666 
BULLETIN OF THE BUREAU OF FISHERIES 
does not fall quite as low in winter, being farther offshore. The date when the tem- 
perature rises to its maximum for the year is also about the same' here^as in the bay — 
mid-August for the surface, late August or early September for the 20-meter level — 
but in 1920 this part of the basin was not coldest until about the last week in March, 
whereas the surface in the neighborhood of Gloucester had begun to warm by the 
end of February, a difference corresponding to the difference in location (p. 694). 
Vernal warming is also generally parallel at these two locations down to the 40-meter 
level; but it can readily be appreciated that any upwelling of the much colder bot- 
tom water at any time from June to October would interrupt the orderly progression 
Fig. 88— Composite diagram of the normal seasonal variations of temperature near Mount Desert Island, at the surface, 
20 meters, and 40 meters, from data for the years 1915, 1920, 1921, and 1923. The curves are smoothed 
of the 40-meter temperature, and it is probable that the very low 40-meter reading 
recorded off Cape Cod for August 22, 1914 (station 10254, 5.75°) is to be accounted 
for on this basis. Lacking data for late September or early October, I can not defi- 
nitely state whether the 40-meter level of this side of the basin warms .to its annual 
maximum at about the same date as in Massachusetts Bay (September) . 
The amplitude of the seasonal variation in temperature is nearly the same in the 
superficial stratum of the basin off the mouth of Massachusetts Bay as within the 
latter — i. e., a range of about 17° to 19° from summer to winter at the surface, about 
10° to 11° at 20 meters, and about 7° to 8° at 40 meters. Unfortunately the only 
