1882. 121 [Davis. 



notably the West Rock range in Connecticut, and the Palisades in 

 New Jersey, are demonstrably intruded, by far the greater num- 

 ber of trap-ridges in the Connecticut Valley at least are outcrop- 

 ping edges of well-proved contemporaneous trap overflows poured 

 out from fissures, not now determinable, upon the unfinished ac- 

 cumulation of sandstone strata, and buried under later deposits. 

 This origin was first stated by Hitchcock in 1833 and 1841 ; again 

 by Dawson for Nova Scotia in 1848; but since then it has 

 generally found but little acceptance. The cause of its 

 neglect seems to be two-fold : in the first place, th'e two sheets 

 already mentioned as intrusions, and so known because they 

 break across the enclosing strata at certain points and bake the 

 rocks on their back as well as below them, are of easy access 

 from two centres of observation, New Haven and New York ; and 

 as the external form of these intruded sheets is essentially the 

 same as that of the other trap ridges of less convenient access 

 and examination, they seem to have been taken as the types, and 

 their explanation has served for the others as well. In the second 

 place, the most important point of observation for the distinction 

 between overflow and intruded sheets is the contact of the trap 

 with the overlying sandstone ; if an intrusion, this sandstone is 

 baked ; if an overflow, the sandstone is not in the least baked, and 

 it may contain fragments from the pre-existent trap below it. 

 But these upper contracts as a rule are rarely visible, the 

 overlying sandstones have generally been worn away below the 

 present lines of drainage, or are covered by soil and drift ; and 

 in their usual absence, the evidence taken from the baked sand- 

 stone on the intruded sheets has been improperly extended to the 

 overflow sheets. But in addition to the character of the overly- 

 ing sandstone, there is another feature of the overflow trap sheets 

 that has good probability of serving to detect them ; this is the 

 vesicular or amygdaloidal condition of their upper surface, the 

 result of the giving off of gas by the molten rock when the exter- 

 nal pressure allowed of such expansion, that is when the trap 

 lava escaped from its deep fissures and flowed over the Tri- 

 assic sea-bottom, and became bubbly on its upper surface after 

 the manner of modern lavas. So far as observed, the trap 

 sheets that are prevailingly vesicular at their upper surface are 

 all overflows ; the intruded sheets are as a rule dense and compact 

 above as well as below. 



