Dust Figures produced by Electric Sparks. 73] 



of plate-glass was used. The corresponding dust figure is 

 reproduced in fig. 1 (PL XXI.). Other obstacles were also 

 used so as to form a square with parts of its sides missing, 

 and thus give other examples of reflexion. The striae at the 

 right-hand upper corner show some distinction between 

 incident and reflected waves. The light diagonal bars at 

 the corners of the square are also noteworthy. 



Fig. 2 is a larger example of a square, practically com- 

 plete. Here again diagonal bars are noticeable, with fainter 

 accompanying bars occupying the whole intermediate spaces. 

 This recalls the effect inside a pentagon obtained by Marsh 

 and Nottage.' 



Figs. 3 and 4 show triangular dnst figures in lycopodium 

 and iron filings respectively. In fig. 3 we have again 

 diagonal bars in the dust figure, but here, the spark being 

 asymmetrical, it is noteworthy that these bars do not tend 

 to meet at the spark but at some other point nearer the 

 centroid of the triangle. In each of the first four figures 

 striae are seen near the corners concave to them (instead of to 

 the spark). 



Fig. 4 is one of the very few dust figures obtained fairly 

 well in iron filings. The confinement of the spark distur- 

 bance probably assisted the formation. The magnetic field 

 accompanying the discharge was probably a vital factor also 

 in the orientation of the filings. Xo success attended the 

 endeavour to obtain dust figures with red lead, which is very 

 dense but non-magnetic. 



Fig. 5 shows an enlargement from a shadowgraph of the 

 dust figure obtained with an inverted crystallizing dish 

 having a gap at one side next the spark. The reflexion 

 from the interior concave mirror gives a focus distant about 

 half the radius from the centre at the side remote from the 

 spark, as might be expected. But the waves after con- 

 verging to a focus at half the radius on one side the centre, 

 might diverge again and produce by reflexion a secondary 

 focus at a quarter of the radius on the other side the centre. 

 And this secondary focus, though not anticipated, is clearly 

 shown in the figure. Again, the waves diverging from this 

 could by reflexion form a tertiary focus at a sixth of the 

 radius on the other side the centre, that is, near the primary 

 focus. And this tertiary focus is clearly shown also. Further, 

 these three foci are in diminishing order of distinctness, as 

 might reasonably be expected. It should be observed that 

 the simple distances just referred to are approximate only, 

 being on the supposition that the spark source was at 

 infinity* 



