34 G. II. Williams — Peridotites near Peekskill, N. Y. 



total disappearance of the characteristic inclusions. The min- 

 eral becomes nearly colorless and consequently nonpleochroic 

 while retaining the compact structure and optical behavior of 

 the unaltered portion. Later there is developed, particularly 

 around the edge of the hornblende, a bright, emerald green 

 substance which on account of its lack of dichroism and very 

 feeble action on polarized light may be regarded as chlorite. 



Next to the hornblende the most important constituent of 

 this rock is the olivine, which is remarkable both for its fresh- 

 ness and for its beautiful inclusions. It is present in rounded 

 grains or in well defined crystals upon which the usual combin- 

 ation of domes, prism and pinacoids may be seen. These crys- 

 tals vary from \ to 2 mm. in diameter. The mineral is quite 

 colorless, with a high index of refraction, and is traversed by 

 irregular cracks along which serpentinization may frequently 

 be seen to have commenced, although in many sections there is 

 hardly a trace of this alteration. The inclusions in this olivine 

 are quite similar to those which have been described in the 

 olivine-gabbro from the island of Mull on the west coast of 

 Scotland by Prof. Zirkel* and by Prof. Cohen in the so-called 

 B/ypersthenite from Palma.f They are black and opaque, having 

 generally the form of minute, rounded grains or long rods ar- 



clusions iu the manner above described, the present writer would take exception 

 on the following grounds: — 



1st. It is by no means true that all crystals of the same mineral from the same 

 locality, or even in the same specimen, always contain these inclusions in equal 

 quantity. In the hypersthene of the Baltimore gabbros, for example, they are 

 sometimes abundant, sometimes wholly wanting, and this is true even of indi- 

 viduals occurring side by side in the same thin section. 



2d. Professor Judd's explanation is inconsistent with the frequent zonal ar- 

 rangement of these inclusions. If they were formed subsequent to the solidifi- 

 cation of their host, we should expect to find them uniformly distributed ; on the 

 contrary in the hornblende above described and in many feldspars they are con- 

 centrated toward the center or grouped in regular zones. 



3d. It does not appear to be a fact that, as Professor Judd suggests, the minerals 

 in which these inclusions are most abundant, are lighter in color or less strongly 

 pleochroic than those without them. In sections of the Baltimore gabbros no 

 difference could be observed by the writer between either the color or pleochroism 

 of hypersthene crystals destitute of inclusions and such as were completely filled 

 with them. 



4th. The minerals in which the inclusions are most abundant are always ex- 

 tremely fresh in their appearance. At the commencement of anything like alter- 

 ation they are the first things to disappear. 



5th. Many inclusions of this class do act upon polarized light, indicating that 

 they belong to definitely crystallized species. 



To the writer there seems every reason to regard the indications which produce 

 " schillerization " as original in their nature. They appear to be composed of 

 substances extruded from the rest during the process of its crystallization as in- 

 capable of forming a part of its chemical composition, not unlike the crystalliza- 

 tion of certain impurities in limestone as various silicates, when the limestone 

 undergoes metamorphism to marble. 



* Zeitschrift der deutschen geologischen Gesellschaft, xxiii, 1871, p. 59, PL TV, 

 Pig. II. Mikrosk. Beschaffenheit, p. 214. 



f Neues Jahrbuch fur Min., etc., 1876, p. 750. 



