306 On the Revolution of the Seasons. [April, 



18°. When you have marked all the heights, join them, and you have 

 the upper, or continuous line, fig. 1 . The lower or dotted line in 

 fig. 1, where it separates from the upper, — is formed from it, by sub- 

 stituting for the increments, equal decrements, so as to be exactly the 

 inverse of it. Where this lower line again changes to a continuous 

 one, it runs parallel (or varies directly) with the upper one, and 

 again, where it changes to a dotted one, becomes the inverse of it. 

 It is this lower line, partly direct, partly inverse with the upper, that 

 appears to be the type of the variation of the seasons. As a proof of 

 this, we subjoin below (fig. 2) the general average line of variation 

 in the prices of corn during the same period. This line was thus 

 formed. The three principal lines, the north-west, the Benares, and 

 the Bengal, were first formed from the average of the different lists. 

 When the maximum and minimum number in each line within 

 the last 85 years (since 1750), were noted, and the difference 

 between them reckoned as the whole amount of variation. This 

 amount was divided into 1000 parts, and, for the actual number 

 in each line, the proportionate parts of the variation were sub- 

 stituted. The average was then taken of the 3 lines, and this is 

 the line expressed in fig. 2, which is there traced upon the paper at 

 the rate of '020 parts of variation for ^th of an inch. The lowest line 

 (fig. 3) is the general average, simply taken, of the principal lines, 

 without any previous division of the variation into centesimal parts. 

 A fourth, or southern line, was in this case included in the average, 

 having been formed from prices at Jubulpoor (two kinds of corn), at 

 Bhopaul (three kinds of corn), at Tndore (two kinds of corn). But as the 

 country in that direction was during part of the time the seat of war, 

 and has been generally subject to unsettled government, and more- 

 over the returns are not numerous, no great dependance can be placed 

 upon it. In fact, the indications given by the north-west series are 

 much more to be relied on than those of the others, owing to the 

 more extensive induction. 



In the last paper on this subject we noticed that there were certain 

 years in which, about the solstices, the perigee of the moon fell on 

 the same day with her maximum declination, either north or south, 

 and that these were commonly extreme years, both of drought and 

 moisture. These years are marked thus in the Chart N.* and S.* 

 according as the declination is north or south, and it would appear 

 on referring to the figures that these are usually the extreme years 

 both of plenty and scarcity. They appear also to be the periods at 

 which the variation changes from direct to inverse. 



