1877.] Surface Geology of Eastern Massachusetts. 585 
but for the bar across its southwestern end — is a well-marked de- 
pression, and has clearly been formed by the erosion of the 
ancient Huronian granite, petrosilex, and diorite, by which it 
is bordered. Near the middle of the southwest side of the 
harbor, visible only at low tide, is a hard, whitish, fine-grained 
sandstone or arenaceous slate. It overlies unconformably the 
banded petrosilex found on this shore: the petrosilex dips steeply 
to the southeast, while the sandstone has apparently, a gentle 
dip in the opposite direction. Interposed between the petrosilex 
and sandstone is a thin stratum of conglomerate, composed of 
pebbles of the former. Obviously, Marblehead harbor was ex- 
cavated before the deposition of this sandstone, which can hardly 
be newer than the Carboniferous period, and is probably coeval 
with the Primordial rocks in the vicinity of Boston. Other 
remnants of the sandstone are scattered over Marblehead neck, 
in such positions as to indicate that the granite and petrosilex 
of which the neck is mainly composed have suffered but little 
erosion since the formation of the sandstone. The removal of 
the sand-rock from the harbor, which it doubtless once filled, 
may have been the work of ice in recent geologic times ; but the 
harbor itself must have had substantially its present form before 
the close of the Palwozoic era. Salem, Beverly, Manchester, 
and Gloucester harbors have also been cut out of Huronian or 
still older rocks ; and, when we consider their. striking resem- 
blance in form and trend to Marblehead harbor, it is difficult to 
avoid the conclusion that they have an antiquity equally great. 
The limitation of fiords to high latitudes and to coasts favored 
with an abundant precipitation of moisture (usually western 
Coasts), that is, to coasts most favorable for the formation and 
development of glaciers, certainly appears a sufficient warrant 
r the commonly accepted opinion that these deep, narrow, 
and ofttimes tortuous channels are ‘the product of glacial erosion, 
e More especially since fiord valleys usually exhibit, in the 
form of moraines and striated and polished rock surfaces, unmis- 
takable traces of the former presence of glaciers, and in some , 
regions are occupied by existing ice streams. Yet this theory 
fails most signally to adapt itself to some phenomena of an im- 
portant and general nature. Mr. James Geike, in The Great 
ce Age, says that some of the fiords of Great Britain are 
known to date back to the Devonian age, and that, though many 
"ay have been deepened by ice action during the last and earlier 
glacial epochs, they were all originated by streams and rivers in | 
