IV
MAIN OBSERVATIONS IN ACUITY

Attention has already been called to the fact that of the colored
incandescent bulbs used in the preceding observations only the red
gave approximately monochromatic light, and this was of very low
intensity. In order to increase the intensity of the red illumination
and to secure more nearly pure colors in the green and the blue, it
was found necessary to adopt a different source of light.

An attempt was made to project a spectruni. by means of a col-
limating lens and prism, following the arrangement used by Abney
in his work on color vision. This method, however, was found im¬
practicable, for the reason that it was impossible to secure a suffi¬
ciently large field of uniform intensity and color tone to meet the
requirements of this study.

Several other devices were tried, as combinations of colored
glasses and gelatines, but none of these yielding satisfactory results,
recourse was finally had to colored liquids contained in absorption
cells. For red a dye was used called poncean red, a strong solution
of which transmitted a color well within the limits of the red,
ranging from 664 jn/* to 626/x/x. The red incandescent bulb and a
glass plate which was subsequently used yielded light within prac¬
tically the same limits of wave-length.

In order to secure a satisfactory green it was necessary to use two
solutions—a saturated solution of copper chloride and a weak solu¬
tion of crystal violet. This yielded an illumination in which green
strongly predominated, without any yellow but with a touch of blue.
The wave-length ranged from 536 /xju. to 475 /x/x, but the blue was of
so low intensity that its effect may be considered negligible.

It was found to be entirely impossible to produce a pure blue of
sufficient intensity to be accurately photometered or to be used in
acuity tests. A solution of copper sulphate oxidized with an excess
of ammonia gave a practically pure blue, if left of full strength, but
as has already been said, the intensity of the light thus transmitted
was too low to be available. It was necessary therefore to dilute the
solution with water, and this extended the spectrum a considerable
distance into the green, the wave-length ranging from 527 fifi to
436 /Aft. To the direct vision, however, the illumination was distinctly
blue, there being no suggestion of the green component.

24