94 



Prof. J. J. Thomson. 



[Mar. 9, 



making this connexion by placing a rubber stopper in the prolonga- 

 tion of the bulb, and pushing the tube L through a hole in this 

 stopper, care being taken to push the tube right through the hole. I 

 may remark in passing that it is very desirable to adopt a form of 

 apparatus which is easily constructed ; the life of these tubes is by 

 no means long, as they are exceedingly liable to crack, especially when 

 cooling, at the end of an experiment. The apparatus I am now 

 describing is one which was designed with special reference to easy 

 construction, and fused joints are done away with in those places 

 where I have found by experiment that this course could be taken 

 without injury to the accuracy of the experiments. I have, however, 

 repeated the experiments, using apparatus in which, all the joints were 

 fused. 



The top of the tube L is fused on to the horizontal discharge 

 tube CD ; this tube is blown out into a bulb in the region where the 

 sparks pass, so that when long sparks are used they may not fly to 

 the sides of the tube. The top of the tube L, near its junction with 

 CD, is encircled by a ring burner K, and this part of the tube is 

 surrounded by an asbestos case ; by these means the steam may be 

 superheated to a temperature of 140° to 150° C. 



The details of the electrodes between which the sparks pass are 

 shown in fig. 3. For the metal parts a, b, it is necessary to use some 

 metal which is not oxidised by the steam, as a very small amount of 

 •oxidation would be sufficient to render the results nugatory. 



Fia. 3. 



c 



I have used as the electrodes, (a) brass tubes thickly coated with gold 

 with their sparking ends carefully rounded off, or (b) tubes made by 

 winding thick platinum wire up into a coil. These tubes are 

 placed in pieces of glass tubing, e, e, to hold them in position. To 

 facilitate the expulsion of air from the apparatus, it is desirable that 

 the metal tubes should not fit so tightly into the glass ones as to pre- 

 vent the steam from passing between the two. If they fit too tightly 

 to allow this, air is apt to lodge between the metal and glass tubes, 

 and if this gets driven out into the delivery tubes F, G (fig. 2) when 

 the sparks pass, it will vitiate the experiment. 



The glass tubes e, e stop short of the places /, g, where the delivery 

 tubes join the discharge tube. The discharge tube is closed at the 

 ends by two pieces of tube, h, which have their ends inside the tube 



