possible Fall of a Meteor into the Sun. 341 
Such extravagant conclusions only show the insufficiency of 
our data, and demonstrate the uncertainty in which the subject 
is involyed, so long as an approximation to the actual tempera- 
ture of the solar surface is wanting. 
Ts it possible to ascertain the temperature of the radiating 
surface of the sun? Ordinary observations give us the tempera- 
ture in the sun and in the shade. Suppose the temperature of 
the sun to be double its present amount, it is probable that the 
absolute temperature in the shade and in the sun would also be 
double the present amount; consequently, also, the difference 
between them. We might thus expect this difference, when due 
precautions are taken, to be a constant quantity, and to be a 
function of the sun’s absolute temperature. 
Suppose a thermometer to be enclosed in a vacuum and sur- 
rounded on all sides by matter having a uniform absolute tem- 
perature ¢: we may consider it to be the centre of a sphere, the 
interior surface of which radiates heat to it, and the balance of 
temperature to be thus maintained by reciprocal radiation,—as 
much power issuing from the thermometer on all sides towards 
the concave surface of the enclosing sphere as enters into it by 
radiation from the concave surface. There is a dynamic inter- 
change of force in constant operation. If the temperature of 
the sphere is augmented one degree, the thermometer rises until 
its radiating power increases to the same amount. If half the 
concave surface remains at ¢, while the other half increases from 
t to t+2°, the rise in the temperature will be the same as before, 
viz. 1°, because the supply to it is the same as if the whole sur- 
face were raised 1°. If ,J,,th of the surface had the tempera- 
ture ¢+1000°, the other parts remaining at ¢, we have 
1 x (+1000°) +999 x ¢ 
1000 
the resulting temperature as before. If +, 9th of the surface 
had the temperature of ¢-+2000°, then 
1 x (¢+2000°) +999 x¢_ 
" 1000 ry, 
is the temperature of the thermometer; and generally, if : of 
= a be 
ix? 
the surface of the sphere had the temperature ¢+ 2°, we have 
(t+ 2°) + (n—1)t 
i Celi ts 
the temperature of the thermometer. Hence 2° =n7°, a simple 
relation, by which we can deduce the temperature of the sun’s 
radiating surface, assuming for the present the non-absorption 
Phil. Mag. 8. 4. Vol. 19. No. 128, May 1860. 2A 
