24 Scientific Proceedings, Royal Dublin Society. 
of fall of temperature in one hour, from 8 p.m. to 9 p.m., for the 
four years 1900 to 1903 :— 
Monruty Normats. 
Dew- Point Relative Cine Wind- Fall of 
Montu. : at Ce at} 48 at Noles pedo 
p-m. p.m. p-m. p- 
és Per cent. Per cent. M. per H. 5 
Janes. cs! ve 49-5 45 33 5°d IL) 
Bebe es 54-0 53 46 5:3 1:6 
March,. . 52-1 59 34 4:3 iLo7/ 
Ape, 5 6 49-1 66 28 4-0 1°5 
Maynareiie 38-0 56 12 4-0 1°6 
June, .. 34-0 61 I7/ 4°5 1:4 
Ji, 2 - 33°3 58 13 4-0 1:8 
ANWR, gc 33°8 48 12 4-6 2°0 
Selon 0 6 38°6 48 20 51 2°4 
Octen ees 40-9 44 30 6:0 3-2 
INOVes ne 41°4 38 18 6°5 2°5 
Dec., . . 50°1 48 39 52 2-2 
Such a table as this can, of course, only be expected to indicate 
in the roughest way the results we are seeking. In fact, it would 
only be admissible if the clouds remained the same in quantity 
month by month, and also occupied the same areas of the sky. 
For clouds may be expected to reflect as well as to absorb heat from 
the Earth; and while the absorption may be at its lowest when the 
zenith is clear, both absorption and reflection will exert their maxi- 
mum efiect when the zenith is clouded. It is pretty certain that 
the disturbing influence of the clouds may at times completely 
override the effects due to either vapour-tension or humidity; and 
yet, in spite of so obvious an objection, it is generally the case that 
averages of this sort are relied upon to demonstrate supposed 
physical processes in the atmosphere. To avoid being misled, then, 
