20b* 



Mr. H. Pender on the Magnetic 



the middle of March and the 1st of May ; also the value of v 

 calculated for each set in the manner above described. 

 Observations were attempted only on those nights when the 

 atmosphere was comparatively dry. A bracket in the table in- 

 dicates that the sets included under it were taken immediately 

 following one another on the same night* 



In connexion with sets 14 and 15, a further series of 

 readings was made, which may be looked upon as an additional 

 proof of the existence of the effect sought. The disks were 

 charged oppositely throughout. First was taken a complete 

 set of readings on reversing the switch S while the disks 

 were at rest, the remainder of the apparatus running as usual. 

 The disks were then set rotating, both positively, and a second 

 series of readings taken ; then one of the belts from the 

 counter-shaft to the disk was crossed, so that I rotated ■+■ 

 and II — , and a third series taken; then the main belt from 

 the motor was crossed, so that I rotated — and II +, a fourth 

 series taken ; finally, the belt from counter- shaft to disk was 

 put straight again, so that both disks rotated negatively, and 

 a fifth series taken. In connexion with sets 16 and 17 a 

 similar series of readings was made, except that here both 

 disks were charged alike. Here are the observed and cal- 

 culated values for the deflexion for the various cases : — 



Table V. 



Speeds. 



Charged. 



Deflexion. 



I. 



IX 



I. 



II. 



Obs. 



Oalc. 











+ 



+ 



4- -9 



00 



+83-2 



+88-9 



± 



+ 



+ 2-0 



+ 2-1 



+784 



-73-4 



± 



+ 



-63-7 



-601 



-79-3 



-74-2 



+ 



+ 



4- 2-3 



4- 2-1 



-83-0 



i-81-9 



± 



* 



4-68 2 



+650 











± 



± 



4- -1 



00 



-86-9 



-888 



± 



± 



-662 



-68-2 



+88-0 



-89-6 



± 



± 



- 1-0 



- -4 



-88-0 



+89-6 



± 



± 



4- 2-0 



4- -4 



-78-3 



+78-3 



± 



± 



4-62-0 



+65-0 



In all of the preceding the surfaces both of the disks and 

 of the condensing-plates were uniform. A series of experi- 

 ments early in the winter had shown that the qualitative 



