HARVEY'S METHOD 



265 



accuracy, so that numerous observations could be taken in spite of the bad weather 

 normally prevalent in the southern ocean. We had also to consider the fact that the 

 greater working height above water on the larger ship would tend to increase the 

 surging strain on the gear during heavy rolling. We therefore 

 decided to have the apparatus assembled for vertical upward 

 hauling only, in conjunction with one of our well-tried single- 

 type release gears. This enabled us to substitute a metal 

 upper cone with rigid bridles for the upper canvas cone with 

 throttling band of Harvey's original model (Harvey, I934rt, 

 p. 762). The circular body carrying the meter we also had 

 made of heavy brass tube, nickel-plated. The weight of the 

 attachment ring and bucket was taken off the silk net by 

 three wire bridles shackled to lugs on the upper cone, and to 

 a 10 lb. lead below the bucket. Fig. i shows the apparatus 

 rigged in this way. 



The additional error introduced by the meter spinning 

 during the interval between the net breaking surface and the 

 brake being again applied was' found not to exceed ±3 % by 

 trials against stop-watch under the most adverse conditions. 

 This is avoided by Harvey's method of the double release 

 gear and allowing the balanced apparatus to fish both while 

 being lowered and while being hauled up. It was felt, how- 

 ever, that the risk of fouling would be so great in all but the 

 calmest weather that this procedure would prove unsuitable 

 for continuous work in the open sea. With the apparatus 

 rigged in the fashion we finally adopted, we were repeatedly 

 able to make routine observations in winds up to gale force, 

 and rarely obtained markedly discrepant meter readings 

 unless there was a bad stray on the wire, when we found, 

 as Harvey had done before us, that an unexpectedly larger 

 volume of water appeared to pass through the net. 



Our meters were made by Messrs R. W. Munro, Ltd., and 

 calibrated by the National Physical Laboratory. Colour 

 standards from Harvey's formula made up in sealed tubes by 

 British Drug Houses were sent out each season, and checked 

 against freshly prepared solutions in shore laboratories when 

 occasion offered. No signs of fading or darkening were 

 observed over the periods for which the standards were in 

 use. It was sometimes found that small southern samples 

 gave a slightly yellower tint than the original Harvey standards, but medium and 

 larger catches always gave a good match. 



Phaeocystis brucei sometimes gave trouble by clogging the filter, until Mr Marr hit 



Fig. I. Semi-diagrammatic 

 sketch of Harvey apparatus 

 rigged ready for use. b. bucket, 

 b.l. brake-line, b.lev. brake 

 lever, bri. bridles, c. collar, 

 c.b. clamping bands, d.s.l. deep 

 sea lead, d.w. dial window, 

 /.link, 7n.b. main body carrying 

 the meter, r.g. release gear, 

 ^.spring, ^.w. silk net,!/. f. upper 

 cone. 



