92 Mr. W. Spottiswoode on Stratified Discharges. [Mar. 22, 



pletely retreated into the hilt, the continuous or true jar-discharge took 

 place. 



With a view to testing experimentally how far the effects here de- 

 scribed were due to quantity and how far to tension, the size of the jar 

 was altered, all other circumstances remaining the same. It was then 

 found not only, as before, that small charges gave stratified and large 

 unstratified discharges, but also that the maximum charge compatible 

 with stratification was greater with a large than with a small jar. 



As a further experiment in this direction, a series of jars were arranged 

 in cascade ; and it was found that the greater the number of jars so 

 arranged, the smaller the charge necessary to insure a true jar-discharge. 

 A charge insufficient to destroy stratification with one jar was sufficient 

 to destroy them when more than one was used in cascade. These results 

 point to tension rather than to quantity as the determining cause of the 

 character of the discharge. 



In fact, having taken a number of jars of the same size, and having 

 ascertained the maximum charge with which one jar could be charged 

 without obliterating stratification, say,. " the critical charge," I found 

 that the critical charge for 2, 3, . . . jars arranged for quantity was 2, 

 3, . . . times that for a single jar ; and, on the other hand, that the 

 critical charge for 2, 3 . . . jars arranged in series was 1 : 2, 1 : 3, . . . 

 of that for a single jar. The illumination, however, was always greater 

 with the larger charges, i. e. with the greater quantity of electricity dis- 

 charged. 



The experiments above described were made first with tubes in which 

 the pressure was moderately high. They were afterwards repeated with 

 lower pressures, and results of the same character as before were obtained. 

 But, owing to the smaller amount of the critical charges, to the greater ex- 

 tension of the negative glow, and to the consequently increased delicacy 

 of the phenomena, the same numerical precision was not attained. But 

 there seems no reason to doubt that the discrepancies might be indefinitely 

 diminished by instrumental refinements. 



The duration of the stratified discharges observed throughout these 

 experiments was exceedingly short, indistinguishable, in fact, from that of 

 the true jar-discharge. When viewed in a revolving mirror, either with 

 or without a slit, they showed no sign whatever of prolonged duration ; 

 and we may thence conclude that, so far as our present instrumental 

 arrangements extend, there is no inferior limit to the duration of dis- 

 charge necessary for the production of striae. 



In connexion with this part of the subject another form of experiment 

 was arranged. Beside the jars hitherto described another was used, 

 having its inner surface connected with one terminal of the tube, and its 

 outer with the other. When this disposition was made, the additional 

 jar acted as a buffer and produced a stratified discharge under circum- 

 stances which would without it have produced a true jar-discharge. 



