A Nigh-Frequency Alternator. 299 



provided that b is small and greater than C. The distance b 

 satisfies this condition, as the thickness of the layer across which 



the cathode-fall of potential acts is less than — . 



It is thus seen that the theory leads to a consistent expla- 

 nation of many of the phenomena which accompany a 

 continuous current through a gas. 



There are several other points on which further comparisons 

 might be made between the theory and experimental results. 

 For example, the field of force between the electrodes at a 

 distance a apart, when a current is passing through the gas 

 at pressure j), ought to be double the intensity at corresponding 

 points when the same current is passing between electrodes 

 at a distance 2 a apart, and the pressure J/>. Experimental 

 investigations are at present being made with a view to 

 obtaining further results which may be easily compared with 

 the theory. The most important of these consists in tracing 

 the changes in the field which occur as the current rises. 

 Some unexpected difficulties arose in making these determi- 

 nations, so that the progress of the researches has been some- 

 what delayed. 



XXX. A High-Frequency Alternator. 



By W. DUDDELL * 

 [Plates V. & VI.] 



IN view of the increasing interest taken in high-frequency 

 alternating currents, especially in connexion with wireless 

 telegraphy, a description of a small alternator designed by 

 the writer in 1900 and of the difficulties encountered in its 

 construction may prove of use to other experimenters on this 

 subject. 



The alternator was made for some experiments on the 

 resistance of the electric arcf. As the experiments pro- 

 gressed, and it was found that with each increase of frequency 

 the solid arc behaved more and more like an ordinary resist- 

 ance, the alternator was altered until finally the very high 

 frequency of 120,000 ~ per second was attained. The 

 inductor type of alternator was chosen ; the final design is 

 shown in plan and elevation in PL V. figs. 1 and 2. 



The inductor, 6 cms. in diameter, was built up of 53 disks 

 cut out of ferrotype plate, the iron being about 0*007 inch 

 thick, enclosed between two plates 0'019 inch thick, the 



* Communicated by the Physical Society : read December 9, 1904. 

 f Phil. Trans. A. vol. cciii. p. 305. 



