340 Mr. R. Sabine on a Method of Measuring 



joined by the discharging wire, and (2) that the time which 

 elapses between the interruption of the circuit b and the press- 

 ing down of the key is not sufficient to allow of any loss 

 through the dielectric. Further on I will show how nearly 

 these conditions may be secured. 



It is, of course, necessary to have every part of the apparatus 

 as well insulated as possible ; and it is desirable that the dis- 

 charging resistance and a shunt across the galvanometer should 

 be so adjusted that all excursions have about the same value. 

 The following are a series of observations made by firing a 

 bullet from a small breech-loading pocket-pistol through two 

 wires set up (as in fig. 1) from one to four feet apart, the 

 muzzle being close to the first wire. The accumulator was insu- 

 lated with mica and shellac ; its capacity was £ microfarad 

 (f= 0*333 x 10~ 6 farad) ; its leakage in fifteen minutes was 

 only 1 per cent., and therefore its insulation resistance about 

 250,000 megohms. The discharging wire had a resistance of 

 100,000 ohms. The constant (fr) was therefore =0'0333. 







Excursion of light. 



Time of flight, 

 calculated 



Calculated 



Experi- 

 ment. 



Bange, 

 d. 







(Mog,5), 



t 



velocity of 



shot, 



d 



F 



C. 



c. 





feet. 



divisions. 



divisions. 



second. 



feet per sec. 



1. 



1 



300 



280 



00023 



436 



2. 



1 



300 



281 



00022 



457 



3. 



1 



300 



280 



00023 



436 



4. 



2 



300 



251 



0059 



340 



5. 



2 



300 



256 



00053 



379 



6. 



2 



300 



260 



00048 



422 



7. 



2 



300 



259 



00049 



408 



8. 



3 



299 



240 



0073 



408 



9. 



3 



299 



245 



00067 



450 



10. 



3 



299 



250 



00061 



496 



11. 



3 



299 



256 



00052 



576 



12. 



3 



299 



246 



0065 



460 



13. 



4 



299 



230 



0-0087 



458 



14. 



4 



299 



229 



0-0089 



451 



15. 



4 



299 



225 



00095 



422 



16. 



4 



299 



231 



00086 



467 



A glance at these results shows that the separate observa- 

 tions were not more different from the mean value than could 

 be well accounted for by the inequality of the charges of pow- 

 der in the cartridges. 



When the time to be determined is that which elapses be- 

 tween making and breaking a circuit, I have found that the 

 apparatus may conveniently be arranged in principle as is 

 shown in the sketch, fig. 2. 



The poles of the battery are not required to be connected by 



