204 ECOLOGY AND LIFE HISTORY OF THE COMMON FROG 



the transformation, was removed in the analysis as irrelevant. Sometimes two 

 collections had been made on a certain date, and sometimes three, so that the 

 analysis was not quite simple, but a suitable scheme was found in Davies (1949), 

 p. 102, Appendix 5D. The analysis was as follows, the figures being logarithms 

 to two decimal places multiphed by 100 to remove the decimals. 



The variance ratio is significant at more than the i per cent. Dagger Lane 

 Pond is not remarkable for the size of the variations, and there is no reason to 

 suppose that similar analyses on other ponds would be any less significant. No 

 doubt remains about the reaUty of the colony differences. 



The numerical data for this table and the other weights of tadpoles have 

 not been pubhshed, for the amount is voluminous, but they can be found in the 

 library at London University in the PhD thesis entitled 'The ecology of anuran 

 tadpoles' (1950). 



(b) The Distribution of Polystoma integerrimum in Tadpoles 



If a population of tadpoles were uniformly exposed to infestation by swim- 

 ming larvae, which entered the tadpoles at random, it might be expected 

 that the number of parasites in the tadpoles would be distributed according to 

 the Poisson series, so that the number of tadpoles with 0, i, 2, 3, . . .n parasites 

 would be in the proportions given by the terms of the expression: 



23 n I 



where e is the base of natural logarithms, and m is the mean number of parasites. 



Using the data from three ponds Upper and Lower Park£eld and Hospital, 

 all in 1948, the numbers calculated from this expression are given in Table i, 

 together with the numbers actually found. 



In each case, the large excesses in the first and last classes give such high values 

 for ;^2 that the corresponding values for the probabihty are beyond the range 

 of the usual tables. It is impossible to regard the way in wliich the parasites 

 are distributed as being due to random causes. 



