698 
BULLETIN OF THE BUREAU OF FISHERIES 
factor in producing the low temperature of the mid-layer of the Gulf of St. 
Lawrence. 
The chilling effect of ice melting in the Gulf of St. Lawrence, and to a greater extent 
of the drift ice melting over the Banquereau-Sable Island Bank region is, in turn, 
brought indirectly to the Gulf of Maine by the cold water flowing westward past 
Cape Sable in spring and early summer (p. 832) ; but no ice, either of Arctic or of 
St. Lawrence origin, has ever been known actually to enter the Gulf of Maine 
though pans (almost certainly from the latter source) do rarely drift down past 
Cape Sable along the edge of the continent or outside it. Consequently, as the 
surface of the open Gulf of Maine never freezes, ice melting in situ plays only a 
very subordinate role in its temperature complex, except in its shallow and more or 
less inclosed bays and among the islands that skirt its northern shores. 
Cape Cod Bay offers an instructive example, on a small scale, of the effect that 
melting ice exerts upon the sea temperature, for more or less ice freezes over the flats 
along its western side nearly every winter. The greatest amount forms during heavy 
blows from the northwest, when it may stretch out 2 or 3 miles from the shore and 
pack several feet high along the beach. When ice has so formed, easterly winds and 
high tides soon disperse it; and, according to the United States Coast Pilot (1912, 
Part III, p. 59), “instances are on record of this ice, and that forming in the shallower 
parts of Cape Cod Bay in severe winters, being driven by the winds out into the bay, 
where it masses into heavy fields or windrows, sometimes as much as 10 feet or more 
thick, making the navigation of parts of the bay unsafe or impracticable at times.’ * 
Unfortunately, no observations were taken in Cape Cod Bay during the ice 
season of the almost Arctic winter of 1919-20, or until April of the succeeding 
spring; but in 1924 a considerable amount of ice formed along the west shore of the 
bay between the 20th and 26th of December, during a spell of very severe weather 
(p. 655), and the temperatures taken by the Fish Hawk on January 6 and 7, 1925, 
showed the effect by a drop in temperature at the near-by station (No. 7) from 
about 4.3°, two weeks previous, to about 0.3°. Ice chilling was also reflected still 
more clearly in the fact that the water was colder just off Wellfleet Bay (station 7) 
than anywhere else in the southern part of the Massachusetts Bay region on that 
date, as is described above (p. 655). 
The sea ice that freezes in greater or less amount among the islands along the 
coast of Maine in all but the warmest winters must also exert a local chilling effect 
on the water as it melts, but no measurements of this have yet been made. 
In severe winters, when much ice forms in Vineyard Sound, most of it reported 
to drift out to the eastward past Nantucket, melting ice must lower the tempera- 
ture of the Nantucket Shoals region indirectly or directly. Here, again, however, 
definite data are lacking. 
Ice is also an effective chilling agent in shallow bays such as Barnstable and 
Plymouth, for the flats, laid bare at low tide, skim over with ice on cold winter days 
or nights, which melts when the tide floods again. This is one reason (active tidal 
circulation is another) why such situations serve as centers for chilling in winter, 
just as they do as centers for warming in summer. 
