THE HOLYOKE SHEET. 459 



portions of the trap as large as a fist being \vra})ped around by tliick flakes 

 of the thin-fissile, sandy shale, as if balls of i)utty had been separated by 

 being folded in thick wads of wet wrapping paper. 



Above this intimate mixture a few angular fragments of scoria are 

 inclosed for a foot or two in the thin-bedded sandstones. This layer can be 

 followed north 10 miles wherever the upper surface of the trap is exposed. 



Another contact of the sandstone upon the trap occurs on the West- 

 field-Holyoke highway, just where it crosses a brook, and this is the most 

 southern point where the trap contains limestone inclusions at its surface. 



President Hitchcock plainly refers to a further effect of the trap farther 

 south on this line, in West Springfield, at a place which escaped my obser- 

 vation, when he speaks of the limestone in contact with the trap being 

 converted to "tripoli" and in part made brittle as glass.' 



MAGMATIO DIFFERENTIATION. 



Many fragments of the traj) which were inclosed in the mud while 

 still molten are bordered with black from the concentration of the iron in 

 feathery groups of twinned octahedi-a of magnetite. This illustrates on a 

 small scale a process which has been the subject of much study — the differ- 

 entiation of a molten magma into a more basic portion, which seeks the 

 cooled outer surface, and a more acid one, which remains at the center. 

 When this process is earned to its limit the centers of the fragments 

 become white and free from iron and iron-bearing minerals, and the frag- 

 ments of white trap described on page 365 seem to have been thus formed. 

 They are found only in this contact layer and in the sandstone immediately 

 above it. 



ORIGIN OP THE CLAY AND MARL DEPOSITS. 



It is hard to explain how, over a portion of the surface of the great 

 sheet, so large a quantity of laminated marl can have been deposited and 

 then become so regularly and deeply intermixed with the trap. It seems 

 most probable that the central currents carried the mud out over the sheet 

 while it was still moving, and filled its brecciated surface, and that the mud 

 flakes sank down at times into the still-liquid trap in such quantity that 

 they were merely indurated and cemented by the small quantity of the 

 diabase. 



'Geol. Mass., 1835, p. 433; 1841, p. 659. 



