JOINTS DEVELOPED BY SHRINKAGE. 437 



surface where the dikes are exposed. The westernmost dike crops out 

 ou the North fork, just below the mouth of Eagle creek. It must reach the 

 same stratum at a much less depth, probably within 2,200 feet of the surface. 



These figures do not appear to be unwarrantably large, and yet when we 

 compare them with what is actually known of the depths of earthquake 

 fissures they seem very deep. Their horizontal extent, however, is not in- 

 compatible with great depth, for one of them is certainly not less than six 

 miles in leugth and from 5 to 8 feet wide. 



Origin of the Joints in the Dikes. — The peculiar jointing in the dike re- 

 quires explanation. It may be accounted for in the following manner : The 

 parallel jointing is developed in these dikes only where the minerals have a 

 most decided flow arrangement, and the jointing is parallel to this align- 

 ment. It is a feature which may be well seen in hand specimens. The 

 direction of the jointing is determined by a sort of slaty cleavage ; but the 

 fissures are actually developed along these lines of minimum cohesion, prob- 

 ably by shrinkage. 



The transverse joints are of a different nature. Generally, but not always, 

 the principal system of transverse joints in the dikes is parallel to the strati- 

 fication of the adjoining shales and sandstones, so that it is evident that the 

 planes of stratification have some influence in determining the position of 

 the principal transverse joints. The different strata touching the dike would 

 vary greatly in porosity. Some being open would take up water rapidly, 

 and the water would be drawn toward the porous strata on both sides of the 

 same plane of stratification. The shrinkage in the dike from the loss of 

 water would produce a strain at right angles to the stratification, and when 

 the strain becomes greater than the cohesion of the dike it cracks transversely, 

 parallel to the strata. Occasionally, as in the large dike on Crow creek 

 (plate 7, figure 1), this system embraces nearly all the transverse fractures 

 of the dike. Others associated with them may be at right angles to and a 

 natural consequence of the first, but there may still be other sets whose origin 

 as shrinkage cracks is not so evident. 



The only other change which has taken place in the dikes since their 

 formation is the deposition of carbonate of lime, which has cemented the 

 sand firmly together, so that the sandstone of the dike usually has greater 

 solidity than that of the beds. This larger amount of carbonate of lime in 

 the dikes is clearly shown by the chemical analyses. 



Distribution of the Dikes, considered as Earthquake Phenomena. — The gen- 

 eral distribution and parallelism of the dikes may be seen in figure 2. Only 

 the dikes 18 inches or more in width have been represented. The various 

 exposures which appear to be of the same dikes have been connected. Only 

 the width of the dikes has been exaggerated. The dotted line to the right 



