1898.] PLANKTON OP THE EA.EROE CHA.TSTXEL. 571 



and fired the trigger of the lower lock-movement, thus slackening 

 the bb chains and allowing the net frame to fall open ; still 

 travelling upwards, as the apparatus was hauled in, the propeller 

 presently fired the trigger of the upper lock-movement, slacked 

 the aa chains, and the net then closed. The whole apparatus was 

 prevented from spinning in its descent, and thus causing the 

 propeller to begin travelling too soon, by being suspended from a 

 swivel which worked on ball-bearings. 



This arrangement worked successfully in about three hauls out 

 of four, the failures being generally due to one or both chains 

 hanging on the hammer, even when the lock-movement had been 

 fired, owing to the great friction of the chains on the hammers. 

 A further disadvantage in the apparatus was the difficult ad- 

 justment of the distance between the triggers, which determined 

 the distance in fathoms for which the net remained open ; th's 

 further had a tendency to vary somewhat with the rate of 

 hauling in.] 



In designing the locking-gear of the 1897 pattern I therefore 

 abandoned the propeller in favour of messengers, which I had 

 originally avoided on the grounds of others' experience with the 

 light messengers of deep-sea thermometers. There seems, how- 

 ever, to be no objection to the use of heavy messengers on any 

 well-stretched rope (hemp or wire) which hangs free of the 

 bottom, and in which kinks are thus avoided by the maintenance 

 of a steady strain. 



Photographs of the whole apparatus are given on pnge 572. 

 Details of the locking-gear are furnished by figs. 3 a, 3 b (p. 569), 

 which are sectional drawings at right angles to one another. They 

 are carefully drawn to scale, about one-seventh of the real size. 



Four vertical pillars of teak l T, connected below by two cross- 

 pieces of the same material T', and strengthened by iron plates at 

 the angles, form a rigid frame ; on to this is screwed a brass 

 casting D, to which a second casting E is screwed. The rope by 

 which the machine is slung passes through a hole in the centre of 

 D into the space R between D and E, and is kept there by being 

 worked into a broad knot. 



Two brass cylindrical rods or pins FF (fig. 3 ) run in two good 

 bearings through D and are rigidly bolted into a cross-piece which 

 carries a third shorter pin a, travelling in bearings through the 

 centre of E. The pin a is passed through the chains from aa on 

 the net-frame, and is kept in place by springs (not drawn) between 

 the hooks shown in fig. 3 a with a pull of about 10 Ibs. If a weight 

 be dropped on to the pins FF, it will overcome the springs, depress 

 the pin a, and let go the chains from aa. 



A second pair of pins GGr (fig. 3 b~) run in bearings through D, 

 and through another casting H which is bolted to TT. They are 

 rigidly bolted to a cross-piece which carries a third pin b, travelling 

 in bearings through the centre of H ; this pin is passed through 



1 Teak is one of the few woods that will resist the enormous pressure at 

 great depths ; less closely grained woods warp and split. 



