OF TEMPERATURE AT TORONTO IN CANADA. 
161 
If we collect the values in column 3 of Table V. into monthly averages, and 
arrange the months in meteorological seasons, we obtain thereby the following 
comparison of the probable variability of the daily temperature at different seasons : — 
Winter. Spring. Summer. Autumn. 
December . 5’6 March . . . 5T‘ June .... 4’0 September . 4'2 
January . . 6'6 April .... 4'6 July 3’5 October. . . 4‘0 
February . . 6’6 May 4’2 August ... 3*0 November . 4*4 
Means . . . 6’3 4*6 3'5 4*2 
Whence we learn that during the winter months the temperature of a day is as likely 
to differ from its mean or normal state 6°'3 or more, as it is to differ less than that 
amount. Regarding 6°’3 therefore as the index of variahiUty in the winter months, 
the index in the spring months is 4°'6, in the summer months 3°'5, and in autumn 4°‘2. 
The probable variability is a maximum in the last half of January and first half of 
February, and a minimum in August ; it is progressive from the maximum to the 
minimum and from the minimum to the maximum, and its amount is about twice as 
great at the maximum as at the minimum. 
The principal variations from the general progression shown in the five-day averages 
of the twelve years, accompanied in each case by the “ probable variability,” are as 
follows : — 
January 1 8 to 22 inclusive 
0 
-2-8 
±1-1 
February 12 to 16 inclusive 
— 2-9 
±0-9 
August 28 to September 2 inclusive . . 
+2-3 
-f~0’5 
October 8 to 12 inclusive 
H-2-9 
±0-5 
November 28 to December 2 inclusive . 
— 21 
+07 
December 18 to 22 inclusive .... 
-3-2 
+0-8 
December 29 to January 2 inclusive . . 
-1-2-4 
+0-8 
Of these, more than half, four, occur in the winter season, two in the autumn, and 
one in the summer. Their connexion with particular winds and with other meteoro- 
logical circumstances will be more properly examined in the general discussion of 
the meteorological observations at Toronto in the volumes of that observatory ; but 
it has appeared desirable to bring into immediate notice the principal instances, 
furnished by the five-day means in twelve years, of such more considerable departures 
from the regular progression of the temperature, as may be supposed to indicate in 
some degree a tendency to periodical recurrence. 
For the convenience of those who study monthly averages, the mean temperatures 
of every month in each of the twelve years have been collected in Table VI., together 
with the average temperature of each month in the whole period. By the comparison 
of these numbers, the amount by which the temperature of any particular month 
during the twelve years exceeded or fell short of the average may at once be seen. 
