1898.] 011 Some Recent Results of Physico-Chemical Inquiry. 651 



ice. The proportionate amount of these " ice-molecules " depends, 

 under ordinary conditions, upon the temperature. On heating they 

 become fewer and fewer ; on cooling they become more numerous. 

 We may regard water at any particular temperature as a saturated 

 solution of such molecules ; when cooled below its ordinary solidifying 

 point it is a supersaturated solution of such molecules, and of course 

 behaves under such conditions like any other supersaturated solution. 



Now any circumstance which effects the transformation of the 

 ice-molecules into the other kind of molecules should be attended by 

 a contraction of volume. When water is heated from 0° upwards, w^e 

 have two distinct volume changes — expansion of the water as such, 

 and the destruction or transformation of the ice-molecules with 

 consequent diminution of volume. Up to 4° the diminution due to 

 the transformation of the ice-molecules is greater than the expansion, 

 and the nett result is contraction. After 4° the ice-molecules 

 become fewer and fewer, and the degree of expansion gradually 

 gains upon that of the diminution in volume due to the alteration of 

 the ice-molecules ; and thence the degree of contraction becomes less 

 and less, until the nett result is an increase of volume and the water 

 seems to behave like any other liquid on heating. It does not, 

 however, follow that all the so-called ice-molecules will have dis- 

 appeared, even at above 8°, for the two distinct sets of molecules 

 may co-exist, but of course in gradually diminishing ratio as the 

 temperature rises. 



It is easy to see how this assumption, which is but an extended 

 form of a very old idea, may serve to explain the " anomalies " above 

 referred to. Take the case of compressibility of water at low tem- 

 peratures. It is unnecessary to remind a Royal Institution audience 

 that ice, even at low temperatures, may be converted into water by 

 pressure ; the classical experiments of Faraday and Tyndall are 

 admirable illustrations of that fact. Now the more ice we thus 

 convert into water the greater the contraction. A given increase 

 of pressure at a low temiDcrature causes a i^resitev contraction than 

 at a higher temperature, because at the lower temperature there are 

 more ice-molecules to be changed. The diminution of volume under 

 compression is like the increase of volume by temperature, made up 

 of two parts, viz. (1) the real compressibility of the water ; and (2) 

 the diminution attending the transformation of the ice-molecules. 

 Probably the water-molecules, as such, behave like other molecules — 

 they contract under pressure, and to a gradually smaller extent as 

 the pressure is increased ; it is only the effect of the increased 

 pressure in changing the ice-molecules, with consequent diminution 

 of volume, that makes the apparent compressibility greater, and thus 

 gives rise to the " anomaly." It should follow, therefore, that at some 

 point of tcsmperature above the freezing point of water there should 

 be a minimum point of compressibility, just as there is a minimum 

 volume. Experiment shows that such a minimum point exists at 

 about 50°. 



