244 bulletin: museum of comparative zoology. 



near Provincetown, the thirty-four fathom temperature was 39°-39.o . 

 A few miles further south, i. e., along the shore off Cape Cod and 

 Nauset light-houses, the bottom readings were : — 



These records, taken with reversing thermometers (Tanner, 1884c), 

 show that the deeper waters of the Gulf were considerably colder in 

 the summer of 1882 than in any other year of which there is record; 

 and, that in that year, as in 1913, the coldest water was not the 

 deepest but in an intermediate zone at 50-70 fathoms. But the sur- 

 face temperature of the Gulf in 1882 was apparently normal, just as it 

 was south of Cape Cod, so far as the readings taken at various light- 

 houses along the coast show (Rathbun, 1887). 



In 1904 Dawson took a few subsurface temperatures at the mouth 

 of the Bay of Fundy, finding water of 44.6°-48.4° in July, at fifteen 

 fathoms; 48° to 52.1° in the middle of August, which agrees very well 

 with our results. And Dawson seems the first to notice how the shoals 

 and banks lower the surface temperature of the Gulf by causing ver- 

 tical circulation (Dawson, 1905, p. 15). 



Observations on salinity previous to 1913 are very scanty and many 

 of them unreliable. Libbey (1891) took a large number of specific 

 gravities in 1889, in the waters south of Marthas Vineyard, with 

 the ordinary floating hydrometer. And although this instrument, as 

 now universally recognized, is not sufficiently accurate to satisfy the 

 demands of modern oceanography, his surface records agree fairly 

 well with those of 1913, when reduced to salinity by Knudsen's (1901) 

 tables. Thus on August 19, 1889, the surface salinity south of Block 

 Island rose from 33.5%o, over the thirty fathom curve to 34%o at the 

 100 fathom curve; in 1913 it was 34%o near by, over the sixty fathom 

 curve. Apparently, then, his instruments do not require the correc- 

 tion which Clark (1912) found necessary to apply to those used on the 

 Albatross. But his subsurface readings yield salinities as high as 

 38.5%o at 100 fathoms, 39%o or more at 500 fathoms. Such values 

 as these are, of course, out of the question in the North Atlantic, where 

 the 500 fathom salinity is known to be about 34.9%o (Murray and 

 Hjort, 1912) being equalled only in the eastern half of the Medi- 



