(376) 
phenomenon. A single remark may show its importance in explaining 
several widely varying chemical processes. First of all, attention may 
be called to the occurrence of water in plutonic rocks. BUNSEN +) 
when carrying on his investigations on the Icelandic rocks suggested 
that liquefied rocks might have absorbed water. Since GAUTIER has 
lately demonstrated the presence of 1 to 2 per cent of water in all 
kinds of igneous rocks, this possibility has well nigh become a 
certainty. Bunsen, however, could only account for the absorption 
by assuming the existence of enormous pressures. The investigations 
and views now communicated lead to the belief that the absorption 
may possibly take place at a very low pressure*); and it is not 
even precluded that afterwards on cooling and solidification, the 
dissolved aqueous vapour did not escape owing to supersaturation 
phenomena assisted perhaps by rapid congealing, or because, on 
cooling, hydrated silicates are deposited first. Owing to our complete 
ignorance as -to the situation of the vapour pressure curves of solutions 
of anhydrous and hydrated silicates when close to their melting 
points, it is not possible to form a decided opinion. 
But we have already got a better insight in a totally different 
sphere of phenomena, namely the oxidation of melted metals. The 
system water + salt is indeed quite comparable with the system 
oxygen + metal and therefore a hydrated salt with an oxide. Of 
late it has been abundantly proved that all kinds of melted metals 
dissolve the gases of the air, oxygen in particular. Whether we must 
consider this to be a simple solution or a solution of the oxide 
makes no difference for in either case the quantity of oxygen dis- 
solved is dependent on p and t. The melting point of the metals is 
lowered thereby even much more than was formerly believed, for 
instance 20° for silver and 16° for copper. With a free supply of 
air the absorption therefore takes place at !/; atm. oxygen pressure. 
A line similar to 4LBFC necessarily exists for the oxygen con- 
taining liquids saturated with solid metal. If now a horizontal line 
drawn for 1/; atm. intersects this line below the maximum it means 
that on cooling to the point # the metal will erystallise and all 
the dissolved oxygen will be expelled. This phenomenon is shown 
for instance by silver; it is known under the name of spitting. 
1) Lieb. Ann. 61. 271. (1849). 
Pogg. Ann. 83. 237. 240. (1851). 
2) It is no objection that the melting temperatures are situated very far above the 
critical temperature of water, as we are dealing here with solutions containing very 
little water whose critical temperature is therefore considerably raised. 
——. ge —— 
