Thermodynamic Properties of Air. 11 



turnings (fine silver thread- tape) ; the whole weighs 30*024 

 grammes. I have found that this number of tubes is more 

 than sufficient to allow perfect equalization of temperatures 

 of the gas and the water. 



The heater is placed vertically along the axis of a rather 

 wide glass tube R', which acts as a receiver of the freezing- 

 mixtures. Near its lower end it is contracted into a narrow 

 neck, in which the end D of the axial brass tube is fitted 

 tightly by means of an indiarubber ring. Another glass 

 cylinder R of larger diameter surrounds the vessel R' ; its 

 bottom a consists of a varnished cork, in a perforation of 

 which the neck of R' is fitted. The glass cylinder and the 

 cork act as insulators against heat. 



When liquid oxygen was to be applied as frigoriflc medium 

 I found it economic to cool the current of air previously to its 

 passing into the copper coil P. For that purpose a small 

 metallic trough / (indicated by dotted lines) was provided, 

 which was filled with a mixture of solid carbon dioxide and 

 ether. 



In order to determine the specific heat of gases at higher 

 temperatures, for instance at -f- 100°, the same form of 

 apparatus could obviously be used with the glass receiver R ; 

 replaced by some form of steam-heating. In my first experi- 

 ments, however, I used a slightly different form (fig. 3). 

 The connexion of the brass tube D F with the cooler, the 

 thermometric arrangement, &c., are exactly like those in the 

 apparatus for low temperatures. Instead of the copper spiral 

 the heating of the gas is accomplished, as in Wiedemann's 

 apparatus, in a wide horizontal brass tube P, filled with copper 

 turnings, and soldered at a right-angle to the tube D F. 

 Both tubes are placed permanently in a brass trough R, being- 

 soldered to it. The temperature of +100° was obtained by 

 boiling water in the trough, with the use of a small gas- 

 burner G. The soldering A of the thermo-electric couple 

 was heated in a separate steam thermostat. On this occasion 

 I found (after having lost much time in search of an error 

 in all parts of the apparatus) that the wires of the couple 

 leading into the thermostat ought to be very fine, otherwise 

 the soldering would not acquire ihe temperature of -f 100°, 

 even in a copious current of steam. 



§ 8. The initial temperature of the gas was determined, as 

 already mentioned, by means of a thermo-electric couple 

 (copper-nickel) in connexion with a hydrogen thermometer., 

 It will be useful to say a few words on the mode of calcula- 

 tion of the temperature. 



