Figure 37. — On facing page and above: Patent drawings of Nollet machine, 1853. 

 From British patent specification 1587, July i, 1853. 



more accurate as the 1880's were approached. Such 

 figures for the AHiance machine during the decade of 

 the 1860's are as follows: 



Reference 

 Cosmos, 1860, vol. 17, p. 427. 

 Cosmos, 1861, vol. 18, pp. 197-200, 



646-647. 

 Annales Telegraphiques, 1862, vol. 5, 



pp. 505-520. 

 Reynaud, op. cit. (footnote 52). 

 Le Rou.\, op. cit. (footnote 55). 





Mum- 



Carcel 





ber 



units 



Tear 



disks 



per hp. 



1860 



6 



55-65 



1861 



6 



65 



1863 

 1866 



85 



90-95 

 65 



Of the preceding figures, probably only those for 

 1866 are adequate for the purpose. (An analogous 

 comparison of electroplating generators can be worked 

 out by determining how much metal was deposited for 

 unit time.) 



Similar measurements performed in the middle of 

 the following decade led to the first careful comparison 

 of the older magnetoelectric machines and the newer 

 dynamoelectric machines. (That such a test first 

 occurred adecade after the enunciation of the principle 

 of self-excitation serves to demonstrate the slowness 

 with which the commercial electric generator 

 developed.) Holmes had suggested the use of the 

 new kind of generator early in 1869, and had even 

 constructed a pair for the South Foreland lighthouse 

 that year.''^ However, despite the fact that the 

 dynamos produced a much brighter light than 

 Holmes' magneto generators in the tests, the Elder 

 Brethren of Trinity House held it to be wiser to choose 



Douglass, op. cit. (footnote 36). 



364 



BULLETIN 228: CONTRIBUTIONS FROM THE MUSEUM OF HISTORY AND TECHNOLOGY 



