R. D. Irving—Divisibility of the Archwan. 243 
also to be seen, however, at its base, on the contact line. 
There are places, as has 
- R. Van Hise, of the U.S. Geological Survey, where one 
mere occurrence of fragments of a lower rock within the mass 
of'a higher one does not always prove a great time gap between 
the two rocks, but merely the relative ages of the two; for any 
rock which reaches the solidified state rapidly may yield of its 
material to overlying sediments as soon as beaten upon by the 
waves of the sea. Thus lavas, and certain forms of chemical 
y the waves of the sea an immense thickness of rock material. 
Moreover, it is not only the gneisses that have furnished frag- 
ments to the overlying series. Fragments of the schists accom- 
panying them are also frequently met with in the slates of the 
upper series, 
It may be said, in passing, that these “recomposed” rocks, 
which occur in these slates, close to the sources of the detrital 
material of which they are made, have quite striking parallels 
a the Triassic sandstones of the Connecticut Valley. 
These sandstones lie in a trough of granite, gneiss, mica schist, 
etc. They are described as nearly always highly feldspathic, 
and little sorted. At certain points the sorting process has 
een so entirely omitted that the indurated detritus is, macro- 
Scopically, difficult to distinguish from the parent roc ro- 
essor - Dana, to whom : m indebted for first calling my ‘ 
attention to these recomposed Triassic rocks, has kindly fur- 
nished me with specimens of the sandstones and of the crystal- 
line rocks from which they are derived. We have made an 
Studied thin sections of these sandstones and have compared 
them with a number of the recom d rocks from various 
points in the Lake Superior Archean region, and have found 
them to be of recisely the same nature and origin. : 
aetna should next be invited to certain occurrences indi- 
cated on the accompanying diagram,* which have been ap- 
* Difficult to show on the small scale of the reduced figure printed herewith. 
