164 William Libbey — Gulf Stream and Labrador Current. 



These two portions of the Gulf stream are therefore seen to 

 have different characters. The lower one, being more steady 

 and constant, is further characterized by the slight changes 

 which take place in it. The upper one, on the other hand, 

 might be said to be characterized by the rapidity of its changes 

 of position. As has been said, the 50° temperature curve is the 

 line which bounds these two portions. 



The shape of this curve beyond the edge of the continental 

 platform is that of the letter S inverted. The lower part of the 

 letter represents the main body or lower portion of the Gulf 

 stream. 



In the year 1889 the lower portion did not touch the edge of 

 the continental platform at any point within the area we were 

 studying. In 1890 this portion of the curve touched both at 

 Block island and at Nantucket in the latter part of the season ; 

 •and in 1891 it touched along the whole edge for the greater part 

 of all the summer months. The change which was thus pro- 

 duced in the temperature at the bottom along this edge of the 

 continental platform was somewhere in the neighborhood of 10°, 

 an item of considerable importance. The effect produced by 

 this temperature change can be seen to l^est advantage by refer- 

 ence to a very interesting problem in biology on which it directly 

 bears. 



In the years 1880 and 1881 a new edible fish was found in 

 considerable numbers in the area we were studying, and had 

 attracted so much attention among fishermen that preparations 

 were made to take it on a commercial scale for the New York 

 and Boston markets during the ensuing season. 



Unfortunately it happened, however, early in the summer of 

 1882, before the fishermen could enter upon their work, that the 

 water from Cape May to Nantucket, in a long crescent-like curve 

 following the continental edge, was covered with the bodies of 

 this fish, dead and dying, in countless millions. From that 

 time the tile-fish (Lophilatllus chamseleonlice'ps) disappeared from 

 this area entirely, and attempts to find the fish since that time 

 have been unsuccessful. The subject, moreover, had become a 

 sort of biologic puzzle. Fortunately the temperature of the 

 water in which the fish was caught had been noted at a number 

 of points. 



In studying over the three sets of profiles for the three years, 

 1889, 1890, and 1891, obtained from our work I noticed the fact 



