why do the girls stir the cocoons with those whisks of peeled
birch?" inquired Pierre curiously. "What are they trying to do?"

"The stirring frees the ends of the filaments, and the brush of twigs
serves to collect them," answered Henri. "When the ends have been caught
in this way they are passed on to the reeler; but if after trying this
method the girls find the ends do not free themselves they put such
cocoons into a different temperature of water, or else toss them out
entirely and leave them to the employees who handle the lots that
require special treatment. They cannot stop here to fuss with cocoons
that fail to wind off readily; not only would such troublesome ones
retard the work, but they would be likely to snarl the others.
Frequently we find cocoons with uneven places in the thread, spots where
the silkworm has been interrupted in its spinning and stopped,
afterward going on with its work and making a lump or knob where the
filament has been joined. Such cocoons wind badly, as you can well
imagine, and they, too, cannot be reeled with the general lot."

"I notice those boys are taking the empty cocoons out of the tanks after
the silk has been reeled from them. Is it necessary?"

"Yes. We never allow the discarded cocoons, or shells as we call them,
to stand in the water with those that are soaking, because they not only
spoil the sheen of the silk on the unreeled cocoons but discolor it,"
Henri replied. "Now let us watch the reeling. Shall we?"

The boys turned toward the whirling machinery.

"I had no idea they reeled so fast," declared Pierre, speaking loudly so
his companion could hear.

Fascinated he stood watching the flying threads pass over the glass
rods.

"The speed of the reels can be regulated, of course," answered Henri.
"It is not often, though, that the filament snaps because the reel is
moving too fast. When the thread does break it is more frequently
because the regular motion of the machinery wears it until it parts.
This cannot always be avoided. All filatures count on some loss by
winding. But the percentage in a modernly equipped filature is very
small. We use the glass rods to prevent the thread from being caught or
roughened. The process of winding cocoons has been so carefully studied
that now our French reelers can turn off silk of fifteen or twenty
fibres and lose only one or two per cent. of it by waste. In Turkey the
loss runs as high as six or eight per cent.; in Syria it