THIRD HYPOTHESIS OF DERIVATION 475 



material. Thus, on Manhattan island, it has been remarked by Mather, 

 Dana, Kemp, and Gratacap, that the crystalline limestone has been often 

 found merging into the adjoining gneiss, of which a good example, 

 called to my attention by Dr A. W. Grabau, may still be seen in the 

 limestone outcrops at One hundred and sixty-fifth to One hundred and 

 seventy-fifth street, along Sheridan avenue, north of Harlem river. 



But the sheets of hornblende schist, on the contrary, are universally 

 characterized throughout the island by their sharply defined lines of con- 

 tact with the inclosing gneiss (see plate 63). Even in regard to the largest 

 exposure of amphibole and serpentine rocks, that at West Fifty-ninth 

 street (figure 9), this fact was long ago expressed with surprise by the 

 earliest observer on record, Doctor Gale, who said : 



" It is quite remarkable that, at the junction of the anthophyllite with the gneiss, 

 it is so sudden that there is no intermixing of the two, but each remains perfecth' 

 distinct, side by side, a stratum of anthophyllite and a stratum of gneiss, and each 

 pursuing its own pecuharity within the space of 3 or 4 inches of its neighbor." * 



This feature, therefore, bears additional testimony to the common 

 genesis of this amphibole rock with that of the adjoining hornblende 

 schist, in opposition to the view of Dana as to the derivation of the 

 former from alteration of limestone, t 



Induration of a contact band. — Contrast is often shown by finer texture 

 in the micaceous gneiss near its contact with hornblende schist (plate 63) 

 than at a small distance ; but the same differences obtain in ordinary 

 beds of gneiss toward their division planes. This has been evidentl}^ 

 due to more ready penetration of pegmatitic material along the central 

 bands (figure 1, plate 63). As that might have been induced by looser 

 texture and greater porosity at a certain distance from the hornblende 

 schist, specific gravity determinations were made on lumps of the two 

 rocks, in distilled water at 21 degrees centigrade, in a series across the 

 contact (shown in figure 2, plate 63), to ascertain whether the density 

 increased toward the contact line (see table on page 476). 



There is here clearly no evidence of density increasing toward the 

 contact line, as by the influence of a dike, but only oscillations caused 

 by ordinary variations in density of layers, by permeation of pegmatitic 

 material, and by degrees of decomposition of the weathered surface. 



The survival of a contact band seemed more probable in the vicinit}^ 

 of a larger mass of the hornblende schist; but its absence from the 

 13-foot layer at West One hundred and sixty-fifth street above the 

 Speedway, from the 20-foot layer at West One hundred and thirty-fifth 

 street near Eleventh avenue, and from the 35-foot layer near Spuyten 



* Mather, op. cit., p. 583. 



fAm. Jour. Sci., vol. xx, 1880, p. 32. 



