SMITH . \rn.IcAT10N OF THE THEORY OK 



during the night with lower temperatures. The highest 

 growth rates per two hours recorded for Hakgala were in fact 

 observed during the day and not during the night. The maxi- 

 mum record for two hours of darkness at Hakgala is 1*4 cm. 

 on April 1st. 6-8 p.m.. the temperature for the beginning of the 

 period being 166° C. (620 F.). This was exceeded by the 

 following records for two hours of light, rain in each case fall- 

 ing continuously or nearly so during the period. 



On April 1st, 4 to 6 p.m.. 1-65 cms. Temperature at 4 p.m. 

 19-3° C. (66-7° F.). 



April 25th, 1.40 p.m. to 3.40 p.m., 2 -55 cms. Temperature 

 at 1.40 p.m. 21-1° C. (70 0° F.). 



April 5th. 11.30 a.m. to 1.30 p.m., 2-27 cms. Temperature 

 at 11.30 a.m. 18-8° C. (66-0° F.). 



No exact proportion is to be expected in these figures, since 

 it is probable, as has been pointed out previously, that the 

 internal temperature of the culm is the true controlling factor, 

 and this, as my observations show, was often considerably 

 removed from the shade temperature of the air during the day, 

 especially if the weather had been sunny, while during the 

 night the two temperatures were as a rule not far apart. They 

 do show that given a sufficient supply of moisture the highest 

 rate of growth occurs at the highest temperatures. 



3. M in i mum Temperature and Night Growth. 



The minimum temperatures were included in all the obser- 

 vations taken between Maroh :>lst and May 14th. These 

 minimum temperatures arc a guide, somewhat rough of course. 

 to tin- average temperature <>f the different nights. It maj 

 happen thai the minimum temperature is a sudden fall from a 

 considerably bighei temperature, and thus would not be a 

 lOUnd indioation of the average temperature of the whole 



' . On the whole, however, the minimum temperature may 



B6 used nfl n n uide to the average temperature during 



the nielli Thin being bo. we should expect it the temperature 



