where there is no deposit should Ije in the shajje ot a cross. If the copper cylinder 
he heated as before and the plate not heated, bnt placed f>n it at an ordinary 
temperature, then there is an open space, square in general form—possibly the former 
cross filled up—formed, and in the centre there always aj^pears a very small whke 
DEFINITE FIGURES BY THE DEPOSITION OF DUST. 191 
obtaining the cross on the square glass entirely free from all fuzz at the centre, is by 
using as a source of heat a metal plate, and placing on it another thin piece <»f metal 
about 1 inch in diameter, and allowing the centre of the glass plate to rest upon it. 
If the copper cylinder under the plate be heated to about 150 C., and the plate 
rests upon it, no deposit occurs immediately above it, and this open space assumes 
the form of a cross (fig. H)- That dust does not deposit on a sufficiently heated 
surface has long been known ; but it is interesting that in this case the portion 
O V., , 
Fig. IG. 
