516 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



Eiglity-one species were estimated in the catches, 

 and in only one case did Loth estimations agree with the 

 numbering. It was possible for three things to happen : 



(1) For both estimations to agree with the numbering; 



(2) for estimation and numbering to give parallel results, 

 but not be alike; (3) for estimation and numbering to be 

 contrary to one another. Only one species agreed in 

 every respect, 13 species gave parallel figures, and in 07 

 species the estimations and numbering were contrary. 

 Thus the personal error forms an additional source of 

 failure in the simple method of estimating a catch by the 

 frequencies, whereas by the counting method two 

 observers will practically agree, if both count the same 

 catch. Thus, for tables to be of any scientific worth in 

 comparisons made to show the dependency on hydro- 

 graphical or other conditions, or of the various forms 

 upon each other, the catches must be made quantitatively 

 and the unfortunately tedious method of counting 

 followed. 



A very detailed account of the apparatus and method 

 of counting has been given by Jenkins (12), but the 

 method as at present carried out for general work will be 

 briefly described here, in order to give the complete 

 procedure. The first work consists in the estimation of 

 the volume and the construction of curves to illustrate 

 this. In most cases, unless the plankton is caught on an 

 expedition lasting some months, the volume estimation 

 will be made on shore. If it is required to estimate the 

 volumes of the catches on board ship, the usual swinging 

 table is required. The catch which has been fixed and 

 preserved in alcohol is allowed to stand, and the alcohol 

 decanted and its place taken by distilled water. The 

 catch in distilled water is now brought into the measuring 

 vessel. If distilled water is not used instead of the 



