the radiant heating effects from terrestrial sources igi 
top of a box, the front of which was open, so that the glass 
screen could be applied to it or not, as required. When the 
screen was not used the box would acquire more heat, and 
radiate it to the bulbs in a small degree ; which affecting them 
in the inverse ratio of their diameters, would diminish the 
ratio of their risings. That this diminution was very trifling, 
and not at all sufficient to account for the observed difference 
of ratio will be evident, because the ist set was made with- 
out employing the box, the thermometers being suspended 
at a distance from any object which could radiate heat to 
them ; and in this set the difference of ratio is quite as con- 
spicuous. This remark applies likewise to the possible com- 
munication of heat by the air. 
(lo.) We must also take into consideration the effect due 
to the glass screen. When we consider the two bulbs as 
heated only by that part of the radiation which is transmitted 
through the screen, the screen may be regarded simply as a 
third body placed near the two bulbs ; and whether it pos- 
sesses a higher or a lower temperature, there will be a ten- 
dency to bring all three to an equality in proportion to the 
difference of temperature, and in the bulbs, dependent on 
their diameters modified by the state of their surfaces. This 
effect arises from simple radiant heat, whilst that derived 
from the luminous hot body, is evidently following a different 
law with regard to the surfaces. It will easily follow from 
what has been already shown, that such a secondary heating 
effect will be of a kind tending to diminish the ratio other- 
wise obtaining between the effects on the two bulbs. If the 
effect were of a cooling nature, the same thing would also 
take place : for I ascertained that the radiating pow^ers of the 
