HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 319 



picture that emerges is not always a very tidy one. Variation in the growth-rate alone must produce 

 many seeming anomalies and the vagaries of the bottom and warm deep currents, and above all those 

 of the surface drift, many more. Retarding winds, for instance, would tend to delay the arrival of the 

 swarms in the east in certain years and assisting winds to advance it. Variation in the volume and speed 

 of the cold deep current must also play a part. It may be, for instance, that in exceptionally 

 cold winters more bottom water is formed at the head of the Weddell Sea than usual, and that the 

 flow of the bottom current itself is in consequence stronger after such winters than after mild ones.^ 

 From this it would follow that the colder the winter the more swiftly the deep larvae would sub- 

 sequently be carried away to the north and east, and conversely, the milder the winter the slower. 



c 



I OOn 



090 



oeo 



70 



60 



50- 



040 



30 



020- 



o 10 



000 



TEMPERATURE 

 DEEP CATCH 



- 100 



200 



3005 



4 00 



500 



1928 1929 I930 1931 1932 1933 1934 1935 1937 1938 



Fig. 8i. Average annual catch of deep larvae and average temperature at 1500 m. 



The dynamics are further complicated by the existence of the warm crosswise intermediate current 

 they must traverse before they reach the surface stream. Where it flows directly below the surface 

 drift the major effect of this current would be to divert the deep larvae from their easterly course, 

 tending in general to carry them farther south. The stronger the flow the greater the diversion, 

 involving sometimes, it seems possible (p. 302), their transference from the Weddell to the East 

 Wind zone. In general, however, its most likely effect would be to carry them away from a region 

 where the surface drift was strong into another where it was weak, or vice versa, and so in the end 

 again tend to advance or retard the eastward movement of the resultant surface population. 



As far as the bottom water is concerned, there does appear to be some measure of correlation 

 between what might be described as the annual volume or activity of this current and the annual 

 abundance of the deep larvae encountered during the spawning season. A graphical representation of 

 this relationship is given in Fig. 81, which shows the average temperature at 1500 m. (taking that as 



1 Although Fofonoff 's (1956) work suggests that both temperature and sahnity of the water sinking into the bottom layer 

 are likely to remain more or less constant, Deacon (1959) has pointed out that the volume of the transfer can vary, and that we 

 should expect more sinking at the end of winter than at the end of summer. We should expect, too, as I have suggested, more 

 sinking at the end of a severe winter than at the end of a mild one. Worthington (1959), discussing the formation of 'the 

 18° water' in the Sargasso Sea, notes that on the basis of existing measurements anomalously cold winters will produce a 

 larger volume of this type of water than usual. 



