The frequency-distribution of igneous rocks. Part II. The laws of distribution in relation to petrogenic theories

Abstract

In Part I of this paper the actual frequency-distribution of igneous rocks as revealed by the latest edition of Washington's collection of analyses was examined, and compared with that of earlier records. An attempt wilt now be made to determine the law of this distribution–a necessary preliminary to any petrogenic application of the results. The silica frequency will be considered since this has been found to give a curve more characteristic of igneous rocks than does any other oxide Now Dr. Harker pointed out that the silica distribution of records published in 1903 did not obey any simple law such as that of the probability curve, and the same conclusion applies to the collection of 1917. However, when the empirical method of Karl Pearson was applied to analyse the curve into component curves of error, fairly simple conclusions resulted.