28 Scientific Proceedings, Royal Dublin Society. 
Our previous formula connecting the relative humidity and 
the fall of temperature applies here also. Since the actual average 
quantity of water-vapour as indicated by the dew-point is pretty 
much the same for each item of relative humidity, it follows that 
the regular decrease in the fall cannot be due to that quantity. 
We see, however, that the velocity of the wind decreases as the 
humidity increases; and it might be thought that this'is a pos- 
sible principal cause of the variation in the fall of temperature. 
Actually it is not so, because an inerease in the velocity of the 
wind after nightfall almost invariably means a temporary elevation 
of temperature; while the low minima occasionally registered 
during the winter occur invariably on still nights. But since 
it is important to determine what the effect of the wind really 
is, the various velocities have been compared with the other 
elements :— 
Manes cae Ea Hee fence nee ee 
‘ 8 p.m. 8 p.m. 8-9 p.m. Clear Nights. 
M. per H. 5 Per cent. Bs 
0—1 09°9 57 2°4 39 
1—2 40-6 57 2°4 37 
2—3 391 5) 2°3 100 
3—4 39°4 53 2°0 118 
4—5 3671 50 1:9 121 
5b—6 36°1 46 1-6 108 
6—7 36°5 39 2°0 43 
7—8 36°6 40 2-1 55 
8—9 36°7 33 2°4 24 
9-10 88°2 37 2°3 16 
10 and over. 371 40 2-4 27 
It seems from this that as the wind-velocity increases from 
zero to about six miles per hour, the fall of temperature will gradu- 
ally diminish ; but for further increments of velocity, the fall will 
gradually increase. A partial explanation of this is to be looked 
for in the nocturnal temperature gradient. Close to the surface, 
