D'ARCY W. THOMPSON 



the maxinuiin attained at this season was nuich higher in 1905 than in either the prec- 

 eding or succeeding year. There is also an indication of a maxinumi in summer about 

 June and July, followed by a minimum 



in late autumn; and this summer 1904- 1905 1906 



maximum was very high in 1906, when 

 it overtopped the maximum of the 

 preceding winter. Here again we seem 

 to have in this double maximum the 

 intermixed effect of two different pheno- 

 mena characteristic of regions on either 

 side ot that with which we are dealing. 

 The winter maximum is characteristic, 

 as we shall immediately show, of 

 regions further to the west or south- 

 west, while the summer maximum is 

 due, so to speak, to a share in that 

 great migration of Cod which pro- 

 duced the well-marked maximum at 

 Shetland in July. The normal maximum 

 for Codling in this area is clearly in 

 autumn, about September, and there is 

 an equally clear minimum in February 

 and March (fig. 17). 



The curve for Codhng^in area XIII deserves a moment's attention, as an instance 

 of the care that must at times be taken to guard against an erroneous conclusion. In 



the four years to which 

 1905 1904 1905 1906 



Fig. 15. Codling. Area XIV. 



t 



t. 



t 



-\ 



S 



■Î: 



k 



:v 



^ 



i: 



■î^': 



s: 



"^2 



it refers there is re- 

 gularly a maximum in 

 September or October 

 for each of the four 

 years to which our 

 statistics refer; but 

 there is present at the 

 same time a high 

 maximum in the spring 

 of 1904, of which no 

 trace occurs in other 

 years. On reference 

 to the original statistics 

 we find that this cusp 

 in 1903 is wholly due 

 to two boats fishing 

 during forty hours in 

 May of that year, and we can accordingly give little or no weight to this apparent ex- 

 ception to the general course of phenomena. 



(*o 



80 



40 



Fig. 16. Cod. Area XIII. 



