VERTICAL DISTRIBUTION OF ADOLESCENTS AND ADULTS 327 



smaller specimens are never found in deeper water. It would be interesting to compare the mean 

 lengths at different depths between 0800 and midday but unfortunately the data are insufficient. 

 There is an even more marked variation in the mean lengths when the larvae are included with these 

 figures (pecked line) showing that there is a very strong tendency for them to inhabit the upper layers 

 during the night. There is probably little or no significance in the reduction of mean length below 



250 m. in this figure. 



Although illumination is probably not the only factor involved in diurnal vertical migration, it is 

 reasonable to assume that it plays the major part, and that any variation in length of daylight would 

 influence the time spent at a particular depth. At 50° S there are approximately 8 hrs. more darkness 



HOURS 



Fig 1 1. The diurnal variation of the numbers of adolescent and adult E. triacantha in the upper 100 m. in winter and summer. 



The number of observations for each 2-hourly period are shown. 



in winter than in summer, and it could be expected that this would affect the time spent i^n the upper 

 100 m by E. triacantha. In Fig. 11 the mean numbers per 20 min. loo-o m. N lOoB haul are plotted 

 for summer and winter at 2-hourly intervals. As illumination is the primary consideration in this 

 case summer has been taken as mid-September to mid-March and winter as mid-March to mid- 

 September. This allows approximately three months either side of the longest and shortest days. 

 From the figure there is not a great deal of difference in the time spent in the surface water by the 

 majority of the animals, but the period during which no adolescent and adult E. triacantha are found 

 in the surface at all is 6 hrs. shorter in winter than in summer. If there were sufficient data to compare 

 winter and summer observations in the Bellingshausen Sea region it would probably be found that 

 there was a greater difference, for the horizontal range of E. triacantha lies very much further south- 

 between about 58° S and 69° S-where there is a difference of more than 13 hrs. between winter and 



"ZsTworkron plankton have at some time been faced by the problem of whether their results 

 are affected by the organisms avoiding the net, in particular whether there is a greater degree of 



