414 H.S.WASHINGTON — IGNEOUS COMPLEX OF MAGNET COVE 



also true that the ionizing action of different .solvents varies much both 

 in degree and kind, dissociation taking place in some and to different 

 extents, while association, or an apparent molecular condensation, takes 

 place in others. 



Owing to the lack of data, especially as regards concentrated solutions, 

 the whole subject is in a vague and uncertain condition, and I may be 

 allowed to express the hope that the physical chemists, by the investiga- 

 tion of this phase of the study of solutions, and more especially concen- 

 trated ones and fused salt mixtures, may soon put us in a position to 

 attack such problems with greater knowledge and consequently greater 

 hope of their ultimate solution. * 



At any rate, it would seem that the application of Soret's principle is 

 by no means as general as has been supposed, and indeed some strong 

 objections f have been recently raised against it. Backstrom's liquation 

 theoryj certainly contains elements of probability, but seems rather 

 applicable to particular cases, such as orbicular granites, than to differ- 

 entiated zonal laccolithic masses or dikes. In such cases as these the 

 substance first crystallized would tend to collect and grow from the 

 rough containing walls, the points of which would serve as nuclei. In 

 the case of one liquid separating from another in which it happens to 

 be insoluble under the existing conditions, we have no evidence, so far 

 as I know, that there would be such a tendency. It seems more likely 

 that the insoluble liquid would separate in the form of drops and 

 schlieren. 



COMPARISON WITH OTHER REGIONS 



A comparison with the rocks of other regions is of great interest, but 

 so much space has been devoted to the preceding discussions that only 

 one or two special points will be mentioned. 



In general, it may be said that the rocks of Magnet Cove offer special 

 analogies with those of Kola, Alno, Greenland, Portugal, Brazil, and 

 Beemerville, New Jersey. These are all regions of foyaitic rocks low in 

 silica and especial^ high in soda. The chemical resemblances will best 

 be seen on an examination of the analyses given in Table III, which 

 are only a few out of those available. At Magnet Cove the most' acid 

 rock is the nepheline-syenite, with 53 per cent of silica; but the pulas- 

 kites of the neighboring igneous area of Fourche mountain may reason- 

 ably be supposed to be derived from the same general magma. Rocks 

 corresponding to these are found at several of the regions above men- 



* It is encouraging to note that Professor Bancroft lias lately (Jour. Phys. Chem., vol. iii, 1899, 

 p. 605) called attention t<> the neglecl of this very important field ol research, 

 t Backstroni, .lour. Geology, vol. i, 1893, p. 773, and G. F. Becker, Am. Jour. Sci., vol. iii, 1897, p. 21. 

 J Backstroni, op. Cit, 



