167 



it still evolves from its outer coating the same number of measures 

 for each unit of quantity as it did at first ; and conversely, the 

 second jar receives as easily the unit of quantity taken in terms of 

 the unit explosions when charged with 20, 40, or 60 measures, as 

 it did at first. The author concludes, that the charging of an elec- 

 trical jar is by a rough analogy rather to be associated with the 

 pouring of an inelastic fluid such as water into an open vessel in 

 measured quantities, which is done up to the point of overflow as 

 easily at last as at the first. 



Having given experimental illustrations of the nature of the seve- 

 ral instruments just adverted to, and shown their accuracy as in- 

 struments of research, the paper proceeds to consider the pheno- 

 mena of what the author, after the learned Mr. Cavendish, deno- 

 minates electrical charge. By the term electrical charge of a given 

 conducting substance, the author understands the quantity of elec- 

 tricity which the body can sustain under a given degree of the 

 electrometer. In pursuing this interesting question, he commences 

 with an examination of the charges of hollow spheres or globes of 

 different diameters. The method of experiment is to place the given 

 sphere in communication with the electrometer, and find by a transfer 

 of measured quantities of electricity the precise number of measures 

 required to bring the index to a given degree of the arc. These 

 measures are obtained by insulated balls or plates of given dimen- 

 sions, brought into contact with the ball of an insulated charged jar 

 carefully prepared and screened from the external air. The author 

 shows how this method of measuring quantity by means of what he 

 terms a quantity-jar may be perfected, so as to be relied on as a means 

 of estimating small quantities of electricity. 



The results of a series of experiments with spheres and plates of 

 equal area led to the deduction, that the charges of these bodies are 

 as the square roots of the surfaces multiplied into the circumferences, 

 and that the charge of a sphere is to the charge of a circular plate of 

 equal surface as 1 : \/2, and the charge of a great circle of a sphere 

 is to the charge of the sphere as 1 : \/4, or 1:2. 



Taking a given surface of 100 square inches, and placing it under 

 various forms, viz. a sphere, circular plate, square plate, rectangular 

 plates of variable extension, a hollow open cylinder, a cube, &c., 

 and subjecting these to the same process of experiment by which is 



