318 SMITH : APPLICATION OF THE THEORY OF 



mine the corvee more closely to give us an idea of the 

 accuracy which the methods used are capable of. The tem- 

 ture both at Hakgala and Peradeniya was taken in the 

 thermometer shed some 150 yards or so away from the plant 

 measured. It was difficuU to obtain the shade temperature 

 nut in t he open near the plant, but on one occasion a thermo- 

 meter reading 1)2-6° F. in the thermometer shed was taken 

 immediately to the plant and there registered 61*7° F.„ a 

 difference of -9° F. It appears, therefore, that the record of 

 temperature may vary by a small amount either way from 

 the actual air temperature near the plant. Then, as has been 

 pointed out. the temperature of the surrounding air may differ 

 somewhat from the internal temperature of the growing shoot. 

 That notwithstanding this, the correspondence between the 

 two carves should be so striking is an indication of the 

 accuracy attainable in a continuous series of measurements, 

 where the error of any one measurement does not seriously 

 affect the general form of the curve. 



Ii will be useful to notice that though the general corres- 

 pondence is dear there are deviations from it of small mag- 

 nitude. Thus, on June 12th at Kandy the rate of growth of 

 Furcraea increases between 2 and 6 a.m. somewhat consider- 

 ably, though the temperature is almost stationary. 



I n flat nee of Grand Period. 

 The general result does not seem to be affected by the stage 



at which the shoot is measured. 



Agave \1 « • r i i — j i at FTa.lrgfl.lq was measured during the earlier 



Oi growth when the shoot was about <i feet high. Fur- 



igantee at Kandy was in its later stages, and had already 



'ii i" blanch. It, was about 25 feet high. .\<jave anicn- 



i at Elakgala was about IS feet high and branches wero 



jusl appearing, while the same species at IVradeniya was 

 aboul !."• feet high and measured before any side branches 

 had i|i|m an (I 



