248 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



meshes of the net. Above the mean the distribution pro- 

 bably represents fairly that of the fish on the sea bottom . 

 But below the mean, the distribution represents mainly 

 the action of the net. 



It is evident also that no combination of nets will 

 succeed in representing the general distribution of plaice 

 on the sea bottom, when the lengths of the fish are below 

 about 15 cms. We might use a shrimp-trawl (an half- 

 inch mesh), and thus obtain such catches as are tabulated 

 on p. 258, or a four-inch trawl mesh, when such catches 

 as are tabulated on p. 102, in the Report for 1909, would 

 be obtained.* But the curve obtained by plotting these 

 figures would differ according to the share which the nets 

 of j, 4, and 6-inch meshes had in obtaining the total 

 catch, and w T e should not know what proportional numbers 

 of hauls by each kind of net ought to be made. It is true 

 that we should be able to determine the mode, since the 

 position of this would gradually change as the absolute 

 number of plaice below IT cms. becomes greater, and, 

 finally, would become stationary, but an infinite number 

 of curves would obviously be obtained, differing in their 

 parameters. 



Much the same difficulties apply, of course, to the 

 use of nets in obtaining samples of plankton. As the area 

 of the mesh changes, the composition of the plankton will 

 also change. A fine enough mesh will capture all, or 

 most of the micro-plankton, but in proportion as the area 

 diminishes, so the net will fail to capture the larger 

 planktonic organisms, since these will be driven out again 

 from the mouth of the net by the reflux of water from 

 the latter, just as large plaice are seldom taken in a 

 shrimp trawl even if they are present on the sea bottom. 

 Only a net of the same length as the vertical column of 



* Reft. Lcmcashire Sea-Fish. Laby. for 1909, p. 102. 



