276 BULLETIN OF THE 



The thin sections belonging to Professor Whitney's collections were 

 prepared by Mr. J. H. Huntington, late Assistant Geologist on the 

 Geological Survey of New Hampshire, whose preparations have been far 

 superior to the other American work on rock sections that I have seen, 

 and are fully equal, if not superior, to the German work sent to thia 

 country. 



All classifications of rocks in vogue at the present time are, with one 

 exception, confessedly artificial, as one can readily see by examining 

 those proposed by Xaumann, Blum, Von Cotta, Zirkel, Dana, Lang, 

 Lasaulx, Ilosenbusch, Roth, and scores of other writers of more or less 

 note. The one exception is that scheme for the Tertiary volcanic rocks 

 given by Baron Richthofen in the Memoirs of the California Academy 

 of Sciences (Vol. I. Part IL, 1868), in which he announces that, be- 

 ginning with the Tertiary age, volcanic rocks succeed one another in 

 massive eruptions in the following order : propylite, andesite, trachyte, 

 rhvolite, and basalt. Having laid down this order, his classification 

 then becomes largely artificial. 



The usual classifications have been based on structure, age, minera- 

 logical composition, chemical composition, and upon almost every part 

 of the rock separately, but never on the entire rock as a unit. But 

 little attention has been paid, when the scheme was based on min- 

 eralogical characters, whether the minerals were foreign, original, or 

 alteration products. If the same minerals existed in two rocks, no 

 matter if the minerals in one were products of the crystallization of 

 the cooling magma, and in the other alteration products, both rocks 

 were classed as the same, even if there existed a difierence of silica 

 amounting to over twenty per cent. 



If we adopt the mineralogical classification, we are met b}' the fact 

 that the feldspars cannot be distinguished from one another, in spite of 

 the method of Des Cloizeaux, and its ingenious modification by Professor 

 Pumpclly, — methods that fail exactly where they are most ncedod. 



Furthermore, our best chemists and mineralogists are not agreed as 

 to what are the species of feldspar, — a point that it would be well to 

 settle before the classification of rocks is made wholly dependent on the 

 feldspathic constituent. 



Quartz, hornblende, augite, monoclinic and triclinia feldspars, and 

 magnetic iron are found extending through rocks of every ago and of 

 every percentage of silica. The quartz and hornblende exist as foj-eign, 

 as indigenous, and as alteration products, — distinctions that should bo 

 noted in employing these minerals in classification. 



