108 J. D. Dana— Origin of the Rocks of the Cortlandt Series. 
A freshwater stream must have emptied into this Cortlandt 
bay near the present channel of the Hudson; for the general sur- 
face of the Highland area and the course of the existing streams 
over its surface have a pitch southward ; but the length of this 
young Hudson River could hardly have equalled ten miles ; 
for these old lands, as the Lower Silurian in its valleys prove, 
stood at a lower level than now. This little stream was the 
chief one that gave aid to the ocean’s waters in the work of dis- 
tributing Archean detritus over the Cortlandt area. Nothing 
could have come down the valleys called Canopus Hollow and 
Peekskill Hollow; for these were for several miles arms of the 
sea in which limestone beds were accumulating. The cut 
through the Highlands now occupied by the Hudson was prob- 
ably begun ina fracture during the making of the Green Moun- 
tains at the close of the Lower Silurian. 
2. Metamorphic origin of the Rocks. 
The following are the principal points in the evidence sus- 
taining the view that the rocks are, to a large extent, meta- 
morphosed sedimentary beds. 
1.) The mica schist or micaceous gneiss in several places 
graduates into the soda-granite along the plane of contact, 
though always rather abruptly. 
(2.) The soda-granite, near its junction with the schist, and 
sometimes remote from it, contains, at short intervals, distinct 
layers of the schist, in positions conformable to the bedding 
outside, and single beds of this kind are in some cases contin- 
uous beds for 200 feet or more. 
e mica schist at Cruger’s in some parts contains beds 
that consist largely of staurolite, fibrolite, and magnetite (all 
infusible species), with abundant scales of silvery mica, a min- 
eral that fuses with great difficulty ; and the layers of schist 
which are in the soda-granite, just north, have a similar consti- 
tution ; as if they owed their resistance to the fusion which the 
rest experienced because of their consisting chiefly of these re- 
fractory materials, 
4, e noryte and chrysolite rocks contain, occasionally, 
similar included conformable beds of schist; and some of these 
are beds of magnetite and corundum, with fibrolite, that is, are 
beds of emery ; and the noryte is sometimes crossed by gneissi¢ 
layers and has occasional planes of bedding parallel to the 
bedding of the limestone near by. 
5.) Since ascending lavas have the motion of a fluid, deter- 
mined partly as to direction of movement by the friction along 
the sides, a layer of schist 50 or 100 feet long falling into it 
wou t remain entire, and parallel or conformable to the 
original schistose rock; and much less could a series of such 
