1902) MiCROSCOPICAL JOURNAL. 119 
tive appears evenly illuminated over its entire area ; (bd) 
with the diaphragm a little closed so as to illuminate five- 
sixths of the central portion of the back lens of the ob- 
jective, leaving a narrow dark outer annulus; (c) with 
the diaphragm successively more and more closed up, so 
as to light respectively about 3, 4, 4, and 4th of the cen- 
tral part of the back lens. Then, on the same object, care- 
fully observe the effect of lowering or raising the sub- 
stage condenser beyond its focal point, an all too common 
expedient for obtaining an evenly-lighted field. In these 
experiments it will be advisable to use some simple form 
of light modifier, such as a piece of signal-green glass, 
and a rather high eyepiece may be employed with ad- 
vantage in order to render differences in the quality of 
the various images more prominently visible. It should 
alsobe borne in mind, and experimentally verified, that 
when the back lens of the objective is nearly filled with 
light, itis extremely sensitive to any alteration of tube 
length or cover-glass thickness, a very slight increase or 
decrease of the former being sufficient to make or mar 
the sharpness of the resultant picture, while any over- | 
sightin making this adjustment will inevitably cause it 
to appear veiled and milky, even with a first-class apo- 
chromatic objective. Indeed, the better the lens employ- 
ed the more sensitive it will be found in this respect. If 
the beginner will only examine a number of miscellan- 
- eous objects under the above specified conditions, there 
can be but little doubt that his conversion to critical mi-_ 
croscopy, even for low-power every-day purposes, will 
be speedy. 
Passing on to the second cause of the avoidance of crit- 
ical methods of research by general observers, especially 
such as habitually employ low amplifications—i. e., their 
strong predilection for an entirely illuminated ‘full- 
moon’ field—it should be realized that if this condition 
be considered a sine qua non, there are several ways of 
