Ill

PRELIMINARY STUDY

The preliminary observations here reported were made not so
much with the hope of securing accurate results as with the purpose
of becoming familiar by actual trial with the important details of
procedure, in order that sources of error might be discovered and,
as far as possible, eliminated before taking up the more exact study
of the problem.

In making choice of the various tests proposed for the study of
visual acuity, the writer was governed largely by the consideration
that the chief function which the eye is called upon to perform under
artificial illumination is that of reading. Characters of the alphabet
were therefore chosen. Even though such a test-object may be open
to theoretical objections, its use would seem to yield results of the
highest degree of practical application, in view of the very general
use of such characters among optometrists.

The apparatus and method of procedure were as follows: At one
end of a long bench conveniently graduated was placed a stand sup¬
porting a Snellen's optometric chart. On this chart the line of type
was selected which should be read by the normal eye under average
daylight illumination at a distance of three meters. This line was
chosen for the reason that it happened to be the most convenient for
the distances and intensities of illumination available. It was the
intention at first to use all the letters in the line for observation, but
a few trials developed the fact that the letters varied considerably in
legibility, and full perception of the line therefore resolved itself
into perception of the least legible letter in the line. This letter
was found to be B, and consequently the attention was centered
upon it, the effort being made to perceive it at each observation with
the same degree of distinctness.

For the illumination of the chart the following sources of light
were used:

Ordinary carbon filament lamps, with clear glass bulbs, with
nominal ratings of 2 and 16 candle power respectively;

Carbon filament incandescent lamp, with ruby bulb, nominally
rated at 16 candle power, but with an actual intensity, as subse¬
quently determined photometrically, of .6 candle power;

Carbon filament incandescent lamp, with blue bulb, rated at 16
candle power, with an actual intensity of 1.1 candle power;

15