848 
BULLETIN OF THE BUREAU OF FISHERIES 
eddying in the extremely complex fashion that may be expected to characterize the 
zone of contact between waters that differ widely in their physical character and in 
their direction of flow. 
Similar alternations between colder (and less saline) and warmer (and more 
saline) bands have been reported on several occasions and at localities widely sepa- 
rated off the eastern seaboard of North America; but in most cases, at any rate, 
these are transitory and rapidly changing phenomena. The westward drift of water 
close in to the upper part of the slope, just described, has, on the contrary, proved 
characteristic of the La Have Bank region; and so long as the dynamic motive for 
this drift persists, the neighboring oceanic triangle off the mouth of the Eastern 
Channel is supplied with slope water from the eastward. By this reasoning, the 
current that flows into the bottom of the gulf via the Eastern Channel draws from 
the slope water manufactured at about an equal depth on the Nova Scotian slope— 
chiefly between Browns and La Have Banks — not from shoaler or deeper strata 
there. 
This conclusion is confirmed by the fact that temperature and salinity proved 
very nearly the same on bottom in the channel (34.7 to 35 per mille and 6° to 7° at 
200 to 250 meters) as at equal depths on the slope between these two banks (34.6 to 
34.9 per mille and about 7° to 8°) in July, 1914, in June, 1915, and again in the spring 
of 1920. 46 
Further evidence that the indraft into the channel is supplied from the east- 
ward, not from the westward, is afforded by the fact that considerably lower tem- 
peratures and salinities have been recorded around the eastern and southeastern 
slopes of Georges Bank (p. 714). In fact, there is reason to believe that the western 
side of the channel is the site of a dominant drift outward from the gulf (p. 974). 
The cold band encountered off the southwest slope of Georges Bank by the 
Grampus in July, 1916, and reported there in other summers (pp. 608, 919) may also be 
credited with an eastern source, because its temperature and its salinity both agree 
closely with that of the slope water that is manufactured in the offing of Cape Sable 
in early summer, as exemplified by the observations taken there in June, 1915, and 
July, 1914 (p. 629; Bigelow, 1922, p. 166). Thus it owes its low temperature indi- 
rectly to the Nova Scotian current (and so to ice melting far to the eastward). 
Why this southwesterly cold current was so much more in evidence along the 
bank in 1916 than in 1889, 1913, or 1914 remains an open question, but it seems 
probable that some westerly movement of slope water takes place along Georges 
Bank to a greater or less extent every spring as the Nova Scotian current floods to 
its maximum velocity and volume. In some years (1889, for instance, and 1916) 
this drift persists into the summer, as seems to have been the state in 1922, also, 
when so many of the bottles set out at the edge of the continental shelf in July 
made long drifts to the westward (p. 882). In other years (exemplified by 1914) 
it seems to be obliterated west of longitude about 68° by July, as the tropic water 
advances toward the edge of the continent. But although so variable, the existence 
of this cool band in some summers is extremely instructive as one of the several 
«The slope water was somewhat more saline at this locality at the end of July, 1915 (Bjerkan, 1919, Acadia station 41), but 
no observations were taken in the channel at the time. 
