254 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1&19. 
case of a very soft gold glass, heating and working the colorless glass 
in the flame will only give a red glass, and no passage through the 
other stages is seen, or, in general, if the glass at any of the stages 
of color be worked in the flame, the change of color is but slight. 
The explanation seems to be that, throughout the mass of the glass 
the temperature never reaches high enough to give the state of 
viscosity for free aggregation of the gold particles, while, on the 
immediate surface of the glass the temperature may be so high that 
the gold is retained in solution, as it is in the highly heated furnace. 
It is very difficult to get a gold glass to behave like some opal glasses, 
in the sense that it will go clear and colorless in the flame and remain 
so when quickly cooled ; but a copper glass can be made in a similar 
manner with tin oxide, which will go quite colorless in the flame. 
Bulbs can be blown from it which remain white on cooling, and the 
gradual "striking" of a deep red color observed on gently reheating 
in the flame. It has been noticed in experiments with some copper 
glasses that, in the initial stages of "striking" the color which 
develops is not red, but a dark neutral tint with a suggestion of olive 
green in it. This may be from copper in the very finest state of 
division in which it can exert visible action on light, or it may be 
due to the presence of traces of oxidized copper in the glass giving 
rise to the well-known dark compounds of cuprous and cupric copper. 
Certainly very dark glasses of rather similar tint can be obtained by 
intentionally allowing some oxidation to take place in the making 
of a red copper glass, or by fusing together a reduced copper glass 
with one in which the copper is fully oxidized. At the same time 
it is worth noting, and is perhaps suggestive, that a chilled and 
colorless gold glass which goes through the stages of very pale red 
to a fine full red on heating in the flame has, after six months' expo- 
sure to the ^-rays of radium, only developed a dusky neutral tint. 
A piece of glass of the same composition, except that there was 
no gold at all in it, has not been affected by the radium. It would 
lead too much into theoretical discussion to consider in what form 
gold, copper, and selenium exist in the respective chilled and color- 
less glasses, and how far they may be looked upon as being in com- 
bination, or merely in so fine a state of division, that they have no 
visible effect on light. The chemical and physical evidence that, 
when they do give color they are in the elementary state seems to 
be fairly conclusive; but this does not necessarily exclude the pos- 
sibility of their being in something very like chemical combination 
in their colorless states. 
Perhaps it is needlessly striving after definiteness to attempt to 
distinguish between bodies being dispersed in an extremely fine state 
of division, partly at least through chemical attraction for their 
solvent, and being held in a loose kind of chemical combination. 
