thus vary in any substance, merely because it has transmitted or in- 

 tercepted it for a greater or less length of time. Hence he thinks it 

 essential to have recourse to some permanent rule from which the 

 results may in all cases be accurately derived, and which, when the 

 phenomena do not correspond, may lead us to the investigation of 

 some other cause. Such a law has been deduced from direct expe- 

 riments, and implies that a body placed in a medium of a constant 

 temperature, becomes heated or cooled in such a manner, that the 

 differences of its heat from that of the medium are in a geometrical 

 progression, while the times of heating or of cooling are arithmeti- 

 cally proportionate. It will readily be perceived in what manner it 

 is practicable to deduce from the two progressions mentioned in this 

 law, a third progression, which will apply to the intermediate steps 

 of any series of observations. 



This law, when adapted both to Dr. Herschel's experiments and 

 to some new ones here described, is found to apply with singular ac- 

 curacy through the three or four first minutes of increasing heat ; but 

 after this period the series manifestly varies, the increase of heat by 

 computation according to the law falling progressively short of that 

 indicated by the thermometers. The author is at considerable pains 

 to explain this anomaly, and at length ascribes it to the heat accu- 

 mulated in the intercepting body, which renders it in a manner a 

 new source of heat, the emanation from which, it must be admitted, 

 cannot but cooperate with the transmitted rays, to raise the ther- 

 mometers near it. 



If the progress of this accumulation of heat be perfectly regular, 

 its effect will be confounded with that of the transmitted rays, as 

 was actually found to be the case when a thin plate of talc was used 

 as an intercepting medium. The cause of this difference is ascribed 

 chiefly to the thickness of that medium, and in some measure also to 

 the weakness of the source of heat. It will scarcely be necessary to 

 explain the operation of these concurrent causes, it being obvious 

 that the greater the bulk of a body, the greater will be the accumu- 

 lation it admits of, and the greater the source of heat, the more rapid 

 will be this accumulation. 



The next object of inquiry is how long an experiment should last 

 for the thermometer to acquire the maximum of heating, that is, the 

 temperature of the source of heat, or medium in which it is immersed. 

 Here the experiments can be made only on direct heat, since the in- 

 termediate body containing accumulated heat, might, and probably 

 does in most cases, continue to emit this heat after the thermometer 

 has arrived at the maximum, that is, the temperature of the source 

 of heat. In the direct heat of the sun this maximum was obtained 

 in little more than 12'. 



The author hereupon examines a number of Dr. Herschel's expe- 

 riments, in which he mentions only the initial and final degrees of 

 the thermometer. After showing what the mean ratio is between 

 the degrees computed for the progression of the differences, and those 

 determined by observation, which he finds is as 13 to 10, he deter- 



