732 Prof. E. H. Barton and Mr. W. B. Kilby on 



Fig. 6 is from an enlarged photograph of one corner of 

 the dust figure confined within a pentagon, the spark being 

 at the centre. This shows very clearly the characteristic 

 corner effect, as though a disturbance were radiating from it 

 after reflexion. 



Refraction. — Several attempts were made to obtain refrac- 

 tion effects by letting the disturbance pass from one powder 

 to another very different in density or in some other property. 

 No success was attained in this direction. 



But it was noticed that something like refraction occurred 

 near the edge of a cover-plate or any other plate used as a 

 boundary if, instead of fitting close down, a small crevice 

 was left between it and the lower plnte. An arrangement 

 was accordingly made on these lines the result of which is 

 shown by the shadowgraph of fig. 7. A piece of glass five 

 inches by one inch with a sloping end was supported on two 

 oblong pieces of card (seen black in the figure) about half a 

 millimetre thick. The refraction of the stride at the front 

 edge of this plate is clearly exhibited. 



Fig. 8 shows a fortuitous example of refraction. It is 

 a photograph of a figure on metal between two inclined 

 boundaries. But the metal was slightly curved, and the 

 boundaries being of plate glass did not fit closely down 

 except near their ends. Consequently, at- the sides of the 

 figure remote from the corner a little effect leaked under 

 the glass plates and showed a refraction of the striae. This 

 figure also shows very clearly the corner effect, for which 

 indeed it was really taken. 



Interference. — The simplest way of obtaining interference 

 is that of using two sparks in series. These were passed 

 through holes in the same glass plate, the resulting effect 

 being shown in fig. 9, which is an enlargement of a portion of 

 a shadowgraph. In the dust figure itself (or the print from 

 it), on the part remote from the sparks, the central bar and 

 three on each side of it may be clearly seen. The central 

 bar consists of stria; almost parallel to the line joining the 

 sparks. The second bars to the right or left consist of striae 

 about perpendicular to those in the central bar. The other 

 four bars or intervening spaces (i. e. the first and third on 

 the right and left) do not present striae, but may be called 

 cellular spaces, since the dust is specialized in formation and 

 quite different from the random positions after merely scat- 

 tering, although no distinct striae are visible. The above 

 features are considerably modified as the region between the 

 sparks is approached. 



But though the above method is the natural one to try 



