90 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
results very similar to the conditions observed over the trench: west 
of Jeffrey’s Ledge. And this is probably the correct explanation. 
Further evidence in its favor is afforded by the fact that diurnal 
changes of surface temperature are not’so great in this region as they 
are further off shore. 
The profiles show that this mass of coast water is fairly sharply de- 
fined from the off shore water east of Jeffrey’s Ledge in July and Au- 
gust, by low temperature and low salinity, in which it agrees with the 
water off the mouth of Massachusetts Bay. And no doubt the con- 
tour of the bottom is largely responsible for this fact by hindering free 
circulation of the water below thirty-five fathoms; because although 
the northern end of the trench is open, and the water there (Station 
22) was salter than it was at Station 11, yet it was so much warmer at all 
depths that the density, depth for depth, was about the same at the two 
stations. Consequently there is no dynamic cause for an active flow 
of water of high salinity into the deep parts of the trench, and the 
latter retains more nearly the conditions of early summer than does. 
the coast water further north and east. 
Temperatures and salinities show that the cold bands of water so 
often observed along the north shore of Massachusetts Bay are evi-: 
dence of upwelling of bottom water, probably due to off-shore winds. 
But in the southern half of Massachusetts Bay, the curves of both 
these factors, taken with the strong tides of this region, show that the 
cool surface water is the result of mixing, rather than of upwelling, 
two forms of vertical circulation which may be perfectly distinct, 
though they are often combined. 
The existence of a band of coast water of very much lower salinity 
than the off-shore water is no doubt the direct result of the vast volume: 
of fresh water poured into the Gulf by the large rivers which empty 
into it, chief of which are the Merrimac, Saco, Androscoggin, Kennebec, 
Penobscot, St. Croix and St. Johns, with a combined water-shed of 
about 45,550 square miles. Unfortunately we have very little data 
on the salinities of the Gulf at any season of the year except in mid- 
summer, but the salinity curves for July and August show that at 
that season, at least, the fresh river water is localized along the 
coast, swinging off shore opposite the Penobscot and off Cape Ann. 
It is true that at that season there is little or no evidence afforded by 
the surface salinities of an influx of river water in the northeast corner 
of the Gulf, although it is there that the greatest volume enters, 1. é., 
from the St. Johns and St. Croix. And although this can be partly ex- 
plained as due to the active vertical circulation in this region, which 
