RELATIONS TO OTHER SCIENCES 601 



One objection is that this law has been verified only for the case 

 where there is but one substance dissolved in the solvent. Guy and 

 Chaperon's law, that gravity cooperates to destroy the homogeneity 

 of a solution, can have no application here, for in that case the heavier 

 basic submagma should appear at the lower levels, and the lighter acid 

 submagma in overlying higher levels, an arrangement which would 

 not explain the contrast between centre and periphery. The explana- 

 tion of the peripheral basicity as the result of the melting of the border- 

 ing country rock is not only difficult to understand, but contradicts 

 innumerable facts, and furthermore denies that any magmatic differ- 

 entiation takes place. 



Brogger has contributed a distinct advance to our understanding 

 of this problem by showing that, in special cases, it was definite 

 stoichiometric combinations and not the isolated materials which 

 moved in these opposing directions, the silica-poor, iron-magnesia-lime 

 silicates having moved in the one direction and the silica-rich, alkali- 

 alumina silicates in the other. Furthermore, he found that these 

 stoichiometric combinations correspond' to the minerals of the 

 eruptive rocks where, as is well known, it is the rule to find the 

 alkalies, aluminum, and calcium associated on the one hand, while 

 on the other magnesium, iron, and calcium usually go together. 

 Thus it comes about that the least soluble combinations are those 

 which gravitate toward the cooling surface, and to this extent dif- 

 erentiation obeys the laws controlling the tendency to crystallize. 

 Harker seems to hold similar views. We have here, certainly, an im- 

 portant exposition, but it merely recognizes a fact and offers no 

 actual explanation of the same. There yet remains the unanswered 

 question, What is the nature of that motive force, on account of 

 which precisely the melanocratic pole has a peripheral position 

 while the leucocratic pole has a central one? We know nothing about 

 the difference in the diffusion constants of the respective [stoichio- 

 metric] combinations. 



Contrasts similar to that between the centre and periphery of the 

 same massif are found in a region where many so-called comple- 

 mentary dykes occur, acid and basic rocks being in close proximity. 

 This association is explained as simply the result of fissure-filling by 

 a differentiated plutonic magma precisely similar in origin to the one 

 above referred to. 



The question is an important one whether the fluid molten magma 

 suffers decrease or increase in volume with its transition to the solid, 

 crystalline condition. Gustav Bischof , Mallet, and David Forbes, as 

 the result of experimental investigations, have expressed the opinion 

 that the mass suffers a contraction of about one tenth, and the devel- 

 opment of contraction fissures (cooling cracks) in the solid lava agrees 

 with this. In the oft-cited experiments by Barus the frozen lava was 



