158 SCIENCE PROGRESS. 



successive intrusions a gradual decrease in the proportions 

 of iron-oxides, magnesia and lime, and a corresponding 

 increase in silica, potash, soda, and alumina, the last two 

 declining again at the acid end of the series, which includes 

 members with silica-percentage ranging between the ex- 

 tremes $7 % an d 76. These relations are made clearer by 

 reducing the analyses to molecular ratios and plotting them 

 graphically after the fashion of Iddings. An examination 

 of the constituent minerals of the rocks, taken in the same 

 order, leads to equally interesting- results. The more basic 

 silicates disappear in turn, while more acid ones come in, 

 the order thus arrived at being as follows : olivine, pyroxene, 

 hornblende, biotite, plagioclase, orthoclase and quartz, mi- 

 crocline. Now this is also the order of crystallisation of 

 the several minerals in so far as they occur together in the 

 same rock, or, more accurately, it is the order in which 

 the several minerals began to crystallise out from the 

 magma, for their periods of crystallisation to some extent 

 overlapped. This order, as Rosenbusch has remarked, is 

 the normal order in plutonic rocks in general, and may be 

 roughly summarised as an order of decreasing basicity. 

 Our authors point out that the earliest intruded rocks are 

 thus the rocks richest in the earliest formed minerals. If 

 this is to be regarded as a fact of general significance, it 

 seems to point to a connection between crystallisation and 

 differentiation in the presumed parent-magma. Such a 

 supposition might be fortified by evidence from other 

 districts, where the law of the more acid rock cutting the 

 more basic seems to hold very widely. 



In this connection we may note in passing the observa- 

 tion of Bayley 7 on the great mass of gabbro in North-eastern 

 Minnesota. This remarkable body of rock exhibits striking 

 variations in composition in different parts, being sometimes 

 very rich in olivine, sometimes rich in augite and biotite, 

 and again poor in augite and very rich in felspar. Never- 

 theless, so far as the investigation goes, the felspars of 

 these various types seem to be of practically the same 

 variety, a basic labradorite. If the differentiation had been 

 effected in the magma prior to the crystallisation of the 



