METHODS OF PLANKTON EESEAECH. 521 



seen to contain the uninjured and living food before it 

 passes into the alimentary canal. Here in this filtering 

 apparatus of the Appendicularian are numerous naked 

 Ehizopods and Gyninodineae, together with smaller 

 skeleton -carrying Rhizopocla, completely absent from the 

 net catches. It is obvious, therefore, that if a complete 

 knowledge of the plankton is to be gained, other methods 

 must be applied. 



It was assumed that the loss of plankton by the use 

 of No. 20 silk in the net was unimportant, and that the 

 real masses of plankton in the sea might be only 2 to 3 

 times greater than the figures given by the net. This 

 would certainly be of no great importance if only the 

 volume or weight of the plankton present was required 

 without any reference to its constitution. If one requires 

 however, the chemical constitution, it is quite incorrect, 

 and this applies further to the qualitative and quantita- 

 tive counting method, because the animals and plants in 

 the catch will occur in quite different relative proportions 

 from the true conditions present in the sea. Since the 

 mesh work of the net itself has a large area, many of the 

 small forms which could pass easily through the meshes 

 will be caught on the net tissue itself ; and this will give 

 a more deceptive appearance of reliability than if these 

 forms had altogether escaped. 



Again, the fractions of these small forms caught is 

 not always the same, because if the sea contains a great 

 number of diatoms (as Chaetoceros), the meshes of the net 

 will be gradually filled up, and the spines interlocking 

 will cause the net to act as a much finer filtering material 

 and hold back many species which would otherwise slip 

 through. This accounts very often for the large catches 

 with the nets, when diatoms are very abundant. By 

 comparison of the net catches with the other quantitative 



