176 Proceedings of the Royal Society of Edinburgh. [Sess. 
from more distant objects or from the sky on the sides of the apparatus 
can be intercepted by means of a small annular piece with wide central 
hole surrounding the bulb, as shown in tig. 3. It should be grooved on 
the outside vertical rim so as to prevent the air heated on it from entering 
the bulb enclosure. Another obvious improvement would be to make each 
of the small disks a of two thin parallel disks with air space between, in 
place of one thickish disk of wood. 
With all these precautions there still remains some secondary radiation 
which, falling on the bulb, causes the thermometer to read too high by 
an amount depending on the size of the bulb and the nature and thickness 
of the glass. No amount of air circulation can check this heating effect. 
For example, if we take two similar thermometers 
and protect one with a silver sheath, and then 
sling them in the usual manner so as to cause them 
to move rapidly through the air, we find that the 
unprotected thermometer may read as much as a 
degree higher than the other. This is due to radia- 
tion entering the unprotected bulb. The best pro- 
tection is, of course, to gild the bulb. This plan, 
though satisfactory for experimental purposes, must, 
I fear, be dismissed as unpractical under ordinary 
working conditions, since it would be impossible to 
keep all the bulbs constantly bright. 
In addition to the ordinary measurements of 
temperature by means of shaded or protected ther- 
mometers, it is important to record measurement 
of radiation of direct sunshine. This is usually done by means of a 
black-bulb thermometer; but there are advantages in the use of an 
instrument first described in my early papers on Thermometer Screens 
already referred to, and used in connection with my researches on 
Dew. (See “ Thermometer Screens,” Proc. R.S.E., vol. xii, 1883-84; also 
paper “ On Dew,” Trans. R.S.E., vol. xxxiii, 1885-86.) This instrument is 
called the radiation box, R.B., and its indications are compared with the 
readings of the black-bulb thermometer, B.B. 
One of the radiation boxes with its enclosed maximum and minimum 
thermometers is placed horizontally On a light wood frame with its surface 
at the same height as the thermometers in the screen. During the day 
the maximum radiation temperature is recorded, and the minimum during 
the night. During summer its average maximum is higher than the black 
bulb in vacuo ; but in winter it is lower owing to the low sun, so that 
