FAULTING. 73 



but the final results proved that in every case the faults assumed from such 

 evidence were not faults, while the ultimate discovery of numerous and important 

 faults was due to careful study of the rocks. 



When by close examination and correlation of facts the complicated and often 

 closely related rocks were satisfactorily separated into stratigraphic units, after 

 numerous unsatisfactory attempts, the most important step toward the elucidation 

 of the geologic history and structure had been taken. But still the most extreme 

 caution was necessary, for while the local geologic column was probably historically 

 correct for the whole district, there were many local gaps and irregularities. As 

 there were several periods of apparently active but irregular erosion between 

 volcanic outbursts and as the distribution of many of the members of the series 

 was limited and irregular it seemed that any member might rest directly upon &ny 

 older one, the intervening ones being unrepresented, while a few hundred yards 

 away the represented succession would be different. For similar reasons it was not 

 possible to reckon upon any constant thickness for any formation; in one place it- 

 might be a few feet thick, in others hundreds. So the ordinary stratigraphic 

 criteria of faulting were very inconclusive. 



SIEBERT TUFF BOUNDARIES. 



The key to the problem, undoubtedly was the determination of the geologic 

 position of the Siebert tuff, which consists of characteristic finely stratified thick 

 beds. In working out the structure the first thing done was to carefully follow 

 the limits of these Siebert tuff areas. It was found that in most cases these were 

 separate; they reappear in different parts of the area mapped and are bounded on 

 several sides by straight lines. This fact immediately suggested the existence of 

 numerous intersecting faults. 



Where a rectilinear boundary of a Siebert tuff area ran transversely to the 

 strike of the beds, a fault was evident, in case the contiguous rock was not 

 intrusive. In the case of a surface formation, like the Fraction dacite breccia, 

 this evidence was conclusive, and parts of the majority of detected faults were 

 followed in this way. Similarly, if a fault was parallel with the strike, and the 

 dip of the tuff would carry it below a contiguous rock (as the Fraction dacite 

 breccia, for example) which was known to be lower in the geologic column than 

 the tuff, the nature of the contact, as due to dislocation, was evident. 



DIKES ALONG FAULT ZONES. 



Another criterion, perhaps not so important, was developed by the discovery 

 that the Brougher dacite sent out dikes along some of the faults, as along the Cali- 

 fornia fault. (See map, PI. XVI.) This showed at once that the dacite reached its 

 present position essentially subsequent to the faulting (a conclusion which was other- 



