168 Prof. O. P. Hubbard on Erosion in New Hampshire. 
umn of lava and secure a uniform dike to the full height of the 
ak. e hardpan would be fissured and the dike found con- 
tinuous through it; but this supposition implies a previous exrcava- 
tion or erosion of the valleys. 
2. If existing valleys were formerly filled with trap by an over- 
flow from the fissure, they are now clear of it and have been re- 
duced to their former condition, i. e., they have been excavated a 
second time as in the former case. 
3. If the present valleys were formerly occupied by continu- 
ous rock, such as constitutes the mountains, then a support is had 
for the molten lava when injected to the height of the peak, so as 
to allow the filling of the fissure and to form a dike; and the pres- 
ent configuration is the result of a subsequent erosion or excava- 
tion. Whether these valleys, therefore, were ever filled with 
diluvium or trap, need not be shown, as we must in that case, at 
a previous time introduce the latter supposition, involving as it 
does only a single excavation. 
There is a possible supposition that the valleys of the White 
Mountains are the result of fracture and subsidence of their areas 
—or of fracture and elevation of the ridges—but I know of no 
evidence of this, and where the rock can be traced across from 
side to side of a valley, the continuity is complete. To sustain the 
third view above, I cite the following facts. 
Mount Pleasant, the third peak south of Mount Washington, 
is about 4500 feet high. Its top is a plain of five or six acres, So 
trap three feet wide, whose course is east and west, which 1s 
worn off entirely smooth and level with the enclosing rock, and 
may be traced entirely across the top. Circumstances did not 
where the sides of the mountain are covered; but by telescopic 
examination from Fabyan’s, which showed distinctly the little 
column of trap fragments piled upon the dike of two feet in 
height, I was led to infer a probable relation between the dike 
and the slides on the west side of Mount Pleasant, which seemed 
to be in the range with it. However this may be, the dike must 
be considered for the reasons already given, as of much greater 
length than the crest of the mountain and of course extending 
down into or across the valleys. 
Though the positive evidence may be wanting of this latter 
point, yet from a party of our students who, in the autumn 0 
1848, descended into the valley on the east side of Mount Pleas- 
ant, from directly opposite the “Lake of the Clouds,” and fol- 
lowed the mountain stream throughout its course of falls and 
rapids, &c., to the valley of Dry river, and down this to its june- 
tion with the Saco in the Notch, two or three miles below the 
