286 DISCOVERY REPORTS 



the path of the deep net in the upper and in the lower layer is only very approximately 

 known. The proportion of the catch, therefore, which properly belongs to each layer of 

 water cannot be ascertained. All that can ever be said is that part at any rate of the catch 

 in the deep net belongs to the lower layer, except at certain stations (marked with two 

 asterisks in Table II a-c, p. 369) where the lower layer approaches the surface to such 

 an extent that the whole of the path of the deep net lies within it. Similarly the catch in 

 the upper net either belongs wholly to the upper of the two water layers under 

 consideration, or else, when the boundary between the layers lies near the surface as 

 at the stations marked with two asterisks in Table II a-c, it must contain a percentage 

 which belongs to the lower layer. The variation in the depth of the boundary between 

 these two layers and the mixing which always takes place across it are, therefore, factors 

 which introduce large but unknown variations into the catches and which make it useless 

 to attempt any exact interpretation of the figures. 



Little need be said about the methods employed in the laboratory during this work. 

 In every case the samples were fixed and preserved in 10 per cent formalin and were 

 analysed by direct inspection and counting, using a binocular microscope and a Petri 

 dish, in which the sample was spread out. When the samples were too bulky for com- 

 plete examination they were fractioned over a card marked in eighths or tenths and 

 a fraction only of the sample was examined. Very often, after a fraction of the whole 

 sample had been examined, it was necessary to take yet another fraction of the Copepoda 

 left after the other organisms had been counted, so that fractions of a fiftieth or a 

 hundredth of the total Copepoda are not uncommon in the analyses. This method of 

 analysis is, again, only approximately accurate when fractions are examined, and the 

 smaller the fraction of the total sample examined the greater the error in the analysis. 



Many of the samples were analysed on board the ' Discovery II ' during the cruises, 

 but a large number of the copepod analyses were repeated on shore. 



HYDROLOGY OF THE AREA 



It is not necessary here to enter into a detailed description of the hydrology of the area 

 covered by these cruises. For the hydrology of both the Falkland Sector of the Antarctic 

 and of the Southern Ocean reference should be made to the accounts published by 

 Deacon (1933 and 1936). It is perhaps desirable, however, to give the very briefest 

 account possible of the hydrological conditions in the area traversed, so far as they are 

 likely to aflFect the catches with which this paper is concerned. 



Antarctic Seas generally are characterized by a cold, poorly saline layer at the surface 

 known as Antarctic surface water. It owes its low temperature and salinity to the cold 

 Antarctic climate and to the melting of pack-ice, and it has an average depth of about 

 200 m. Antarctic surface water streams away from the Antarctic Continent and its 

 surrounding pack-ice in a northerly and easterly direction and becomes part of the 

 general eastward movement around the Southern Hemisphere known as the West Wind 



