Yoi.. 8, 1922 
CHEMISTRY: WASHBURN AND N A VI AS 
3 
The inversion temperatures, coefficients of expansion and volume 
changes on inversion were obtained with a dilatometer holding 60 grams 
of pov/der and filled in vacuo with a mixture of H2SO4 and K2SO4 as the 
expansion liquid. The indices of refraction were measured with a 
petrographic microscope, using the Becke line method. The values given 
represent the upper and lower limits, that is, no part of the material 
showed an index outside of the limits given. 
A study of the data given in the table brings out the following facts: 
(1) The density of raw flint and chalcedony is appreciably lower than, 
but close to, that of quartz, the difference being not greater than that 
which might arise from moisture and other impurities in the raw material. 
The density of the calcined flint is slightly lower but close to that of 
tridymite. The density of the calcined chalcedony is distinctly lower 
than that of silica glass, the lightest known form of silica. 
(2) The indices of refraction of the raw flint and chalcedony are sub- 
stantially the same and while close to that of quartz are distinctly lower. 
The index of the calcined flint is close to that of cristobalite, while the index 
of the calcined chalcedony is definitely lower than those of cristobalite 
and tridymite but agrees very well with that of silica glass. As far as 
these two properties are concerned, therefore, the calcined chalcedony 
might be considered as a form of silica glass, while the calcined flint might 
be either cristobalite or tridymite. 
(3) Calcined chalcedony shows an inversion point at about 220° and 
calcined flint one at about 227°. The cristobalite sample studied was pre- 
pared from purified flint by a 7 hours calcination at 1500° with sodium 
tungstate. It showed an inversion temperature of 244° while the sample 
of tridymite prepared by a similar calcination for 150 hours at 1300° 
showed an inversion temperature of 117°. 
The inversion temperature of cristobalite has been shown by Fenner 
to depend upon the history of the sample studied and the values obtained 
for the calcined chalcedony and flint are therefore not inconsistent with 
the hypothesis that these materials are cristobalite. Such a hypothesis 
is evidently supported by the values obtained for the volume change on 
inversion. The results obtained with the dilatometer, therefore, clearly 
rule out tridymite and if the calcined chalcedony and flint are to be identi- 
fied with any of the known forms of silica, cristobalite seems to be the 
only possibility. But if these materials are to be classed as cristobalite 
it is evident that density measurements or indices of refraction alone or 
together are insufficient for the identification of the material, since, es- 
pecially in the case of the calcined chalcedony, both values are very much 
lower than any of those recorded for cristobalite. 
Since the density and index of refraction of the calcined chalcedony 
seem to indicate that it should be classed as a kind of silica glass, while 
