[5] 



THE BOTTOM OF THE GULF STREAM. 



1049 



cold climate. It is evideut that the " warm belt " is here a comparatively 

 narrow one along the bottom, wedged in between tbe cold waters of the 

 inshore plateau and the still colder waters that cover the outer and 

 deeper part of the Gulf Stream Slope. The actual breadth of this warm 

 belt varies, however, according to the stee])ness of the slope and in con- 

 sequence of variations in the currents. Just south of jMartha's Vine- 

 yard, as will be seen by the ma]), the slope appears to be less rapid 

 than it is either to the eastward or southward, and consequently there 

 is here a broader area occupied by the warm belt, especially between 

 the 65 and 150 fathom lines. Probably this warm belt finally narrows 

 out and disappears from the bottom before reaching the coast of Nova 

 Scotia. Wo have hitherto obtained no evidence of such a belt off that 

 coast from temperature observations and the character of the fauna. 

 Therefore it is probable that the cold water of the greater depths there 

 mingles directly with that of the inshore plateau. Southward the warm 

 belt coutinues to the Straits of Florida, and beyond, the dei)th of the 

 water characterized hy idrntical temperatures gradually increasing as 

 we go south. At Cajjc Hatteras this belt becomes very narrow, owing 

 to the abruptness of the slope, and ai)proaches nearer to the shore; but 

 off the Carolina coasts it spreads out over a wide area, which is inhab- 

 ited by a rich fauna, similar to that investigated by us off Martha's 

 Vineyard. Many of the species are already known to be identical. 



In the following summary table are shown the usual range of varia- 

 tion and the approximate average temperature at the bottom in the 

 more characteristic zones of depth, beyond 20 fathoms, in summer : 



From this table and from tbe diagrams a few of tbe published temperature obser- 

 vatious, which were abuoimally high, have beeu excluded, because they were prob- 

 ably erroneous, owing to a disiilacenicnt of the index, or some other acfrident. 



A singular feature of the serial temperatures taken at many stations 

 is illustrated by Plates IV aud V. In tweut;s-niue localities out of 

 thirty-six, where sufficiently full series of temperatures were taken, the 

 temi)erature was lower at 20 to 30 fathoms than at 50 fathoms. Usually 

 the temperature falls pretty regularly from 5 to 30 fathoms. It then 

 rises often three or four degrees, and sometimes eight to ten degrees, 

 at 50 fathoms, falling again at 100 fathoms, but the temperature at 100 

 fathoms was often higher than at 30 fathoms. In some cases, as shown 



