PETROGRAPHY. 349 



next oldest constituent. In general, solidification begins with 

 the crystallisation of the ores and accessory constituents, then 

 follows the formation of the coloured silicates (olivine, mica, 

 augite, hornblende, etc.), then that of the felspathic minerals, 

 and finally that of free silica. In the rocks of eruptive flows 

 the more basic constituents crystallise out before the less basic, 

 so that at any period during the consolidation the sum of the 

 constituents already crystallised out from the magma is more 

 basic than the remaining portion of the magma. 'Mr. Teall 

 assumes that in the rocks with a large or medium amount of 

 silica, the dissolved constituents represent a so-called " eu- 

 tectic " mixture, and as such can. remain unchanged at a 

 temperature which is below their melting-point. But if they 

 do not occur in the definite eutectic relations, the overplus 

 of substances continues to separate out until . the eutectic 

 mixture is attained. 



In an important memoir (1887) on the crystallisation of 

 igneous rocks, Lagorio classified the porphyritic flows accord- 

 ing to the amount of silica in five grades, and gave the results 

 of chemical analyses of the ground-mass. He arrived at the 

 conclusion that the separation of the minerals in an eruptive 

 magma depends almost entirely on the chemical composition 

 of the magma, as well as on the affinities and internal move- 

 ments within the mass ; whereas pressure and high temperature 

 exert only a subordinate influence. 



Iddings in 1889, in a paper on the same subject, expressed 

 views in many respects similar to those *of Lagorio, but 

 ascribed greater importance to the influences of pressure and 

 temperature in regulating the rate and processes of cooling ; 

 he thinks the local conditions of pressure and temperature 

 mainly determine the structural differences which often exist 

 at different portions of a continuous mass of eruptive rock, 

 and explain why a superficial portion may display porphyritic 

 structure while the deep-seated portion is granite-grained. 



There .are abundant examples of transitional rock-varieties 

 in eruptive bosses and sheets. As far back as 1852, Delesse 

 showed that the Ballon d'Alsace in the Vosges mountains 

 consists of hornblendic granite in its central portion and its 

 summit, but towards its peripheral portions passes into syenite 

 and finally into diorite. More recently, in 1887, similar facts 

 were demonstrated by Barrois in his brilliant account of the 

 eruptive rocks in Brittany. The researches of Barrois have 



