158 DISCOVERY REPORTS 



With the exception of the rather small numbers of Furcilia 5 and 6 in the 70-cm. 

 vertical nets, the catches at St. 461 provide the only direct indication we have of vertical 

 distribution in the later stages of development. Yet if the younger larval stages are not 

 susceptible to the changes resulting from a vertical migration from warm to cold water 

 and vice versa, it is unlikely that the older larvae would react more readily. It has been 

 shown above that the younger larvae migrate upwards from depths greater than 250 m., 

 involving passage from the intermediate warm water to the colder surface water. The 

 eurythermal and euryhaline qualities of E. superba are fairly high, and it is considered 

 that the presence of a discontinuity like that at St. 461 is not adequate to explain the 

 preference of the bulk of the animals for the surface. 



(iii) I feel compelled to select the third as the most important factor in the distribu- 

 tion of the larvae. It has been shown above that the younger larvae from Calyptopis i 

 onwards have a quite well-defined diurnal vertical migration. In the last of the diagrams 

 in Fig. 73 it was shown that the vertical migration is not pronounced and that the 

 animals tended to remain at the surface whatever the time of day. Examination of the 

 time of year when the larvae were taken, from which that diagram was derived, shows 

 that about nine-tenths of the total were taken early in their second season— that is to say 

 after the southern winter. 



The winter and all it implies is, I am sure, the factor which leads to interruption of the 

 rhythmic migrations of the euphausians. Prolonged darkness, overcast skies, low alti- 

 tude of the sun and consequent increased reflection of its rays from the surface of the 

 sea, snow falling in the very cold water and hardly mehing, but predominantly the 

 presence of pack-ice, all render it less necessary for the larvae to seek out the deeper 

 water in daytime. 



This influence of ice has been previously recognized. Russell {he. cit., p. 231) men- 

 tions the researches of Damas and Koefoed (1907) into the distribution of copepods in 

 the Greenland Sea. They observed changes in the depth distribution according to 

 region. " They found that species of the intermediate and deep layers were met with in 

 the ice covered western part of the Greenland Sea at much higher levels than in the 

 eastern uncovered portion. They further, on an examination of the hydrographic data, 

 concluded that this difference was not occasioned by currents and had no relation with 

 salinity or temperature." 



Damas and Koefoed indicate that the light intensity is the important factor. 

 Romer (1904, p. 72), cited in the same paper by Russell (p. 256) states : " In the warmer 

 ice-free waters of the west coast (of Spitzbergen) the difference between the scarceness 

 of plankton in the surface layers in the daytime and its abundance in the evening and at 

 night is much more marked than below the thick pack-ice of the eastern region. It is 

 most probable that this is due to the strong cutting off of the light rays by the ice." It is 

 likely that similar conditions produce like results in E. superba and explain the abund- 

 ance in which it is found in the vicinity of ice fields. 



In addition, however, to pack-ice affecting light intensity there is another way in 

 which it probably helps in the formation of shoals of krill. If we assume that in ice-free 



