DEVELOPMENT OF EUPHAUSIA SUPERB A 107 



dissimilarities which must accelerate or retard the growth rate in accordance with 

 favourable or unfavourable conditions. 



Comparison of size with locality of occurrence gave no satisfactory result, nor is there 

 any clear suggestion that larval development periodically commences sooner in one area 

 than another. There is perhaps a hint that in the South Georgia region this may be so, 

 but it is so ill-defined in the data as to be of little importance. 



Notwithstanding the great variations in average length which exist at any one time, 

 the figures as a whole show an upward trend from early in the summer season to the end 

 of it, so that by June the larvae have attained an average length of io-6 mm. The June 

 figure is the only record for that period of the year, but it is from a reasonably large 

 number of measurements and may be considered fairly representative. There is not 

 another record until the second half of August when the length of a small number of 

 larvae in a different season is io-02 mm. The averages from August onwards are of small 

 numbers of Furcilia 6 and adolescents and for these a better indication of what happens 

 is obtained from the results of the i-m. net samples set out above. 



It may be concluded then that the larvae generally reach a length of about 10 mm. by 

 the time the winter sets in and that this size has been achieved in a period of six months. 



AVERAGE GROWTH RATE 



Fig. 30 is constructed on the data shown in Figs. 27 and 29 above. It attempts to give 

 some impression of average growth rate during the first year or so of larval life. From 

 November to August the figures represented are the half-monthly average lengths of all 

 the first season larvae taken during that period. From August to January the graph is a 

 repetition of Fig. 27. 



The interpretation of the curve of average growth is straightforward in this second 

 portion, but the first part requires some comment. The figure for the first half of No- 

 vember is derived from a single individual. The figures for the second half of December 

 and second half of January are from four and five measurements respectively. In the 

 beginning of January and February numbers have increased to 368 and 112 respectively, 

 and in the second half of February and first half of March they are 1966 and 6602. The 

 next three half-months have roughly comparable figures in the thousands. 



The average lengths during this period show no well ordered progression in size. 

 In January and February these figures are very much greater than in the first half of 

 March, where there is a large population with greatly reduced average length. In the 

 second half of March the average length increases again by about 4 mm. The position of 

 the first portion of the average length curve, as shown in the graph, is suggested by these 

 figures. The large number of very small larvae in the first half of March counteracts the 

 effect of the lesser numbers of greater length before and after that time and makes the 

 upward tendency of the growth curve less steep. 



It is possible that the conjectured earlier development of larvae in the South Georgia 

 region, referred to in the previous section, is responsible for the differences between 

 the January-February period and March. In the former of the two periods the average 



14-2 



