300 
sparks, a method described elsewhere in these Proceedings. A second is a 
considerable increase in the two capacities, to obtain waves of greater inten- 
sity. A third isa modification of the sound gap, or rather a disposition of 
screens about the sound spark in order to obtain waves from the same spark 
of both great and small intensity. These waves are photographed on the same 
plate, enabling one to determine their relative velocities. A few of the results 
are given in this preliminary paper. 
The details of the sound gap and screen are shown in Figure 31. A heavy 
spark is passed between the platinum terminals P-P. This produces a 
cylindrical sound wave shown in section at S, S. G is a cylindrical metal 
sereen, which I shall call a grating, concentric with the spark axis, and having 
longitudinal slits or apertures O, O, cut in it, as shown in the figure, thus 
forming a sort of grating. The grating is so placed that it intercepts but one 
end, the left end in the figure, of the cylindrical wave, the right end or half 
spreading out the same as if the grating were not in use. I shall call this wave 
the main wave. Some of the energy of the left end of the wave is reflected by 
the grating, but some of it passes through the apertures which thus become 
sound sources, the waves spreading out in every direction from these sources. 
I shall call these waves wavelets, 
