i66 DISCOVERY REPORTS 



head of the davit it would be difficuh both to read and to adjust. For the lead from the 

 forward machine on the forecastle head the anchor davit is employed, with a removable 

 frame carrying the accumulator, and here, as shown in Plate XIII, fig. 2, the recording 

 sheave is slung from the davit head. The sheaves used as fair-leads are all inter- 

 changeable: they are 8 in. in diameter and are made of cast steel with a brass bushing. 



In the machine used for vertical plankton hauls it was hoped that the accumulator 

 would tend to reduce inequalities in the speed of hauling due to the rolling of the 

 vessel, and that with its help an even rate of i m. per second could more easily be 

 maintained. As will be seen, however, from the diagram, the arrangement adopted 

 involves the use of a number of sheaves, and it was found by experience that, if the 

 wire was to be paid out rapidly without over-riding the sheaves, it was essential to use 

 a comparatively light spring on the accumulator. With the greater strain of hauling this 

 spring was always in full compression, and though a moderate amount of compensation 

 for rolling was afforded by the engine itself (which slowed automatically as the vessel 

 rolled to starboard), the accumulator gave no assistance. But though the accumulator 

 thus had little value for the purpose for which it was primarily intended, it proved of 

 benefit in other ways, for the wire could be paid out more rapidly and, except in the 

 roughest weather, it indicated by a sudden extension the exact moment when the 

 messenger eflFected the closure of the net. It was thus easy to observe whether the 

 apparatus had worked properly and whether the vertical haul had been taken to the 

 desired level. With the machine used for the Nansen-Petterson water bottle the ac- 

 cumulator was useful in the same way : it indicated the time of arrival of the messenger 

 and hauling could begin at once. 



As originally constructed the arrangement for spreading the wire as it was wound 

 on the drums consisted of a sheave (Fig. i) running freely on a shaft whose length 

 was equal to the breadth of the drum, the sheave being moved from side to side 

 by a forked lever. In the forward machine this fitting was replaced, at an early date, 

 by a small carriage with three rollers (one horizontal and two upright) moved by a 

 worm gear. A similar method would have been advantageous with the two other 

 machines, for when being paid out, the wire showed a tendency to jump the sheave. 

 The sheave at the bottom of the accumulator was also loose on its shaft and was 

 free to move laterally for half the distance of that used for spreading the wire. By 

 this arrangement a straighter lead was maintained, but the wire was still liable to 

 over-ride the lower sheave, and to prevent this happening two cheek-plates with small 

 rollers were fitted. 



The Recording Sheaves used in the 'Discovery' to indicate the amount of wire paid 

 out are of two kinds. For the heavy wire ropes carried on the main winch a large 

 sheave was fitted on either side of the winch-house (Plate IX (46); pi. X, fig. 2), 

 lying in the path taken by the wire in its passage from the outer pedestal fair-lead to 

 the stern fair-lead. Each sheave is of brass, i-| m. in circumference, and actuates a 

 Hardinge revolution counter through bevel gearing. Close contact is ensured by two 



