IS 



ON THE VARIATIONS OF TEMPERATURE AT TORONTO. 



[18*53 



A. M. The range from the minimum to the maximum in the 

 day is greatest in July (18°. 2,) and least at the end of De- 

 cember" (5 ° .2). The daily range has but one maximum in 

 the year, which is in July ; not as at Prague, where June and 

 July have a less range than the months immediately preceding 

 and following them, and where consequently there are two max- 

 ima, a phenomenon attributed to greater prevalence of the clouds 

 in June and July. 



[The following are the mean daily ranges of temperature' for 

 the several months ; they must be carefully distinguished from 

 the ranges of the whole, months separately, these latter being the 

 differences between the highest and lowest temperatures occuring 

 in each month. 



o o o 



January 5.78 May 15.13 September 13:80 



February 9.38 June 15.69 October 10.88 



March 1023. ...July 18.21 ....November 6.68 



April 11.97 August 16.27 December 5.98 



Mean daily range on the average of the whole year 1 1°.4 1 .] 



It may be desirable to add a few words on the assistance to 

 observers of tables which &trnish corrections to the mean tem- 

 perature of the day for every hour of every day in the year, such 

 as the table spoken of. Besides their direct use at the station 

 itself, they have a useful bearing within a reasonable distance from 

 the station, on the selection of observation hours in the many- 

 cases in which it may not be possible to observe at hourly or 

 two-hourly intervals, by affording a ready means of estimating 

 the amount of error to which a deduction from any limited com- 

 bination of hours is subject. If we desire, for example, to seek 

 the observation hours within the command of a single observer, 

 which may give the best approximation to the mean temperature 

 of the day, and to that of the month, and of the year, as well as 

 to climatic difference (i. e., the difference between the hottest and 

 the coldest months), we find that, of homonymous horn's, the best 

 pairs at Toronto are 9 1 ' — 9" and 10 h — 10^ 10 h — 10h being the 

 better of the two ; but that S' 1 — 8 h , which is a combination fre- 

 quently adopted bv observers, does not suit so well at Toronto as 

 either 9 h — 9" or l*0 h — 10 h . 



For the purpose of combining with an approximate mean tem- 

 perature of the day an approximation to the hottest and coldest 

 hours of the day, and to the horn's of maximum and minimum of 

 other meteorological elements — three equidistant observations are 

 frequently adopted in preference to a binary system, and the 

 hours of 6 a.m, 2 p.m., and 10 p.m., appear to be usually pre- 

 ferred. These hours are still within the command of a single 

 observer, though we often find substituted for them the non-equi- 

 distant hours of t a.m., 2, p.m. and 9 p.m., doubtless because 

 they suit better the convenience of observers. In comparing the 

 mean temperatures in the different months derived from 6' 1 , 2 h , 

 10' 1 , or 7 b , 2 h , 9 h , with the full complement of twenty-four hours, 

 we find that the approximation to the mean temperature obtained 

 by V h , 2 h , 9 h , is not quite so good as by 6 h , 2' 1 , 10 h ; and that 

 either of the triplets gives a less correct mean temperature than 

 10 h — 10 h : 6-2-, 2, 9£, would appear a more suitable combination 

 as far as regards approximation to the mean temperature. 



Three equidistant observations in the twenty-four hours are the 

 utmost that can be perseveringly maintained bv a single observe]'. 

 When there are two or more observers there is no difficulty in 

 multiplying the times of observation so as to comprehend all the 

 objects that may be desired, each in the manner and by the means 

 which are most suitable to it, and will be most satisfactory. But 

 as the work of observation at by far the greater number of meteoro- 

 logical stations is usually carried out by a single observer, and as 

 this is likely to be always the case, it should be a primary object 

 with meteorologists who are furnished with sufficient means, to 



form tables of corrections to the mean duly temperature for every 

 hour of the clay, upon the basis of a sufficient number of years of 

 observation, to be used at the respective localities, or within the 

 distances to which such tables may be severally applicable by 

 persons whose means or convenience may restrict them in respect 

 to the number and choice of hours of observation. 



With such a table, the choice is disembarrassed of its chief dif- 

 ficulty, that of selecting houis which by their combination will 

 give an approximate mean temperature for the several months 

 and for the year; and the observer is left free to give a preference, 

 independent of such consideration, either to the hours when the 

 phenomena change least rapidly, and consequently small irregu- 

 larities in the times of observation will be least injurious, or to the 

 hours which will furnish the best approximation to the daily 

 maxima and minima of the meteorological elements generally,, 

 viz. : of the temperature, the tension of vapour,, the pressure of 

 the gaseous atmosphere, and the force of the wind ; or to the hours 

 which will have the most effective bearing upon other points of 

 meteorological or climatic interest, to which the observer's atten- 

 tion may be directed." 



Having thus obtained the necessary corrections for all the 

 the hours; they have been applied to the observations individu- 

 ally, during the next 6 years in which the hourly system of obser- 

 vation w;ts abandoned. The daily means of the whole 12 years 

 were then collected into monthly averages, and Bessel's formula 

 was used to obtain the mean normal temperature for each day of 

 the year. These are entered in table IV along with the differ- 

 ences between these normals and the observed means of each day 

 for the whole 12 years — being in fact the non-periodic variations 

 for these days — and also a column with the average of these 

 differences, or the average non-periodic variation for each day. 

 On this Colonel Sabine remarks.' 



"We may learn consequently from this column the average non- 

 periodic variation in twelve years of any particular day of the 

 year which may be surmised to be subject to some special phy- 

 sical peculiarity, causing it to be warmer or colder than the 

 general progression of the temperature in the part of the year to 

 which it belongs. An example of its application may be given 

 by the reply which the values in this column furnish to the ques- 

 tion, whether the three days of May (the 11th, 12th, and 13th,) 

 which Madler has stated to be characterised, on the average of 

 86' yeais of observation at Berlin, by a depression exceeding 2° 

 Fahr., when compared with the general march of the temperature 

 at that season, undergo a similar depression in North America. 

 On a reference to the month of Ma] 7 in the fable, it is seen in the 

 final column that on the average' of 12 years from 1841 to 1852, 

 the 11th of May was 0-1 below, and on the 12th and 13th of 

 May respectively 3°-l and 2-°4 above- the general mean of the 

 temperature in those years. It may be seen also that the average 

 non-periodic variation in the five days from the 8th to the 12th 

 of May inclusive, is in the same 12 years, 1° - 1 above, and in 

 the five days from the 13th to the 17th inclusive, 1°-1 above, the 

 general mean of the temperature. The meteorological observations 

 at Toronto during these 12 years do not therefore support the 

 supposition that the depression of temperature on the 11th, 12th 

 and 13th of May, observed at Berlin, is a general and periodically 

 recurring phenomenon over the whole globe, such as would be 

 occasioned by a partial obscuration of the Sun's disc by the inter- 

 vention of a periodical stream of aerolites ; but they tend rather 

 to indicate that the depression observed in Europe may have 

 been a partial phenomenon, having a local cause." 



From the tables mentioned, I have computed by interpolation 

 the normal values for each hour of observation at Toronto on 

 every day in the year, the differences between which and the ac- 



