suspension up to a standard volume, filter, 

 and measure the volume of the fluid removed 

 from the plankton by the filtering process. Of 

 these two the former appears preferable: for 

 example, a sample of zooplankton was meas- 

 ured 10 times by the first method, and the 

 statistics were range 52-69 ml., mean 69 ml., 

 and standard deviation 6 ml.; statistics for 10 

 measurements of the same sample by the 

 second method were 36-80 ml., mean 60 nril., 

 and standard deviation 15 ml. 



Replicate measurements of this plankton 

 sample were made also by the colorimetric 

 method of Sutcliffe (1957). They were highly 

 inconsistent, and several tests were made to 

 determine possible causes. The method de- 

 pends upon the absorption of the supernatant 

 liquid of the plankton sample, before and after 

 the addition of India ink. The factor that 

 apparently contributed mainly to the incon- 

 sistency of the results was the minute invisi- 

 ble debris that remained in the supernatant 

 liquid after visible plankton and debris had 

 settled: readings of transmittance taken after 

 the sample had stood overnight were always 

 several units higher than readings taken after 

 the sample had stood, say, half an hour, till 

 visible debris had settled. In one test a dif- 

 ference of 1.1 units of percent transmittance 

 changed the computed plankton volume by 28 

 ml. The method is probably too sensitive for 

 plankton work. 



Fron-i previous experience it is known that 

 the dominant types of small eastern tropical 

 Pacific plankton (i.e., excluding large salps 

 and medusae) are copepods, euphausiids, 

 chaetognaths, pteropods, and siphonophores, 

 though other forms may predominate occa- 

 sionally. To pick these organisms from a 

 plankton sample is tedious and difficult. Work 

 has been done on several selected Expedition 

 SCOT stations, but to reduce time taken only 

 very small aliquots (2-3 ml.) have been 

 picked. 



Copepods, for example, are segregated from 

 an aliquot, being counted at the same time. 

 A imethod of determining the volume of such 

 a small quantity of plankton is given by 

 Tranter (1960). The equipment consists of: 

 a piece of burette tube about 25 cm. long, 

 graduated in intervals of 0.05 ml., with a 

 stopper at one end; a tube of gauze (monel 

 metal, 80 meshes per inch, strengthened with 

 strips of solder, plugged at one end by a 

 brass disc) t?iat fits inside the burette; and 

 a funnel through which the plankton is intro- 

 duced into the gauze tube. The plankton is 

 put inside the tube and washed with alcohol. 

 The tube is dried by tapping and rolling on 

 filter paper. The tube, with the plankton 

 inside, is dropped into the burette tubing 

 which contains alcohol to a known level. The 



change in level is measured and, given the 

 displacement of the empty gauze tube, the 

 volume of the plankton is found. 



In practice certain difficulties were en- 

 countered. In an aliquot of about 3 ml. of 

 plankton usually more than 1,000 copepods 

 are found. It proved difficult to get them into 

 the gauze tube, because they plugged the funnel 

 and their bristles caught in the mesh. 



A modification of Tranter's method was 

 developed. A glass tube, with a removable 

 filter at each end, was used instead of the 

 gauze one. The plankton was introduced into 

 the tube (with or without a funnel). It still 

 occasionally plugged the glass tube, but much 

 depended on the way it was put in. When all 

 the plankton was in the tube the liquid was 

 allowed to drain through either end as desired. 

 The filters occasionally became clogged. Later 

 the tube with the dried plankton was dropped 

 into a burette containing a known level of 

 isopropyl alcohol. This part of the process 

 had to be done carefully and slowly, to ensure 

 that no air bubbles were retained by the 

 plankton. 



The method used is not entirely satisfac- 

 tory, but it has been used to produce some 

 results that were in urgent demand. With 

 either Tranter's method or the above-men- 

 tioned modification of it, the problem of 

 leaching by the alcohol may make replication 

 of results difficult. An investigation of the 

 replicability of results is proposed. 



Estimates of copepod number and volume, 

 per 1,000 m.3 of water strained by the net, 

 are now available for 20 Expedition SCOT 

 stations, together with estimates of mean 

 volume per copepod for each station. The 

 latter range from 0.0004 to 0.0018 ml. and 

 are more or less normally distributed. 



It was surprising to find that the copepod 

 volumes were generally less than half the 

 previously measured total volumes. Possible 

 explanations are (a) that draining of liquid 

 from the plankters is more efficient with a 

 homogeneous part of the sample than with the 

 whole, especially where alcohol is used as 

 an aid to drainage; and (b) that the plankton 

 becomes shrunken with time, the elapsed 

 time in this case being about 2 years. How- 

 ever, the measurements within each series 

 can probably be properly compared with each 

 other even if between-series measurements 

 are suspect. 



Collection and Measurement of Micronekton 



(M. Blackburn and A. 0. Reith) 



Examinations of tuna stomach contents, in 

 the eastern tropical Pacific (F. G. Alverson, 



30 



