by equal surfaces of Cylinder and Plate Electrical Machines. 29 1 



theless satisfied of the fact, and having subsequently constructed 

 many electrical machines of various kinds and sizes, the supe- 

 riority of cylinder over plate machines was continually forcing 

 itself upon my attention. About sixteen years since, I fitted 

 up a small cylinder machine for the purpose of examining some 

 of the phsenomena of electrical excitation. As the experiments 

 were undertaken principally for my own practical guidance, I 

 did not publish them, but during some recent investigations with 

 my new induction coils, my attention has been frequently again 

 directed to the subject, and I have thought that some account of the 

 experiments might not be altogether unacceptable to electricians. 



My object in the present paper is to point out some of the 

 phsenomena attendant on the action of the electrical machine, 

 together with some circumstances by which they are modified. 

 To ensure accuracy, I have lately repeated the experiments with 

 several machines, and amongst them the lOi-inch cylinder before 

 alluded to, and with corroborative results. 



The standard which I have employed for comparing these re- 

 sults, has been a discharge of a given length from a Ley den jar, 

 using the same jar in all experiments that were compared with 

 each other. I once attempted to establish a common unit-mea- 

 sure which would be accessible to all, and which would afford a 

 means by which electricians could compare their own results with 

 those of others, but did not succeed to my satisfaction. Leyden 

 jars, although they might contain a given amount of coated sur- 

 face, did not seem adapted to the purpose, as their variable thick- 

 ness would modify considerably the quantity of electricity requi- 

 site to produce a discharge of a given length ; thus two jars of 

 equal size, in the hands of two different individuals, would give 

 very different results. I therefore thought of employing coated 

 plates of glass, whose thickness might be always ascertained by 

 their weight per square foot, but found, to my disappointment, 

 that their retentiveness of residual charge was so considerable, 

 that it was perfectly impossible to ensure accuracy of result, even 

 in a consecutive series of experiments with the same plate. As 

 much as 10 or even 20 per cent, of the original charge sometimes 

 remained as a residue, which would furnish ten or a dozen minute 

 discharges over a period of more than half an hour. This disad- 

 vantage, which appears peculiar to crown and plate-glass, since I 

 find it to exist in Leyden jars made of the same material, induced 

 me to abandon the use of the plates, and content myself with 

 using the Leyden jar and a graduated discharging electrometer. 

 Ttic number of turns of the machine requisite to produce a dis- 

 charge of a given length being multiplied by the area of glass 

 rubbed at a single revolution, gave the amount of surface required 

 to be submitted to friction to produce a given effect. 



U2 



