CHEMICAL RELATIONS AND SERIAL CHARACTER 403 



increases, the former rise regularly with it, while the latter fall as 

 regularly. This behavior, however, is well known and calls for no 

 comment. 



SERIAL CHARACTER 



The diagram makes more evident another feature, already shown by 

 the analyses, namely, the serial character of the six types. They all 

 vary continuously in one direction, with scarcely a break or abnor- 

 mality of any kind. Indeed, these six rocks are an excellent example 

 of a rock series as defined by Brogger.* So well marked is the series, 

 and so regular is the variation from the center outward, that our belief 

 in the integral and contemporaneous character of the various rock types 

 of the mass is changed almost into a certainty. It is scarcely possible 

 that such a regular variation would be found in any complex made up 

 of successive intrusions, still less that they should have been intruded 

 so nicely in their proper order. Some kind of differentiation of the 

 igneous magma as a whole is the only possible explanation, and it 

 seems to me that this complex of Magnet Cove, with its beautifully 

 developed zonal structure, its series of interesting types, and the way in 

 which the component rocks fit into their places in the series, both 

 chemically and mineralogicall)'', forms one of the most striking in- 

 stances in favor of the differentiation hypothesis which has yet been 

 found. 



It will be seen in the diagram that there is a rather large gap in the 

 center between Si0 2 = 41.75 and 49.70 — that is, between the ijolites and 

 the syenites. As the six analyses represent all the main types observed, 

 it seems hardly possible that another exists which fits into this gap. 

 It suggests rather that the main course of differentiation has split the 

 magma into the ijolites and the syenites, and that their varieties are 

 due to a secondary differentiation of these main differentiates. The 

 ijolitic and syenitic rocks are then to be regarded as the melanocratic 

 and leucocratic complementary divisions of the series. 



THE DIKES 



General trend of the dikes. — Before discussing the problem of differen- 

 tiation and comparing the rocks of Magnet Cove with those of other 

 regions, a few words may be devoted to the dikes, in the attempt to see 

 where they fit into the general scheme. As a preliminary it ma} 7 be 

 stated that the majorit} 7 of them show a general east-and-west trend — 

 that is, between northwest and southwest. This is approximately par- 



* Rrogger : Emptivgesteine des Kristianigebietes, vol. i, 1894, p. 169. 



