"All this is amazing to me," said Pierre. "I thought the men just took the end of the thread and wound it off without any trouble."

Henri shook his head.

"No, indeed. I wish it were as easy as that. A reeler needs judgment, judgment, judgment at every turn. Not only must the floss be removed from the outside of the cocoon before it is reeled, but also the first part of the filament, which usually is weak and too fine for use, must be wound off until the firm, strong thread is reached. You see, the caterpillar has to work a little while before it gets under way and does its best spinning. All that poor filament on which it experiments or gets started must be broken off and saved for embroidery floss, since it is fit neither to be woven into broad goods or twisted into sewing silk. The reeler begins to wind where the end of the filament becomes strong. He then must combine enough fibres of the same size and strength to make a thread uniform in size. And this is not so easy as it sounds, since there is great diversity in the coarseness and fineness of the filament on the cocoons. He cannot always put the same number of filaments together. In addition to this fact he is often required to reel silk of various sizes. The coarseness or fineness demanded all depends on the purpose for which the silk is to be used. But always each kind must be of uniform size throughout."

"And how does he join the fresh cocoons to the others? By tying?"

"No. He runs them in so skilfully that the joining cannot be detected. Every moment he must be on the watch to add new filaments when he sees any of his cocoons giving out. As one cocoon takes the place of another its filament blends unnoticed in the thread. You can see that it would never do to join a lot of new ones all in the same spot."

"I suppose the cocoons run in uneven lengths, anyway, don't they?" ventured Pierre. "Scarcely any of them would contain the same number of yards of filament even were they all to be started together."

"Precisely. So the reeler keeps adding fresh fibres, being careful always that his thread is running uniform all the time. If he uses fibres of fine quality there must be more of them; if coarser fibres not so many. He can't turn out thread that is thick in one place and thin in another."

"That is what you or I would do," laughed Pierre. "Or at least I should. I never could reel so fast, either."

"It winds better fast," replied Henri. "It has not so much time to loosen or get caught. It just has to keep moving right along. If we can get cocoons soon enough so they can be wound off before the moth has time to come out, instead of having them cured first, they reel far more easily. The curing affects the silk. Of course in most cases it is unavoidable, for it would require very quick work for our agents to buy up the products of outlying silk-raisers and get them to us before the chrysalis matured. We should be taking a big chance of having our silk ruined, since one never can predict exactly how long it will be before the moth will come out. Varying conditions bring different results. It is a pity, however, that they have to be cured. Still, the curing has one advantage—it decreases the weight of the cocoons about twenty-five per cent."

"I didn't realize that curing caused shrinkage."