excess of tension; and the excess of power required, especially on some Monday mornings, must be sought in some other cause than contraction. I think I shall show you that when you use an excess of power on a Monday, or any other morning, you are exerting such power in stretching all your spindle-bands; that you are thereby making conditions for bad work, and that you do not get back to good even yarn until after you have readjusted almost all your bands.

Let us assume certain conditions, to wit:
:-

A hot, damp dog-day.

A cool, damp September night.

An oil on the spindles that does not become fluid until the bearings are heated four or five degrees.

In the hot, damp day your bands all become moist; they remain moist all night, and in the morning are all in condition to be stretched when pulled upon, although while at rest they may contract a very little.

In the cool night the oil ceases to be fluid, partly on account of the coolness, partly because, when at rest, in an even temperature, it is less fluid than when worked.

In the morning every bearing sticks; you apply your power and you need an excess, not only to overcome the inertia of the metal, but also to create the heat to make your oil fluid. The moment you apply your power, you stretch every band to its utmost: your works run badly; you lose twist. You find fault with the cotton and the roll-covering; but is not the fault to be divided between the oil and the cotton band? Am I not right in my inference, that you do not get good work again until you have re-adjusted all your bands?

What are the remedies?

1st, Put more steam-pipe in a spinning-room and not less than in other departments, as is now your practice.

2d, Let heat into your spinning-room on cool nights succeeding hot dog-days before starting work.

3d, In winter, keep on the steam over Sunday, and do not start spinning in a cold room even on a dry day, because even a dry cotton band will stretch.

4. If there is any virtue in the composite band made of linen and cotton submitted to-day, try that.

In other words, the conditions of good spinning must be to have the bearings in condition to start with the least excess of power, and to use a band that will not stretch.

In many mills, I am aware these rules are partly or wholly established already; but in many more some or all of them are not practised, especially in the dog-days. If I were in charge of a mill, I should test these observations by overheating my spinning-rooms two hours before starting. I would not put less steam-pipe in a spinning-room than in other departments, as is sometimes the practice.

Having observed the facts in respect to the stretching of cotton banding, especially that protected by a solution of paraffine wax which stretched worse than any other, it occurred to me that a composite band might be made to serve a good purpose. I had no special reason, except that I knew experiments had been made with worsted bands, nor did I know that cotton and linen had ever been used before.

I sent to the Silver Lake Company for some cotton bands with linen core, and the same with worsted core, and afterward had some bands made of cotton and linen twist, and some cablelaid bands of cotton and linen. Mr. Woodbury will give you the result of the tests of these bands.

It appears that the same humidity that causes cotton to stretch causes linen to shrink; and therefore, if the two fibres. are combined mechanically in the right way, you have a compensation band.

Professor Ordway explains the reason. The fibre of cotton is a spiral, twisted ribbon. Humidity relaxes the twist, and causes each fibre to extend in length.

The fibre of flax is a straight, jointed cylinder. Humidity swells it laterally; and, if the fibres are twisted together, the expansion of the fibre hardens the twist, and shortens the strand.

Here seems to be a sound principle; but the difficulty comes in fastening the ends. The band is hard, and the knot is an obstruction. If you cannot fasten without a knot, you may at least use a more open V in your whorl, as the danger of slipping will be less in this band than in cotton if it is put on with a uniform tension.

I trust, however, that Mr. Draper has solved this difficulty of fastening the ends of the bands. If he has not, he knows the dreadful consequences with which he is threatened.

We desire to say that the matters treated in this statement are suggestions from our partial tests, and are not yet complete.

To prove them will require longer time, and a series of minute observations with the spinning-frame. We lost a good deal of time in attempting to apply dynamometers, and in proving that our anti-frictional machines were not suitable for our purposes.

I now wish to call your attention to another of our observations, while testing oil on our small frame, which may mean something and may not.

If the rotary motions of a spinning-frame impart a sideway vibration to the frame or its attachments, such a vibration must result in the wearing away of the bearings of the spindles on one side. I understand that this wear has frequently happened, and that in some special cases there has been a very rapid deterioration of the spindles, steps, and bolsters from such wearing on one side, and of course such wear implies the requirement of more and more power, with increasingly bad results in the quality of the yarn.

Our little frame is six feet long, and contains twenty-four Sawyer spindles on one side. In order to support the thermometers by which we measure the heat of the steps, we attached a board five feet and a half long, five inches wide, and seven-eighths of an inch thick, distant five inches from the frame, and fastened to it at each end.

The rotary motion of the machine imparts to this board a very curious sideway vibration, having a certain rhythm according to the speed of the machine. From a dead point as to the whole board, the vibration rises to a maximum in a certain time, which vibration is marked by a double curve from a central dead point; after reaching its maximum, it recedes until the whole board is still, then repeats, and so on. It is clear that the spindles must be affected by the same causes that vibrate the board, and that they must be worn away on one side. It seems to us that a remedy may be found for this mode of vibration.

NOTE. Since making this observation on our experimental frame, I have observed the working of many frames in several different mills, and I find a very great difference in this matter of lateral vibration. In some frames it is scarcely perceptible, in others excessive: it does not appear to follow the weight of the spindle, or the full or empty bobbin; neither is it to be attributed to the drag of the traveller. Our experimental frame had neither bobbin nor traveller attached. It has been suggested that the cause of vibration is in the motion of the spindles themselves, and cannot be

avoided; but it seems hardly possible that so many spindles could work such uniform rhythm in the vibration, and I cannot but believe it has something to do with the cylinder or other operating parts of the machine; and that if the spindle-rail were cut off from the frame by a buffer of lead, wood, or rubber, it might prevent the action of the vibrating force, whatever it may be.

Mr. BIRKENHEAD. Is there any loss of speed in the stretching of the band?

Mr. ATKINSON. That Mr. Woodbury can tell you, when he speaks of the experiment he has made on bands.

REMARKS OF C. J. H. WOODBURY, ESQ.

MR. PRESIDENT, -While making some experiments upon the power requisite to operate a spinning-frame under certain different conditions, I found that the power required to drive the frame varied so much with the tension of the bands, that this varying circumstance, for purposes of experiment, must be eliminated, by having an equal tension to the band turning each spindle, that this tension must be known, and preserved constant. These requirements seem equally important for economy of production. To turn the spindles with a constant band-tension just sufficient to produce the desired twist without slipping on the whorl, and, on the other hand, avoiding an excessive stress which would cause unnecessary friction, I believe would lead to a saving of power greater than that claimed by any spindle-maker for his spindle over those of his competitors. That is, the difference of power, caused by correct and false band-tension, is greater than that between any two rival spindles. By weighing the band-tension in various mills, it was found that the practice of tying bands lacked uniformity. As an example of this variation, in one mill the bands on a single coarse frame are reported to vary from one to sixteen pounds; in another mill, on finer work, taking a number of spindles at random, the variation was from one-half to two and a half pounds; and in a third mill the band-tension was between the limits of one-fourth to five pounds. Now this haphazard method seems hardly mechanical.

Is it not worth while for each spinner to learn the proper band-tension required for his special work, and then endeavor to keep within those limits? The changes constantly occurring in the cotton bands used, present an obstacle to carrying out these plans.

In accepting the invitation of your Board of Government, I do not expect to impart any new information upon the question of changes in length of bands, except so far as to give matters of detail connected with accurate measurements of these alterations in a number of bands. These measurements were taken by the following method: Two upright posts were firmly fastened against the wall, about ten feet from each other the samples of banding were extended between these, one end being fastened to a hook, and the other passing over a leading pulley to a weight. In order that the fibre of which these bands were composed would be as free as possible to expand and contract, it was desirable to use a light weight. By experiment, I found that with a weight of two and a quarter pounds, the band deflected only one-tenth of an inch from a straight line. A scale against the wall connected the tops of the upright posts; and a sliding framework carried two plumblines (fine wires), hung at right angles to the banding. When the samples had been under tension of the weight about two hours, this framework was slid, until the wires coincided with the division one hundred inches on the scale; then the corresponding length was marked on the banding with red cotton, and this point, also one hundred inches from the extreme end of the sample, was used to measure the dilatation in length, by sliding the framework until the plane of the wires coincided with the mark, then reading the distance on the scale above. In this manner I have taken about forty-five hundred readings, all of which were measured upon the same scale, and under equal conditions of accuracy, during a period of three and a half months. The apparatus was simple and rude in general construction, but it was sufficiently accurate to indicate a variation of less than one-fiftieth of one per cent in the length of the samples of banding.

It must not be understood that by this method I assume to reproduce the exact circumstances to which a spinning-band is subjected when in operation, but simply to maintain the bands subject to a constant tension, under the conditions given above, which, being applied in an identical manner to each sample, yield comparative results, giving the correct relations of these samples to each other when acted upon by the same specific causes.

A wet bulb hygrometer hung near the banding, and the