182 THE ANTARCTIC MANUAL. 
If R be the term of cooling in still air, or the fundamental term 
of cooling, and r be the occasional term of cooling when the air is 
moving with any velocity v, then Leslie gives the following expres- 
sions for this velocity : 
20 R--r 
30) Or 
v= in feet per second ; 
R-r 
or, y= 4} in miles per hour. 
Converting into metrical units, we have 
R-r 
v = 2°032 
in metres per second. 
It is right to observe that R in Leslie’s equations is the term of 
cooling of his flask of water when suspended “ owt of doors, on a calm 
evening.” 
Difference between a Calm Indoors and a Calm Out-of-Doors.— 
The preceding equations give the diminution of the term of cooling 
produced by sensible wind as compared with a calm, both being 
out-of-doors. It does not appear that Leslie distinguished between 
a calm indoors and a calm out-of-doors. 
Returning to Table XI. at page 127, we find the term of cooling of 
the thermometer to be in round numbers 65 seconds in the still air 
of a room. A number of observations were made with the same 
instrument in the open air in very calm fine weather. The method 
of the “ half-fall”’ as exemplified in Table XIII. was used, the excesses 
of temperature used being 12°, 10°, 8° and 6° C. The experiments 
were made in Edinburgh on September 16, 1894, during very fine 
anticyclonic weather. All the afternoon the air was perfectly calm. 
The smoke from chimneys went straight up and indicated no hori- 
zontal component of motion, 
In the following table the “half-falls” for different initial 
excesses at different times during the afternoon are given, which 
show the extent of their agreement. The mean “half-fall” is then 
101 
=" 
This table shows well what great differences may exist in the 
calmness of calm air. Nobody doubts that such motions do occur; 
otherwise it would be impossible for the permanent difference of 
temperature which is found generally to exist at different elevations 
in the atmosphere to be maintained; and it is interesting to have a 
means of gauging them. 
In order to obtain a standard of measurement, a number of 
converted into the term of cooling by multiplying by 
