234 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
THE AUTOCHROME PROCESS. 
We are indebted to the Messrs. A. and L. Lumiere for this process, 
which gives such beautiful results. 
The process is based on the same principle applied in Cros’s and 
Ducos du Hauron’s methods, but instead of using three different 
screens, three negatives, and three superimposed dyed positives, one 
single plate serves for all these—screen, negative, and positive. 
This is accomplished by the use of a mosaic of starch granules of 
the three fundamental colors. 
The starch is sifted through very fine sieves, and the part taken 
which has grains of a diameter of from 10 to 12 microns. These uni- 
form grains are divided into three portions and dyed part orange- 
red, part green, and part violet. They are then mixed so that the 
resulting powder has a grayish color and does not show the tint of 
any of its component parts. This powder is then spread out on glass 
plates which have been covered with a coat of gelatine. The single 
layer of spherical grains thus obtained is flattened out by pressure so 
that the edges of the grains touch as far as possible, the small inter- 
stices which would allow the passage of white hght filled by very 
fine particles of carbon, and the screen thus formed covered by a 
thin coat of waterproof varnish for insulation and _ protection. 
Finally on top of this is applied the photographic emulsion, which 
of course must be panchromatic; that is, sensitive to all the light 
rays of the spectrum. In actual practice, it has been impossible to 
get an emulsion that is not slightly more sensitive to the violet end 
of the spectrum, so that in exposing the plates a yellow screen of a 
carefully chosen shade must be placed before the objective to counter- 
act this oversensitiveness. 
The plate thus prepared is exposed in an ordinary camera with the 
glass side toward the lens. The light rays therefore have to pass 
through the polychrome screen before they strike the sensitive emul- 
sion, so that they affect the emulsion only behind the granules which 
transmit their particular color. Thus rays coming from a green 
portion of the object photographed would attack the silver salt only 
behind the green granules, leaving it untouched behind the red and 
violet. In developing after this exposure, the affected silver salt is 
reduced and obscures these granules, leaving the red and violet gran- 
ules uncovered and transparent. As a result the plate, viewed .as a 
transparency, after this first development shows instead of green in 
a green portion of the object the complementary tint, carmine red, 
formed by the combination of the unobscured red and violet granules. 
If the plate were fixed here, this is the result we should obtain. In- 
stead of fixing, however, the plate is immersed in a bath of acid per- 
manganate, which dissolves the reduced silver, but does not affect 
