514 PROCEEDINGS OF THE AMERICAN ACADEMY. 



The next work of importance is that of Berthelot, ^^ whose method 

 (a modification of Andrews's) is so well known as to need no descrip- 

 tion. The vaporizer was placed immediately over the calorimeter and 

 heated by a ring burner, the tube for delivering the vapor passing 

 through the ring. His apparatus, although ingenious, has grave de- 

 fects, some of which have been pointed out by Schiff ^^ ^nd by Kahlen- 

 berg.^* Superheating of the vapor, the most glaricg defect, was 

 suspected by the former critic, but the experimental demonstration 

 was first made by the latter. Another source of error is direct radia- 

 tion from the burner into the calorimeter, for which, however, an ap- 

 proximate although unsatisfactory correction may be applied. There 

 was no device to prevent the vapor which came off from the liquid be- 

 fore it reached the boiling point from gaining access to the condenser. 



Passing over the work of Schall, ^^ who used the method of Favre 

 and Silbermann, we come to the work of Robert Schiff, ^^ who sought 

 to avoid the causes for error in the Berthelot apparatus by removing 

 the source of heat from the neighborhood of the calorimeter, and inter- 

 posing a small silver trap just before the entrance of the vapor into the 

 condenser. By this ingenious device, particles of unvaporized material 

 that might be carried over by the current of vapor, and most of the 

 liquid resulting from a premature condensation of the vapor, would be 

 caught and prevented from going into the condenser. Some of the 

 earlier experimenters had sought to prevent prematurely condensed 

 liquid from reaching the condenser by making a sharp bend in the 

 vapor delivery tube, so that condensed liquid would run back into the 

 vaporizer, but Schifif's device has the decided advantage that the trap 

 can be placed much closer to the condenser, so that the amount of con- 

 densation between this point and the condenser will be smaller. That 

 this was a distinct improvement is apparent in his results ; they were 

 among the best that have been made. While the method of Schiff is 

 undoubtedly better than any similar methods previously used, it is 

 objectionable in the case of liquids of higher boiling point, as Kahlen- 

 berg has shown. The danger here arises from the fact that a substance 

 of high boiling point condenses in the trap to such an extent that it 

 finally overflows and runs into the condenser. 



Another objection, heretofore not mentioned, lies in the fact that the 

 vapor, just before it reaches the calorimeter, passes through a zone, 



" Berthelot, C. R., 85, 646. 



" Schiff, Liebig's Ann., 234, 338 (1886). 



" Kahlenberg, Jour. Phvs. Chcm., 5, 215 (1895). 



" Schall, Ber. Chem. Ges., 17, 2199 (1884). 



" Schif!, 1. c. 



