4.2.6 Reviews—The Geophysical Laboratory. 
classification of which he was one of the four authors. He still 
keeps to the old groups of rocks—sedimentary, igneous, and meta- 
morphic—which have long figured in textbooks; no doubt the wisest 
course to follow, since, although that arrangement is logically 
indefensible, petrologists have as yet come to no general agreement on 
a better one to replace it. The discussion is clear and well expressed, 
and is rendered very much easier to follow because of the many 
well-drawn diagrams. In an appendix the most important rock- 
forming mineral species are briefly described, and a copious index 
brings the book to a close. 
Ill.—Tse Geropuysicat Lasoratory. 
(]\HREE further papers, dealing with the equilibrium of the systems 
occurring in igneous rocks, have recently been published. In 
the first of these! the system albite-anorthite-diopside is considered 
and the equilibrium conditions determined. This system includes 
mixtures which are comparatively close to certain simple types of 
igneous rocks such as diorite, gabbro, and basalt, and the results, 
therefore, are of great interest in petrology. Of the three binary 
systems concerned, two are eutectiferous and the third conforms to 
Roozeboom’s “‘type [’’. The ternary composition-diagram consists 
of two fields, diopside and plagioclase, separated by a single boundary 
curve, and there is no eutectic point. The latter fact practically 
precludes the possibility of any of the above-mentioned rocks 
representing a eutectic, and the replacement of diopside by complex 
pyroxene mix-crystals renders the possibility still more remote. 
This point is of great importance in the theory of erystallization- 
ditferentiation, while at the same time it negatives the suggestion 
that ‘ophitic’ structure represents a eutectic one. The composition 
of the plagioclase which crystallizes out varies not only with the 
original composition of the liquid but also with the rate of cooling. 
If the latter is not sufficiently slow for the continual attainment of 
equilibrium, zoning tends to occur, there being a certain rate at 
which the tendency is a maximum and beyond sairoln it diminishes, 
until in very rapidly cooled melts it may be absent. 
The second paper? deals with the system lime-alumina-magnesia, 
and, as might be expected, is of more-importance in the study of 
Portland cement clinker than in petrology. The chief points of 
interest are the discovery of a new form of alumina (B-A], O.), which 
appears to be monotropic with respect to corundum (a—Al, O,), and 
the existence of an almost continuous series of solid solutions between 
the latter and magnesium spinel. The presence of comparatively 
large amounts of alumina does not seem to affect the optical properties 
of the spinel to any appreciable extent. 
The third paper® contains a critical review of previous work on the 
1 N. L. Bowen, ‘‘ The Crystallization of Haplobasaltic, Haplodioritic, and 
Related Magmas’’: Amer. Journ. Sci., ser. IV, vol. xl, pp. 161-85. 
2 G. A. Rankin & H. E, Merwin, ‘ The Ternary System Ca O- Als O3s-Mg 0”: 
Journ. Amer. Chem. Soc., vol. xxxviii, pp. 568-88, 1916. 
3 J. Johnston, H. BE. Merwin, & EH. D. Williamson, ‘‘ The several Forms of 
Calcium Carbonate”? : Amer. Journ. Sci., ser. IV, vol. xli, pp. 473-512, 1916. 
