FOOD, HIBERNATION AND MIGRATION II5 



complicated matter, wliich needs more study, and I feel that it is an 

 over-simplification to pay attention only to wind, which may not 

 exist in the important place. It should also not be overlooked that in 

 my studies there was a significant correlation with wind direction, 

 even though I refused to consider that this was the whole story. 



In my observations, I found that frogs tended to follow ditches even 

 if these were dry, and I suggested that the animals were following the 

 lines of air drainage at night, when it is common for air to flow from 

 the tops of hills, where my pond was situated, to the valley below. 

 Moore (1954), in his study o£B. hufo, found that many toads came down 

 from the hills to his pond in the valley, following particularly the 

 simken roads. A map of the district shows that the pond was almost 

 surrounded by hills, with narrow outlets to the lower country some 

 distance away. It is interesting to speculate on the effect of a layer 

 of cold air at hUltop level. It seems possible that air currents from the 

 warmer valley floor might indeed pass up the deeper parts of the 

 side valleys, being pushed up by the denser air above. There is 

 obviously much that we do not yet know about ampliibian migration 

 or the fine structure of air in a complex situation. Many different 

 facts suggest that smell is the guide, and there is notliing known 

 against the hypothesis, except this matter of wind direction in some 

 circumstances, which may be after all wrongly based. No reasonable 

 alternative has been suggested, but this does not add up to proof. 



The Stimulus for the Beginning of Migration 



What is it that starts the frogs on the migration? Before we seek 

 the answer to this question, it seems necessary to consider whether the 

 stimulus of migration is the same as the stimulus that initiates breeding, 

 once they have reached the ponds. I think that in the description of 

 the events I have just given, there is evidence that the two may be 

 different. In some years, the frogs breed as soon as they arrive. If this 

 always happened, it might be very difficult to distinguish between a 

 migration and a breeding stimulus. But it does not always happen. 

 In some years, frogs begin to migrate for weeks before they breed, 

 and although animals of both sexes are in the ponds, there is no sign 

 of the definite breeding behaviour until, as if at a signal, tliis starts in 

 full vigour. This must surely imply that the stimuli are different, and 

 the most probable explanation of breeding immediately on arrival is 

 that when two separate events take place near in time, there will be 



