﻿Triassic formation of the Connectictct Valley. 345 



erally found acceptance on account of the great thickness of 

 strata that it involves. It implies indeed a disturbance of im- 

 probable simplicity, and is directly excluded by the presence of 

 numerous strike faults, always with upthrow on the east, 

 whereby single beds are enabled to do more than double dut}', 

 and a not immoderate thickness suffices to account for the 

 breadth of the formation. Many of the smaller faults are 

 almost directly visible: the evidence on which the existence of 

 the larger ones rests is of the same character as that which is 

 accepted to prove the occurrence of faults in such a region as 

 Tennessee ; namely, the repetition of corresponding series of 

 beds. It is true that some members of the series here are 

 sheets of trap ; but when these sheets are found to be con- 

 temporaneous overflows, they at once take rank as conformable 

 members of the formation, and gain a high value as indications 

 of structure from the success with which they resist erosion 

 and the prominence that their outcrops therefore maintain. To 

 deny the existence of the larger faults requires one to admit 

 the frequent recurrence of a definite complex sequence of for- 

 mative conditions repeatedly causing the deposition of a series 

 consisting of sandstone, amygdaloidal trap, limestone, shales, 

 heavy trap, sandy shales, lighter trap and sandy shales, in this 

 regular succession and of closely corresponding thickness, mem- 

 ber for member. This is so inherently improbable that it 

 would not be admitted for a moment, if the trap-sheets were 

 sedimentary rocks, like the ridge-making sandstones of Ten- 

 nessee ; and it is difficult to see how the eruptive origin of the 

 sheets lessens the improbability of the supposition. To admit 

 the existence of the faults allows a natural explanation of these 

 repeated sequences, reduces the otherwise surprising thickness 

 of the formation to a moderate measure, and simplifies the 

 eruptive action of the period by correlating nearly all the trap- 

 ridges as the repeated outcrops of a very few overflow-sheets. 

 If the fault-lines thus indicated were disorderly, and could not 

 be shown to be related to one another, very complete evidence 

 might be demanded in proof of them ; but they are most sys- 

 tematic in their position. In any single district, they are all 

 nearly parallel to one another ; south of Hartford, their trend 

 is north-northeast or northeast; north of Hartford, they follow 

 closer to the meridian. The upthrow is on the east, with in- 

 significant exception. It does not seem illegitimate to regard 

 so systematic an arrangement as a valuable corroboration of 

 the evidence that first indicated the existence of the faults, and 

 even to use it as a means of extending their discovery to local- 

 ities where the evidence from sequences could not be fully 

 applied for want of sufficient outcrops. It has in this way 

 been made at least extremely probable that all the large trap 



