196 Professor Silvanus P. Thompson [May 8, 



glass to show a greenish fluorescence near the kathode end. In the 

 seventh tube the luminous column has subsided into a few greyish- 

 white nebulous patches, the dark space round the kathode has greatly 

 expanded, and the glass of the tube has now begun to show a yellow- 

 green fluorescence. The exhaustion has been pushed so that only about 

 ^TTWU ^^ ^^^^ ^^ *^® original air is present. In the eighth and last 

 tube only one or two millionths of the original air have been left, 

 with the result that the tube now oflfers an enormously increased 

 resistance to the passage of the discharge. All the internal flickering 

 nebulosities have vanished ; the tube looks as though there were no 

 residual air within. But now the glass itself shines with a fine 

 yellow-green fluorescence which is particularly bright in the region 

 around the kathode. Were the exhaustion to be carried much further 

 the spark from this induction coil would no longer pass, so high 

 would the resistance become. All these successive stages up to the 

 last can be shown in one and the same tube attached to a modern 

 rapid air pump. But for the proper production of the high vacua of 

 the last stages, where electric shadows are alone produced, nothing 

 short of a mercurial pump, either in the form invented by Sprengel 

 or in that used by Geiesler (or one of the recent modifications) will 

 suffice. 



The phenomenon of fluorescence of the glass, which manifests 

 itself when the exhaustion has become sufficiently high, was known 

 in a general way as far back as 1869 or 1870. The tube next to be 

 shown is a modern reproduction of a tube used at that time by 

 Hittorf, of Miinster. It dififers from the tubes last shown by having 

 a bend in it. Hittorf observed that when such a tube is ex- 

 hausted sufficiently highly to give at the kathode the characteristic 

 greenish-yellow fluorescence, this greenish-yellow fluorescence re- 

 fused to go round the bend. It might appear at one end or the other, 

 according to the direction in which the discharge was being sent, but 

 would not go round the bend. The efl'ect was as if the discharge 

 went in straight lines from the bit of wire that served as kathode to 

 the walls of the tube. Indeed shadow effects were observed by him, 

 and by Wright, of Yale, and afterwards independently by Crookes, 

 who greatly extended our knowledge of the facts. We may take this 

 fact, that the fluorescence caused by the kathode will not go round a 

 corner, as the starting point of the memorable researches of Crookes 

 on radiant matter a score of years ago. 



Before you are several tubes which illustrate the researches made 

 by Crookes. The first is a simple glass bulb into which are sealed 

 the two electrodes — the anode, by which the current enters, ter- 

 minating in a bit of stout aluminium wire ; the other, by which 

 the current leaves, called the kathode, terminating in a small flat 

 aluminium disk. The glass bulb was itself highly exhausted — how 

 highly we shall presently see. From the flat front surface of the 

 kathode, when sparks are sent through the bulb, a sort of back- 

 discharge takes place in a direction normal to the surface. This 



