PHYSICAL OCEANOGRAPHY OF THE GULF OF MAINE 
879 
Series O: Bottles Nos. 1952 and 1981 to 2000, set out on July 18, 1926, by 
T. E. Graves, on a line running 107° from Cape Neddick, Me., for 9 miles, 1 bottle 
every one-half mile. First bottle (No. 1952) put out at 8.17 a. m.; last bottle (No. 
2000) at 10.44 a. m. 
No. 
Distance 
set out 
from 
Cape 
Neddick 
Where found 
Date, 
1920 
Inter- 
val 
1982 
Miles 
l l A 
Kenwood Bridge, Salem, Mass 
Aug. 4 
Days 
17 
1985 
3 
10 miles southeast by south from Thatchers Island, Mass 
Aug. 3 
16 
1987 
4 
do 
do..... 
16 
GENERAL DISCUSSION OF THE RECOVERIES 
With the Bay of Fundy experiments as a guide, it was natural to expect a con- 
siderable number of the bottles released in the Gulf of Maine on the several lines off 
Mount Desert, Cape Elizabeth, and Cape Ann, in 1922 and 1923, to be picked up in 
the Massachusetts Bay region. This, however, did not prove to be the case. Not 
a single bottle from any of these series has been found anywhere between Cape Ann 
and the southern elbow of Cape Cod, and only five of them south of Kennebunkport. 
It is therefore evident that the dominant surface drift was not the same in the sum- 
mers of 1922 and 1923 as it was in 1919, but drifts of the 1919 type were recorded 
for series L and O, as described below. 
The most striking aspect of the experiments carried out in all these summers is 
that more than 30 per cent of all the recoveries of bottles put out north of the south- 
ern angle of Cape Cod have been from the Bay of Fundy and Nova Scotia, which 
(if these were the only data available on the circulation of water in the gulf) would 
obviously suggest a drift from south and west to north and east. However, as we 
have just seen, the bottle drifts of 1919 and of 1926, on the contrary, point to an 
anticlockwise current skirting the shores of the gulf from northeast to southwest, 
and salinities (p. 910), temperatures (p. 918), and the distribution of the plankton 
(p. 923) all point in the same direction. It therefore becomes necessary to reduce 
these apparently contradictory lines of evidence to a rational order, which may best 
be done by analyzing the results for the years 1922 to 1926 regionally, not chrono- 
logically, to test whether they prove consistent, one with the other. The dominant 
sets of the surface water are shown rather clearly for the southwestern part of the gulf 
by the lines off Cape Ann, in Massachusetts Bay, off Cape Cod, and in Vineyard and 
Nantucket Sounds. These, therefore, may be considered first, leaving until later the 
study of the more puzzling drifts of the bottles set out in the northern side of the 
gulf. 
SOUTHWESTERN SERIES 
These bottles were set out off Cape Ann, in Massachusetts Bay, off Cape Cod, 
and to the southward of the latter. 
The Cape Cod line of July, 1922 (line B), proved, in some ways, the most instruc- 
tive of all, for out of these 600 bottles, 131, or 22 per cent, were picked up within 
