of Edinlnmjh, Session 1883-84. 
677 
these exceptional conditions for making a number of test observa- 
tions with the screens. 
For a standard of comparison, the fan apparatus already described 
was used. The first thing to be done was to test the action and 
efficiency of the draught produced by this fan. In this apparatus 
the vacuum produced is less than ^ of an inch of water, and the 
current consequently is not very strong. I therefore first compared 
its action with that of a more powerful instrument, the fan of which 
is driven by multiplying gear, and which easily gives a vacuum of 
1 inch of water, and draws a perfect hurricane of air over the ther- 
mometer. Working these two instruments alongside of each other 
in the open air on a bright day, there was not found to be any 
difference in the readings of their respective thermometers ; showing 
that for all practical purposes the draught in the simple apparatus 
gave as correct results as the powerful draught of the more com- 
plicated fan. The only advantage of the powerful apparatus was, 
that it brought the thermometer more quickly than the other to the 
temperature of the air. 
The readings of the fan apparatus were now compared with those 
given by the sling thermometer. The result was, that whenever 
there was any sunshine, the fan thermometer read lower by one 
degree than the sling, and if the sun was bright the difference was 
more than one degree, I shall presently show why these two 
methods of testing the temperature of the air should give different 
results, and why the sling should read higher than the fan arrange- 
ment ; also why the lower temperature given by the fan is nearer 
the true temperature of the air than that given by the other. 
In the following experiments with the screens the fan apparatus 
has been used as a standard, as it is, I think, more correct than the 
sling method, and also because it is much more easily worked. 
Assuming then the fan apparatus as our standard instrument, it w^as 
fitted up alongside of the screens on a lawn, in a position well pro- 
tected from any little wind that might be blowing, so as to get the 
most trying conditions possible for the screens acting correctly. The 
amount of wind during most of the experiments was very small, and 
generally from the east. At times there was scarcely a breath of 
air, and during the calm periods frequent readings were taken. In 
conducting the experiments, the fan was kept in constant action, 
