260 " BULLETIN OF THE 



is somewhat uneven, but not more so, I imagined, than would result 

 from the washing in of sand and mud to fill an irregular hollow ; this 

 exposure is about twenty feet square ; the second is a much larger trian- 

 gular patch of sandstone (D) at the mouth of Fall River, giving excel- 

 lent contact specimens, very little weathered, retaining the usual texture, 

 softness and color of the sandstone directly to the trap surface, and 

 holding small scraps of trap clearly separated from the mass below (see 

 fig. 34). A wooden dam gives passage across Fall River near its 

 mouth, and directly opposite (S. E.) there is a small exposure of sand- 

 stone overlaid by trap breccia, with contact hidden. Farther up this 

 valley, by climbing up its steep southeast bank opposite a road bridge, 

 a larger exposure (E) is found some eighty feet over the stream. The 

 sandstone is gray and micaceous, and dips a little steeper than usual ; 

 its contact with the igneous rocks is hidden, but after a blank of say 

 five feet there is a bank of tufa containing fragments of hard trap. The 

 tufa is deeply weathered, so that I could get no good specimen with 

 fresh surface ; it shows parallel lines on its weathered front, which seem 

 to indicate bedding, as they were conformable with the sandstone strata 

 below. The trap fragments here are round or oval, not angular as by 

 the dam ; some are as much as two feet in diameter, with a hard, dense 

 surface, but vesicular within ; they may be volcanic bombs. Twenty 

 to thirty feet will probably cover the thickness of this tufa bed ; next 

 comes the dense trap of the second posterior range, well exposed at its 

 southern end in a rocky point on the river bank ; passing around this, 

 we come to the third exposure (F) of overlying sandstone, showing the 

 same features as the first. This trap also is not apparent as a ridge 

 farther southwest, but points to a sandstone island in the river, and 

 may, like the first posterior ridge, end about where the river cuts it. 

 From here up to the falls, one fourth of a mile, beds of shale, sandstone, 

 and conglomerate are well exposed in ascending section, but no more 

 trap appears. The close conformity of the trap and sandstone through- 

 out this series is noticeable. 



This section is described and figured by Hitchcock (d, 423 ; e, 653, 

 here copied, fig. 7). Numerous specimens of foot-prints have been 

 found in the neighborhood. Emerson's recent article on the Deerfield 

 Dyke and its minerals appeared after the above description was written : 

 it confirms the overflow origin of the trap, but speaks of only two trap 

 ridges, and regards these as originally one, now separated by a fault in 

 the Fall River valley, I do not consider the evidence offered in favor 

 of this view as fully excluding the idea of separate overflows of limited 



