288 



REPORT — 1879. 



Condenser A =0-3324 mfds. 

 „ B =0-3261 „ 



C =0-0995 „ 



) 



By comparison with glass condenser- 

 Thompson's method. 



! 1 

 - O-^^ L ^ direct measurement, the time of oscillation 



" p -0-1026 1 of the needle being 16-8 sec. approximately. 



The discrepancy in the two values of is due to the great absorption noticed 

 in this condenser. 



The value of B is also rather too large from the same cause. 



The surfaces of forty-nine plates of diameter 0-127 mm., and separated by an 

 average distance of 1-623 mm. from the larger plates, were discharged in this case, 

 giving for v, without corrections, the value 296 x 10 6 — m. per second. The cor- 

 rection for the greater induction through the pieces of shellac is + 3 x 10 6 , and for 

 the edges of the plate — 1 x 10 6 , giving for v the value 298 x 10 6 . 



The condenser was then taken to pieces and set up again. In the second case 

 the larger plates were separated by pieces cut from a plate of flat plate glass, and 

 the smaller plates resting on them were insulated by three very small fragments of 

 shellac. 



The distance of the surfaces was determined in this case differently. Fifty 

 small plates were used. The 150 slips of glass separating the larger plates were 

 cleaned with nitric acid, brushed with a camel's hair pencil, and piled on each 

 other on an inclined metre scale, and the total height measured several times. 



The small plates were brushed in like manner, placed on each other, resting on 

 a plate of flat glass, and another large stout plate placed on them, with a weight of 

 about 3 kilos. The distance between the glass plates was measured with a pair of 

 inside calipers at four points on a circle, concentric with the pile of small plates. 

 From the data thus obtained the average distance apart of the inducing surfaces 

 was estimated. The result was as follows : — Fifty plates, diameter 127 mm., 

 distant 2-362 mm. from larger plates. Capacity, 0-005007 mfds., 



Whence . 



Correction for edges 

 „ shellac 



Final result 



v = 299-5 xlO 6 

 -1-0x10* 

 + 0-3 x 10 s 



298-8 x 10 6 — say 299. 



A further slight correction should have been made for the connecting wires 

 and for the small quantity of solder on each plate, but the corrections will not be 

 sensible within the degree of accuracy obtained by the measurements. 



The final result, 298,000,000 metres per second, agrees closely with that 

 obtained by Professor Ayrton, whose method (nearly) is adopted, and with the 

 best value of the velocity of light determined by direct experiment. 



A great many measurements of condensers A and B were next made by the 

 ballistic method. I give only the result of each set of experiments here, leaving 

 the figures for a supplement, to show the sort of accuracy obtained in observation. 



9 

 10 



Sets 6 and 7 I have put down, but they may fairly be neglected, I think, 



