light was sensibly diminished all over the picture. This veiling of the
high lights was a serious disadvantage in another important particular,
inasmuch as it lessened the contrast between the lights and shadows of
the picture, thereby robbing it of some of its charm and deteriorating
its quality.

Since that evening I have endeavored, by a series of experiments, to
find out some means by which this opalescence might be got rid of in the
most convenient manner. Cementing the transparency to a piece of plain,
clear glass with Canada balsam, as suggested by Mr. Woodworth, I found
in practice to be open to two formidable objections. One of these was
that Canada balsam used in this manner is a sticky, unpleasant substance
to meddle with, and takes a long time--nearly a month--to harden when
confined between plates in this manner. The other objection was of
extreme importance, namely, that, in consequence of commercial gelatine
plates not being prepared on perfectly flat glasses in all cases, I
found that, after squeezing out the superfluous balsam and the air
bubbles that might have formed from between the two plates, they are
liable to separate at the places where the transparency is not flat,
causing air bubbles to creep in from the edges, as you may see from
these examples. I, therefore, have discarded this method, although it
had the effect desired when successfully done.

I have hit, however, upon another way of utilizing Canada balsam, which,
while retaining all the good qualities of the former method, is not
subject to any of its disadvantages. This consists in diluting the
balsam with an equal bulk of turpentine, and using it as a varnish,
pouring it on like collodion, flowing it toward each corner, and pouring
it off into the bottle from the last corner, avoiding crapy lines by
slowly tilting the plate, as in varnishing. If the plate be warmed