accompanying the Electric Spark. 297 



From this fact, and also that the circles can be produced 

 without any clearing of the powder, I am led to conclude that 

 the production of the circles is due to an entirely different 

 cause to that producing the space, but that both are necessary 

 accompaniments of the disruptive discharge. 



12. Discussion of Results. — When these figures were first 

 obtained it was imagined that they were due to the sound 

 which accompanied the spark, and in fact gave a graphical 

 representation of the sound-waves. This view was to some 

 extent confirmed by finding that the same substance gave the 

 same number of waves in the same space. If this be so, 

 however, then there ought to be no difference in the numbers 

 obtained with different powders ; for the sound of the spark 

 produced by a coil is remarkably constant, and, if such a term 

 can be rightly applied to it, it is of unvarying pitch. This 

 view, however, is also disproved by the measurements ; for if 

 it be taken that the lines show the phases of the undulatory 

 movement, then the longest distance between similar phases 

 is that given by powdered chalk, viz. ^ of an inch. But the 

 upper limit of vibration producing a sound-wave may be taken 

 as equal to 40,000 per second, giving a length for the wave 

 of as nearly as possible one third of an inch, a distance 

 thirteen times greater than the longest given by these expe- 

 riments. It is therefore evident that the vibrations producing 

 these curves and those producing the sound of the spark are 

 not identical. 



From the fact that the number given by a mixture is inter- 

 mediate between those of its constituents, we should be led to 

 conclude, since the density of a mixture follows the same 

 law, that therefore this is the factor which regulates the num- 

 ber given by any particular substance. If this be the case, 

 then the powders should arrange themselves in a series, with 

 those of the highest density at the top and those of the lowest 

 at the bottom. A glance at the tables of results will, however, 

 show that no such order is observed. Thus we have silica 

 giving a higher number (88) than the heavy oxides of lead 

 (64) ; magnesia alba, a substance of very low specific gravity, 

 appearing high up in the list ; and tannin, also of low specific 

 gravity, on the other hand, appearing low down. What con- 

 ditions, if any, regulate the number of these lines I have 

 hitherto been unable to determine. Before any generalization 

 can be made it is necessary to examine a very much larger 

 number of substances, and experiments are being conducted 

 with this purpose. At present, indeed, I am not disposed to 

 attach much importance to the absolute length of any one of 

 these undulations, but rather to the great similarity of the 



