and Attached Water. 117 



tenacity that they will not relinquish it wholly and imme- 

 diately when heated to or even above 100° C. Again, acetate 

 of potassium in solution and subjected to heat, shows continuity 

 of the liquid condition between strong solution and anhydrous 

 fusion. Further, if nitrate of ammonium be dried before a 

 brisk fire, or in vacuo over sulphuric acid, it has no associated 

 water. But a solution of that salt in water passes, on 

 evaporation over the flame, from the state of solution to that 

 of fusion without intermediate solidification. Nay, under such 

 treatment it may begin to decompose before it has become 

 anhydrous. This has, indeed, led some experimeters to 

 conclude that in the dry crystalline state it contains a mole- 

 cule of water. This has arisen from the two facts — first, that 

 it is hygroscopic in moist air, and, second, that one of the 

 products of its decomposition in the dry state is water. The 

 lowering of the boiling-point of a liquid by admixture with 

 even a very little of another liquid having even a higher 

 boiling-point is a phenomenon of the same order. 



§ 255. Geological bearing o/§§ 249-254. — Just as in the 

 selective formation of what in my last memoir were described 

 as salt-alloys we may have the artificial type of the genesis of 

 manv primary rocks and metamorphic modifications, so in the 

 wonderful solubility in or miscibility with water of such 

 alloys and of some salts at high temperatures we may have a 

 no less clear type of the formation of certain volcanic rocks 

 and an explanation of some of their peculiarities. The 

 function of water in affecting rocks has been subjected to 

 a most exhaustive examination by Daubree. Water, in 

 both its solid and liquid form, is a rock. Under pressure 

 the limit of temperature is not known to which it may be 

 heated without decomposition, when in contact with bodies 

 saturated with oxygen. Granting that water may have a so- 

 called critical temperature, and range above it when it is 

 heated with bodies which have no physical relationship to- 

 wards it, still at high pressures it will be compressible as a 

 vapour to a density at least as great as that of liquid water ; 

 and until actual decomposition ensues the physical relation- 

 ship of the water-molecule with the rock-molecule will 

 remain possibly unchanged. It is true that at the very highest 

 temperatures water appeared to be decomposed ; but this is 

 only when it is unconfined. 



Whether the earth's temperature be vastly greater towards 

 its centre than it is near the surface, or whether the observed 

 increase with depth be confined to a mere skin of the earth, 

 there seems to be no reason to suppose that water may not, 

 and does not, exist at the earth's very centre. 



