1894.] On some Phenomena in Vacuum-tubes. 231 



termined number of bright and dark bands, as it appeared to me this 

 would be the first step in examining the laws which govern their pro- 

 duction. In this I succeeded at a very early stage. But, in order to 

 make the proofs more convincing, other experiments were entered 

 upon which brought to light a number of new points of considerable 

 interest; arid many of these experiments appeared to throw con- 

 siderable light as to the origin of the bands. 



The object of this paper, therefore, is, first, to show the methods 

 by which a definite number of bright and dark bauds can be pro- 

 duced in a vacuum-tube ; and, secondly, to describe a number of 

 interesting phenomena which have a bearing on the production of the 

 bands iu general. 



The first step is to describe the apparatus employed in making the 

 experiments. A small direct-current motor is coupled directly to the 

 smallest sized Pyke and Harris laboratory alternator. The alternate 

 current produced can be made to vary its E.M.F. according to the 

 speed given to the motor and also by varying the exciting current. 

 The E.M.F. of the alternate current can be made variable from % to 

 100 volts, and the maximum current which the machine is intended 

 to give is 3 amperes. The pressure of this current is raised by 

 means of a Pyke and Harris oil transformer or with a Salomons and 

 Pyke combination transformer. Sometimes one form of transformer 

 is used and sometimes the other. It is usual to give the exact 

 electromotive forces of the current employed, as well as its periodi- 

 city, but in the following experiments it is not necessary that this 

 should be known accurately, from the very nature of the experiments, 

 because, as already mentioned, the experiments were started with ex- 

 ceedingly slow alternations and a very low E.M.F., which were 

 gradually raised, the phenomena being watched throughout as they 

 varied. 



It is essential that the speed of the motor should be under complete 

 command, and that the speed should be made variable during any 

 experiment. To attain this end I employ a Kelvin rheostat and a 

 Wirt rheostat. 



All the tubes employed, so far as their exterior form is considered, 

 may be regarded as practically of one type; long tubes of varying 

 lengths and large diameters, many of them containing small devices 

 which consist of little glass disks, glass rods, and other arrangements 

 for modifying the nature of the electric discharge, and, for want of a 

 better name, I term them " deflectors." Although, perhaps, the 

 word "moderator" is more applicable, it has another meaning, and 

 its employment here might give rise to confusion. 



In many of the experiments to be described the bands appear to 

 be repelled from some part or parts in the tube towards the elec- 

 trodes. I feel some difficulty in selecting language suitable for de- 



