ON THE ORIGIN OF CERTAIN SILICEOUS ROCKS. 



I. NOTES ON ARKANSAS NOVACULITE. 



An examination of the residues of specimens representing 

 both the Ouachita and Arkansas (commercial) types of novac- 

 ulite confirms Griswold's observations and conclusions made 

 on thin sections. Ground in an agate or porcelain mortar, the 

 greater part of the rock is easily reduced to a fine slime that 

 readily floats away in water. A considerable portion, however, 

 consists of granules which, though made up of fine grains, 

 require considerable rubbing to reduce them to slime. Even 

 when, in the final stages of the grinding, a soft pestle is 

 employed to avoid breaking up the grains of quartz and other 

 minerals, a considerable number of the aggregated granules 

 remain, thus showing a strong cohesion of the constituent grains. 



The amount of quartz in simple grains that can be positively 

 identified by the polarization colors is not large, and for the 

 most part is not distinctly detrital. Such grains are frequently 

 enclosed in the composite ones or have composite silica adhering 

 to them, and often appear to be more completely developed 

 individuals of the aggregated grains. In many cases, however, 

 they appear to be detrital grains with adherent secondary silica. 

 Unmistakable detrital grains are represented by numerous 

 zircons and rare fragments of tourmaline and garnet. 1 These 

 are exceedingly fine, and both in character and amount seem to 

 confirm Griswold's conclusion that the deposit took place in 

 still and comparatively deep water into which only a small 

 amount of very fine detritus would be floated. 



The great bulk of the silica of these rocks appears to be 

 secondary, and to account for its peculiarities the replacement 

 hypothesis of Rutley seems to offer less difficulty than that of 



1 Wichmann found garnets in the novaculites of the Marquette, Mich., district in 

 1879. Q. J. Geol Soc, Vol. XXXV, 1879, pp. 156-164, J. C. Branner. 



366 



