492 Royal Society, 



ture is greatest; but that at all the stations the increase of the range 

 is most rapid in the first or second hour alter sunrise ; and that its 

 extent at the hours from 7 to 9 a.m. is not exceeded at any subse- 

 quent hour at Hobarton, the Cape and St. Helena, whilst at Toronto 

 the great enlargement takes place even earlier, the hours of 6, 

 7 and 8 a.m. being exceeded by none, though they are equalled by a 

 second increase at noon and the two following hours. This second 

 enlargement is perceptible at the same hours at Hobarton and St. 

 Helena. 



With reference to the relative positions of the several months in 

 each of the vertical lines, or at the different hours, it is observed that 

 certain months, which are found congregated at the one extremity 

 of the range during the early hours of the morning, undergo a 

 transfer towards the opposite extremity at a subsequent period of the 

 day; thus the months June, July, August usually occupy one 

 extremity of the range, and November, December, January the 

 other extremity, in the morning hours, and until from ^^ to 10** 

 A.M., when each of the two groups is respectively transferred 

 towards the opposite extremity to that which it previously oc- 

 cupied. The period at which this transfer takes place is somewhat 

 earlier at Toronto and St. Helena than at the Cape of Good Hope 

 and Hobarton. The comportment of the two equinoctial months, 

 March and September, at the Cape of Good Hope and St. Helena is 

 pointed out as presenting a remarkable contrast to that of the two 

 solstitial groups which have been described, and at the same time a 

 still more remarkable contrast to each other, March being at almost 

 all the hours on the West, and September on the East, of the mean 

 line. 



In conclusion the author points out one or two practical conside- 

 rations suggested by the facts under notice : — 



1. That as, in the Annual Variation represented in the plate, the 

 same months occupy positions on opposite sides of the mean line at 

 different parts of the twenty-four hours, the mean annual variation, 

 or that which is shown by the mean values in each month taken from 

 all the observation hours, must be merely a residual and not an abso- 

 lute quantity ; and that consequently natural features must be more 

 or less masked in deductions in which only mean values are brought 

 into view. In fact, as has been shown in the published volumes of 

 the observations at St. Helena and Hobarton, the mean annual va- 

 riation at those stations is so small as to be scarcely sensible. But 

 when we resolve these mean results into their respective constituents, 

 viz. the annual variation at each of the observation hours, there is 

 then at once disclosed to us an order of natural phenomena, very 

 far from inconsiderable in amount, systematic in general aspect, and 

 apparently well deserving the attention of those who are occupied 

 in the delightful and highly intellectual pursuit of tracing the agen- 

 cies of nature. 



2. We perceive in the variations of the position of the several 

 months in the annual range, the necessity of paying regard to the 

 period of the year, as well as to the period of the day at which ob- 



