PHYSICAL OCEANOGRAPHY OF THE GULF OF MAINE 
899 
a very uniform track, which for the Cape Ann and Cape Cod lines veered unmistak- 
ably through southeast, east, and northeast (p. 889). The time intervals are consist- 
ent with this, also, the great majority ranging between 70 and 105 days, irrespective of 
which line the bottle in question was launched from. For any of the bottles from the 
Cape Elizabeth line to have reached the southern shore of the Bay of Fundy by the 
alternative route via the coast of Maine and through the Grand Manan Channel 
would have involved a drift from north to south across the Bay of Fundy directly 
contrary to the dominant set established there by Mavor’s (1922) experiments with 
drift bottles, as well as by measurements of currents (p. 861). Such an explanation 
would also be contrary to the time intervals, for the two bottles that went from the 
offing of Cape Elizabeth to Grand Manan and to The Wolves (Nos. 43 and 88) and 
were not reported until 103 and 104 days after release, while two others, set afloat 
near by (Nos. 99 and 105), were reported from the Nova Scotian side of the Bay of 
Fundy in 80 to 98 days. 
By this reasoning the bottles that went to Penobscot Bay from the inner end 
of line A, and to the coast of Maine farther to the eastward, may safely be credited 
with essentially the same route as those that reached this same sector of the coast 
from the outer end of this line, circling anticlockwise at first toward the Bay of 
Fundy, to return westward again. The time intervals between release and recovery 
(80 days for No. 65, picked up at Jonesport; 63 days for No. 98, reported near Swans 
Island; and 103 days for No. 87, found at Matinicus) favor this interpretation. 
The general uniformity, both of localities of recovery and of time intervals, for 
the outer two-thirds of line A, indicates a well-developed, dominant set of the anti- 
clockwise sort just outlined. This, however, seems hardly to have affected the sur- 
face water within 15 miles of the land at the time, judging from the regional disper- 
sion of the returns from the inner end of line A and from the fact that the time 
intervals between release and recovery vary widely for these, quite independent of 
the distances which this group of bottles made good. Thus we find intervals rang- 
ing from 25 to 77 days for 7 bottles that were picked up in the Casco Bay region, 
15 to 30 miles from the points of launching, and 5 to 72 days for 5 bottles recovered 
along the southern side of Cape Elizabeth after journeys of 8 to 23 miles. One was 
found at Monhegan Island (35 miles) in 47 days, but another, reported from Danis- 
cove (25 miles), was not found until 75 days had passed. 
Of course, little stress can be laid on the time interval for any one bottle, because 
there is no knowing how long it may have lain on the shore, overlooked; but our 
general experience suggests that if bottles are not reported comparatively soon after 
stranding they are either broken or buried in windrows of seaweed and never after 
heard from at all. Consequently, when time intervals vary widely for bottles 
drifting only a short distance to a coast as frequented as the Casco Bay region is, 
contrasting with uniformity of intervals for bottles journeying right across the gulf, 
it is obvious that the former did not follow as definite a set as the latter. On the 
whole, the regional distribution of the localities of recovery for the inner end of this 
Cape Elizabeth line trends eastward across Casco Bay, pointing to an irregular 
eddying drift in that direction as involving the mouth of the latter. Cape Elizabeth, 
however, seems to have bounded this eddy on the south at the time, witness the 
several strandings to the south of the cape (fig. 181); the fact that one bottle, set 
