402 DISCOVERY REPORTS 



paper (Laws, 1959 a) and it was shown that the samples taken in the Antarctic in the period of 5 months 

 from October to February may be considered to be representative of the progress of foetal growth, 

 but that differential migration out of the area affects the validity of the samples from March onwards. 

 A similar differential migration into the area in spring means that very early embryos (less than one 

 month post-conception) will tend to be absent in antarctic samples. The effect of this would be most 

 marked in the earlier months, prior to November, for which in any case very few length records are 

 available. 



It is estimated that, in addition to the 956 foetuses recorded in Table io, little more than 50 (some 

 5%) would have to be added to the lower size groups of the frequency distributions for October and 



J JASONDJ FMA 



MONTHS 



Text-fig. 30. Mean curve of foetal growth and monthly foetal length frequencies of southern 

 hemisphere fin whales; class interval 0-2 m. 



November, and the higher size groups of the March and April samples, to overcome this bias. The 

 effect of this small error on the frequency distribution of pairing, which is now to be calculated, means 

 that the pairing frequencies for the early months of the breeding season and for the late months will be 

 slightly under- rep resented. This is thought to have a negligible effect on the shape of the pairing curve. 

 The effect of individual variation in growth-rates must also be considered. Zemskiy (19500) states 

 that male and female foetuses grow at different rates. Kimura (1957, p. 113) studied the difference 

 in length between twins of different sexes, but found that in 57 % of cases there was no difference, 

 and the remainder were almost equally divided between pairs in which the female was larger and pairs 

 in which the male was larger. No distinction as to sex need, therefore, be made in studying foetal growth. 

 We are concerned here with the average growth of large numbers, so individual variations in the 

 growth-rate need not be considered in drawing up an average growth curve. When foetal lengths are 



