688 BITLLETIN OF THE BtTEEAU OF FISHERIES 



conditions are approximated in the deep bowl off Gloucester. By analogy, therefore, 

 we might expect the mean annual temperature of the bottom water of the latter to 

 be lower than the mean annual temperature of the air over the neighboring land, 

 quite independent of any possible chilling by northern sources. And such, by our 

 observations, is the case, the mean bottom temperature of 4° to 5° at 70 to 150 meters 

 depth in this sink being 3° to 4° below the mean annual temperature of the air at 

 Plymouth and Gloucester, on the two sides of the bay, or at Concord, Mass., some 20 

 miles inland. '^ We have not taken readings enough in the deep trough between 

 Jeffreys Ledge and the Isle of Shoals to establish the mean annual temperature as 

 closely there, but such data as are available point to a mean annual value of 4° to 5° 

 at 100 to 150 meters for this locality, about 3° lower than the mean annual air tem- 

 perature at Portland, Me. (7.3°). 



Near Mount Desert Island, which may be taken as representative of the 

 coastal waters of eastern Maine, the mean annual temperature of the bottom water 

 (close to 5° to 6° at a depth of 40 to 50 meters) is about 1° cooler than the mean 

 temperature of the air at Bar Harbor near by, but nearly the same as the air at St. 

 Johns, New Brunswick, and at Eastport, Me. Mean temperatures of 4° to 5° at 

 depths of 100 to 175 meters in the Bay of Fundy for the year November, 1916, to 

 November, 1917,°^ again prove 1° or 2° lower than the mean annual temperature of 

 the air at St. Johns, New Brunswick, on the one side of the Bay, or at Yarmouth, 

 Nova Scotia, on the other (5° to 6°) . 



The foregoing comparison warrants the tentative generalization that in those 

 parts where regional interchange of water is most hindered by submarine barriers 

 the mean temperature of the bottom water averages about 1° to 3° lower than the 

 mean annual temperature of the air over the neighboring lands, a rule applying 

 whether vertical circulation be active, as in the Bay of Fundy, or weak, as off Glouces- 

 ter. The mean annual bottom temperature at equal depths also proves decidedly 

 uniform in such situations in the two sides of the gulf. In the open basin of the 

 gulf the deepest water averages warmer, a fact discussed in a subsequent section 

 (p. 691). In short, it is not necessary to invoke more than a slight influence on the 

 part of the Nova Scotian current, if any, to account for thermal differences between 

 bottom water and air no wider than those just quoted. 



Brief analysis will, I think, convince the reader that this conclusion applies 

 equally to the cold mid layer that usually persists through the summer in the basin 

 of the gulf. The presence of a cold layer of water of this sort in the mid depths, 

 with higher temperatures below as well as above it, has sometimes been classed as a 

 sure criterion for Arctic water. This, however, is not necessarily the case. True, 

 such a state characterizes the polar seas in summer (Nansen, 1902; Helland-Hansen 

 and Nansen, 1909; Knudsen, 1899; Matthews, 1914); and wherever such a layer is 

 colderthan —1° in summer, as it is in the Labrador current and in the extensions 

 of the latter around the slopes of the Grand Bamks (Matthews, 1914; Fries, 1922 and 

 1923; E. H. Smith 1922 to 1924a; Le Danois, 1924 and 1924a) we have positive evi- 

 dence of Arctic water, for nowhere else does winter cooling alone cause temperatures 

 as low as this in the open sea on either side of the North Atlantic south of latitude 60°. 



•' The mean annual temperature Is higher (about 10°) at Boston than at most other stations around the bay. 

 " Calculated from data tabulated by Mavor (1923). , 



