142 



curling around and across the latter, and extending westward along 

 the slope for a distance of more than 120 miles. 



No examination was made farther to the northwestward; yet it is 

 probable that Arctic water did not extend much farther west than 

 this beyond the Tail. There it spread out in a pool 250 meters 

 (137 fathoms) deep and 120 miles wide, connected with the more 

 active supply around the Tail by an icy cold neck. 



Two weeks later. May 2-4, we observed that this mass of Arctic 

 water had slu-unk noticeabty in size from its outhne as shown in 

 sketch No. 4. (wSee profile No. 8.) On the east slope the cold cur- 



4-5 



rent from the north had been reduced to a narrow stream restricted 

 to the continental edge about the 100 meter (55 fathoms) depth. 



A proportional shrinkage from the April state was noted at the 

 Tail, and especially 30 miles to the westward along the slope. There 

 was an unmistakable movement of the Arctic water inshore over the 

 bottom of the Bank during this period. On the south, the Gulf 

 Stream had increased in strength as shown by tlie increase in tempera- 

 ture and salinity. This conclusion is based upon the general assump- 

 tion that in a continuous transference of water such as an ocean 

 current the rate at which changes take place in the physical character 



