246 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1919. 
evidence could be discovered of the early stages of orderly arrange- 
ment of particles toward the crystalline form. So far, etching glass 
surfaces with hydrofluoric acid has failed to reveal any of the net- 
work of crystalline structure filled in with vitreous material which 
is sometimes described as representing the texture of glass. Tearing 
the surface of glass by letting a film of strong glue dry and contract 
on it is also stated to reveal a crystalline network. Both the etching 
and this method give markings, very like a network, no doubt, but 
the figuring of the surface seems to be more correctly ascribable to 
surface tension. Nothing which could be called definite evidence of 
crystalline structure is visible with any kind of illumination under 
the microscope of such surfaces, but this is not to deny that the 
texture of some glasses may be that of a network of crystalline com- 
pounds inclosing vitreous bodies. 
Certain facts, from which it appears reasonable to conclude that 
many glasses have something of a crystalline nature in them, have 
been obtained from a study of the phosphorescence of various glasses 
and other vitreous compounds exhibiting different degrees of readi- 
ness to pass into the crystalline state on heating. Much of the work 
was done about 20 years ago, but more recent experiments have not 
modified the conclusion then formed that a truly vitreous body ex- 
hibits no phosphorescence in ultra-violet light or X-rays or under 
cathodic discharge. Nearly every glass shows some phosphorescence 
and some show it very strongly, as, for example, the glass from which 
X-ray bulbs are largely made, and which gives the well-known green 
glow when the tube is in use. If some of this glass be fused and 
very rapidly chilled, as, for example, by making a Rupert’s drop from 
it, the glass is practically nonphosphorescent so far as its surface is 
concerned. A very little distance below the surface the chilling was 
not sudden enough to prevent some change of the truly vitreous 
to an attempt at crystalline structure, so that just below the surface, 
as shown by broken pieces of the drop, the glass exhibits phos- 
phorescence. The tail of such a Rupert’s drop, if heated below the 
temperature at which the thin thread of glass bends, is found to be 
strongly phosphorescent, and the glow under cathode discharge can 
be seen to fade slowly away toward the part which was not heated. 
Many observations with vitreous borates and silicates have shown 
similar phosphorescence, appearing more and more strongly as the 
vitreous bodies are made to approach the crystallizing stage. There 
does seem, therefore, reason to state that, given a body which in its 
crystalline state exhibits phosphorescence, it will not do so when it 
is in a truly vitreous state, and to infer that if a glass be phos- 
phorescent there is something of a crystalline nature in it. It would 
not be right to come to the conclusion that a glass showing no phos- 
