THE PHYSICAL AVERAGE DENSITY 401 



tion of many rocks, disturbing features which would not occur at depth 

 and which are also eliminated by the normative method. 



Finally, it is the general rule that a magma crystallizing under plu- 

 tonic, deep-seated conditions forms mineral aggregates which differ from 

 those which would be formed by the same magma solidifying under 

 effusive, surface conditions. Some examples of this have been given on a 

 preceding page, and the observation is one of the commonplaces of petrog- 

 raphy. Its application to the present matter lies in the fact that the two 

 different mineral aggregates — that is, the rocks — would have markedly 

 different densities, so that they would not be mutually comparable for 

 the purposes of investigations in isostasy, in which we must deal with 

 rocks, the majority of which are plutonic. 



It follows from this very brief exposition that the results to be ob- 

 tained by the physical method for densities can not be regarded as capable 

 of forming a very safe base for generalizations as to isostatic compensa- 

 tion. I have, however, computed the average specific gravity 22 of 1,849 

 rocks, of which specific gravity determinations were found along with the 

 analyses in Professional Paper 99 of the United States Geological Survey. 

 The specific gravities of weathered or otherwise altered rocks were not 

 considered, but those of fresh rocks, represented by bad as well as by good 

 analyses, were included, as the quality of the analysis had no bearing on 

 the matter. Many more determinations might have been obtained by 

 search through the literature, but time was lacking for this, and it was 

 thought best to confine the investigation to the same set of data which 

 served for the calculations by the normative method. 



It was found that there were far fewer specific gravity determinations 

 than analyses; indeed, the specific gravity determinations used numbered 

 only 1,849, while there were altogether about 8,500 analyses, good and 

 bad, of fresh rocks. It was also found that there were great differences 

 between the various countries as to the numbers of analyses, both absolute 

 and relative to the numbers of analyses. For some countries, notably 

 Germany, Switzerland, Italy, and Australia, specific gravity determina- 

 tions were fairly numerous, while for others, as the United States, France, 

 Norway, and Sweden, they were very few. It is to be regretted that this 

 feature of igneous rocks, even though it be a minor one, is so frequently 

 neglected. Peccavi ! 



22 In all these cases the specific gravity, not the true density, is given, frequently with 

 no mention of the temperature at which the observation was made ; so that reduction to 

 density is impossible. The figures given here are, therefore, specific gravities. The error 

 involved will not he very great, but it is one to be added to the list of errors already 

 set forth. 



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



