1874 Physics. 131 
microscope indicates that this is caused by the force and velocity of impaé@ ; 
it is not a grinding process at all, but a battering action, similar to that of 
leaden bullets against a block of granite. A polished glass surface exposed 
for an instant to the sand-blast shows an aggregation of points of impad, 
from which scales of fractured glass have broken away in an irregular radial 
direction. It appears as if a pellet of glass had been driven in by the collision 
of the sand, and the wedge-like action thus set up had driven away the sur- 
rounding glass. All these spots or indentations, when tested by the polari- 
scope, show a coloured halo round each, proving that the glass surface is 
under strain and ready to yield to further fracture. The action therefore is 
not so much due to the hardness of the striking particles as to the force and 
velocity of impact. This is sufficiently great to destroy the cohesion of the 
surface of the material operated upon. ‘The external layer is carried against 
the under stratum, and the material is crushed and disintegrated by a portion 
of its own body. : 
Mr. William Webb, whose minute writings on glass are so well known 
to microscopists, disputed the reality of the finer bands of lines on 
Nobert’s test-plate, on the ground of the tool cutting away the surface 
of the glass, and leaving, as he showed by certain diagramatic sections, irre- 
gular jagged furrows. Dr. Woodward, of the Army Medical Museum, Wash- 
ington, in reply, reminds the readers of Mr. Webb’s paper, that there is a 
physical reason which compels us to believe that the first fifteen bands, at 
least, of the nineteen-band plate are composed of real and distiné lines, and 
that the distance apart of these lines must approximate very closely to what 
was intended by Nobert. When the bands of Nobert’s plate are illuminated 
by oblique light, and are looked upon from above with a low power (too low 
to show any of the lines), each band appears as a smooth coloured stripe. 
From the known wave-length of the colour seen, and the angle of the incident 
pencil, the distance which the lines of any band must a@tually be apart can 
be computed by the well-known formula for the spectrum of gratings enunci- 
ated by Frauenhofer, and the distance thus obtained agrees with that at which 
Nobert ruled the lines. On the other hand, the angle of the incident pencil 
being known, and Nobert’s given distance being assumed to be true, a table of 
wave-lengths for the different colours may be calculated, and the wave-lengths 
thus deduced agree substantially with those computed by other means. 
Nobert has discussed the whole matter in the 58th volume of ‘ Poggendorff’s 
Annalen” (1852). His discussion leaves, in the opinion of Dr. Woodward, 
no room for the possibility of a doubt of the objective reality of the lines up 
to the fifteenth band. Dr. Woodward calls attention to the fa& that this 
reason is altogether independent of our ability to resolve the lines with the 
microscope. In fad, it enabled Nobert to know that his plates were corre@ly 
tuled long before the resolution of any but the coarsest bands had been 
effe@ted. Asno spectral colour is obtained in the bands finer than the fifteenth, 
the formula of Frauenhofer cannot be applied tothem. In fac, the formula 
demonstrates that if these bands are actually ruled, as claimed, they can give 
no spectral colour. Dr. Woodward has no hesitation in expressing the opinion 
that the four higher bands (sixteenth, seventeenth, eighteenth, and nineteenth) 
have also an objective reality. He bases this opinion upon the comparison 
of their optical appearances, as seen with the best glasses, with the appear- 
ances of the lower bands (especially those from the ninth to the fifteenth). 
These appearances are quite the same in both cases, and, as similar results 
follow similar causes, he infers the existence of real lines in the four higher 
bands, since he knows they exist in the others. Dr. Woodward has recently 
examined two new test-plates by Nobert,—the first ruled for Professor Barnard, 
of Columbia College, the second for the Army Medical Museum,—in which 
the maker has attempted to rule lines twice as fine as those of the nineteenth 
band. These plates have twenty bands. The first ten correspond respectively 
to the first, third, fifth, seventh, ninth, eleventh, thirteenth, fifteenth, seven- 
teenth, and nineteenth of the old plate. The lines in the second group of ten 
bands purport to be ruled at the following distances apart:—The eleventh 
band, ,y$y5th of a Paris line; the twelfth band, y,},5,th; and so on up to the 
