SEA-FISHERIES LABORATORY. 



205 



curves in Fig. 4 has been obtained : the figure is really 

 the graph of Table XXII, with some modifications sug- 

 gested by a critical survey of the data. The heavy, 

 continuous lines are the graphs of the series for Age- 

 Groups 0, II, III, and IV, as they are set out in 

 Table XXII. The graph for Group I, on the other hand, 

 is constructed from the following series of measurements : 



Mean length. 



No. 



% 



6-5 



4 



0-71 



7-5 



13 



2-30 



8-5 



8 



1-42 



9-5 



11 



1-95 



10-5 



25 



4-42 



11-5 



65 



11-50 



12-5 



82 



14-52 



13-5 



63 



1115 



14-5 



50 



8-85 



15-5 



51 



9 03 



16-5 



47 



8-32 



17-5 



61 



10-80 



18-5 



51 



903 



19-5 



22 



3-89 



20-5 



7 



1-24 



21-5 



3 



0-53 



22-5 



2 



0-35 



Totals 



565 



10001 







It will be obvious, from a consideration of Table XXII that 

 the means of Groups I and II lie very closely together 

 (about 18 and 20 cms.). Now the difference between the 

 mean length of Group O and that of I is considerable, 

 and so also is the corresponding difference between II and 

 III, and III and IV. Why, then, should I and II be so 

 close together ? The reason is, of course, that the figures 

 for Group I in Table XXII do not represent a real "sample" 

 of the fish " population,'' which is being investigated, 

 and the consideration of this point is of some importance 

 since it points to a source of error which is inherent in all 



