i62 DISCOVERY REPORTS 



connecting continuous lines are intervening periods between half-months in which 

 larvae were found, and during which periods it may be assumed that they would be 

 likely to occur. 



The figure shows that from the second half of January onwards to the end of April at 

 any rate, stages from Calyptopis i to Furcilia i are to be found. There is a single record 

 of a Calyptopis 2 in the first half of November which, I consider, must be regarded as an 

 exception. There is no general incidence of early larvae until the second half of January, 

 and their absence in the May observations of the circumpolar cruise indicates that the 

 stages up to Furcilia i are passed through before the coming of winter conditions. 

 Furcilia 2 and 3, like earlier stages, occur in the second half of January and are the 

 youngest forms found in May ; their time range is therefore four and a half months. The 

 first records of Furcilia 4 and 6 are not obtained until the second half of February. 

 Furcilia 5 has no record before the second half of March, but it must, of course, occur 

 at a time preceding the first occurrence of the ensuing stage. Furcilia 4 is found in June, 

 so that like Furcilia 2 and 3 it has a time range of four and a half months but with its 

 first record a month later than the previous stages. 



Furcilia 5 has a considerably longer time range of occurrence, for from March on- 

 wards it is found until the end of the southern winter in September. It should be 

 pointed out, however, that Furcilia 5 occurs only in extremely small numbers in the 

 August-September catches (see Appendix I, Sts. WS 259, 261-263, 266, 268 and 286)_ 

 Furcilia 6 has by far the greatest period of occurrence. The first record is in the latter 

 half of February and the last in the first half of January, a period of eleven months ; but 

 from November to January (see Appendix), where numbers in the samples are large 

 enough to give a reliable indication, Furcilia 6 is a very small and insignificant con- 

 stituent of the krill population. 



In the diagram the occurrence of adolescents is traced from the first record in June 

 to the time in January when the euphausian becomes of such a size that determination 

 of the state of maturity and distinction from the previous year class is an anatomical 

 problem not to be dealt with here. 



The extended period during which Furcilia 6 is represented is in part due to the inter- 

 vention of winter, during which growth is retarded and developmental processes slowed 

 down. But there is no doubt that the extended period of occurrence is to some extent 

 apparent rather than real. There is no clear line of demarcation between Furcilia 6 and 

 adolescent; the character used to recognize the former stage, namely the presence of 

 three postero-lateral spines and one terminal spine on the telson, is one which can be 

 possessed by larvae within a very wide range of size. If, as is the case, it is possible for 

 E. siiperba to be sexually mature while still lacking its full structural development, it is 

 quite likely that forms recognized as Furcilia 6 may in fact be more properly classed as 

 adolescent. In time distribution, as in development, later stages become less capable of 

 narrow definition. The significant point ecologically is that this extension of the time 

 occurrence of well-advanced krill ensures a constant food supply for the whales in 

 Antarctic waters. 



