289 



The probable errors of these sums, considered individually, arc 

 quite large, and their agreement from year to year is not sufficient 

 to justify the belief that we have attained to a satisfactory expression 

 for the connection between the temperature and the date of flowering. 



Flowering of rye and winter wheat — Harvest of rye, winter wheat, 

 and spring barley. — A new investigation, based on an increased number 

 of stations for the vears 1886 and 1887, gives for the rate of retarda- 

 tion of these epochs the following figures : Flowering of rye, -1.2 days 

 per 100 meters; flowering of winter wheat, 4; harvest of rye, 4.5; 

 harvest of winter wheat, 4.3; harvest of spring barley, 4.2. We can, 

 therefore, as before, take 4 days as an approximate value for all these 

 phenomena. 



The mean temperature at the time of flowering is determined, both 

 for daily means and for daily maxima, as follows: 



Again, the average numbers agree well from year to year, but the 

 individuals from which they are derived have a wide range. 



The sums of the mean daily temperatures, less 5° C, counting from 

 December 1 for the winter rye and wheat, but from March 21, for the 

 spring barley, are as follows : 



/S'w»i.s- of tcjiipcrature. 



From the flowering to the harvest, on the average of these two 

 years, rye has received 1,048 — 364=684° C., and winter wheat 1,236 — 

 682=554° C, but on the average of eight years, 1880-1887, the sums 

 of temepratures, less 5° C., have been, for rye, from December 1 to 

 the flowering, 477° C, and from flowering to harvest, 607° C. ; for 

 winter wheat the numbers are, respectively, 708° to 549° C. From 

 the beginning of vegetation up to harvest the numbers are: For rye, 

 1,084° C; winter wheat, 1,256° C; spring barley, 1,103° C. The.se 

 results can be considered as having definitely established the fact tlmt 

 2667—05 M 19 



