DISTRIBUTION OF ANTARCTIC MACROPLANKTON 71 



sub-samples a rough and ready method has been found to be best. The catch is put in 

 a large Petri dish and the water strained off until the sample has the consistency of a 

 loose paste. It is well mixed up and the dish is shaken until the sample lies in a flat, 

 even layer. It is then divided into quarters (or some other fraction), and a quarter is if 

 necessary removed to another dish and further divided. The method is crude, but if 

 carefully done can give surprisingly accurate results. An Einar Lea apparatus can be 

 used, but with large organisms it is not more accurate than the method described 

 above. 



In the analysis of these samples no hard and fast line of procedure can be followed. 

 The method must be suited to each sample, and the worker must endeavour to satisfy 

 himself that he has not missed any of the rarer forms, and that he has estimated the 

 approximate numbers of each species present. The analysis of a large sample of perhaps 

 20,000-30,000 organisms may occupy several hours, especially if there is a great variety 

 of species. The smallest samples can be analysed in about 20 minutes. 



The object of plankton work is to determine the nature of the plankton in a given 

 area from the analysis of samples taken at selected points in that area. The inferences 

 thus drawn are liable to certain errors which may arise from the distribution of the 

 plankton itself in the water, from the method of collecting the samples, or from the 

 method of analysing the samples. The subject has been dealt with by Hardy and 

 Gunther (1934), whose remarks apply in a large degree to plankton collected by the 

 N 100 B, and need be discussed only briefly here. The following are some points to be 

 borne in mind: (i) Heterogeneous distribution of the plankton organisms may give a 

 false impression of the fauna of an area. Thus allowances must be made for species 

 which tend to a specially patchy distribution, (ii) Active avoidance of the net must 

 sometimes take place, especially by such species as Euphausia superba. I have been able 

 from the ship's side to watch a net being towed a few feet below the surface and passing 

 through a shoal of this species. The Euphausians could clearly be seen to leap backwards 

 out of the way of the approaching net. (iii) Variations in the speed of towing and depth 

 of the net have already been mentioned, (iv) Where a number of species of many 

 groups have to be quickly identified, complete accuracy is not always to be depended 

 on. (v) Big catches of shoaling species such as E. superba and Salpa may swamp the 

 sample and make an estimation of the other species almost impossible. Such catches 

 must sometimes be disregarded except for the Euphausians or Salps themselves. 



It may seem that some of these factors must lead to serious errors in the analyses, 

 but Hardy and Gunther (1934) have shown that fluctuations in the numbers of the 

 smaller plankton organisms are so great that an error of 50 per cent has little sig- 

 nificance, and Hart (1934) points out, in connection with the analysis of samples of 

 phytoplankton, that "in practice differences of 100 per cent and over are the smallest 

 that can be regarded as of much significance". The same applies, possibly with even 

 greater force, to the catches of the N 100 B, for although it may be too much to say that 

 the important features of a sample are those which may be recognized at a glance, it is 

 at least probable that the most obvious features will be the important ones, and that a 



