EATE OF GROWTH OF SOME SEA FISHES. 253 



Beile-Isle, 1888. 



15 to box to 8 to box. 



10th September to 30th November. 



16-6 grms. to 30 grms. = 14 grms. 



12-5 cm. to 15-0 cm. = 2-5 em. J ^^^' 



Le Ceoisic, 1888. 



12 to box to 9 to box. 



20 grms. to 27 grms. = 7 grms. 



140 cm. to 14*5 cm.=0"5 cm. 



QUIBEEON, 1884. 



12 to box to 6 to box. 



22 grms. to 40 grms. = 18 grms. 



13-0 cm. to 17-7 cm. = 47 cm. 



Keenevel, 1869. 



12 to box to 7 to box. 



20 grms. to 34 grms. = 14 grms. 



14'0 cm. to 16"5 cm. = 2*5 cm. 



15th August to 15th October. 

 61 days. 



15th June to 21st September, 

 98 days. 



24th July to 22nd November. 

 121 days. 



Tliese are sufScient to show the irregularity of the results ob- 

 tained in this way. I have copied the figures as given by Pouchet. 

 Pouchet thinks that the growth of a pelagic species like the 

 pilchard ought to be very uniform, because they inhabit water of 

 uniform temperature, and their microscopic food is always sufficient. 

 He thinks that the inequality of the above results must be due either 

 to the character of the records from which they are derived, or to 

 the fact that the differences of size in the records represent dis- 

 placements of the fish and not the growth of stationary fish. There 

 is doubtless much reason for this judgement, but I am sorry to be 

 obliged to point out that the calculation of these results offered as 

 samples is marred with serious errors. Thus, the period chosen at 

 Belle Isle for 1888 is 10th September to 30th November, and it is 

 said that the record for the latter day is eight fish to the box, 

 whereas the table for 1888, from which the figures are taken, ends 

 at November 14th, and gives no figures for November 30th, the 

 figure 8 occurring on October 27th. 



The figure 12 as the number per box at two places is translated 

 into weight 20 grms., length 14'0 cm. ; at another into weight 22 

 grms., length 13-0 cm., as though the length of the fish decreased 

 as the weight increased. The lengths given as corresponding to the 

 weights do not agree with the graphic on another page of the 

 report. 



Pouchet's industrial records occasionally show two features which 

 are worthy of attention, namely, the appearance of smaller fish in 

 July, and the appearance of still smaller specimens in September 

 and October. This ' poisson de Juillet ' is represented by the 

 figures 11 to 13, so that its size is 13*4 to 14"0 cm., and its weight 

 19'2 to 22*7 grms. Its occurrence is indicated by Pouchet in the 



