1898.] PIAITKTON OF THE lAEEOE CHANNEL. 573 



the chains from bh on the arms of the net-frame, and is kept in 

 place by springs (rubber loops) between the hooks shown in 

 fig. 3 h, with a pull of about 10 lbs. If a weight be dropped on to 

 the pins GG, it will overcome the springs, depress the the pin b, 

 and let go the chains from bb. 



The apparatus is worked thus : — The whole machine is lowered 

 with the locking-gear in the position drawn in figure 4, the chains 

 aa held on the pin a, but not carrying the weight of the net 

 and frame ; the chains bb held on the pin b, and holding the net- 

 frame tightly closed by its own weight. When the machine is at 

 the bottom of the zone which it is desired to study, the first 

 messenger is despatched down the rope ; this, being small, drops 

 into the nest N, striking on the pins GG, and fi-eeing the chains 

 bb ; the net-frame falls 6 inches, and opens, the weight being 

 caught with a jerk on the chains aa. 



The machine in this condition (fig. 5) is hauled upwards for a 

 hundred fathoms ; the second and larger messenger is despatched, 

 which, striking on the pins FF, frees the chains aa ; the net 

 falls 15 inches, the weight is caught again on the chains bb ; the 

 net-frame closes, and can be then hauled in-board -without any 

 admixture with the fauna of higher zones (fig. 6). 



The chains of course are not let go altogether, as the net and 

 frame would then be lost ; each chain has a large link in it to go 

 over its pin, and beyond this a short length by which it is bolted 

 to T or a shackle-bolt in the centre of T'. 



Chains aa. Chains bb. 



Prom net-frame to pin 33 in. 23 in. 



From pin to T 9-5 in. .... 



Prom pin to central bolt of T' 12 in. 



The messengers used in 1897 were clumsy and unnecessarily 

 heavy, and will not be described here. Probably weights of 4 lbs. 

 for the smaller and 6 lbs for the larger are amply sufiicient on 

 rope : smaller weights would do on wire, since the friction is less. 



The apparatus was tested in 1897 on H.M.S. ' Eesearch,' but, 

 unfortunately, owing to heavy weather, we were only able to spend 

 one day in the deep water of the Paroe Channel ; the apparatus 

 was tried four times, and seemed to work perfectly. The only 

 improvement which suggested itself was that a weight should be 

 hung from the axle C into the middle of the net, heavy enough to 

 prevent the net in its descent from washing up into the machinery 

 (which happened once, but without serious consequences) ; the 

 additional weight at this point wiU also serve to shut the net- 

 mouth more closely, and can also be arranged to prevent the sides 

 of the net compressing the contents when closed. Should the 

 first messenger strike FP before GG, the net would simply come 

 up empty, having been open only for a few seconds. 



Weight of net-frame 16j lbs. ; of locking-gear and chains 33 lbs. ; 

 of messengers used in 1897 (7g-|-10) IT^ lbs. ; of messengers for 



Pboo. Zooi. Soc— 1898, No'xXXVm. 38 



