Inventors at Work, with Chapters on Discovery

by George Iles

TELAUTOGRAPH, INTERIOR.

Machines Cannot Directly Imitate Hands: A Task Must be “Coded.”

Only a few machines deal with writing or its duplication, most machines perform quite other tasks at first wrought by the hands. Inventors have always gone astray when they have sought to imitate a hand process with anything like precision. On this point Sir John Fletcher Moulton, of London, says:—“Doubtless you have often had to send a message by telegraph to some distant country to which the rate charged per word is high. You write your message as tersely as may be, but even thus its length is formidable. You resort to your telegraphic code. It tells you that if you will change the phraseology of your message you can by a single code-word represent a whole phrase. You thereupon set to work to recast your message so as to make it capable of being expressed in code-words. When you have done so, you have not improved it as a message. It is less terse and less naturally expressed. If you were writing and not telegraphing, you would prefer to use it in its original form. But as now expressed, each of the phrases of which it is composed can be sent over the wires in the form and at the price of a single word, and the cost of the whole is but a fraction of what would have been the cost of the message as originally framed. It has been cast in a form suitable for cheap telegraphing. Just so with the inventor. He has to find a series of operations which, in their totality, are equivalent to the series of the hand worker. But each of these operations in itself need not be such as would in hand labor be suitable or even practicable.

“It is necessary and sufficient for him that they are suited to the new conditions, so that they can be well and easily done by mechanism, and that, taken as a whole, they produce the same result as the series which he is paralleling. He is re-writing the series in terms suited to mechanism just as the message was rewritten in terms suited for telegraphing. The meaning of the message must remain the same, but the terms used to express it are no longer those most naturally used in writing or speaking, but are those which can be telegraphed at least cost.

Sewing Coded in a Machine.

“To make my meaning clear, let me revert to the familiar operation of sewing. The hand process is plainly unsuited for mechanical reproduction. How is it to be translated into an equivalent cycle suitable for mechanism? In other words, how is it to be ‘coded’? This case is interesting, inasmuch as we have two independent solutions worked out at different dates and widely different in nature. The earlier invention imitated the hand cycle very closely. The thumb and finger of the right hand in the human being were replaced by pairs of pincers capable of taking hold of the needle and letting it free again, but to avoid having to follow the intricate movements of the human fingers in the operation two pairs of pincers were used, one on each side of the work, which passed the needle backwards and forwards through the fabric one to the other. Following out this idea the needle was pointed at both ends with an eye in the middle, and, as in hand sewing, it carried a moderate length of thread. The pair of pincers which held the threaded needle advanced to the fabric and passed through it to the other pair which took it and retreated so as to draw the thread tight and form the completed stitch. To form the next stitch the work was moved through the proper distance and the same process was gone through, the line of movement of the needle always remaining the same.

“There is not much ‘coding’ here. The new cycle imitates the hand-worker so faithfully that it benefits little by the advantages of mechanical action. As in hand work it can only sew with moderate lengths of thread, and must therefore have the needles re-threaded at intervals. Its superiority over hand labor is therefore so slight that it is doubtful whether such a sewing machine could ever have competed with, much less replaced, hand work. But it has one great merit. The needle mechanism is capable of being re-duplicated almost without limit, and the movement of the work which is necessary to direct the stitches for one needle will serve equally well for any number of needles working parallel to it. Hence the machine that would have failed as a sewing machine has survived and proved useful as an embroidery machine. The work is stretched between two rows of pincers and moved by the workman according to the stitches of the pattern. Each stitch is repeated by each of the parallel needles which work side by side at convenient distances, and thus as many copies of the pattern are simultaneously produced as there are needles. Each is a perfect facsimile of all the others, and as each copies faithfully the errors of the workman, this machine is entitled to the proud boast that its productions possess all the defects of hand work—an essential we are told of artistic beauty.

“What is the cause of the comparative failure of this attempt at a sewing-machine? It is evident that it is due to the retention of the feature of the hand operation by which the needle is passed from one holding mechanism to the other. The inventors of the modern sewing-machine on the one hand decided to work with a needle fixed in its holder and never leaving it throughout the operation. It at once followed that the needle and thread must, on the back stroke, return through the same hole through which they had entered the fabric, so that no stitch could be formed unless some obstacle were interposed to the return of the thread. Here the two famous and successful forms of the machine parted company. Both placed the eye at the point of the needle that the stroke might not be needlessly long, but while the lock stitch machine used a second thread to provide the necessary obstacle, the chain stitch machine availed itself of a loop of the original thread for that purpose. Thus in the lock stitch machine the substituted cycle became as follows:—

(1) The work is moved under the needle for the new stroke.

(2) The needle (which has an eye at its point through which the thread passes) pierces the fabric carrying with it the thread.