280 



SMITH 



October and November, in the height of the north-east 

 monsoon, show the smallest number of growing species. 



A question which affects our treatment of these results is 

 whether the phenomena observed in any year are to be corre- 

 lated with the weather of that particular year or with that of a 

 long series of years. We know that leaf-fall in temperate 

 climates, though it occurs with regularity sometime during 

 autumn, yet varies as to its actual date with the kind of 

 weather experienced. If the autumn is fine and warm, the 

 leaf -f all wiU be later, while if there are early frosts and rough 

 weather, it will be earlier. It is not known how far such 

 temporary variations of weather in the tropics affect the 

 growth periodicity, nor, in fact, whether they do so at all. 

 The writer did not stay in Ceylon long enough to obtain 

 records for more than one year, so that there is no evidence 

 as to how far the records of that year represent the usual 

 occurrence of the phenomenon. It cannot be stated how far 

 the records reflect only the general periodicity estabhshed as 

 a result of average climatic influences , or to what extent they 

 show the influence of the rather special variations of rainfall 

 during that particular year. 



It is possible that the large number of growing species in 

 February and the fairly large number in September represent 

 the two normal growth periods of the year, corresponding to 

 the two drier periods ; and that the fairly large numbers in 

 December, 1905, and June, 1906, are not usual, but represent 

 the effect of the abnormally dry weather of those two months 

 in the particular years in question. On this view the practical 

 absence of growth in October may also be due to the exceed- 

 ingly wet weather of that month in 1906. The following are 

 the figures of rainfall for December, 1905, to November, 1906, 

 with the number of rainy days in each month : — 



Rainfall m Inches. 



