32 SCIEXCE PROGRESS. 



first explicitly pointed out in the Western States, and it 

 appears to hold not only in the Great Basin and on the 

 High Plateaux but also in Mexico. Turner has shown, 

 however, that the volcanic rocks of the Sierra Nevada 

 district do not conform to this order, the earliest Tertiary 

 eruptions being of rhyolite, followed by basalt, then by 

 andesites, and finally by a later group of basalts (lo). 



Reviewing the history of volcanic eruptions at various 

 periods, we find many cases which fall under the law of 

 increasing divergence, and the number is considerably in- 

 creased if we admit the possibility of reversion at a later 

 stage to the original type, and again if we make allowance 

 for the partial or total suppression of one of the diverging 

 branches by the rocks of that branch being intruded instead 

 of extruded. For instance, the Bala volcanic series of 

 Carnarvonshire, beginning with some rocks of andesitic 

 composition, consists mainly of a great succession of 

 rhyolites, no basaltic lavas being developed. The basic 

 branch, however, is sufficiently represented by innumerable 

 intrusive sills of diabase associated with the lavas and by 

 laccolitic intrusions of very basic hornblendic rocks at a 

 lower horizon. Again, Geikie has pointed out the existence 

 at the summit of the rhyolitic group of an andesite, con- 

 stituting the latest volcanic outpouring in the district, and 

 this may perhaps be regarded as marking a new accession 

 of the original intermediate magma. Similar examples 

 might be multiplied, but we also find many cases which 

 cannot, without undue forcing, be brought under the law of 

 increasing divergence, and we are bound to admit that any 

 generalisation taking cognisance of the observed succession 

 of volcanic rocks only is based on a partial view of the 

 facts. 



Some years ago Brogger established clearly the order of 

 succession of the Devonian igneous rocks, mostly of plutonic 

 and intrusive types, in the Christiania basin. It is in general 

 an order proceeding from more basic to more acid, beginning 

 with olivine-gabbro-diabase and passing through progres- 

 sively more acid members to granite ; but the latest rocks 

 in the district are again basic. More recently the same 



