356 TJic A)nerican Geologist. Juno, looo 
A rather marked exception to this general rule was re- 
cently observed on the surfaces of the rocks forming the bot- 
tom at certain points among the islands ol¥ the shore of the 
town of Tuftonborough, on lake Winnipesaukee, N. H. Not- 
withstanding the great purity of the waters, it was found that 
the rocks of the bottom showed distinct evidences of differen- 
tial weathering, the diflference in relief between the harder and 
the softer parts being as great, if not greater, than in the same 
rock above the level of the waters. 
The country rock is a rather coarse biotite gneiss, carrying 
in places a considerable amount- of muscovite, and some horn- 
blende, in addition to the biotite. It usually possesses a dis- 
tinct, and often very marked lamination which in general has 
a northwest-southeast trend. It is cut by numerous strings 
and dikes of pegmatyte, mostly small and composed mainly 
of feldspar. It is these strings and dikes which have been 
left in relief by the removal of the surrounding rock, the re- 
sultant ridges running in all directions without reference to 
the lamination of the gneiss, and by their crossing and re- 
crossing, forming a sort of network. Not only is the relief 
rather more pronounced than in similar ledges upon the land, 
but it is also more sharply defined. In amount it varies from 
a fraction of an inch in the case of the smaller strings, up to 
two inches or more in the case of the larger strings and 
dikes. The differential weathering was observed at all depths 
down to some ten feet, this being about the limit of distinct 
vision on a calm day. The relief was fully as strong at the 
greater as at the lesser depths. The ledges show no striation, 
though traces of the heavier grooves may sometimes be seen. 
The surfaces are not polished, but are rough and granular. 
In seeking a cause for this dififerential weathering, the pro- 
cess is largely one of elimination. The erosion can not have 
been pre-Glacial as the situation of the ledges is commonly 
such that they must have been subjected to the full force of tht 
moving ice. That such was the fact is attested by the gen- 
eral outlines and contours of the submerged ledges, and by the 
presence of distinct, though partially obliterated, glacial 
grooves. 
The weathering is certainly a sub-aqueous phenomenon. 
The level of lake Winnipesaukee, though controlled within 
