64 RegnanW s Researches upon the Dilatation of Gases. 



densation, and that its value was 0.375. This coefficient was 

 adopted by all philosophers, and employed in their calculations, 

 until M. Rudberg, a few years since, showed its inexactitude. 

 By a series of experiments made with care, Rudberg sought to 

 show that Gay Lussac's coefficient was too high, and that its 

 true value was comprised between 0.364 and 0.365." 



Regnault gives at length the experiments of Rudberg, and de- 

 scribes the instrument made use of; he then goes on to give a 

 description of his own experiments, which were performed in four 

 different methods, the first of which was similar to Rudberg's. 



By the first method, the dilatation of air between 0° and 100° 

 Cent, is shown to be — 



"The mean furnished by these fourteen experiments is 1.36623. 

 The difference between the two extreme numbers is 0.00140, that 

 is to say, a little more than T „V o °f tne quantity that was meas- 

 ured. The numbers furnished by these fourteen experiments are 

 all higher than the mean, 1.3646, that Rudberg obtained in his 

 experiments, made by a similar process. I believe that this dif- 

 ference arises from the fact, that in the experiments of Rudberg 

 aspiration of the external air took place ; and it appears to me 

 difficult to operate after his manner without this cause of error 

 presenting itself, and it is evident that it must have escaped his 

 attention, for he does not speak of it. 



"The errors produced by this aspiration become more and more 

 sensible, as the volume of air operated on is smaller. It was 

 some time before I succeeded in avoiding this aspiration, and I 

 am persuaded that it exercised a sensible influence upon my first 

 experiments, making the numbers too small ; and what confirms 

 me in this opinion is, that from the moment that the aspira- 

 tion became impossible, I never obtained a number lower than 

 1.3658." 



In these experiments, it was necessary in one step of the pro- 

 cess, to immerse a glass tube having a bulb at one extremity into 



