•260 report— 1879. 



the wire, bursts across the paper or air space in preference to passing through the 

 apparatus, and thus escapes to earth. 



An important modification of this plate-discharger has been made by Dr. Werner 

 Siemens, who, by serrating or grooving with a pointed tool the opposing faces of 

 the two plates at right angles to each other, converted them into a conductor, which 

 was supposed to be one composed of an infinite number of opposing points. The 

 remarkable action of points in facilitating discharge is well known, and their intro- 

 duction into lightning protectors occurred very early in the annals of telegraphy, 

 by Mr. 0. V. Walker, F.R.S. 



Messrs. Siemens' arrangement, very pretty in theory, never carried conviction of 

 its value in the mind of the author, because protectors so prepared never singled 

 themselves out as evidently superior to others that were not so prepared ; and while 

 the intersection of the grooves certainly formed mathematical points, they did not 

 form physical or mechanical points, and it is upon the action of this latter kind of 

 point that such remarkable electrical effects are produced. 



Dr. Warren De La Rue having very kindly placed his well-known battery of 

 11,000 cells at the disposal of the writer, he prepared four plate protectors, identical 

 in dimensions, excepting that two were serrated and two were not. The two plates 

 were separated from each other by narrow ebonite washers, - 01 inch thick. The 

 upper plate was placed in connection with the positive pole, and the lower plate 

 with the negative pole. The number of cells was increased until a continuous 

 current of electricity flowed. 



1. Plain Plates. 



No. of Cells. Effect Produced. 



1,000 . . Slight sparks just commencing on completing circuit. 



1,080 . . Sparks evident. 



1,200 . . Sparks frequent and abundant. 



1,500 . . Continuous arc. 



2. Serrated Plates. 

 No. of Cells. Effect Produced. 



1,000 . . Sparks just commencing on making contact. 



1,080 . . Sparks evident. 



1,200 . . Sparks frequent. 



1,500 . . Continuous arc, but fitful. 



2,000 cells in each produced a continuous stream of electricity. The effect with 

 1,500 cells was decidedly more marked with the plain plates than with those 

 serrated. The experiments were extremely pretty, and very decided in their 

 character. 



Hence it appears that grooving is not only of no use, but that it rather deteri- 

 orates the value of the protector. 



These experiments confirm very decidedly the accuracy of the figures obtained 

 by Dr. Warren De La Rue and Mr. Miiller on the striking distance between two flat 

 discs given by them in their paper read before the Royal Society {Phil. Tram., vol. 

 169, 1877), where it was shown that 1,200 cells struck across -012 inch. Here 1,000 

 cells struck across - 01 inch, which agrees perfectly with the curve produced by 

 those observers. 



It is the practice in the Post-Office Telegraph Department to keep these plates 

 apart by thin paraffined paper, - 002 inch thick, so that the air-space is really much 

 thinner than that experimented upon, and the striking difference of potential only 

 250 volts. 



Messrs. De La Rue and Miiller have shown that for points and various kinds of 

 surfaces opposed to each other plain surfaces act the best for potentials less than 

 1,500 volts, and that points are only efficient for high potentials. Now, as it is 

 doubtful whether atmospheric electricity causes much higher potential in telegraph 

 wires than 1,000 volts, it is clear that plain surfaces are the most effective for 

 protecting apparatus. It is quite certain that such plates, plain and smooth, 



