20O DISCOVERY REPORTS 



rolls to starboard, that is, away from the side of operation, the wire must be allowed 

 to run out easily ; as it rolls to port the drum must be checked by the brake or the wire 

 may become slack, over-ride a sheave and possibly kink. The danger of this happening 

 becomes less the further the net goes out, because the strain on the wire is increasing 

 with its own increasing weight ; in hauling in the strain completely removes the danger. 

 It has already been explained in a previous section (page i66) that modifications were 

 made to the apparatus to prevent the wire from over-riding the sheaves when being paid 

 out, and that the spring accumulators were of considerable assistance in mitigating this 

 nuisance. But when all possible improvements had been made, the danger of kinking 

 still existed and great care was always necessary in paying out the wire. If a hitch 

 occurred the strain outboard was taken up by means of the brass clamp shown in 

 Fig. 26. 



Reference has already been made, on page 183, to the defects in the original design 

 of the net which caused failure of operation. The fouling of the closing mechanism by 

 the wire bridles or throttling rope was most likely to occur when the net was allowed 

 to run out too fast. Even now that these difficulties have been overcome, and under 

 favourable conditions of sea, a limit should be set to the speed of descent; for at great 

 depths the wire by its increased total weight may fall faster than the net itself if allowed 

 to run out at full speed. This, with the rolling of the ship, was another cause for pre- 

 mature releases with the old type of gear (see page 192). 



As soon as the net reaches the required depth, the drum is stopped and put into gear, 

 and the messenger, unless a haul to the surface is being made, is attached to the wire and 

 held by a piece of line until the moment for despatch. The net is wound up at the 

 speed of I m. per second. An even and exact rate of winding may be maintained, after 

 a very little practice, by regulating the steam valve whilst watching the metre-recording 

 sheave in conjunction with a stop watch. The times that the particular messenger in use 

 takes to fall to the difi^erent depths required has been calculated from previous trials. 

 Thus when the haul is (say) from 750 to 500 m., the messenger, which is known to fall 

 500 m. in 160 sec, is released from the surface when the metre wheel records that the 

 net, in its upward passage, has reached 660 m. The upward haul is continued without 

 a break, so that the net meets the messenger at 500 m. and is closed. After this the 

 net, no longer fishing and ofl^ering less resistance, may be hauled more quickly, but 

 never more than 4 or 5 m. per second if damage to the net is to be avoided. In the 

 haul from 1000 to 750 m. the messenger, which takes 240 sec. to fall 750 m., 

 must be released when the dial reads 990: i.e. 10 sec. after the haul is started. At 

 greater depths it may be necessary to release the messenger and wait before starting the 

 haul : thus from 3000 to 2500 m. one must wait 5 min. A table of dial readings and times 

 is easily drawn up for any particular type of messenger. Careful winding should ensure 

 accuracy of release to within i or 2 m. in the shallower hauls and not more 

 than 10 m. in the 1000-750 m. haul. When the ship is drifting before a wind so 

 that the wire is off the vertical a slightly longer time must be allowed for the falling 

 of the messenger. It has already been explained (p. 166) that the spring accumulators 



