142 THE MICROSCOPICAL NEWS. 
other additional series in the angles between the perpendicular 
groups. Insect scales and diatom valves exhibit these phenomena 
in the greatest variety. 
This method of direct observation of pencils of light coming 
from the object enables us to determine by experiment what part 
is played by diffractive phenomena in forming the image of the 
structure in question. A suitable test-object being placed in focus, 
and the light being suitably regulated by diaphragms placed imme- 
diately above the objective, as closely as possible to its upper focal 
plane, for the purpose of excluding at will one or another portion 
of the groups of rays exhibiting diffractive effects, the image of the 
preparation, as formed by those rays only which were not so shut 
off, could be readily observed with the ordinary ocular. The imme- 
diate result of experiments carried out in this manner was as 
follows, it being first premised that every trial was made with very 
correct low-power objectives (114 to 44 inch) and corresponding 
weak amplification: Higher powers, an immersion lens of } inch 
in particular being used only to control the results obtained already 
with coarse objects, by experiments on the finer diatoms. The 
preparations for all decisive trials were of such a kind that their 
structure was accurately known beforehand, system of lines 
oSEeee in glass, whose linear distance varied from ;$, inch to 
sso inch; similar groups of lines ruled on silvered glass, the 
Byes coating being immeasurably thin; groups of lines crossing 
each other without any difference of level were obtained by laying 
upon each other two glasses, the surface in contact being separately 
ruled. 
The facts thus ascertained are— 
(i.) When ad/ light separated from the incident rays by diffraction 
was completely shut off by the diaphragm, so that the image of the 
preparation was formed solely by the remaining undiffracted rays, 
the sharpness of outline at the confines of the unequally trans- 
parent parts of the field was zo¢ affected, provided the opening of 
the diaphragm remained sufficiently large, so that no diffraction 
arising from the reduction of its opening should occasion any visible 
lowering of the “necessary amplification;” nor will the clear 
recognition of separate structural particles be sensibly hindered, 
provided that not more than 30 to 50 of such particles are found 
in ;. inch.* But the more this number is exceeded, so much the 
more of detail disappears; so that when the fineness of detail 
reaches too parts to the millimeter (that is, when their interspace 
is only z#5p inch) nothing remains visible except a homogeneous 
25 
* The definition of number is here uncertain, because the exclusion of dif- 
fracted rays, whose diffraction is slight, can only be obtained by using a dia- 
phragm pierced with small openings. 
