4 
CHEMISTRY: WASHBURN AND NA VIAS Proc. N. A. S. 
the dilatometric data show that it behaves as a kind of cristobaHte, some 
evidence of a more crucial nature seemed necessary before a certain clas- 
sification of the material could be made. Recourse was therefore had to the 
determination of the X-ray spectra of the various materials as being the 
most conclusive evidence that could be secured. Samples of all the ma- 
terials shown in the table were sent to Dr. A. W. Hull of the General Elec- 
tric Company who kindly consented to photograph the X-ray spectra. 
On receipt and comparison of these photographs the following results were 
obtained. Bach spectrum was found to consist of some 25 or more dif- 
ferent lines. The spectra of the raw chalcedony, the raw flint, and the 
quartz were absolutely identical, as were also the spectra of the calcined 
chalcedony, the calcined flint and the cristobalite This evidence is of such 
conclusive nature that we are forced to conclude that the calcined flint 
and chalcedony are composed of cristobalite crystals, while the raw mate- 
rials are composed of quartz crystals. 
How then are we to explain the discrepancy in the density and refrac- 
tive index data? The following hypothesis seems to be in accord with all 
of the known facts. Raw chalcedony and flint are colloidal quartz, the 
individual colloidal crystals being of microscopic or perhaps sub-microscopic 
size. On calcining the materials the included water is driven out and the 
individual crystals are transformed in situ into cristobalite. The resulting 
cristobalite crystals would evidently be loosely packed together since in 
the absence of any flux, there is no opportunity for increase in crystal size. 
Each cristobalite crystal is thus a colloidal particle made up of a compar- 
atively small number of molecules. The spaces between these colloidal 
particles of cristobalite must be of sub-microscopic dimensions, so small, 
in fact, that it is impossible to force a liquid into them. Many of them 
are possibly completely enclosed. Such a structure would readily ac- 
count for the very low density displayed by the material. As far as the 
refractive index is concerned we know that properties such as melting 
point and vapor pressure depend upon crystal size and it is conceivable 
that the surface tension forces might also be great enough to distort the 
crystal form sufficiently to influence the index of refraction. This the- 
ory would seem to be borne out by the fact that the calcined chalcedony 
was not optically homogeneous since different parts of it showed different 
indices of refraction, the index ranging between 1 .456 and 1 .470. It is quite 
possible that many of the crystals are so small that they approach in size 
the molecules of the liquid phase and the mixture would thus be a colloidal 
solution of crystals dispersed in their own liquid. 
In order to further test the above hypothesis as to the nature of the flint 
and chalcedony an experiment was made to ascertain whether the density 
of the calcined chalcedony could be raised by very fine grinding, since if 
closed pore spaces were present some of these might be disrupted and opened 
