Mat 1901.] BOTA^'ICAL SOCIETY OF EDINBURGH 



77 



from 8th to 18th, and 4-42 inches from 23rd to 31st 

 August. It is surely reasonable to suppose that the 

 effects of this heavy rainfall had not worn away, 

 and that they account for the slight over-average 

 growth of the 12th October period, the dryness of 

 which causes the high maximum and minimum 

 thermometers. 



Table III. 



1888 

 1889 

 1890 

 1891 



Growth. 



25-1 



4-1 

 1-1 

 5-1 



35-4 



Thermometers. 



Max. 



1-8 

 0-8 



4-1 

 0-9 



6-0 



Raix. 



Sex. 



12th 

 October 

 • to 12th 

 November. 



In 1888 there is a good growth, but very little rain 

 over the average, so the nights had not been much cooled 

 down. But the good growth of this period is no doubt 

 largely due to a heavy rainfall just a day before measuring 

 day, 3 inches being measured at 9.30 a.m. on 11th 

 October. 



The rather over-average growth of 12th October to 12th 

 November 1891 is accompanied by an over-average of all 

 weather factors. Xow the rainfall of this period was 

 distributed as follows: — 1'58 inch from 12th to 31st 

 October; 0"27 inch 1st to 6th November; and 3-34 

 inches on 8th November. Thus for the first twenty- 

 seven days there was an under-average rainfall, and, 

 in consequence, an over-average temperature, both by 

 day and night. The heavy rain of 3 '3 4 inches does 

 not perceptibly affect the mean temperature, but does 

 affect the growth. 



We may perhaps draw the corollary that rainfall affects 

 growth within four days. 



