METHODS OF PLANKTON RESEARCH. 551 



large number, and hence, also, the attempts made by fish 

 hatcheries to save a greater number of the embryos by 

 rearing them through the early stages. 



Then, again, since unfertilised plaice and cod eggs do 

 not remain pelagic and other dead eggs fall to the bottom 

 (when their death is not due to their being devoured), the 

 floating eggs capable of being caught must be but a small 

 proportion of the number actually produced. Hence the 

 number 110*6 is much too high, as a calculation of the 

 number of eggs per square metre lost by the capture of 

 the adult fish, and cannot be compared directly with the 

 number 370, which although the actual number of eggs 

 fished, represents only a portion of those produced. The 

 calculation has assumed that the relation between the 

 number of eggs floating in the sea and the fishes that 

 produced them is the same as that between the number 

 of eggs in the ovaries and a mature fish. 



In conclusion, it may be repeated that for a scientific 

 quantitative study of the plankton, the complete apparatus 

 and the Hensen method of counting must be employed. 

 It is quite obvious that a certain amount of inaccuracy 

 will occur with the use of each piece of apparatus, and the 

 numbers must be considered approximate only ; but since 

 the errors will average the same for each catch, they do 

 not invalidate the results for purposes of comparison. 

 It is quite another matter, on the other hand, if the 

 plankton is found to be not so equally distributed that 

 the small samples taken will give reliable results for the 

 whole areas. It is not to be expected that under varying 

 hydrographical conditions the plankton will remain the 

 same; but, at the present time, very little is known of 

 the actual relations. Again, it has been pointed out 

 several times in this paper that the results of recent 

 plankton work have very often shown sudden and striking 



