388 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
advance of winter. And the data show that such was the case, for 
on the next visit, December 4, we found that the water was not only 
appreciably colder at all depths, but very nearly uniform from sur- 
face to bottom, the surface temperature having fallen to 6.6°, with 
46.1° at 20 fathoms and at the bottom (30 fathoms). Probably the 
slight excess of heat at the surface over the deeper layers was the result 
of diurnal warming, the day being sunny and calm, with an air temper- 
ature at noon of 46°. The salinity was 32.56%, on the surface and at 
25 fathoms, 32.61%, on the botton, 2. e., practically the same as at the 
preceding station (November 20). The density at the surface (at 
the temperature in situ) was 25.38, at 25 fathoms 25.39, at 38 fathoms 
(bottom) 25.42. The fact that the surface water was slightly less saline 
than the subsurface layers is no doubt to be explained as the result 
of recent rains. 
The next station was made on December 23, on a bright sunny day, 
with a brisk northwest wind, and air temperature, in the shade, at 
noon, of 36°. Considerable cooling of the water proved to have taken 
place during the three weeks since our last visit, and the fact that this 
was the first station at which there was no change of temperature at 
all with depth, the reading being 4.5° from surface to bottom, shows 
that convectional overturning, together with tidal currents, now kept 
the water thoroughly mixed. The water samples proved especially 
interesting, for while the salinity, like the temperature, was uniform 
from surface to bottom, it was considerably higher than any previous 
reading in Massachusetts Bay, 2. ¢., 32.74%p,, good evidence that there 
must have been an accession of salt offshore water, the origin of which 
is discussed (p. 400). 
On January 16, 1913, at the same locality, the water had cooled to 
41.7° at the surface; 41.5° at 25 fathoms; 42.1° at 38 fathoms, an im- 
structive series for the fact that the lowest temperature was at the 
mid-level shows that the convectional overturning, now a constant 
phenomenon only interrupted by diurnal warming of the surface, was 
most active in the upper 25 fathoms, foreshadowing the time when cool- 
ing would be so rapid at the surface that the latter would be constantly 
cooler than the bottom. The slight excess of warmth (.2°) of the 
surface over the 25 fathom reading, was no doubt the result of diurnal 
warming during the preceding two or three days, which were unseason- 
ably warm. The salinity was 32.81% . at the surface; 32.86%p at 28 
fathoms; 32.94%, at 38 fathoms (bottom); a considerable rise since 
the previous stations. The difference in salinity between surface and 
bottom was probably in part evidence of an inshore flow of salt bottom 
