through Rarefied Gases. 3G1 



taiiiing the metallic electrode and two cylinders connnunica- 

 ting with each other by narrow channels and about G centims. 

 wide. 



27. I have already mentioned that a continuous transition 

 may be established from positive to negative light. This can 

 be shown in the following way. I have said that a ra])id di- 

 minution in the cross section in the direction of the positive 

 current produces a negative glow at that point. If in a tube 

 presenting such rapid changes, the variation in the cross sec- 

 tion is gradually diminished^ so that the tube gradually as- 

 sumes the shape of a tube of uniform width, the negative 

 glow gradually assumes the shape of a positive stratification, 

 and the dark space which always forms the boundary to the 

 negative glow gradually passes into the dark space which 

 separates the different stratification. 



The same result may be obtained wdth one and the same 

 tube by varying the pressure of the gas. If the variation in 

 cross section is not too rapid, the negative glow extending 

 from the narrow into the wider part gradually loses the cha- 

 racteristic properties of the negative glow on exhausting ; and, 

 finally separating entirely from the narrow part and passing 

 into the wider parts, it forms the first positive stratification. 



I may state these results in the following words ; but as a 

 matter of precaution, I limit the statement to the longer stra- 

 tifications, excluding the short and oscillating ones, which can- 

 not easily be subjected to accurate observation : — 



Each stratification of the positive glow corresponds to the 

 so-called negative glow which surrounds the negative elec- 

 trode ; the stratified positive discharge is therefore only a suc- 

 cession of negative glows. 



28. We may also look at the negative glow as a simple 

 stratification. 



This new way of looking at the negative light allows us to 

 draw many conclusions, especially if we combine with it another 

 law. 



29. This law may be stated in the following words : — The 

 interval between the ends of two stratifications for a given 

 pressure and intensity of discharge depends entirely on the 

 number of the stratification as counted either from the pri- 

 mary or secondary negative pole. I assume here that the tube 

 is of the same width throughout. 



The intervals between the diiferent stratifications need not 

 be the same. The interval, for instance, betw^een the negative 

 light and negative end of the first positive stratification is 

 always larger than the interval between two positive stratifica- 

 tions, if the tube is of the same width throughout. 



