664 
Proceedings of the Iloyal Society 
the thermometer is exposed, is that from a small surface of the 
ground under the open end of the tube, it has little effect, and the 
thermometer shows nearly the true temperature of the air. The 
screen shown in. fig. 2 is also provided with non-conducting con- 
centric tubes / and g, surrounding the thermometer bulb the same as 
in fig. 3, as these tubes prevent heat being conducted inwards 
towards the thermometer. 
It may be thought that in these screens the thermometer is too 
well protected, and that it will follow too slowly the changes in 
the temperature of the air. The changes must no doubt be to 
a certain degree sluggish. This, however, does not seem practically 
to interfere with the readings so far as has been observed. This 
sluggishness will prevent sudden local gusts of hot air from affect- 
ing the readings. An addition has, however, been made to one 
of the screens to cause it to act more quickly while the wind is 
blowing. It is shown in section at at the bottom of the draught 
tube in fig. 2. It is simply a wind deflector, made of a circular 
piece of wood shaped as shown in the sketch, and having a number of 
thin vertical radial plates fixed on it. When the wind blows it 
enters the radial passages, and has its course deflected upwards and 
past the thermometer. In this way the wind keeps up a circulation 
in the draught tube while there is no sunshine. Experiments have 
also been made to get some satisfactory form of wind-ventilating 
apparatus, which could be fixed near the to}') of the draught tube, 
to cause a rapid upward current when the wind is blowing. 
The advantage of putting the wind circulator near the top of the 
tube, is that when placed there the air does not come into contact 
with any surface before it reaches the thermometer, and therefore, 
theoretically at least, the arrangement is more perfect than the 
other. The slight heating got by having it underneath the tube 
is found practically to be of little importance. 
To combine the full advantage of wind-circulation with that of 
the draught tube the screen shown in fig. 4 has been devised. As 
before, ah is the draught tube. To the lower end of this tube is 
attached a small circular louvre box c, and to protect this box 
from the sun and rain an umbrella-shaped shade is placed over it. 
This shade consists of two circular discs of wood d and e, and a metal 
covering /, all fixed parallel to each other, and with air spaces 
