of Edinburgh, Session 1884-85. 
203 
instruments, and of which the explanation is not obvious. As the 
readings are somewhat difficult to take accurately — the temperatures 
being liable to change before both thermometers can be read, and 
any error in the clean bulb is doubled by the process of reduction — 
these peculiarities probably arise from errors of observation, and it 
will be necessary to use the instruments for a considerable time 
before anything definite can be said. The chief objection to the 
instrument is that the black coating of the bulb is liable to be 
chipped; but as the adjustment of the amount of black surface does 
not appear to be very delicate, this will probably be easily got over. 
The readings of wet and dry bulb hygrometers have been taken 
throughout in the fan apparatus and in the two Stevenson screens. 
The dew points, as calculated from these readings by means of 
Glaisher’s Tables, present some interesting peculiarities — but as yet 
sufficient material has not been collected to enable any very definite 
conclusions to be drawn. 
It was to be expected that the fan apparatus would give results 
more nearly approaching the truth than the others, as the operation 
of drawing the air through the wooden tube can scarcely affect the 
amount of moisture in that passing over the bulbs, — at least when 
the air is comparatively dry, — and the free circulation produced by 
the fan prevents the water evaporated from the wet bulb from 
remaining in its immediate vicinity, and so indicating a greater 
amount of moisture than is really present in the atmosphere. Again, 
the passage of a large quantity of air through the apparatus should 
enable us to get a better idea of the amount of moisture in the air 
around. 
The dew points given by the fan apparatus are in all cases con- 
siderably below those given by the other instruments, being in one 
case more than 5° lower than the Stevenson. As a general rule, the 
curve of dew points for the improved Stevenson is the highest of 
the three, but it is in most cases an almost exact reproduction of 
that given by the fan apparatus. The curve for the ordinary 
Stevenson usually lies about midway between the two — on the 
whole nearer to the curve for the improved Stevenson than to that 
for the fan apparatus. On some days this curve resembles the 
others only in its general outline — the maxima and minima being 
greatly rounded off, and sometimes only hinted at. This effect is 
VOL. xm. 
p 
