428 Mr, SpoUiswoode [May 21, 



the positive and the negative ends of the discharge displays itself. 

 This dissymmetry increases with the progress of the exhaustion. 



But passing over, for the present, these features of our subject, I 

 wish to draw your attention to a peculiar condition of the discharge 

 which, having studied with much care, we have turned to account as 

 our special method of research. It is as follows : If a continuous source 

 of electricity, such as a Holtz machine, be used, and the terminals of 

 the tube be connected with the main conductors, or poles, in the usual 

 way, the discharge will pass through the tube in a condition which to 

 all direct observation appears to be continuous in respect of time ; 

 although the researches of Mr. De La Eue, and of others, alike point 

 to the conclusion that the discharge is in every case discontinuous 

 and disruptive. In this condition the discharge is indifferent to the 

 presence of a conductor, or even to that of a charged body, such as an 

 electrophorous, or a Leyden jar. If the latter be brought near enough 

 to discharge itself on to the tube, the luminous column will, it is 

 true, exhibit a momentary flutter, but will show no other sign of 

 susceptibility. This momentary flutter is, nevertheless, worthy of 

 being noticed, as it will reappear at a later stage of our investigation. 



If, however, one of the connections between the machine and the 

 tube be broken by a small interval of air, over which the discharge 

 must always leap in the form of a spark, the luminous column imme- 

 diately becomes sensitive to the approach of a conductor. This 

 break in connection, or air-spark, may be made either in the wire 

 leading from the positive, or in that leading from the negative pole 

 of the machine — in other words, in that leading to the positive or to 

 the negative terminal of the tube ; and it will be convenient to speak 

 of these two dispositions as the positive air-spark and the negative 

 air-spark arrangement respectively. When an air-spark is used the 

 discharge will be described as intermittent ; when it is not used it 

 will be called continuous, although, having reference to a remark 

 made above, the latter term can strictly be used only in a qualified 

 sense. 



This condition of sensitiveness is that which was mentioned above 

 as having been the subject of our special study. The general fact of 

 sensitiveness in an electrical discharge had been noticed by previous 

 observers ; but its connection with intermittence, and the laws which 

 regulate it, do not appear hitherto to have attracted the attention 

 which they deserve. 



In order to examine this condition of the discharge, let us begin 

 with a tube of moderate exhaustion, which presents a column of light 

 from the positive terminal through the greater part of its length ; 

 then a blank sj)ace ; and lastly a halo of light enveloping the negative 

 terminal. If an air-spark be now introduced into the j^ositive part 

 of the circuit, the column will lengthen and a23proach the negative 

 terminal, and it will at the same time contract laterally, and become 

 narrower. If a conductor, such as the finger, be now made to approach 

 the tube the column is repelled ; and if the conductor be brought still 



