404 Prof. R. Clausius on the Behaviour of Carbonic Acid 



these differences were really of so much importance as to jus- 

 tify such an alteration of the formula. This question, upon 

 closer consideration of the matter, 1 believed must be answered 

 in the negative. 



For the differences change their signs in quite a striking 

 manner : at 13 0, 1 the greatest difference is negative, at 

 31 0, 1 positive, at 64° negative, and at 100° again positive. 

 So frequent a change of sign does not make for the suppo- 

 sition that the cause of the differences lies in the formula, but 

 rather that it is to be sought in errors of observation; and in 

 the case in question such errors are very well conceivable, even 

 with the most careful observation. The volume of the highly 

 condensed carbonic acid was measured in capillary tubes. 

 Now, when it had become so small as to amount to only one 

 four- or five-hundredth part of its original magnitude, errors 

 might easily occur in the reading, which, though absolutely 

 very small, were relatively great enough to cause, in the for- 

 mula for p, of which the value very rapidly changes with v 

 when the values of v are small, differences from the quantities 

 in the Table. 



Besides, the air-manometer used for the determination of 

 the pressure consisted of a capillary tube, in which the air, 

 when the greatest pressures occurred, occupied volumes so 

 small that a slight error of observation must have exerted a 

 very great influence upon the pressure deduced from the ob- 

 servation. 



Further, it is to be remarked that Andrews, in deducing the 

 pressure from the data supplied by the air-manometer, started 

 from the hypothesis that air follows Mariotte's law up to the 

 greatest pressures employed in the experiments, which reached 

 more than 200 atmospheres. But this, it is well known, is 

 not the case ; at such pressures considerable deviations take 

 place. I at first tried to take advantage of the observations 

 of Cailletet and Amagat on the compression of nitrogen, in 

 order to correct "the pressure-quantities deduced from the ma- 

 nometer-indications ; but I found that the results of their 

 observations are not sufficiently accordant with each other to 

 be employed with safety for such a correction. Hence I have 

 simply quoted in the Tables Andrews's values of p. 



Lastly, I must call attention to a peculiar distinction which 

 appears in the differences between the observed and the cal- 

 culated values of p : namely, in the older series the differences 

 are almost all positive, and in the newer almost all negative. 

 This also makes for the hypothesis that the differences origi- 

 nate rather in the circumstances affecting the experiments 

 than in the formula. 



