162 A CENSUS OF THE SEA. [PART II 



the possibility of deducing an average. But Biddulphia is 

 very irregular and the average catch probably does not possess 

 much value. There was always a large number of the diatom 

 Chaetoceros present, but possibly a considerable proportion of the 

 individuals of some of these species which entered the mouth of 

 the net escaped again through the pores. The averages for 

 Chaetoceros, then, are probably minimum ones. 



Now, considering Hensen's much-quoted postulate, we expect to 

 find that the plankton is uniformly distributed in a sea where the 

 salinity, temperature, &c., are also uniform. If we consider (say) 

 the distribution of the diatom Goscinodiscus coiicinnus, at the fifteen 

 stations investigated in August we will find it to be extremely 

 irregular. But before we conclude that the postulate of uniformity 

 is fallacious we must determine whether or not physical uniformity 

 prevailed over the area included by these fifteen August stations. 

 If the reader will refer to Cleve's plankton chart (Fig. 29) he 

 will see that this diatom is the predominant organism in a plankton- 

 type which is restricted to a well-marked zone of the North Sea. 

 Here then is the explanation of the lack of uniformity in the 

 catches made by the Poseidon : they were made in sea regions 

 characterized by different phj^sical conditions. In an ideal scheme 

 of plankton investigation there would be half a dozen steamers 

 making simultaneous quantitative collections during three weeks of 

 each of the four sample months of the year. These vessels would 

 cover the North Sea with a network of lines of observation and 

 both plankton and h^^drographic work would be done at each 

 station. Then we should find that there were a number of zones 

 each characterized by water of a certain mean salinity and temper- 

 ature. Over each zone we should expect a tolerably uniform 

 plankton content. Then all the hauls taken on each zone would 

 be grouped together and the average of these would be taken. 

 We should thus find, not that there was an average plankton 

 content for the whole North Sea, but that there were a number 

 of sub-areas, and that over the whole extent of each of these sub- 

 areas, the plankton would be so similar that it would be possible 

 to deduce an average value for its composition and density. Further, 

 by the application of the principles outlined in the Appendix we 

 should be able to state how much this average deviated from the 



