9 



come across specimens which in tank conditions have not cast in 

 the same season in which hatching took place. The exceptions 

 may be said to point to a relative independence of the germ cells, 

 but the fact that the procedure which has been sketched above is 

 almost universal^ followed indicates clearly that there is an inter- 

 relationship between the three processes. The point receives 

 corroboration from a consideration of the migrations. 



It may be said, as was pointed out in a previous report,* that 

 it is generally true the shorter the migration to the north the less 

 the time. But it is evident when the results are inspected in 

 detail that migration is related to reproduction. For example, 

 in the present experiment. No. 172 migrated to about the same 

 region, Banff, in 8J months. The example now described took 

 nearly 3| years. Many recaptures have been made at Dunbar 

 and the neighbouring coast two or three months after hberation. 

 On the other hand, No. 428 was caught at Dunbar after bemg 

 free about two years and four months. Many more instances of 

 the same kind could be quoted from our experiments to prove 

 that while in the same season a time -distance relationship is pro- 

 bable the impulse is concerned with the periodical ripening of 

 the ovaries. 



The distance of the migration of the mature females may be 

 said to vary from 20 to 150 miles, and in all cases the distance 

 ma}^ be covered in one season. The variability in this respect is 

 probably due to the degree of the impulse. So far as our records 

 can guide us only large crabs migrate to the further hmits. The 

 numljer of the ova and the size of the ovaries increase with age. 

 The development of these probably requires a longer time in the 

 older crabs, and there may be a relationship between the time 

 and degree of development of the ovary and the time and distance 

 of the migration. If there is, as in every probability there is, an 

 internal secretion concerned in the im])ulse it probably increases 

 with the increase of the organ. 



It is becommg evident therefore that growth, ecdysis, repro- 

 duction and migration are all intimately associated. 



* Lor. eif., 1913. 



