PYROCLASTICS OF HEMLOCK FORMATION. 143 



represents a portion of the glass boiinding tlie vesicles. Here and there is 

 a fragment with a more or less perfect vesicle remaining. The few mineral 

 fragments found — feldspar — were angular, but quite fresh. The rocks show 

 no intermixture of rounded fragments, and they are consequently regarded 

 as volcanic dust deposited through the air. These ash beds show a delicate 

 banding of finer and coarser-grained fragments. In a single slide a grada- 

 tion can be traced from a moderately fine grained sand composed of distinct 

 volcanic fragments into a verj' fine grained mass composed of hornblende 

 needles, biotite, chlorite, epidote, and sphene, cemented by what is prob- 

 ably quartz, perhaps having associated with it some feldspar whose charac- 

 ters could not be determined. 



Relations of tuffs and ash (dust) beds. — The pyTOclastics sccm to predominate in the 

 northwestern part of the district in the neighborhood of the small town of 

 Amasa. Special opportunities for observing the relations between the tuff 

 and the ash beds are offered by the third cut of the Chicago, Milwaukee 

 and St. Paul Railway west of Balsam, Michigan. Gradation can be traced 

 from coarse tuffs to delicately banded fine tuffs. The average thickness of a 

 single ash bed probably does not exceed 5 feet. In the same exposures the 

 tuff beds are from 50 to 100 feet thick, and even moi'e. 



VOLCANIC CONGLOMERATKS (TUFFOGENE SEDIJIENTS, REYER). 



That certain of the pyroclastics have been brought together and 

 rearranged by the agency of water is made clear by their characteristic 

 structure. Such rocks are the volcanic conglomerates. In very many 

 respects they are strikingly like the various eolian deposits, tuffs, etc., 

 described above. They agree with them in color. The same varieties of 

 volcanic rocks are represented that are found in the tuffs. They are true 

 basalt conglomerates. 



The pebbles are ver}- commonly sharph' outlined by accumulations 

 of epidote grains on the periphery. Some of the fragments have a reddish- 

 brown to purplish-black color, and stand out strongly from the green 

 matrix. Such pebbles are found to contain large quantities of magnetite, 

 the oxide being in beautiful sharp crystals and in absolutely fresh condi- 

 tion, forming a sharp contrast to the altered condition of the fragments in 

 which it occurs. In one case in which the main mass of the fragments 

 now consists of chlorite and epidote, magnetite occurs in large quantity, 



