304 PHENOLOGICAL OBSERVATIONS, CANADA, 1899. MACKAY. 



as to the average time of first flowering and when flowering 

 was beginning to become common. In some counties the 

 observations were so full that thirty good stations could be 

 selected for averaging, ten from the sea coast, ten from low 

 inland settlements, and ten from high land settlements. These 

 average dates or phenochrons of flowering are arranged in 

 parallel columns for the sake of comparison. In some counties 

 only twenty satisfactory stations for averaging, and in others 

 only ten, were found, as can be seen at a glance from the 

 tabulation of the figures. 



The average phenochron for each plant's first flowering and 

 flowering becoming common is calculated for each county, and 

 the mean of the two series is finally taken for comparison 

 with the similar general phenochron for the same phenomenon 

 in 1898. These general phenochrons are plotted on the accom- 

 panying diagram so as to show their curves through the 

 Bounties of the Province arranged in a linear series beginning at 

 the west and south, and proceeding to the east and north. 



This order of the counties will be uniformly followed in 

 future plottings of the phenochronic curves, for the greater ease 

 of comparing those of one year with those of another. Last 

 year the counties were arranged in the order of their most 

 general phenochrons. Were the same rule followed this year 

 the positions of some of the counties would be changed ; but if 

 the positions of the counties remain fixed the configurations of 

 the phenochronic curves will illustrate the variations very 

 clearly from year to year. 



On a future occasion I propose to plot the phenochrons of 

 the same /phenomena running through the counties of the 

 province for two or three consecutive years, in order to study 

 the character of the annual observations, or the peculiarities of 

 climate or flowering. Unfortunately, we cannot be sure of 

 the degree of variation originating in the latter causes until we 

 are sure of a uniform system of correct observations symmetric- 

 ally distributed. 



