298 Prof. Balfour Stewart and Mr. W. L. Carpenter. 



kept as it is the second set be pushed to the left of the first one 

 division, and the third to the left of the second one division the 

 sum of the three 12-y early series so placed being taken to repre- 

 sent 41. In like manner for 43 we should push the second 

 12-y early series of 42 one division to the right of the first, and 

 the third one division to the right of the second. 



The method which we have now described applies both to the 

 Inequalities around 24 days of which we have spoken, and to those 

 around 26 days which we are also about to exhibit. Here, however, 



one day in a year means ^-^- or '0712 day. In other respects the 

 oiw* SH 



treatment is the same for both. 



Temperature-range Observations and their Treatment. 



11. We are indebted to the Meteorological Council, the Kew Com- 

 mittee, and to Messrs. Kingston and Carpmael, the late and the 

 present Directors of the Toronto Observatory, for 36 years of the 

 Diurnal Temperature-ranges at Toronto, extending from the begin- 

 ning of 1844 to the end of 1879. We have also obtained 24 years of 

 the Diurnal Temperature-ranges at Kew, extending from 1856 to 

 1879, for which we are indebted to the Kew Committee. These 

 observations have been split up into series of 12 years each, the 

 first two series of Toronto and the last series of Kew being coter- 

 minous with solar series, the last series of Toronto and the last of 

 Kew being, however, 12 years later than the last solar series. 



It thus appears that a different series of years is represented in 

 each case. It would, of course, have been better for our purpose if 

 precisely the same series of years had been represented by the three 

 sets of observations ; but as this is not the case we have not scrupled 

 to make the comparison implied in this research between the whole 

 body of observations as they stand, believing that in so doing we 

 shall make the best use of the materials placed at our disposal. 



12. Here it will be necessary to allude to a method of treatment 

 which we have thought it desirable to apply to the temperature-range 

 observations, but which we have not found it necessary to apply to 

 those of sun-spots. 



Having obtained the yearly Inequalities corresponding to 24 days 

 in exactly the manner already described, we have to some extent 

 smoothed or equalised our series of differences. The primary series 

 (A) has been converted into another series (B), also of 24 sums, each 

 sum of (B) being the mean of four consecutive sums of (A), and the 

 series (B) has been converted by a similar process into a series (C), 

 each sum of which is the mean of four consecutive sums of (B). It 

 is the series (C) which is ultimately made use of. 



