158 Mr. J. P. Harrison on Lunar Influence ove7' 



time of opposition, had there been ground for supposing that 

 a similar cloud-dispelling power was prevalent, though it had 

 been overlooked, at other periods of the lunation. The link 

 that was wanting to connect the two phsenomena seemed to be 

 supplied by results obtained by Mr. M. J. Johnson, the Rad- 

 cliffe Observer at Oxford, who, I found, had not only noticed 

 that clouds disappeared at other times besides full moon, but 

 had made special observations connected with the subject in the 

 years 1844, 1845, and 1846. Dm-ing this period the action 

 was found to commence about the fourth or fifth day of the 

 moon's age, and recurred with intervals up to the fourth or fifth 

 day before the conjunction*, — thus marking at both extremities 

 of the lunation the very points at which minimum mean tempe- 

 rature had been observed, and rendering it a matter of the high- 

 est probability that the depressions in the curves of mean tem- 

 perature were connected with the greater serenity of the sky, 

 and that the periods of high mean temperature would also be 

 found to depend, on an average of years, in some measure upon 

 the amount of cloud. Other circumstances which had come 

 under my own notice, or had been collected from various sources, 

 seemed to point to the same conclusion f. 



8. Supposing, then, that it were proved that the two phseno- 

 mena were connected, it is evident that the effects which have 

 been observed might be atti-ibuted to one of two causes, or to 

 both. High mean temperature, for example, might be due to 

 heat extricated upon the condensation of vapour into cloud and 

 rain — or to the law of radiation already referred to, by which in 

 certain conditions of the atmosphere heat is retained in the lower 

 strata of the air, more especially when the sky is entirely covered 

 with thick clouds. It is well known that opposite effects occur 

 during a clear state of the atmosphei'e, even in summer luna- 

 tions, when the mean temperature of the hottest day will often 

 be reduced by the action of terrestrial radiation at night to a 

 degree that could a priori have hardly been conceived possible. 



9. The two-hourly observations of the amount of cloud which 

 were taken at Greenwich, day and night, from 1840 to 1847, 

 provided the only means of testing the fact of the dispersion or 

 absence of clouds at different periods of the moon's age which 

 were attainable. Valuable, however, as these observations were 

 for ordinary meteorological purposes, they were not so well 



* In other years the cloud-dispelling power is found to be exerted 

 earlier in the lunation — seldom later than the day mentioned in the text. 

 At and before new moon the sky is often concealed by a thin veil of cloud, 

 much as on the day of the eclipse in March 1858. 



t Mr. Nasmyth also has noticed the phseriomenon of the dispersion of 

 clouds about the foiu-th day of the moon's age. 



