CORRELATION OF COMPOSITION AND TEXTURE. 91 



importance to the differences in this respect which may exist between the 

 various groups. Still, we have the investigations of Daubeny, Deville, and 

 Mallet, which are so far concordant that they indicate decisively the exist- 

 ence of a true relation. The acid rocks have decidedly higher melting tem- 

 peratures than the basic rocks. Many blast-furnace slags approach the vol- 

 canic rocks in constitution, and the great amount of experience gathered in 

 iron-smelting amply confirms the same relation so far as the cases are fairly 

 comparable. We may, with considerable confidence, state as an approximate 

 truth that the melting temperatures of volcanic rocks have a direct ratio to 

 their acidity. 



The textures of volcanic rocks are no doubt due in part to peculiar- 

 ities of chemical constitution. The vitreous character of the rhyolites, the 

 coarse, harsh texture of the trachytes, the compact, fine-grained texture 

 and peculiar fracture of the andesites and basalts are surely in due a 

 great measure to their constitution, but how or why we do not know. 

 There is, however, another serise in which texture is ordinarily spoken of, 

 and to which high importance is attached, and this sense takes account 

 of the degree or extent to which the groundmass of a rock is crystallized. 

 By far the most important difference between a volcanic and a non-erup- 

 tive plutonic rock, so far as pure petrographic considerations are concerned, 

 consists in the fact that the plutonic non-eruptive rock is wholly crystal- 

 line, while the volcanic rock is only partially so. Otherwise the two kinds 

 might be quite indistinguishable — might consist of the same constituents. 

 This distinction, depending upon the extent of crystallization, however, is 

 of great importance, since it arises in all probability from causes associated 

 with the genesis and geological evolution of the rocks themselves. The 

 nature and properties of the silicates are such, that under the conditions 

 ordinarily existing their crystallization is attended with difficulty and pro- 

 ceeds very slowly. An indispensable requisite for crystallization is mobility 

 of molecules inter se, and for this mobility a liquid condition of the magma 

 is essential. But the silicates possess the following peculiarity : at a tem- 

 perature sufficiently high to render them very liquid crystallization is im- 

 possible ; at a temperature just low enough for crystallization, they are 

 exceedingly viscous and the mobility very much impeded. The crystals, 



