679 
of Edinburgh, Session 1883-84. 
be expected to act as correctly as the other, because the louvre 
boards of the screen become heated by radiation, and the air passing 
over them gets warmed before arriving at the thermometer. 
In the construction of screens it is necessary to shade the thermo- 
meter as much as possible from all radiation. It ought not to see 
anything out of its screen. A small amount of exposure even to 
the grass underneath shows its effect in an increase of temperature. 
This could be observed when testing the screen fig. 4. If the bottom 
was taken out of the box, to allow of a freer circulation, the tem- 
perature always rose a little. On the other hand, if the louvre box 
was made too close by the addition to it of complicated louvre 
boards, the flow of the air was retarded, more heated surface was 
given to warm the air before it arrived at the thermometer, and in 
this case also the thermometer read too high. The arrangement 
shown in the sketch is the one which as yet has been found to give 
the best results. The thermometer is so protected that no radiation 
from objects outside can fall directly on it, while the free circulation 
of the air is but little checked. Perhaps an improved form of box 
for this screen might be made of double louvre boards, but as this 
would increase the outside diameter of the box, and consequently 
the amount of surface over which the air has to pass ; it would also 
greatly reduce the velocity of the air when no wind was blowing, 
and only the draught tube acting; it is therefore very doubtful 
whether any advantage would be obtained. 
Prom the experiments on the temperature of large and small bodies 
already detailed, we have seen that large bodies become more 
highly heated than small ones, when exposed either to the direct rays 
of the sun, or only to diffused radiation. Prom this we get some 
guidance for the construction of thermometer screens. As large 
bodies are less cooled by air passing over them than small ones, 
it follows that the louvre boards in the thermometer box ought to 
be as large as possible, because the air in passing over large louvres 
will carry less heat into the box than if it passed over small ones. 
The large size of the louvres will also prevent heat being conducted 
to the inside of the box. Again, if we wish to lessen the radiation 
from any particular area, wq must reduce all objects exposed in 
that direction to as small dimensions as possible, and coat them with 
some substance which is a good absorber of radiant heat. Blackened 
