108 THE AMERICAN MONTHLY [Apr 
Now stain again in the methyl blue, wash in dilute sul- 
phuric or nitric acid, then again in alcohol, and when the 
cover-glasses are dried with a piece of filter-paper or blot- 
ting-paper they can be mounted in balsam. It seems a 
tedious process, but when understood is very easy. I of- 
ten examine spleen for tubercle bacilli by this method in 
fifteen minutes. A splendid double stain can be purchased 
from Messrs. R. & J. Beck, of Cornhill, London, for one 
shilling per bottle, which saves great work, but the solu- 
tion must be warmed before use.—-J. Swift Walker, M. D. 
Logwood is not a good stain for bacteria. Some of the 
aniline stains, such as methyl] blue or gentian violet, give 
much better definition. A preliminary ftxation by heat or 
absolute alcohol is also desirable. Mr. McGhie would then 
have no hesitation in mounting them in Canada balsam, 
as the staining would be very pronounced. 
SCALES OF CLOTHES-MOTH.—This slide may be consid- 
ered a trivial one to sendaround; but though the scales are 
not rare, they exhibit much beauty of marking in the way 
of stri# and villi when examined under moderately high 
powers, besides making a charming dark-ground slide un- 
der Zinch or 1-inch objectives. I have included the slide, 
however, principally because I think these scales exhibit 
better than any others the evolution of the insect scale 
from the simple hair, or rather the probable lines on which 
it took place. The piece of wing on the same slide shows 
well the distribution of the scales; flatest on the centre 
of the membrane, and shading off into bundles on the 
edges. In the nervures a crooked system of vessels is per- 
ceptible, and these may be traced right through to pedi- 
cles of the tufts of bristles at the wing’s point, the func- 
tion being, I believe, to supply the scales with the liquid 
which, according to Dr. Royston Pigott, is found between 
the upper and lower membranes of the scales. Iam writ- 
ing without the book, but think this is so. There is cer-. 
tainly, as can be clearly seen with a good objective of wide 
