for the sake of really
having a picture which is more than superficially like the original;
but in this way can be gained much real knowledge of technique which
cannot be gotten so easily otherwise.

Study your original carefully before and while working on your own
canvas. See how it was done if you can (and you can), and do it in the
same way, touch for touch, stroke for stroke, color for color. Use a
large brush when he used a large brush; if the original was done with
a palette-knife, use yours; and particularly never use a smaller brush
than the painter used on the picture you are copying.

The same thing holds as to processes. If your original was painted
solidly, with full body of color, do so on your copy. Never glaze nor
scumble because _you_ can't get the colors without. Your business is
to try to get the same qualities _in the same way_. And any other
manipulation is not only getting a different thing, but shirking the
problem. Because, if you can't get the effect in the way he did, you
certainly won't get the _same one_ any other way. You are not
originating, you are not painting a picture, you are copying another
man's work; and common honesty to him, as well as what you are trying
to learn, demands that you shall not belie him by stating on your
canvas implicitly, that he did the thing one way, when as a matter of
fact his canvas shows that he did it another way.

This may seem commonplace, because one would think that as a matter of
course any one would naturally make a copy this way. But this is
precisely what the average person does not do when copying, and I have
found it constantly necessary to insist upon these very points even to
advanced students.

So in the pigments, the vehicles, the tools, and even the canvas if
you can, as well as in the handling of the paint and the processes
used, follow absolutely and humbly, but intelligently, the workmanship
of the picture you copy, if it is worth your while to do it at all.

In making copies it is not usual to make the preliminary drawing
freehand. It takes time that may better be given to something else,
and often it is not exact enough. When a painter has made careful
studies which he wishes to transfer to his canvas, they may have
qualities of line or movement, or of emphasis or character which the
model may not have had. These studies, probably, are much smaller than
they will be in the picture. The same things may be true of the
characteristic