340 SMITH : U'l'LKWTloN OF THE THEORY OF 



hotter part of the day. This rapid growth was followed by a 



gradual falling off during the night, the lowest reading for the 

 night being recorded early in the following morning when the 

 minimum temperature for the night was also recorded. The 

 temperatures of 15° C. recorded on the morning of January 

 loth and 16T° C. recorded on the morning of January 11th, 

 are much lower than the usual minimum temperature at 

 Peradeniya. It is only on a few nights in December and Janu- 

 ary that such temperatures as these are reached, so that the 

 range of temperature through the night is considerably lower 

 than that to which these plants usually are exposed, and it is 

 under such circumstances of course that the temperature 

 would be expected to be the limiting factor in growth. 



The same plants were measured again on July 6th to 7th 

 in the south-west monsoon, when the range of temperature 

 was much more uniform, being considerably higher at niglit 

 and lower in the day. 



Plate 24 (middle, on right) shows the curve for Capparis 

 Roxburghii from July 6th at 11 a.m. to July 7th at 9 a.m. 



The day was dull on the whole with high humidity through- 

 out, and we do not get any record of actual contraction during 

 the day. Growth was fairly steady throughout the night and 

 contrasts strongly with the steady fall in growth- rate recorded 

 lor January 9th to 10th and 10th to I lth. It is not easy to 

 -ay whether the temperature, being now about 5 C. higher 

 throughout, is in excess, and thus brings in the humidity as 

 limiting factor through the night as well as through the day, 

 01 whet her here again t he water supply being in excess during 



the night bringe m the temperature as the limiting factor. 



The temperature ourve shows such a slight fall during the 

 nighl t hat it would be difficult to detect its effect on the growl h. 



On t he win ile. perhaps, the growth shows a tendency to respond 

 to the increased humidity of the later part of the night, but 



cannot be stated with oertainty, the result chiefly im- 

 portant being the oon1 rasl wil b the steadily falling night curve 

 of January 9th to 1 1 1 h when the temporal ure was considerably 



