Medical Officers of the Army of India.

43

or gentian violet. The first are remarkably large plasma cells, protoplasmic bodies
with a clear, comparatively transparent central region, which includes the nu-
cleus and sometimes a vacuole, surrounded by a thick coating of coarse granules,
so that when the cell is seen in a contracted state and deeply stained with violet,
it appears as an inky-black granular mass. These plasma cells are particu-
larly noticeable in sections treated by long immersion in methyl-violet and sub-
sequent washing out withalcohol alone; but, on the other hand, if treated by the
Gram process with iodine they lose the stain almost completely. With safranin
their granular coat stains of a reddish-brown colour, which again renders them
very distinct from the surrounding cells. The unequal size of the granules in
these cells, their evident density, and the above-mentioned fact that they lose
their stain under iodine treatment, is sufficient to prevent them being mistaken
for bacteria. Then, in the second place, there were to be seen in all the speci-
mens numerous examples of the hyaline, or colloid, globular bodies described by
Cornil and other writers as characteristic of rhinoscleroma. There are bodies
composed of one or more, sometimes numerous, perfectly transparent non- refract-
ing globules, associated in most, and possibly in all, cases with a nucleus, and
held together by a thin membrane, which is sometimes very indistinct and some
times certainly appears to be wanting entirely. The globules are not stained
at all by carmine (when pure and free from aniline adulteration, which is not
always the case) or by hæmatoxylin, but take a yellow stain with picro-carmine,
a rather brownish red, but deep, stain with safranin, and, last of all, a very deep stain
with any of the aniline violets. They are most effectively shown by staining the
sections in either gentian or methyl violet, the former by preference, and then wash-
ing out in iodine solution in the manner known as the Gram process. By this pro-
cess the globules remain deeply stained, when everything else in the section has
been washed clean; and the iodine must have the effect of fixing the stain in them,
for, if the sections are washed in alcohol alone, without first passing through iodine
solution, the hyaline globules are apt to part with the stain in great measure. If,
therefore, of two sections of the same tissue, one is washed in alcohol alone and
the other "grammed" with iodine, the first will show the plasma cells as above
described, while the second will show only the hyaline globules. It is remark-
able though that in almost all sections a very careful search will detect some
hyaline globules free from stain. With safranin many more of the globules
avoid taking the stain than with violet dyes, so that if two sections are taken
from a specimen and one is stained with safranin and the other with gentian-
violet, on the Gram principle, the latter will always appear at first sight to con-
tain more of the hyaline bodies than the former, the unstained globules escaping
detection, except under a very careful examination of the slide. The number of
distinct globules associated in each body varies enormously. There may be a
solitary globule which is then generally of comparatively large size (as much
as 1/1000th of an inch) with its nucleus flattened out like a half. moon on one

G 2