904 BULLETIN OF THE BUEEAU OF FISHERIES 



the vicinity of Cape Sable, and one or two found along the coast of Maine well to 

 the eastward, in each instance. In the case of the drifts that cross the gulf, this track, 

 I believe, is now definitely proven to approach the Bay of Fundy from the south or 

 southwest, by the evidence just detailed. 



The relationship which distance traveled bears to time interval between release 

 and recovery also argues for a circuitous route for the bottles that went to Nova 

 Scotia from the Mount Desert line, because the average distance for all of them, in 

 a direct line, would be only about 85 miles, though the times range from 62 to 88 

 days for 8 of 10" (averaging 70 days). Evidence of this sort must, of course, be 

 used with discrimination, because there is no knowing how long a bottle lies on the 

 beach before it is noticed. When the results prove reasonably consistent, however, 

 some trust can be put in them. In the present connection we have as a standard 

 for comparison the Nova Scotian drifts from the lines set out off Cape Elizabeth. 

 The distance (in a dhect line) is only about one-half as great from Mount Desert 

 to Nova Scotia as from Cape Elizabeth. The two lines of 1923 were set out only 

 one day apart, and there is no reason to suppose that bottles from one line would be 

 consistently overlooked while bottles from the other would be soon found. Conse- 

 quently, it is reasonable to assume that some of the Mount Desert bottles would have 

 been found a month or more before the first were reported from the Cape Elizabeth 

 line, unless they had journeyed by a very circuitous route. Actually, however, the 

 first four recoveries for the former were on October 7 to 9 ; the first three of the latter 

 on the 9th and 10th. Allowing the one day's difference in the dates when the two 

 series were put out, we have the rather surprising fact that the time intervals for 

 these two groups, launched almost 100 miles apart, were the same almost to a day, 

 though the strandings were scattered along more than 20 miles of coast line between 

 Yarmouth, Nova Scotia, and the mouth of the Bay of Fundy. 



The time intervals for the Nova Scotian drifts as a whole, from these two series, 

 also correspond much more closely than the difference in direct distance would have 

 suggested as probable, averaging about 75 days for the Cape Elizabeth series (extremes 

 of 56 to 111 days; p. 875) and about 70 days for the Mount Desert group (62 to 151 ; 

 p. 874). 



The percentage of recoveries is not only of the same general order of magnitude 

 for the Mount Desert line as for the Cape Elizabeth line of 1923 (respectively, 28 

 and 19 per cent), but the Nova Scotian and Fundian returns formed almost the same 

 proportion of the total for the former (36 per cent of the total returns) as for the 

 latter (42 per cent) . 



The most reasonable explanation for this correspondence between the two series, 

 and the only explanation that fits all the facts just outlined, is that the journey to 

 Nova Scotia covered about as long a distance for the Mount Desert bottles as for 

 the Cape Elizabeth bottles, and that the former drifted southwestward at first, to join 

 the general route of the latter group from west to east across the gulf. 



Bottle No. 1584, set adrift about 25 miles out from Mount Desert Island and 

 picked up at Haycocks Harbor, on the north shore of the Grand Manan Channel, 

 93 days later, probably followed the same general track as the bottles that went to 



"Three others (Nos. 1530, 1651, and 1657), which were not picked up until 133 and 151 days had passed, may have lain 

 unnoticed on the beach or drifted in and out along the shore with the tides. 



