FIN AND BLUE WHALES 297 



In Text-fig. 10 these monthly length frequencies have been smoothed (in groups of three), converted 

 to percentages, and smooth curves drawn through the points thus allowing a more direct comparison 

 to be made. It can be seen that the monthly length frequencies show a normal unskewed distribution, 

 which is compatible with linear growth, in the early months (October and November). The later 

 curves (December and January) are negatively skewed, suggesting an increasing growth rate. On 

 this assumption the curves for February and March should also be negatively skewed, but the 

 February sample shows an almost normal distribution and the March curve is positively skewed, 

 that is to say the trend is reversed. This suggests either that the slope of the foetal growth curve has 

 greatly decreased again (which would result in the extension of the gestation period over more than 

 12 months), or that some large foetuses are missing from the February sample and more are missed 

 in March and April. The first explanation is ruled out by other considerations discussed above, and 

 we are left with the second possibility. 



Text-fig. 1 1 has been constructed to demonstrate the changes which may be expected in the toetal 

 length frequency curves during the year. It has been assumed, for the purpose of discussion, that 

 the frequency of conceptions is described by a normal curve extending over, say, 6 months (shown in 

 the figure as an inset), that the average curve of growth in length is that given in Table 4 and Text- 

 fig. 8, and that there is no variation in the growth rates of individual foetuses. The diagram has been 

 constructed by plotting the average growth curve (thick line) and drawing in six similar growth curves 

 displaced by intervals of i month (dotted lines). The space between these curves then extends over 

 6 months. This follows the practice of Mackintosh and Wheeler (1929, fig. 146). The frequency curve 

 of foetal length has then been drawn in for each month by plotting the assumed conception frequencies 

 against the average foetal lengths taken from the curves for each of the 6 months of conception (see 

 inset). The frequency curves of foetal lengths constructed in this way for different sampling months 

 demonstrate the transformation of the shape of the monthly length frequency curves, though not their 

 magnitude. 



The curve of conception frequencies used here is an arbitrary one and there are reasons for supposing 

 that in the fin whale it does not show a normal distribution but is negatively skewed, i.e. the mode is 

 earlier in time. This would have the effect of shifting the modes in the estimated foetal length fre- 

 quencies for monthly samples to slightly higher values. The skewness of the curves of, for example, 

 December, January and February would be slightly less in a more realistic model. 



The shapes of these model curves are very similar to the sample length frequency curves for October, 

 November, December, January and February (Text-fig. 10), and follow the same trend of increasing 

 skewness and increased length range. The length frequency curve for the March sample shows a 

 reversal of this trend and there are insufficient observations for April. 



It has been suggested above that from March onwards the larger foetuses are missing from samples 

 because females approaching full term migrate out of the Antarctic area, so as to arrive in the breeding 

 area for parturition. Let us assume that the adult fin whale travels some 3000 nautical miles from the 

 Antarctic feeding grounds to the subtropical breeding areas (from about 65° S. to about 15° S.); 

 unfortunately we have no direct information on the duration of this migration. A fin whale and a 

 blue whale each recovered some 2000 miles from the place of marking had travelled, in the Antarctic, 

 at minimum average speeds of about 1-3 and 1-7 knots, respectively (Brown, 1957), but these speeds 

 are undoubtedly well below the average speeds on migration. There is more useful information about 

 the humpback whale, which is, however, a slower animal than the fin whale. Aerial observation of 

 six migrating adults which were followed for distances of 3-20 miles showed that they travelled at 

 speeds of 3-6-6-5 knots, averaging about 5 knots (Chittleborough, 1953). One 43 ft. male humpback 

 marked (no. A 137) on 7 July 1953 off East Australia {ca. 35^ 10' S., 150° 35' E.) was recovered 



