4J:tJ Messrs. V. Crcniieii and H. Ponder on 



open, any electric oscillation through the coil will suffer little 

 clamping. 



To ascertain if such oscillations played any part, we placed 

 Pender's coil in a box of brass 2 mm. thick, entirely closed. 

 This diminished considerably the perturbations caused by 

 charging and discharging the disk at rest ; but the magnetic 

 effect due to the moyement remained of the same order 

 to within 10 per cent., i. e. as close as we could obserye under 

 the unstable conditions of the experiment. 



Howeyer, in Cremieu's negatiye experiments, the iron box 

 in which the disk turns can arrest every magnetic oscillation 

 tending to traverse the coil, as this box constitutes a perfect 

 screen, which is not the case with the brass box in which 

 we inclosed Pender's coil. In fact, it is well known that 

 damped wayes, such as those which occur around a body 

 whose charge is rapidly varying, traverse without con- 

 siderable alteration even very thick conducting screens. To 

 verify this point, it would therefore have been necessary to 

 put a magnetic screen around Pender's coil. 



The following method of procedure, however, is much 

 simpler. If the effects observed by Pender are due to 

 an oscillatory phenomenon, they must certainly be modified 

 by any change in the circuits serving to charge or discharge 

 the disks. 



We placed in these circuits liquid resistances. They did 

 not cause the effect to disappear ; they merely diminished 

 considerably the perturbations at rest. 



Again, we placed in parallel with the disks a variable 

 capacity. The capacity of the disks was 200 C.G.s. 

 electrostatic units. The capacity in parallel was an air- 

 condenser, and could be varied by sixths from 166 to 

 1000 C.G.s. This time the effect in movement was con- 

 siderably diminished. The diminution was approximately 

 proportional to the capacity in parallel with the disks. At 

 the same time we noticed a diminution of the potential 

 assumed by the disks. (In all our experiments we used 

 for the source of charge M. Bouty's high-potential storage- 

 battery, capable of giving as much as 14,000 volts.) This 

 potential was measured by an electrometer connected auto- 

 matically to the disks at the moment when they were 

 charged. However, there was not a proportionality between 

 the diminution of voltage and diminution of the deflexion. 

 The deflexion dropped to one third its original value, whereas 

 the decrease in voltage was only 10 per cent. However, we 

 discovered that the disks also no longer became completely 

 discharged, and that the charge remaining on the disks was 

 proportional to the capacity in parallel. 



