diffusion of the salt one way or the other.

General Considerations.

§ 36. Maximum density of Water.-It was shown that brines of various strength, when mixed with water, absorbed heat and expanded. Let us look upon ice as the cryohydrate of water. Water shrinks as it loses heat till it reaches 4° C. At this point ice is formed, which, however, is dissolved in the water. A solution is obtained having a temperature of solidification below 4° C., namely at 0° C. At 0° C. the ice and the water solidify together, producing the compound body or cryohydrate called ice, which is thus a cryohydrate of water. The expansion from 40 to 0° is due to the greater and greater amount of ice which the water holds in solution, and which expansion is greater than the contraction of the water due to the diminished temperature.

§ 37. Variation of media.-There can be no doubt that the discovery of an enormous number of new bodies of definite composition will reward those who labour in this field. Taking water as the medium for solution, there appears to be no doubt that every soluble salt has a definite cryohydrate; so that in this direction alone the number of new bodies awaiting discovery and description may be estimated at half the number of bodies already known. If we vary the medium, employing, say, alcohols or hydrocarbons as solvents, the number of new compounds will be again indefinitely increased; so that it is fairly within the

truth to assert that the number of known bodies may soon be doubled.

§ 38. Geological.-The behaviour of mixtures of salts will again offer a new chapter for study; and I suppose we may expect that much light will be thus thrown upon some of the most obscure geological questions. For though we have been considering above cryohydrates (that is, compounds of water solidifying below the freezing-point of water), there can be no discontinuity separating the medium water with its peculiar temperature of solidification from other media of very different melting-points. We know already, indeed, very many instances in which the mixture of two bodies has a lower melting-point than either of its constituents. What must happen, then, if a mass of molten rock, such as a silicate, is saturated at a high temperature with another silicate? When the mixture cools, the second may separate out in the solid form, perhaps as quartz, perhaps as felspar, or what not. Anon, at a certain lower temperature, solidification takes place between the medium and the dissolved rock in definite proportion-definite, though perhaps not necessarily in chemical ratio, but presenting that mineralogical ratio which is so striking, and which has not hitherto been satisfactorily explained.

§ 39. Constant temperatures.-Perhaps one of the most interesting aspects of the experimental results is the establishment of fixed temperatures below zero. With the exception of the melting-points of a few organic bodies such as benzol, and the boiling-points of a few liquids such as liquid ammonia, sulphurous acid, and carbonic acid, and the rather ill-defined temperatures to be got by various freezing-mixtures, there are no means in the hands of physicists for obtaining and maintaining with certainty and ease a fixed temperature below 0° C. Now, if we surround a body with one of the solid cryohydrates described above, the body is kept at a corresponding temperature as long as any of the cryohydrate remains solid, and this with as much certainty as the temperature 0° C. can be maintained by melting ice. We thus command temperatures between -23° and 0° C. with the greatest precision.

§ 40. Invitation to others.-I need scarcely point out that the field of inquiry which has been here opened is far too large to be satisfactorily examined by one worker. It is notably at the commencement that the collaboration of many workers is most beneficial, so that fundamental errors may be quickly corrected. On this ground I respectfully invite my fellowlabourers to this branch of inquiry.

I have received through a considerable part of this inquiry much valuable assistance from my friend Mr. F. H. Marshall.


II. On Aniline Derivatives.


HE following results in connexion with aniline derivatives were obtained during the course of an investigation for which the substances that will be referred to were required.

Separation. When chloraniline, bromaniline, &c. are prepared by acting on an anilide with chlorine &c., the function has usually a double period; so that mono-, di-, and tri- derivatives are generated in presence of each other. In order to separate these I proceed as follows. The mixed derivatives are immersed in a very large excess of aqueous hydric chloride (1 vol. common fuming chloride to 9 vols. water) and heated to nearly 100°, with frequent stirring, for about an hour in a loosely covered vessel; the whole is then allowed to cool down until the next day. The clear liquid contains only mono- and di- derivatives, the insoluble portion di- and tri- derivatives. The latter is submitted to repeated hydrochloric treatment as before, until the supernatant clear liquid no longer gives any precipitate with ammonia; it then consists of tri- derivative only-contaminated, indeed, with some black tarry products. This derivative can be purified by distillation per se, or from strong aqueous hydric chloride or potash-lime. The clear liquids are united and precipitated with ammonia during twenty-four hours, a large excess of ammonia being avoided. The precipitate is then washed, rapidly evaporated with hydric chloride to dryness on the water-bath, redissolved (or at any rate well stirred) in hot water, and left to cool thoroughly: the insoluble portion consists of di- derivative, and must be filtered off. The filtrate is again evaporated to dryness and stirred with hot water &c. Three evaporations to dryness are necessary, and usually sufficient; and the final solution contains mono- derivative only, which yields but an inappreciably small amount of insoluble residue when so evaporated. The mono- derivative can be purified by distillation from aqueous soda in a current of steam; the di- derivative by distillation per se, or by successive crystallizations from naphtha and spirit.

When aniline is intended to be converted into chlorine, bromine, or iodine derivatives, it should be dried and purified by cohobation for a few hours with about one eighth to one sixteenth of its weight of mercuric chloride, bromide, or iodide respectively. Subsequent fractional distillation easily furnishes a very pure product. Only aniline so purified is referred to in the following experiments.

Preparation.—(a) Aniline is cohobated with glacial hydric *Communicated by the Author.

« PreviousContinue »