ewith (83). A less intelligent management of
the stitch it would be hard to find. The needlestrokes, far from helping
in the very slightest degree to explain the folding over of the petals,
directly contradict the drawing. The flower might almost have been
designed to show how not to do it; but it is a piece of old work, quite
seriously done, only without knowing. The embroidress is free, of
course, to work her stitches in a direction which does not express form
at all, so as to give a flat tint, in which is no hint of modelling; but
the intention is here quite obviously naturalistic. The rendering below
(84) shows the direction the stitches should have taken. The turn-over
of the petals is even there not very clearly expressed, but that is the
fault of the drawing (very much on a par with the workmanship), from
which it would not have been fair to depart.

[Illustration: 84. MORE EXPRESSIVE LINES OF STITCHING.]

A more clever fulfilment of the naturalistic intention is to be seen in
Illustration 76. The drawing of the doves is in the rather loose manner
of the period of Marie Antoinette; but the treatment of the stitch is
clever in its way--the way, as I have said, rather of painting than of
embroidery, giving as it does the roundness of the birds' bodies but no
hint of actual feathering, such as you find in the bird in Illustration
85. There, every stitch helps to explain the feathering. By a discreet
use of what I must persist in calling the same stitch (that is,
satin-stitch and the variety of it called plumage-stitch) the
embroiderer has rendered with equal perfection the sweep of the broad
wing feathers and the fluffy feathering of the breast. It is by means of
the direction of the stitch, too, that the drawing of the neck is so
perfectly rendered.

[Illustration: 85. SATIN AND PLUMAGE STITCHES.]

The direction of the stitch is varied to some purpose in the head in
Illustration 80, where the flesh is all in straight upright stitches,
whilst the hair is stitched in the direction of its growth.

The five petals on the satin-stitch sampler (Illustration 36)--to
descend from the masterly to the elementary--show something of the
difference it makes in what direction the stitch is worked. It matters
more, of course, in some stitches than in others; but in most cases the
direction of the stitch suggests form, and needs accordingly to be
considered.

It scarcely needs further pointing out how the direction of the st