= 86.387
Iron oxide. = 0.424
Lime. = 0.244
Potash. = 4.806
Organic matter. = 0.507
Water. = 7.632
------
Total. 100.000

Apart from the question of its singular mode of origin, however, and its
remarkable and anomalous physical properties, tabasheer is of much
interest to mineralogists and geologists. All the varieties hitherto
examined, with the exception of the peculiar one from the Andes, are in
composition and physical characters true opals. This is the case with all
the Indian and Java varieties. They consist essentially of silica in its
colloidal form, the water, lime, potash, and organic matter being as
small and variable in amount as in the mineral opals; and, as in them,
these substances must be regarded merely as mechanical impurities.

The tabasheers must be studied in their relations on the one hand with
certain varieties of the natural semi-opals, hydrophanes, beekites, and
floatstones, some of which they closely resemble in their physical
characters, and on the other hand with specimens of artificially
deposited colloid silica formed under different conditions. Prof. Church,
who has so successfully studied the beekites, informs me that some of
those remarkable bodies present singular points of analogy with
tabasheer.

By the study of thin sections I have, during several years, been
endeavoring to trace the minute structure of some of these substances. In
no class of materials is it more necessary to guard one's self against
errors of observation arising from changes induced in the substance
during the operations which are necessary to the preparation of
transparent sections of hard substances. Unfortunately, too, it is the
custom of the natives to prepare the substance for the market by an
imperfect calcination, and hitherto I have only been able to study
specimens procured in the markets which have been subjected to this
process. It is obviously desirable, before attempting to interpret the
structures exhibited, under the microscope, to compare the fresh and
uncalcined materials with those that have been more or less altered by
heat.

Tabasheer would seem, from Brewster's experiments, to be a very intimate
admixture of two and a half parts of air with one part of colloidal
silica. The interspaces filled with air appear, at all events, in most
cases, to be so minute that they cannot be detected by the highest powers
of the mic