FOOD, HIBERNATION AND MIGRATION IO3 



left Middle Tottcridgc before spawning started there. Another niale, 

 E12, was first tagged in Middle Totteridge on March 23 rd, but this 

 frog did not leave the pond, for he was found again there in full 

 breeding activity on April 4th. It is interesting that, at this date, 

 breeding activity in Large Totteridge was almost over, and on the next 

 evening, April 5th, Large Totteridge was silent, although there was 

 one male left, and a female was seen migrating to it. Only 1 50 yards 

 away, the season was just starting in Middle Totteridge, twelve days 

 after Large Totteridge, in spite of the fact that the frogs in each pond 

 were evidently drawn from the same general population. 



In Table 5, 1 show the summarized observations and also the weather 

 of the periods. In the paper, I proceeded to analyse the results with 

 caution, for I did not believe that although there was already statistically 

 significant evidence for migration up-wind, the true position would 

 show itself to be so simple. Large Totteridge, as I have said, is situated 

 on the north slope of a ridge, and all migration apparently took place 

 uphill or on the level. At the time I wrote the paper, there was little 

 evidence about the direction of air currents close to the ground. I 

 considered that these might be the most important, because in Large 

 Totteridge, and possibly many other ponds, air currents under certain 

 meteorological conditions flow as water does, and drain down ditches, 

 whether these contain water or not. There is now more evidence 

 on such matters, although much still remains obscure, and I can now 

 proceed with a more balanced analysis. 



The comparison between March 1932 and March 1934 is instructive, 

 because in those two years the spawn date was almost the same (82 in 

 1932 and 83 in 1934). In 1932, except for four frogs on the 7th, and 

 two on the 19th, there was no migration until the night before that on 

 which spawning started, the 22nd March, when there was a great wave, 

 and another even larger occurred on the 23rd. In 193 4> however, 

 migration started on the 12th, and was much more continuous. There 

 were three large waves of migration — on the nights of the 14th, the 

 1 6th, and the 24th, this being the spawning date. Most of die frogs 

 reached the pond long before they spawned. 



Minimum temperatures occur in the small hours of the morning, 

 but frogs are on the move as soon as darkness falls, so that the tempera- 

 ture that directly matters to them is somewhere between the maximum, 

 occurring in the early afternoon, and the minimum early next morning. 

 This being so, it is not easy, on the temperature data, to see why the 



