^ 



** 



4 On Certain Meleorolosrical Coincidences. 



Of these columns, the first^ third and fourth are mere tran- 

 scripts of the printed observations; with the hours of (T'^ change 

 (morning or afternoon) supplied to the first column. The second 

 and fourth belong to the comparison. It is to be observed, far- 

 ther, that the supposed rule claimed no more (ban the distinction 

 between dry and (as it is called) falling weather. Thus, the ex- 

 istence of nimbus or rain-cloud on any quarter d^-j^ provided it 

 did not rain ^ might be taken as coinciding with the weather on 

 a term-day which was clear. But 1 have not so taken it, in fact ; 

 and have reckoned the existence of nimbus as equivalent to rain, 

 and noted it as accordant or discordant, as the case may be^ 

 However hardly this may strain the supposed rule, I have no 

 doubt of the soundness of its being insisted on. On the other 

 hand, the propriety of considering the weather on 18th Aug., < 

 1843, as coincident with the weather on 25ih July, 1843, for 

 instance, will, I suppose, be equally admitted. 



With this exposition of the principles and manner of the com- 

 parisons, of which throughout the extract above is an average 

 sample, the aggregate results may now be stated. 



Of the 31 lunationSj then, contained in group (a), 



r 



16 agree throughout, 

 8 miss once J 

 4 miss iwicCf 

 2 miss thrice^ 

 1 disagrees throughout. ^ 



31 in all. 



If I were endeavoring to sustain the supposed rule, coute qui 

 coute^ as a matter of argument or opinion, instead of honestly 

 seeking for information, the uniformity of bisection in this last 

 series and the correspondence between the numbers and charac- 

 ters of the discordances, might be plausibly urged according to 

 the received doctrine of probabilities as being of great weight. 



Of the 12 lunations in group (6), 



7 agree throughout, 

 5 miss once. 



# 12 in all. 



Of the 6 lunations in group (c), 



5 agree throughout, 

 1 misses twice. 



6 m all. 



The single lunation, classed (t/), which contains but one quar- 



ter-day, misses. 



4^ 



# 





^ 



