202 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1944 
light corpuscles continuing to increase while the speed of the waves is 
proportionately retarded. If the light is refracted normally to the 
surface, however, it does not bend, but tends to cause a shortening of 
the optical path although the wave length is shortened regardless. It 
is only when it is refracted obliquely to the surface that the light is 
bent, the greater the obliquity of the incident ray and the denser 
the medium, the greater the bending of the angle of the cone of light 
and the shorter the wave length. It might therefore seem desirable to 
obtain as great an angle of refraction as possible. However, shorten- 
ing of the wave length is not in exact proportion to the amount of 
bending except in the case of the diffraction grating. And regardless 
of how great a change there is in its angle, the numerical aperture of 
the light, or angular aperture as it is more properly called, remains 
constant. 
In order, then, that the cone of light be large enough to supply the 
aperture of the objective with sufficient hight to produce an accurate, 
bright, and enlarged image of the specimen, it is first necessary that 
the specimen be refracting or emitting light of an adequate quantity, 
since both magnification and resolution are largely dependent upon the 
amount of light which the objective utilizes and receives into the tube 
of the microscope and since such hight as the objective does receive 
should be only that emitted by the specimen. It is obvious, therefore, 
that it is of primary importance for the specimen itself to be amply 
illuminated. This would seem to depend entirely on the actual light 
source, yet no matter how powerful a light source is employed, it is of 
little avail unless the condenser is of sufficient quality and aperture 
dimensions to accommodate the light which it receives from the source. 
If, for instance, the numerical aperture of the objective is 1.25, the 
width of the cone of light emanating from the specimen should com- 
pletely fill this aperture in order for the fullest powers of the micro- 
scope to be realized. Now, since the condenser supplies the light to the 
specimen, it stands to reason that it, also, should have a numerical 
aperture of at least 1.25. However, if the condenser and specimen 
slide are separated by air, the condenser can provide light of only 
1.00 N. A. to the specimen since, according to a law of optics, no aper- 
ture greater than 1.00 N. A. (this being the refractive index of air), 
can pass from a denser medium into air. To remedy this situation, 
an immersion fluid is placed between the top of the condenser and the 
lower side of the specimen slide as well as between the specimen and 
the objective lens. 
Since no optical medium has an index of refraction greater than 
3 and no immersion fluid an index of refraction greater than 1.7, to 
increase resolving power further, then, might it not be feasible to 
widen the apertures of the objective and condenser lenses, thus afford- 
