a small Fraction of a Dynamo-electric Current. 559 



dynamo-electric machine which acts as an ordinary electro- 

 magnetic engine, performing some useful mechanical work, 

 i. e. pulling the punkhas, lifting messages, producing a draught 

 of cool air, &c. ; or the current is made to pass through a gal- 

 vanoplastic apparatus in connexion perhaps with the survey or- 

 general's office, &c. Now connecting the negative pole* of such 

 a dynamo-electric machine to earth, the positive pole to all the 

 lines terminating in a telegraph-office, while the two poles are 

 permanently connected by the resistance r through which the 

 current produces the useful work above mentioned, then it 

 will be clear, without demonstration, that all the lines so con- 

 nected can be provided with signalling-currents (which are 

 exceedingly weak as compared with the main current) by 

 simply tapping the main current, and without perceptibly re- 

 ducing it, i. e. without influencing the useful work performed 

 by the main current through r. 



Supposing that the useful work performed by the main cur- 

 rent repays all the expense connected with the erection and 

 working of the dynamo-electric machine, then obviously this 

 would be a method which supplied the signalling-current for 

 nothing. This might be an inducement for telegraph-admini- 

 strations to introduce the electric light, since they would get 

 the signalling-currents into the bargain, and the costly and 

 cumbersome galvanic apparatus might be dispensed with. 



An example will show this more clearly. For instance, a 

 Siemens dynamo-electric machine of medium size can easily 

 be made to produce through an electric arc a current of 30,000 

 milli-oersteds, while not more than 3 milli-oersteds are re- 

 quired to work the Siemens's polarized relay with engineer- 

 ing safety. Supposing that the sent current is made equal to 

 twice the current which is required to arrive, we have the fol- 

 lowing calculation for Calcutta office : — 14 long lines termi- 

 nate at Calcutta ; hence 14 x 6 = 84 milli-oersteds would (as 

 a maximum) have to be tapped off from the main current of 

 30,000 milli-oersteds. This represents a loss of only 0'28 per 

 cent., which is so little that even the most sensitive eye would 

 be unable to detect any variation in the light. Hence in this 

 case we should feed the telegraph-lines with currents which 

 actually cost nothing, as the electric light alone would repay 

 all expense. 



During my recent light experiments in London, it was ex- 

 perimentally established that the current in milli-oersteds 

 which a dynamo-electric machine is able to produce can be 



* In India we use positive signalling-currents. 



