1821. 2 Causes of 'Calorific Capacity, Latent Meat, %c. 439 



TABLE continued. 



I need make no comment on the very striking, I might almost 

 say perfect agreement between the numbers observed and com- 

 puted from the theorem I have given. While Dr. Ure's theorem 

 is often 2-i- inches, and towards the end as much as 4 and G inches 

 in error, in no one instance does mine deviate 2^- inches ; and 

 were the observations perfectly correct, 1 think it would never 

 err at the most more than about an inch. For about the tem- 

 perature 285°, when the error is the greatest, the experiments 

 disagree with one another so much that little dependence can be 

 placed on them. Thus the difference between the tension at 

 285'2° and that of 287*2°, two degrees above it, amounts to only 

 2-6 inches; while the difference for 1*4° beneath it is equal to 

 4*5 inches. Now T4° : 2° :: 4-5 inches : 6'6 inches. Therefore, 

 taking the difference between the two lower temperatures as 

 correct, and following only the ordinary rate of common propor- 

 tion, the difference between the tensions of 285-2° and 287-2° 

 should be 6-6 inches ; and we know, according to the laws of 

 the increase of tension, the difference should exceed this if the 

 first difference be correct, yet Dr. Ure's experiments give only 

 2*6 inches. Might we not consequently venture to say that the 

 greater part of the deviation, 2-12 inches, at the temperature 

 285-2° does not belong to the theorem but to the observation? 

 However, I frankly confess, that for want of having the whole of 

 Dr. Ure's observations when I investigated this theorem, there is 

 a possibility, or, perhaps, a probability, of my not having deter- 

 mined the arbitrary constants to a very great accuracy ; and, 

 therefore, it is possible the theorem might not be perfectly con- 

 sistent with truth in every part of the table; yet I think it may 

 be safely depended on throughout the whole of Dr. Ure's range 

 of observation to an inch, or at most an inch and a quarter. 

 Even to a tension of 600 inches, or 20 atmospheres, I think its 

 error would scarcely exceed 5 or 6 inches. In the higher 

 ranges, where Dr. Ure's theorem begins to diverge from experi- 

 ment very rapidly, the agreement may be said to amount to a 

 coincidence. For a few degrees above the boiling point Dr. 



