Observations Conducted on Sailing-Ships 65 



On a second occasion — in Lat. 48° 27', Long. 15° 38' W. — the 

 work of collecting samples of water, and filling from the water-bottle 

 tubes of water for subsequent gas analysis, from depths 1,570, 1,400, 

 1,000, 700, 500, 300, and 100 fathoms respectively, occupied eight and 

 a half hours. 



On another occasion the time occupied in using the closing-net at 

 1.400 1,600, 1,000, 700, 500, 300, 100, 50, and 25 fathoms occupied 

 twelve hours. 



These, of course, are operations with the closing-net, which has to 

 be lowered tightly closed, opened under water by the messenger sent 

 down the line, and closed again under water by another messenger 

 before hauling up ; and these successive operations occupy much more 

 time than would be required for the mere manipulation of nets which 

 it was not so particularly desired to open and close at specified 

 depths. 



There may be a difference of opinion as to the advantages of 

 horizontally towing-nets, such as the author's large closing-net, or 

 Garstang's smaller apparatus, or of vertically actuated closing-nets, 

 such as Fowler's ; but I believe the results are practically the same 

 with horizontally towed nets. 



The drift of the ship of not more than a mile an hour is quite 

 sufficient, and though the tow-net does not remain long at the depth to 

 which it is primarily lowered, because by the drift of the vessel it is 

 constantly rising, it does not rise greatly in the space of fifteen 

 minutes and, moreover, jn - actice proves that if there is much angle on 

 the wire, the messengers for opening and closing cannot descend, and 

 therefore the result is negative. In vertically acting nets the apparatus 

 is raised through so many fathoms, and the result is the collection of the 

 fauna between, for example, 500 and 400, or 400 and 300 fathoms, and 

 so on. The actual quantity captured by either vertically or horizontally 

 actuated nets is probably very little different. Unfortunately, as we 

 cannot see what is going on at any considerable depth, we can only 

 say that the capture has been made approximately at the depth stated. 



While the use of closing-nets gives very interesting and sometimes 



9 



