with Mercury contained in Tubes. 515 



The experiments on this region are difficult, because even 

 quartz tubing will not withstand the battering and explosive 

 action of the sparks. Above AC the life of a quartz tube is, 

 as far as one can tell, unlimited ; below AC, i. e. when the 

 luminosity ceases while the mercury thread is broken, the life 

 is short, and becomes shorter the nearer we approach the axis 

 of electromotive force. Another experimental difficulty is 

 that the lamps are always liable to go out, and cannot be kept 

 alight long enough to be studied in detail. The observer may 

 easily over-estimate the regularity of the luminous phenomena 

 met with in this region if he judges by the appearance of the 

 stroboscope and not by the image of the sparks shown, after 

 reflexion from the revolving mirror, on the screen. Even 

 when the stroboscope is showing a well-defined frequency 

 the appearance on the screen may be very irregular. If the 

 stroboscope rings are marked out in equal angular parts of 

 black and white, so long as the periodicity is fairly regular, 

 the disk may appear still ; the mirror is a much more severe 

 test of regular periodicity. 



It is only when very narrow tubes are employed that the 

 periodicity is sufficiently well marked as to allow of con- 

 tinuous study of the effects in the rotating mirror. With 

 wider tubes the effects are irregular, though sufficiently 

 periodic to allow of frequency determinations with the 

 stroboscope. Even with these tubes, however, the lamp on 

 starting (either by external heating or by the use of a large 

 current) exhibits very regular phenomena. These initial 

 regular phenomena may be regarded as constituting the 

 typical fan effect photographed in tigs. 9, 10, and 11 (PI. VI.). 



In fig. 9 the time signals shown faintly near the lower 

 edge of the picture are *006 second apart. The tube was of 

 soda glass (9 cm., 'I mm. bore). The current was obtained 

 from a dynamo, the electromotive force being 108 volts, the 

 resistance being 37 ohms. Horizontal lines on the picture 

 represent opacity or refractive defects in the tube. On the 

 negative faint vertical marks are seen towards the right of 

 each fan pattern due to small changes in the current owing 

 to the electromotive force of the d3 T namo varying as the 

 armature revolves. This point will again be referred to. 

 It will be seen that the luminous interval occupies about a 

 sixth of the whole period. Some photographic experiments 

 undertaken in order to record the motion of the mercury 

 in the fan effect, apart from the luminosity, have not proved 

 successful ; but from the occasional presence of a necklace 

 pattern in a series of fans seen in the revolving mirror 

 one may be assured that even in the typical fan effect 



