280 
A similar elaborate study of the harvest of rye gave the following 
results : 
(1) Retardation for altitude is approximately four days per 100 
meters, with some indication that the correct figure is rather less 
than this. 
(2) The date of harvest reduced to sea level begins with the 5th of 
June in southern France and ends with the 25th of July on the 
northern border. 
(3) The sum total of temperatures computed by the three methods 
A, B, C, above mentioned, shows that whatever method be finally 
adopted as the best, these sums are less for rye than for winter wheat. 
A similar study for spring barley shows the following results: 
(1) A retardation of four days per 100 meters of altitude suf- 
ficiently harmonizes the observations. 
(2) A retardation of thirty or forty days in the date of sowing 
has no appreciable effect on the date of harvest, which varies from 
the 20th of June in southern France to the 14th of August on the 
northern boundary. 
(3) The sum total of temperatures from sowing to harvest is too 
variable to be determined. 
A similar study of the flowering of the narcissus (Narcissus 
pseudonarcissus) shows that the retardation of the date of flowering 
is at the rate of four to five days per 100 meters, and four days can be 
adopted without notable error. 
A study of the currant (2?2bes rubrum) shows that the retarda- 
tion is between three and four days per 100 meters. The sum total 
of heat from December 1 up to the date of flowering, as deduced by 
the second and third methods, but under three different assumptions— 
i. e., that the initial temperature is 4°, 6°, 8°, respectively, seems to 
show that 4 is the proper figure for this plant. 
A study of the flowering of the hlac shows that a retardation of 
four days per 100-meters best satisfies the observations of both leafing 
and flowering. The latter begins in southern France on the 22d of 
March and ends in northern and eastern’ France on the 6th of May. 
The calculation of the heat required for leafing shows that the most 
accordant results are obtained when we take the sum of maximum 
daily temperatures above 4° C. and count from the date of the last 
heavy frost, which sum is about 360° C. For the flowering, on the 
contrary, we have to take the sums of the mean daily temperatures, 
counting from 4° C. and from the same date of frost, which sum is 
then 350° C., while the sum of the maximum daily temperatures 
would have given 695° C. 
A study of the leafing and flowering of the horse-chestnut (4’s- 
culus hippocastanum) shows that the retardation of four days per 100 
meters also satisfies these observations. The dates of leafing, as 
