1898.] 



PXABrETO>' or THE FAEKOE CHANKEI,. 



1017 



The procedure was then as follows : — The not and machinerj', 

 weighted up to 100 lbs., weie lowered overboard, and a number of 

 fathoms run out, slightly greater than that of the sounding in the 

 case of 'the lowest horizon ; the angle made by the line when taut 

 was approximately measured, and a calculation made from Traverse 

 Tables in the ordinary way as to the depth which the net had 

 reached. As I have pointed out already \ this, the usual method, 

 is most fallacious ; for the towing-line does not form the hypo- 

 tenuse of a right-angled triangle (as presupposed by this method), 

 but an unknown catenarj', which is practically uncalculable except 



ClIAUT OF TilE FaeEOE OiIANKEL, 



Showing the collecting-stations of H.M.S. ' Eesearch ' in 189fi and 1897. The 

 contour-lines have been roughly plotted from the Admiralty Chart and 

 from the soundings taken on these cruises : they are dotted whei-e the 

 soundings are far apart. (Station 20 (1897) is N. of Station 13.) 



by tedious experiment in order to obtain the necessary data. The 

 fallaciousness of this method was brought home to me by striking 

 bottom at 398 fathoms (Station 16 a i) with 450 fathoms of warp 

 out, though by quadrant and traverse tables the net should only have 

 reached 300 fathoms. Fortunately all the details of the previous 

 hauls had been kept ; and as there was sufficient evidence, from 



' Proc. Zool. Soe. 1898, p. .508. 



Peoc. Zooi. Soc— 1898, No. LXYII. 



6^ 



