10 A. W. Witkowski on the 



a silver collar r, fixed tightly in the brass tube D F, just 

 above the end of the glass tube C. 



In the axis of s, some 3 or 4 millims. from its mouth, 

 there is fixed one of the solderings B of the thermo-electric 

 couple, composed of very thin wires of copper and nickel. 

 The wires pass freely through the brass tube D F ; near its 

 upper end they are cemented by shellac into a capillary glass 

 tube, the end of which can be seen in fig. 2, protruding over 

 the end of D F. The capillary is fixed here by a cork, 

 tightened by means of marine glue. In order to protect the 

 cement against the cold the end F of the brass tube is dipped 

 in water contained in a small glass beaker (indicated by dotted 

 lines). One of the wires of the couple (copper) leads to the 

 galvanometer G, the other to the heater, where the second 

 soldering A is in contact with the bulb W of the hydrogen 

 thermometer. 



This construction of the connecting piece was devised 

 in agreement with the remarks of § 5. The silver tube s, 

 being a good conductor of heat, assumes a temperature which 

 differs very little from that of the heater ; the gas is thus 

 conducted into the cooler through a channel of nearly its own 

 temperature. At the same time the polished silver tube is a 

 bad radiator, and therefore it prevents effectually any inter- 

 change of heat between the heater and the cooler, with the 

 exception of that due to the air current. 



Secondly, the silver tube s forms part of the thermometric 

 arrangement, constructed upon the principle of aspiration. 

 The air passes through the relatively narrow channel s with 

 a sufficiently great velocity, and it communicates its tempera- 

 ture very quickly to the soldering B of the couple. In 

 order to shield the soldering against the radiation of the 

 cooler S, the temperature of which is generally very different 

 from that of the gas, it is advisable to place the soldering at 

 some distance from the mouth of s, so as to reduce as far as 

 possible the cone of rays passing from B to S. This radiation 

 is also intercepted by a small silver bridge, put across the 

 mouth of the tube s, which serves at the same time to keep in 

 place the soldering B, which is tied to it by a bit of cocoon 

 fibre. 



The cooler S is made entirely of silver ; it is quite similar 

 to that employed by E. Wiedemann in his researches on the 

 specific heat of gases. Four vertical silver tubes are connected 

 by short pieces of silver tubing, alternately near their upper 

 and lower ends. They are filled with compressed silver 



