162 QUICKSILVER DEPOSITS OP THE PACIFIC SLOPE. 



in composition or in the physical conditions to which the several masses 

 have been subjected. A few tenths of 1 per cent, of carbon in iron changes 

 its fusibility and texture enormously and trifling quantities of silica or 

 alumina cause immense variation in the fusibility of the normal bisilicate, 

 iron blast-furnace slag. It is well known that when furnace men desire a slag 

 which fuses less readily than this they do not dare to add eitlier alumina 

 or silica, because either raises the melting point so rapidly. Lavas, which 

 are natural slags, must be affected in a similar way by these or by other 

 substances, such as titanium. 



Granitoid and porphyritic texture. Wllilc tllC obsidiaUS of Clcar Lakc aud of tllO 



Rossberg have evidently remained amorphous because of their peculiar 

 chemical composition, it by no means follows that liad tliey been cooled 

 sufficiently slowly they might not have crystallized. On the contrary, the- 

 ory and experiment alike point to the supposition that vitreous substances 

 will always crystaUize if they have sufficient opportunity. This is gener- 

 ally admitted. 



It is often supposed to be merely an extension of the acknowledged 

 tendency to crystallization to maintain that, if glassy magma is only cooled 

 slowly enough, the result will be a mass which is not merel}' holocrystal- 

 line, but of granitic structure 



The difference between typical granular texture and porphyritic text- 

 ure, however, is a very different matter from the distinction between holo- 

 crystalline and glassy structure, a fact which appears to have escaped the 

 attention of many lithologists. The conclusion to be drawn from granular 

 structure is that various minerals crystallized simultaneously, while the 

 larger mineral constituents of porphyries have evidently crystallized in 

 advance of the groundmass surrounding them. 



If a substantially homogeneous fluid cools very slowly indeed, the 

 tendency will be for some of the resulting compounds to crystallize in ad- 

 vance of others, and therefore to attain a considerable size and good crys- 

 tallographic development. This follows both from theory and experiments 

 familiar to every chemist. If the cooling of such fluid is continued at a 

 very slow rate, the interstices must fill with other crystals the growth of 

 which will be interfered with by mutual opposition and the obstruction of 



