302 Dr. H. Herwig's Investigations on the Conformity 



The difference between the figures entered in columns a and b 

 is manifest. We may remark that it seems impossible that the 

 volumes as calculated by Zeuner should be correct, as being- 

 larger than those entered under column c. Since, however, the 

 figures in columns b and c are derived from the theoretical va- 

 pour-densities, possibly the preparations examined by Regnault 

 may have possessed a somewhat different final vapour-density, 

 which would account for the discrepancy. In the first figures 

 the difference (in the case of alcohol) is, of course, too large to be 

 accounted for by this explanation. Besides, we may suppose 

 that the specific volumes calculated on the basis of the mechanical 

 theory of heat are obtained in such a circumstantial way that 

 such a variation in the fundamental data of observation (the 

 maximum tension, the total amount of heat, and the heat of 

 the fluid) as, according to Regnault's investigations, would lie 

 within the errors of observation might evoke a proportionately 

 important variation in the final result. For these reasons no very 

 trustworthy conclusions can be drawn from the comparison given 

 above in Table V. 



§8. 



As to the whole question of the superheated state of vapour, 

 I have already remarked that as yet I have not been so fortunate 

 as to derive a precise law from the course of the products^ which 

 lie between saturation and the gaseous condition. 



But some interesting conclusions may be obtained from the 

 examination of the other limit, which separates the condition of 

 vapour deviating from the laws of the ideal gas (which for short- 

 ness I will designate exclusively as the superheated condition) 

 from the gaseous condition. Of course the experiment of fixing 

 these limits accurately has never been made with success, since 

 the product pv varies too slightly in the vicinity of its constant 

 condition to allow of the differences falling without the errors of 

 observation. Generally, from the occurrence of Mariotte's law 

 (*. e. from the constancy pv for each temperature), we can only 

 infer that the eventual small differences of pv are not perceptible 

 by our instrumental measurements. But, on the other hand, if 

 the examination always shows an undoubted fluctuation beyond 

 these limits, then a further determination can be instituted to at 

 least some degree of approximation. And in the case before us 

 the observations suffice to exhibit a very unexpected result. 



Considering now the particulars of V 1 (i. e. of the volumes for 

 which the vapour first enters into the gaseous condition at dif- 

 ferent temperatures), Table II. shows for chloroform, and, still 

 more, Table III. for bisulphide of carbon, that these volumes do 

 not constantly diminish with increasing temperatures, as one 

 might have supposed, but from a certain point increase with the 



