PETROLOGICAL ABSTRACTS AND REVIEWS 577 



to obtain data as to the order of crystallization, the occurrence and distri- 

 bution of micropegmatite, and, if possible, to arrive at some definite con- 

 clusions as to the origin and causation of the observed phenomena. 



The author recognizes three kinds of micropegmatite: (o) intergrowths 

 developed in the regular course of crystallization, (6) centric or radial 

 forms developed in connection with and radiating from the edges of the 

 feldspars, and originating in the reactions between the various minerals in 

 contact with it, and (c) the form of micropegmatite presented by the matrix 

 in which the other minerals are imbedded. From 182 slides examined, 

 representing 90 different rocks, it was found that type a occurred in 23.3 

 per cent, & in 68.8 per cent, c in 13 .3 per cent, and absence of any form of 

 micropegmatite in 17.7 per cent. 



In examining the sequence of crystallization the author found that any 

 two minerals, whose periods of crystallization overlap, even where one of 

 them is crystallizing at an abnormal period, may appear in micropegmatitic 

 intergrowth; and the relative frequency of occurrence of any single pair 

 in such combination is in proportional relation to the general coincidence of 

 the crystallization periods of the particular minerals entering into the 

 combination. Besides the usual intergrowths of quartz and orthoclase, 

 tfie author found intergrowths of quartz and hornblende, orthoclase and 

 hornblende, plagioclase and hornblende, sphene and hornblende, magne- 

 tite and hornblende, biotite and hornblende, quartz and biotite, orthoclase 

 and biotite, plagioclase and biotite, quartz and plagioclase, orthoclase and 

 plagioclase, quartz and microcline, microcline and plagioclase, quartz and 

 sphene, plagioclase and sphene, orthoclase and sphene, biotite and sphene, 

 quartz and epidote, and epidote and orthoclase. 



All the phenomena are most readily explained on the theory of the origin 

 of granite by crystallization from solution. The process is not always 

 progressive, but at certain stages reversions come in and partake of the 

 nature of re-solution of parts that had already become solid. 



In conclusion the author gives in detail the various stages that appear 

 to be indicated in the crystallization of a granite, and considering that the 

 normal biotite granites have been derived by differentiation from a hom- 

 blendic magma, he gives the following stages for all granites: (i) At a 

 temperature probably somewhat above 1,100° C, hornblende began to 

 crystallize. (2) In supersilicated homblendic magmas, quartz, orthoclase, 

 and an acid plagioclase, or any combination of them, may appear along 

 with the last of the hornblende, the appearance of these minerals in this 

 connection being apparently due to their solubility in the originally liquid 

 hornblende. (3) Partial resorption of these minerals, and in particular of 



