10 



SUPPLEMENT. 



were sunk depended in some degree on the amount of slack-rope payed 

 out, and also on the strain upon the dredge-rope due to the resistance 

 encountered hy the dredge when dragging; it cannot, therefore, he said 

 that strictly determinate depths were examined hy that method, even 

 assuming that the nets gathered nothing while heing lowered and hauled 

 hack. 



"It occurred to me that hy using an apparatus in connection with a 

 line and lead, payed out vertically as in sounding, and hy dragging verti- 

 cally, instead of horizontally as formerly, there would be at least as much 

 certainty with regard to depths as in the old method, and that simple 

 mechanical devices could be invented to satisfy the conditions of the work. 

 The scheme has been stated in my volume on 'Deep-Sea Sounding and 

 Dredging' (p. 145, foot-note), as follows: 



"'Our plan is to trap the specimens by giving to a cylinder, covered 

 with gauze at the upper end and having a flap-valve at the lower end, a 

 rapid vertical descent between any two depths, as may be desired; the 

 valve during such descent to keep open, but to remain closed during the 

 processes of lowering and hauling back with the rope. An idea of what 

 it is intended to effect may be stated briefly thus: Specimens are to be 

 obtained between the intermediate depths a and b — the former being the 

 uppermost. With the apparatus in position, there is at a the cylinder 

 suspended from a friction clamp in such a way that the weight of the 

 cylinder and its frame keeps the valve closed; at b there is a friction buffer. 

 Everything being ready, a small weight or messenger is sent down, which 

 on striking the clamp disengages the latter and also the cylinder, when 

 messenger, clamp, and cylinder descend by their own weight to b, with 

 the valve open during the passage. When the cylinder frame strikes the 

 buffer at b the valve is thereupon closed, and it is kept closed thereafter 

 by the weight of the messenger, clamp, and cylinder. The friction buffer, 

 which is four inches long, may be regulated on board to give as many feet 

 of cushioning as desired.' " 



The trap was first tried in Narragansett Bay, and soon after was used 

 for the second time at sea, several improvements having been made in 



