PKOCEEDINGS OF UNITED STITES NATIONAL MUSEUM. 133 



This anal} sis differs from the one actually performed only in fractions 

 of percentages in the alumina and soda, and I think may be assumed 

 as being nearly correct. The amount of iron in the two analyses is 

 identical, but th^e is some difference in the state of oxidation, which 

 can readily be supposed to be the result of secondary actions that have 

 talcen place in the rock analyzed. 



It becomes very easy now to see how extremely diversified the fekl- 

 spathic element may be in rocks of this nature. The molecules mr^y 

 arrange themselves in very diversified ways, while the rocks remain 

 identical in composition. Circumstances of cooling might cause anor- 

 thite to separate in a nearly pure condition, when there would be a com- 

 pensating acidity in the remainder of the feldspathic element. This is 

 a much more satisfactory explanation than thar which I offered at the 

 time I demonstrated the presence of anofthite in the West Kock dia- 

 base,^ for as my analysis showed, the presence of the auorthite did not 

 modify the ultimate composition of the rock. On the other hand pure 

 anorthite might be entirely absent and its molecules might enter into 

 combination with the molecules of the potassium and sodium feldspars, 

 to form one or more intermediate species, as in the Jersey" City diabase, 

 and much diversity' might exist in this feldspathic element in different 

 localities without the slightest change taking place in the ultimate com- 

 position of the rock. 



I regard this work as of some importance, since it completes our 

 knowledge of the normal composition of a rock which has a great dis- 

 tribution and very uniform characters, and shows that this rock is more 

 complex in composition than had been suxiposed. Besides it has been 

 common to consider what feldspar enters into the com[)Ositiou of ba-^ic 

 rocks like this, rather than what feldspars. An exquisite balance of 

 composition and circumstance would be necessary to crystallize such a 

 rock with a single feldspar, and we have reason to be convinced that 

 massive rocks are rarely simple as regards their feldspathic constituent.^ 



It has also an important bearing upon the micioscopic determination of 

 feldspars by means of oi)tical properties. The method proposed by 

 Pumpelly, and further developed by Fouque and Levy, is used lor the 

 determination of the species of feldspar by seeking for the greatest 

 angles which elasticity planes make with twinning planes, in the zone 

 with axis perpendicular to the twinning plane. There is of course a pos- 



1 Am. Jour. Sci. 1875, vol. ix, j). 189. I concluded that a minute change in the com- 

 jjositiou of the I'ock would be sufficient to allow of the formation of anorthite, which 

 on account of its infusibility would first crystallize from the rock mass, as was evi- 

 dently the case. 



2 Fouque has demonstrated the complexity of the feldspar in S.'intorin andesites, 

 Santorin et ses Eruptions, page 3Ci(). 



I have examined the basic rucks at Feekskill, on the Hudson, recently described uy 

 Professor Dana, American Journal of Science, vol. xx, page 194. The feldspathic ele- 

 ment was easily separated into two parts, one of which was a pink variety of ande- 

 iste and the other white orthoclase. 



