THE NORMATIVE AVERAGE DENSITY 



385 



portion of the rocks of the observable crust, assumedly of the order of 

 magnitude of 10 or 20 miles thick. 



Having thus obtained the norm of the average rock, it is reduced to 

 density by the common method of treatment. The simplest procedure is 

 to multiply the percentage amount of each normative mineral by its 

 density, water being included with a density of 1, if desired, the sum 

 giving directly the approximate density. This was the method adopted 

 by Iddings in his publications referred to below, but it does not lead to 

 strictly accurate results, because we are dealing here, not with weights, 

 but with volumes. The relative densities of the several normative min- 

 erals -should, therefore, be reduced to specific volumes by taking their 

 respective reciprocals, and each of the respective percentages of the nor- 

 mative minerals multiplied by its specific volume. The resultant average 

 specific volume is changed to density by dividing it into 1 — that, is taking 

 its reciprocal. 



With the average rocks, of which the chief constituent minerals do not 

 differ greatly in density, the differences between the two methods are not 

 very great in many cases. On the whole, however, the method by specific 

 gravities leads to somewhat higher figures than does that by specific vol- 

 umes ; and this is especially true of the more femic averages, such as 

 those of the ocean floors, in which almost equal amounts of light feldspars 

 and heavy pyroxene, with considerable very heavy magnetite, are present. 



This use of the norm for calculating the density of an average rock 

 originated with Iddings, who talked it over some years ago with his col- 

 leagues in the quantitative classification. He gives briefly a few of his 

 results, without detailed description of the method, in his work on vol- 

 canism, and later published a short account, the last paper to issue from 

 his pen during his life. 9 He compared the calculated and observed densi- 

 ties of several rocks and found close agreement, except when much biotite 

 or hornblende were present. Eskola 10 finds, however, that there is but an 

 insignificant difference between the calculated normative density and the 

 observed specific gravity in the case of amphibolites, and that the differ- 

 ence may be in either direction. 



In the present connection it is of interest to note that, in his "Problem 

 of volcanism" (page 125), Iddings pointed out the application of these 

 calculated densities to the problem of isostasy and noted that "the corre- 

 spondence between the general character of large series of igneous rocks 

 and the general demands of the theory of isostasy respecting the existence 



9 J. P. Iddings : The problem of volcanism, 1914, pp. 121-125 ; and Am. Jour. Sci., 

 vol. 49, 1920, p. 363. 



10 P. Eskola : Bull. Com. Geol. Finlande, no. 44, 1915, p. 143. 



XXV — Bull. Geol. Soc. Am., Vol. 33, 1921 



