390 GEOLOGY OF THE YELLOWSTONE NATIONAL PARK. 



to repeat any of the petrographical description of the rhyolite in this, its 

 most notable, exposure. There is, however, an occurrence of columnar 

 cracking in an inconspicuous locality which is of considerable petrological 

 interest on account of its bearing on the general question of the production 

 of prismatic parting, and which may properly find a place in this chapter. 



The fine columnar parting on the east side of the canyon north of 

 Agate Creek are described in Part I. Here the columns are 80 feet 

 long and several feet in diameter. But the prismatic cracking to be 

 described occurs in a small gulch on the west wall of the canyon a mile 

 south of Deep Creek. The rhyolite forming the upper part of the plateau 

 at this place is lithoidal and porphyritic, light purplish gray, with a rough 

 fracture. It is jointed in broad blocks which weather into rough granite- 

 like masses. This passes downward into more distinctly columnar, denser 

 rhyolite, which is exposed on the south side of the gulch in a beautifully 

 columnar cliff 300 feet high. The vertical columns are so regular in shape 

 that they may be easily mistaken at a little distance for basaltic ones. 



On the north side of the gulch there is a ledge of columnar basalt 100 

 feet thick resting upon gravel and andesitic breccia, . Immediately overly- 

 ing the basalt is the bottom contact of a younger flow of rhyolite. The 

 lowest portion is tuffaceous and yellow, passing- upward into denser material, 

 and this into dark-gray, brownish, porphyritic glass, which has cracked in 

 thin, straight prisms, some quadrangular, others irregularly shaped. The 

 prisms vary in size from 4 to 8 inches long and from one-third to three- 

 fourths of an inch thick, and thicker. This glass grades into lithoidal 

 purplish rock, which js also cracked into small prisms and columns. Some 

 of these are 3 or 4 feet long and 4 inches thick; others are very small, 

 about 6 inches long, and irregularly shaped, like prisms of starch. They 

 are more or less curved, and occur grouped together in the form of large 

 blocks, whose relation to the original form of the whole mass was not made 

 out, They appear to have resulted from a shrinkage within these blocks. 

 The jointing planes which constitute the faces of these prisms or columns 

 intersect the many porphyritical crystals of quartz and feldspar, producing 

 smooth faces, which show that the groundmass of the rock was rigid when 

 the cracking took place. The small columns, in their shape and arrange- 

 ment, are precisely like those formed in dried starch. 



