of Heat by Gaseous and Liquid Matter. 89 



vapour was introduced into the second chamber, its absorption 

 there being also determined. Finally the absorption exerted by 

 the two chambers acting together was determined, both of them 

 being occupied by the gas or vapour. 



The combination here described enabled me to check the expe- 

 riments, and also to trace the influence of the first chamber on 

 the quality of the radiation. In it the heat was more or less 

 sifted, and it entered the second chamber deprived of certain con- 

 stituents which it possessed on its entrance into the first. On 

 this account the quantity absorbed in the second chamber when 

 the first chamber is full of gas, must always be less than it would 

 be if the rays had entered without first traversing the gas of the 

 first chamber. From this it follows that the sum of the absorp- 

 tions of the two chambers, taken separately, must always exceed 

 the absorption of the tube taken as a whole. This may be briefly 

 and conveniently expressed by saying that the sum of the absorp- 

 tions exceeds the absorption of the sum. 



Table VI. — Carbonic Oxide. 



Length. 



A 



Absorption per 1 



A 



00. 



1st 



2nd 



1st 



2nd 



Both 



chamber. 



chamber. 



chamber. 



chamber. 



chambers, 



2-8 



46-6 



6-8 



12-9 



12-9 



8-0 



41-4 



9-6 



12-2 



12-9 



12-2 



372 



10-7 



12-2 



12-9 



15-4 



34-0 



10*9 



12-2 



13-4 



17-8 



31-6 



11-1 



12-0 



13-3 



36-3 



13-1 



12-6 



10-3 



13-4 





Table VII. 



— Carbonic 



Acid. 





2-8 



46-6 



8-6 



13-8 



13-3 



8-0 



41-4 



9-9 



12-7 



13-0 



12-2 



372 



110 



11-4 



130 



15-4 



34-0 



11-8 



12-1 



13-9 



238 



25-6 



11-7 



11-4 



13-1 



23-8 



25-6 



11-2 



11-2 



12-6 



23-8 



25-6 



10-4 



10-5 



12-0 



36-3 



131 



11-6 



10-0 



12-3 



Various causes have rendered these experiments exceedingly 

 laborious. Could I have procured a sufficiently large quantity 

 of gas in a single holder for an entire series of experiments it 

 would not have been difficult to obtain concurrent results, but 

 the slight variations in quality of the same gas generated at dif- 

 ferent times tell upon the results and render perfect uniformity 

 extremely difficult to obtain. The approximate constancy of the 

 numbers in the third column is, however, a guarantee that the 



