GEOPHYSICAL LABORATORY. 145 



The method of differentiation by sinking of crystals seems, then, to afford a 

 promising explanation of the subalkaline series of igneous rocks. With the 

 alkaline rocks many features are more obscure, but the formation of the hy- 

 drous molecules which enter the micas in the later members of the subalkaline 

 series strongly suggests that the alkaline types belong in this late stage of 

 great concentration of the volatile ingredients of the magma, especially water. 

 The mica molecules, for example, are very closely related in type to the mole- 

 cules of nephelite, the most characteristic mineral of the alkaline rocks. 



On the basis of these deductions from experimental results, it is concluded 

 that all igneous rocks could be derived from basic magma, and the hypothesis 

 is advanced that all igneous types actually have been derived from basaltic 

 magma by the process of crystallization-differentiation. 



(3) The correlation of potassium and magnesium, sodium and iron, in igneous rocks. 



H. S. Washington. Proc. Nat. Acad. Sci., 1, 574- .578 (1915). 



The fact that these pairs of elements are correlated, or tend to vary together, 

 in igneous magmas irrespective of the silicity, is briefly discussed and some of 

 the evidence given. The correlation is shown in the minerals of igneous 

 rocks. Thus the sodic pyroxenes and amphiboles are very high in iron and 

 very low in magnesium, while on the other hand the colored potassic micas 

 usually show veiy high magnesium and low iron. It is pointed out that in 

 igneous rocks there are no potassic pyroxenes or amphiboles, and no sodic 

 micas. The evidence from the rocks themselves is based on a collection of 

 nearly 10,000 analyses. A number of examples of petrographic provinces are 

 given, and many analyses are plotted which bring out clearly the correlation. 



(4) A simple device for the graphical solution of the equation A = B.C. Fred. E. Wright. 



J. Wash. Acad. Sci., 6, 1-5 (1916). 



In this equation, which is essentially a simplified form of the proportion 

 A:B = C:D, the letters may represent numbers, or powers of nmnbers, or 

 functions of variables, sines, cosines, tangents, logarithms, exponentials, etc. 

 The graphical method of solution adopted is based on similar triangles, each 

 function A, B, and C, being represented on a scale so chosen that the resulting 

 curves are straight lines. A device is suggested for the mechanical solution of 

 this general equation. At one corner of a sheet of 1 mm. coordinate paper, 

 50 cm. square, attached to a small drawing-board, a straight-edge is fitted 

 into a socket and can be rotated about the corner as axis. Along the two 

 adjacent margins of the paper a strip of millimeter paper is pinned, and on it 

 the scale A or B is marked. The straight-edge functions as the hypothenuse 

 of the similar right-angle triangles employed in the graphical solution. The 

 device is accurate to about 1 part in 1,000. 



(5) A geological protractor. Fred. E. Wright. J. Wash. Acad. Sci., 6, 5-7 (1916). 



By means of this protractor angles of dip and strike can be plotted as with 

 the ordinary protractor; in addition, slope angles (angles of apparent dip for 

 any angle of dip of stratum and for any azimuth of vertical section) can be 

 read off directly ; the protractor can also be used as a hand goniometer for the 

 measurement of crystal angles. 



(6) The ternary .system CaO-AloOs -MgO. G. A. Rankin and H. E. Mervvin. J. Am. 



Chem. Soc, 38, 568-588 (1916). 



No ternary compound stable in contact with the melt was found, therefore 

 the system involved only the equilibrium of the components and the binary 

 compounds SCaO.AloO^,' SCaO.SAlaOs, CaO.AlaOa, SCaO.SAlaOs, MgO.AlaOs- 

 A new form of AI2O3 was described, but its relation to corundum (the only 



