1836.] On the Revolution of the Seasons. 283 



information of the two kinds above-mentioned, possesses over the 

 results of a solitary rain-guage, however carefully kept, viz. that they 

 are an index, imperfect as they may be, to what has happened over a 

 large tract of country ; whereas the rain-guage can only inform us 

 as to one particular spot, and rain-guages in general are so much 

 affected by peculiarities of situation, that the results afforded by any 

 one singly, must be considered as liable to doubt. To obviate this 

 objection, I have placed together in a table all the different series 

 procurable, of a date posterior to 1820, (see Table No. 1.) Most of 

 them are to be found in the different Nos. of the Journal ; and the 

 localities are between Dacca, (E. Long. 90°,) and Delhi, (E. Long. 78°,) 

 between Nagpur, (Lat. 21°, North,) and Delhi, 28° 40' (N.) To these 

 are added the observations at Madras, which I have obtained through 

 the kindness of the Astronomer there ; at Macao, in China, (Journal, 

 July, 1832,) and at Edinburgh, (see Brewster's Philosophical Journal, 

 passim.) In Table No. 2, are given the only three series that I 

 have for the years between 1800 and 1821 . The two first (Madras and 

 Macao) are merely the preceding parts of the series given in Table 1 . 

 The last from Carlsruhe, in Sweden, is given in the Edinburgh Philoso- 

 phical Journal for 1821, there quoted from the Bibliotheque Universelle 

 for November, 1820. The original appears to be given in French 

 inches and lines, and I have not reduced them to English measure, 

 as the doing so would not affect the question at issue, viz. whether 

 some years of the lunar cycle are more rainy than others. Now to 

 make a more correct comparison of the different years, we must first 

 reduce the numbers given in Table 1. to a common mean. Thus, we 

 have Dacca for eight years, (1827, 1828, 1829, 1830, 1831, 1832, 

 1833, 1834,) and the mean of the different sums given is 72*80 inches ; 

 at Calcutta, for the same eight years, the mean is 60*37 inches. The 

 difference between these, 12*43 inches, we may reasonably suppose to 

 be owing to the localities. Subtract, therefore, from each item of the 

 Dacca series, the mean difference 12*43; the remainders will be 

 reduced to the mean of Calcutta*. Proceed in a similar way with the 

 other series, only of course where the climate is drier than that of 

 Calcutta, the mean difference must be added, and not subtracted. 

 The series in Table No. 2, may be included in the comparison by 

 treating them in a similar way, and then considering them only accord- 

 ing to their position in the lunar cycle. Thus, if we take 1821 for 

 the first year of the cycle, 1 803 (or the eighteenth year before that) 

 may also be reckoned as the first year ; 1 802 and 1 820, will of course 

 be the last years. Place the whole in columns numbered according 

 to their distance from 1802 and 1820, and an average may be taken 



* The more correct mode would be to multiply the Dacca series by 60 ^Z. — Ed. 



r * ' 72-80 



2 o 2 



