920 
BULLETIN OF THE BUREAU OF FISHERIES 
(fig. 53; isotherms for 10° and 12°) suggests an eddying movement, drawing warmer 
water inward over the bank on the western side; but in other summers the cool 
drift extends much farther westward. Bottle drifts, for example, place 1922 in this 
category (p. 883); and Libbey (1891 and 1895) records it in longitude 70° to 71° in 
the summer and early autumn of 1889. 
In another chapter (p. 585) I have tried to make it clear that the areas of low 
and high surface temperature, which characterize various parts of the Gulf of Maine 
in summer, are due chiefly to tidal stirring — most active over the shoal banks and in 
the northeastern part of the area generally, least so in the basin off Massachussetts 
Bay. Tidal stirring also plays a part in holding the surface temperature somewhat 
lower along the western margin of the gulf and around the shore of Massachussetts 
Bay than a few miles out at sea; but the gradation also points to some movement 
of the surface water eastward, away from the shore, under the impulse of the 
prevailing southwestern winds, an event with which bathers on our beaches have 
long been familiar (p. 588), and which takes part in the development of the western 
side of the Gulf of Maine eddy. The evidence (by bottle drifts) of a westerly set 
from the Nova Scotian side and from the Bay of Fundy along the coast of Maine is 
also borne out by the extension of surface water colder than 14° westward past 
Penobscot Bay in August (figs. 46 and 47) over depths so great that tidal stirring, 
in situ, is not active enough to be responsible, per se, for surface values as low as 
those actually recorded there. 
The 40-meter charts for July and August (figs. 52, 53, and 54) also suggest a 
similar westerly drift by the isotherms for 8° and 9°, though at this depth the water 
moving in that direction from the Nova Scotian side is warmer than that which it 
replaces off the coast of Maine — not colder, as it is at the surface. or discussion 
of this bathymetric difference, see p. 608). 
The mutual relationships of waters warmer and colder than 9° were especially 
suggestive in August, 1913, as locating the vortex of the anticlockwise eddy about 
60 miles south of Mount Desert and Penobscot Bay (fig. 52). The correspond- 
ing chart for 1914 (fig. 53) is not so easy to interpret in this respect, the picture 
being complicated in the western side by a pool of water cooler than 6°, which 
probably owed its low temperature to vertical stirring or to local upwelling in the 
mid depths. 
None of the summer charts for temperature reveals any dominant movement of 
warm water into the gulf from offshore at the surface, nor do the 40-meter charts for 
the summers of 1914 or 1915, but some circulatory indraft of this sort is suggested 
on the 40-meter chart for 1913 (fig. 52) by temperature, just as it is by salinity 
(p. 782), by the warm (>10°) tongue in the eastern side of the basin, with lower 
temperatures on either hand, to which the reader’s attention has already been called 
(p. 608). 
At first sight the distribution of temperatures at 40 meters prevailing in July, 
1914 (fig. 53), might suggest a drift into the gulf from offshore across the eastern end 
of Georges Bank, but a closer analysis makes it clear (p. 617) that in this case unity 
of temperature had a local significance only, being an adventitious result of the fact 
that vertical mixing was most active on the northern part of the bank. 
