AND THE OTHER PLANETS OF THE SOLAR SYSTEM. 631 



however, it is sufficient for us to know that none of them can be effective unless the 

 temperature of the atmosphere decrease continually as we ascend from its lower to its upper 

 regions. 



Having made these general remarks, I will now proceed to the brief consideration of two 

 or three hypothetical cases, which will lead us immediately to the questions which form the 

 more especial objects of this communication. 



3. Conceive a hollow spherical shell, or envelope, the interior of which contains nothing 

 but the thermaniferous medium. It will be convenient with reference to our ultimate object 

 to suppose its dimensions indefinitely large. If the temperature of the interior surface of the 

 shell were maintained at a uniform temperature (t ), a thermometer placed at any point within 

 the envelope would indicate the temperature t , which would also be that finally assumed by 

 any body placed within the shell, provided the body were not perfectly diathermanous for the 

 kind of heat proceeding from the shell. It will be convenient to designate the particular kind 

 of heat here supposed to radiate from the envelope by H . Again, let us conceive heat of a 

 different kind (H^ to radiate uniformly over the interior space, and in such quantity as to 

 elevate the temperature at every point to t Q + t l ; but suppose the body (which I will call A), 

 placed within the envelope to be perfectly diathermanous for this particular kind of heat (Hi). 

 Its particles would absorb none of this heat, and therefore would not have their own proper 

 temperature (t ) at all affected by it, but a thermometer placed within the body would not 

 now indicate that temperature. For, its bulb not being diathermanous for either kind of heat, 

 the temperature indicated by it would be raised above t by the heat of the kind H l absorbed by 

 its bulb, till this quantity should be equal to that emitted by the bulb to surrounding particles 

 of the body. The indicated temperature would, therefore, be intermediate to t and t n ■+ t v 

 A similar effect would be produced by the body A on the thermometer placed without it, but 

 sufficiently near to be within the sphere of its cooling influence. 



4. Taking another imaginary case : Suppose the general temperature t of the envelope to 

 be derived as in the first instance from H Q alone ; and conceive the body A to be a spherical 

 shell, the center of which may be supposed, for greater distinctness, to coincide with that of the 

 general envelope. Conceive also the center of A to be a source from which heat (H t ) emanates 

 uniformly in all directions, and such that it can be transmitted through the matter of the shell 

 only by conduction, in the same manner as H . The temperature (t,) of the internal surface 

 of the shell will rise till its excess over the temperature (r 2 ) of the outer surface of the shell 

 shall be sufficient to enable the shell to transmit in a given time a quantity of heat equal to 

 that which falls on its inner surface in the same time from the central source ; while t 2 must 

 exceed the temperature (t ) of the external space sufficiently to enable the same quantity of 

 heat to radiate in the same time into surrounding space. In this state the temperature will be 

 steady, decreasing from Tj to t 2 at a rate depending on the quantity of heat transmitted in a 

 given time, and on the conductivity of the matter of the shell ; and since the rate of decrease 

 would be independent of the thickness of the shell, it is manifest that the greater that thickness 

 the greater must be the excess of tj above t 2 . 



Vol. IX. Paet IV. 81 



