194 



Part III. — Twenty-second Annual Beport 



The increase in length in the ninety-five days from 28th December to 

 1st April amounted to about 15mm. (§ inch); on the 139 days from 

 14th November to 1st April, to about 20mm. (| inch). It will be 

 noticed, as pointed out in previous reports, that the larger fishes of an 

 early series grow more rapidly that the smallest, i.e., the variation in the 

 sizes of the individual fishes of the group — due primarily to a difference 

 in the time of hatching, early or late — becomes more pronounced, which 

 is one of the causes of the coalescence of the older generations or groups. 

 It will also be observed that, so far as these data go, growth was more 

 rapid in December and January than in spring. This might be expected 

 from the higher temperature of the bottom water in the depths where the 

 witches lived during the former months, growth being closely related to 

 temperature. Unfortunately, no observations have yet been made with 

 sufficient frequency to enable the temperatures at these depths in the 

 northern waters to be approximately stated for the various months of the 

 year. Off the Firth of Forth, in thirty fathoms, according to the 

 Gaiiancfs observations, the mean bottom -temperature in the months 

 referred to were — November, 49° F. ; December, 49'2° ; January, 41*7° ; 

 February, 41-3°; March, 40-1°; April 43-3°. 



"With regard to the sizes and growth of the witches of older series, 

 above two years, there is more difficulty, owing to the coalescence of the 

 groups, and the different rate of growth of the males and females after 

 sexeral maturity is attained ; and there are not yet sufficient observations 

 on the older males and females to make the matter clear. In most of the 

 collections, as may be observed from the table, there is a general absence 

 of specimens between the second and third groups, and it is not certain 

 whether this gap is natural, i.e., that it is caused by there being really no 



