Electrical Discharge. 355 



The battery consisted of five jars of similar dimensions, B, 

 fig. 8, each containing 5 square feet of coated glass. When fully 

 charged, it readily melted and fused into balls 15 feet of fine 

 ii'on wire. 



The intensity of the accumulation was valued by means of 

 a statical electrometer of great accuracy, the action of which 

 depended on the attractive force directly exerted between two 

 small circular planes, and reducible to a known standard of 

 weight. 



The platinum wire employed in the thermo-electrometer to 

 measure the heating effect was of sufficient diameter to com- 

 pletely transmit the whole of the charge. I was not ignorant 

 of the precautious necessary to be observed in this respect (13), 

 aud 1 have fully considered them in my memoir in the 12th vol. 

 of the Edinburgh Philosophical Transactions. 



In estimating the quantity of electricity accumulated, I re- 

 sorted to three different methods, — 1st, the revolutions of the 

 plate of the machine as indicated by a divided circle, and an 

 index fixed on the axis of the plate; 2ndly, by insulating the 

 battery in the way already mentioned (17) ; Srdly, by the " unit 

 jar," also before mentioned (18). 



2-1. In reply to the remarks (Poggendorff's Annalen for 1841, 

 vol. lii. p. 318) that " I have no clear idea of the theory of the 

 instrument I employed," &c., I have to observe, that I did not 

 think it requisite in these experiments to consider the first tem- 

 perature of the wire, the specific heat of the air in the ball of 

 the thermometer, and other small elements of that kind, so per- 

 fectly calculated by M. llicss in his formula 



t=(;-')(^' + 0(k+i>- 



Any correction which might arise out of such elements in the 

 results of the expei'iments in the way I conducted them would 

 be extremely small, and much less than the errors of observation 

 inseparable from the ex})eriment itself. Indeed we deceive our- 

 selves greatly in physical inquiries, when we attempt to reach a 

 degree of refinement inconsistent with the kind of experinients 

 in which we are engaged. It often serves only to complicate 

 the calculation, and introduce new sources of error into our ex- 

 perimcintal deductions. The first terms of M. Riess's formula, 

 just quoted, could have no relation to my method of manipula- 

 tion. With respect to the last, I have to observe, that if we 

 attcntiv('ly consider the nature and mode of operation of the iu- 

 strumcnt itself, we shall at once j)erceive that its indications 

 depend on the momentary expansion of a small cylindrical 

 coluuui of air immediately in contact with, and surruuiuling the 



