313 
rocks of Glen Tilt, say!: ‘as a rule the one type passes gradu- 
ally into the other — granite merges into diorite — but oc- 
casionally one rock veins the other, and when this takes place, 
the granite is always seen to be the younger of the two’. 
Brôccer, discussing the relative age of larvikite and lardalite, 
mentions that in some places the former rock proves to be 
decidedly older than the latter, while in other places the 
contact relations are such as to suggest that both rocks may 
have been partly fluid (or viscous) at the same time”. In the 
batholites of South Greenland contact relations of this kind are 
very common. 
Now, if the final consolidation of both rocks takes place 
after the intrusion of both rocks, the development of contact 
modifications etc. will depend upon the temperatures of conso- 
lidation: and the magma with the highest consolidation temp- 
erature will consolidate first, and perhaps become ‘veined’ by 
the other. The problem of the chronological sequence of 
abyssal rocks cannot, therefore, be solved without regard to 
the temperatures of consolidation of the different rocks. 
Of these temperatures there are very few exact data. It 
must be taken for granted, that each rock has a considerable 
temperature-range of crystallization, and that the lower point of 
this range, especially in acid rocks, is very different from the 
melting point of the pure mineral, and is largely dependent 
upon pressure and the amount of volatile substances which 
they contain, and hence no precise statements can be given. 
There is, however, no doubt that under like conditions acid 
magmas crystallize, as a rule, at temperatures lower than those 
of basic magmas”. The most important exception to this rule 
is that magmas which are rich in alkalies and in iron must in 
many cases be assumed to crystallize at a lower temperature 
1 Op. cit. p. 107. 
? Erupt.-gest. 4. Kristianiageb. Ш, р. 38. 
ЗА. Harker, Natural History of Igneous Rocks, 1909, р. 186. 
