of Edinhurgh, Session 1883 - 84 , 
691 
nearly Avith those got by slinging. From other experiments made 
in different states of the weather, it is found that there is not muoli 
advantage got by slinging the thermometers, either silvered or clean, 
in sunshine, unless there is extremely little wind at the time. Almost 
exactly the same degree of truth can he got by simply shading the 
thermometer from direct sunshine, and from the sky radiation. 
Another experiment to test the heating effect of the sun on the 
sling thermometer was made in the following way : — An ordinary 
thermometer was placed near the mouth of the suction tube of the 
powerful fan apparatus. First the thermometer was carefully 
shaded from all radiation. The thermometer under these conditions 
agreed with the thermometer in the other fan apparatus. The open 
mouth of the tube was now directed towards the sky, so as to allow 
part of the sky radiation to fall on the bulb. The effect was to 
cause this thermometer to read a little higher than the other, and 
when the tube was directed so as to allow the sun’s rays to fall on 
tlie bulb, its indications Avere more than a degree above the other 
fan thermometer. This shoAvs us that even a poAverful draught of 
air Avill not compensate for the exposure of the thermometer to the 
sun, and confirms the conclusion Ave have come to that the readings 
of the ordinary sling thermometer are at least one degree too high. 
Temperature Ohservations without Screens. 
We have seen that, in taking the temperature of the air, there are 
different methods of operating. First, Ave may surround the ther- 
mometer with a screen to catch all radiant heat, and prevent it 
falling on the thermometer, as is done in the Stevenson screen 
and in the one shown in fig. 4, The disadvantage of this method 
of Avorking is that the air in passing over the louvre boards gets 
heated before entering the screen, and coming into contact with the 
thermometer ; and though in the screen fig. 4 the louvre boards are 
not heated by direct sunshine, and the circulation is kept up 
independent of the Avind, still it can scarcely be expected to give 
correct readings, but must always be a little too high. A second 
method of operating is to cut off all radiation as far as possible, 
isolate the thermometer from all conducted heat, and bring to it air 
which has never been in contact Avith any solid, and has not there- 
fore got heated. This method has been employed in constructing 
