490 



Mr. J. C roll's Further Remarks on 



the whole 3J feet of slope derived from temperature. I do not 

 know upon what grounds he believes that the difference in sali- 

 nity is so great between the equatorial and Atlantic columns. 

 Certainly the researches of the ( Challenger'' expedition do not 

 warrant any such conclusion. It is true that there is an excess 

 of salinity in the surface-water of the North Atlantic; but it 

 does not extend to any great depth. This superior salinity of 

 the warm upper stratum of the North Atlantic, it may be ob- 

 served, is an additional evidence that the water is of Gulf-stream 

 origin. 



Through the kindness of the Hydrographer of the Admiralty 

 I have been favoured with all the observations made in the 

 1 Challenger > of the specific gravities of the Atlantic at interme- 

 diate depths, between surface and bottom. From these observa- 

 tions it will be seen that there is scarcely any sensible difference 

 between the mean specific gravity of the equatorial and the two 

 Atlantic columns. 



The following Table shows the mean specific gravities of the 

 three columns. 



Atlantic. 



Equatorial. 



Depth, in 

 fathoms. 



I. 



Lat. 38° 3' N. 

 Long. 39° 19' W. 



II. 



Lat. 20° 58' N. 

 Long. 22° 57' W. 



Depth, in 

 fathoms. 



III. 



Lat. 1° 22' N. 

 Long. 26° 36' W. 



IV. 



Lat 3° 8' N. 

 Long. 14° 49' W. 



Specific gravity 

 at 60°. 



Specific gravity 

 at 60°. 



Specific gravity 

 at 60°. 



Specific gravity 

 at 60°. 



Surface. 

 100 

 150 

 250 

 400 

 500 

 1500 



1-02884 



102685 

 102732 

 1-02658 

 1-02642 

 1-02609 

 1-02600 

 1-02620 



Surface. 



50 



90 



100 



200 



300 



400 



1500 



1-02616 

 1-02630 

 102627 



1-02591 

 102658 



1-02643 

 1-02620 

 102610 

 1-02629 

 1-02613 



10267/ 

 1-02641 



1-02607 

 1-02618 



1-02608 

 1-02607 



102618 



Mean... 



1026211 



= mean of 

 column A. 



1-02623 



= mean of 

 column B. 



Mean ... 

 Mean of 

 two co- 

 lumns 



1026181 

 I 1-026202 = 



1-026223 



f mean of equa- 

 ls torial column C. 



The mean specific gravity of the equatorial column, as proved 

 by the two soundings III. and IV. in the Table, is 1-026202: 

 and 1*026211 of sounding I. of the Table may be regarded as 

 the mean specific gravity of the North -Atlantic column A ; for 

 the observations were made at a place on the same latitude and 

 only about two degrees to the east of that column. Consequently 



