408 THIRD REPORT — 1833. 



For the winter quarter. 



By calculation {m = 4-42) . 51*4 79-6 

 By observation 51 "5 79'5 



For the summer quarter, 



By calculation (m = 4-14) . 55-5 80-8 



By observation 5^*7 82'6 . . , ^a 



■c 4.1 • . > to 100. 



r or the spring quarter, * 



By calculation {m — 4-92) . 45-9 77-2 



*By observation 48 75" 1 



For the autumn quarter, 



By calculation {m = 3-71) . 59-2 82'9 



By observation 57-8 84-3 



It is therefore probable that the Westminster results obey the 

 • same constant relation to height as those of York. 



8. But it is evident that the values of the variable coefficient 

 are very different ; that its maxima and minima are perhaps 

 not quite in the same periods of the year as at York ; and 

 that the range of variation in its value is very much less. 

 From M. Arago's determination of the relative quantities of 

 rain falling on the Observatory at Paris and in the court 28 

 metres below, as given by Professor Forbes, in his Report on 

 Meteorology, 50"47 : 56*37, the relative mean value of m at 

 Paris = 1 -24, while at Westminster it was 4*23, at York 2*29. 

 It must be owned that these discrepancies with other obser- 

 vations as respects the quantity of the diminution of rain up- 

 wards are somewhat discouraging, and probably will, for a 

 considerable time, deprive the most exact local determinations 

 of this quantity of a general application. This, indeed, could 

 hardly be expected, since the whole quantity of rain is so va- 

 riously dependent on circumstances of physical geography, 

 that centuries have been found insufficient to determine the 

 general law and ascertain the numerical constants of local 

 climate. Yet, on account of the remarkable regularity of the 

 progress of monthly temperature at York, and some very ob- 

 vious relations between the quantities of rain collected and 

 the mean temperature of the period, I will venture to state 

 what seem unavoidably to suggest themselves as probable in- 

 ferences. 



9. First, it is obvious that the diminution at the upper sta- 

 tions is greatest in the cold season, and least in the warm 

 season, and therefore the coefficient is in some way inversely 

 dependent on the temperature, or on some effect of this tem- 



* March very anomalous. 



