488 



BEPORT — 1878. 



end to further investigation, 

 these three days were : 



Final mean 



The mean values obtained for " v " on each of 



29-74 ohms June 18. 

 29-95 „ „ 23. 

 29-72 „ „ 25. 



2980 



It will be observed that the greatest difference between any one of the three daily 

 means and the final mean is only about half per cent. The final number 298*0 

 million metres per second (and which represents the mean of ninety-eight discharges 

 of the air condenser) may, I think, be regarded as correct to, at any rate, one per 

 per cent., and is exactly equal to M. Foucault's velocity of ligbt. 



In the astatic combinations I employed prior to June 18 I used eight needles, 

 and weighted the lower set of needles with pieces of brass, so as to give it a barrel 

 shape, but the results were unsatisfactory, as there was either not sufficient deli- 

 cacy or else too much damping. Consequently, all the numbers obtained prior to 

 June 18 have been abandoned. On June 18 were made the first set of expe- 

 riments with the forty-magnet astatic combination ; the idea of this arrangement 

 being to make an approximately spherical mass of little magnets all slightly sepa- 

 rated, from one another, and all previously magnetised to saturation. As it would 

 have been too difficult to make this entire sphere all of magnets, I finished it oft 

 with segments cut from a little wooden sphere. But the half Napierian logarithmic 

 decrement was 0-12095, and we thought this too high. Consequently, in the inter- 

 val from June 18 to June 23, I took this astatic combination down and replaced the 

 segments of the wooden sphere by segments of a small leaden hemispherical shell. 

 This diminished the half Napierian logarithmic decrement to 0-07825, and with a 

 periodic time of 39'5 seconds I obtained most consistent results. But on the other 

 hand, the close agreement of the results obtained on June 18 and on June 25, 

 leads one to conclude that the wooden segments were quite satisfactory, and that 

 replacing them with the leaden shell was unnecessary. 



The table on the accompanying sheet gives the value of the most important 

 constants employed. The battery consisted of 382 perfectly new Daniell's cells in 

 series, and the galvanometer had a resistance of 20,000 ohms. All resistances were 

 compared with a new German-silver wire box, recently received from Messrs. 

 Elliott, London. 



Determination of the Number of Electrostatic Units in an Electromagnetic one. 



June 18. The lower set of needles was weighted with segments cut from a 

 small wooden sphere. 



June 23 and after. The lower set of needles was weighted with segments 

 cut from a small leaden spherical shell. 



Number of magnets in astatic combination, 40. 



Number of new Daniell's cells iu series, 382. 



* The distance between the upper condenser plate and the guard ring was 

 slightly increased by diminishing the size of the plate to avoid the possibility of 

 loss of electricity. 



