410 H. V. Gill — Electric Discharge in Geissler Tubes. 



near one end. After a few sparks had taken place the interior 

 was strikingly divided throughout its whole length into little 

 heaps quite close together. When the tube is looked at from 

 one side, at the moment when the spark passes it is noticed that 

 the powder rises a considerable distance at the heaps and then 

 falls back. When the tube containing the powder is narrow, 

 the lines of powder almost reach to the top. The result sur- 

 prisingly corresponded with what we had looked for. In order 

 to obtain a permanent record the following simple means was 

 employed : A strip of photographic paper was placed in the 

 tube with a small amount of powder on it. When the sparks 

 had taken place the powder was thrown into position. The 

 whole was then exposed to the daylight : the part occupied by 

 the powder then appeared as white in a black background. 

 Fig. 1, Plate IV, shows a typical example. (This way of ob- 

 taining records is very useful, and has succeeded well in the 

 case of other phenomena, such as the magnetic lines of force, 

 as exhibited by iron tilings.) 



The following are some facts about this experiment. 



The greater the intensity of the spark, and the greater the 

 capacity of the condenser, the farther apart are the lines. Thus 

 we obtained a series of results with condensers of varying 

 capacities. This law is well illustrated by the series of pho- 

 tographs. 



It did not appear to matter whether the end farther from the 

 spark was open or closed — the spark end was open in every case. 



The waves could be reflected from the farther end of the 

 tube into a second tube in which the same result was obtained. 



That this result is due to air waves is clear from the 

 researches of Mach. We confirmed it in a simple way. Some 

 lycopodium powder on photo paper was placed in an ordinary 

 straight lamp chimney, one end of which was 2'8 cm in diameter, 

 the other 3'6 cm . A piece of paper was stretched tightly over 

 one end and tapped sharply; the vibrations thus produced 

 caused the powder to divide up into heaps a few millimeters 

 apart. As we should expect, these heaps were farther apart 

 when the paper was over the larger end of the tube. In order 

 to assure ourselves that the formation of these heaps was not 

 due to the current of air caused by the spark as distinct from 

 the vibrations, we replaced the paper in the last experiment by 

 a bit of cloth ; on tapping the cloth no movement whatever 

 was produced in the powder, as would have been the case had 

 the puff of air caused by tapping the cloth been the cause of 

 the division of the powder into lines. 



We have dwelt on this experiment, as we think it has not 

 been made before. We describe also some interesting varia- 

 tions which were probably suggested by the plates in Mr. J. 



