stick as a bobbin on the spindle, at a speed of fifteen or twenty thousand revolutions per minute, which was accomplished with reasonable steadiness. Orders poured in rapidly

ALLEN SPINDLE.

from these exhibits; and, the facilities of Messrs. Fales & Jenks' extensive works not being sufficient to supply the demand, those of Geo. Draper & Sons, at Hopedale, were also brought into requisition, and taxed to the utmost. The "ne plus ultra" of spindles seemed to be reached.

Practical running, however, developed incidental troubles. The packing, being continually beat upon by the bolster as the spindle vibrated, became thin, and in some cases torn in two; and under these circumstances the bolster was rapidly worn out. At other times the packing was insufficient to keep the bearings from revolution, and in that case the spindle gyrated and became inefficient. About the year 1881, the Rev. George H. Allen of Ayer, Mass., in experimenting to overcome these difficulties, adopted positive stops to prevent the rotation of both the bolster and the step of the Rabbeth spindle. He also omitted the packing, and made the step loose within the bolster, and extended the spindle into the step instead of having it move on the step, as in the regular Rabbeth structure.

The Allen invention was an important one, and has been partially adopted in all the Rabbeth spindles since made. It has also gone into use as modified by Messrs. Butrick & Flanders of Holyoke, and James H. McMullan, agent of the

Saco Water Power Machine Shop, at Biddeford, Me. This spindle is now the standard spindle built by that corporation. Mr. Allen sold his patents to the proprietors of the

RABBETH AS MODIFIED BY DRAPER AND KILBURN.

Rabbeth spindle, and his principle was applied to that spindle in a manner patented by John Kilburn, Esq., agent of the Lawrence Manufacturing Company at Lowell. This made a

substantial improvement in detail in the Rabbeth spindle, by lessening the wear of packing, and obviating the need of frequent replacement of parts. In the latter part of 1883, an improvement was made by William F. Draper, the author of this paper, which still further lessened the wear of packing, and reduced to a minimum the one remaining objection made by spinners; namely, that a tight band, by compressing the packing, would draw the spindle out of the centre of the spinning ring, and thus affect the quality of the yarn and increase its breakage in spinning. By this improvement, the bolster was provided at its upper end with a head which nearly fitted the case, and at its lower end the motion was limited by an extension of the step within it. By this arrangement, the amount of compression of the packing could be regulated, and the lateral movement of the spindle and bolster brought to the lowest terms necessary for steadiness in running.

Somewhat later Mr. George Draper made the bearing of the spindle tapering, for the purpose of saving power. This Rabbeth spindle, so modified, was known in the trade as the Rabbeth D 49," and was the standard spindle for three years. More than a million of them are in use.

During this time it was discovered by the Messrs. Draper that absolute steadiness could not be obtained without regulating the looseness in the bolster itself, that is, between the revolving spindle and its bearings. To accomplish this, Messrs. William F. and G. A. Draper experimented, and the results of their several experiments were patented in March, 1888. The spindle being tapering, and its bearing also tapering, it was evident that the looseness of fit could be regulated by raising or lowering the spindle or bolster, and that a mechanical arrangement for accomplishing this to the extent desired would enable the fit to be controlled. This they accomplished by making an external screw thread on the top of the step, and a corresponding internal one in the lower end of the bolster. The engagement was loose, so as to permit the lateral movement of the bolster, independent

of the step, as in the original Rabbeth. A turn or partial turn of the bolster raised or lowered it, and slots or notches.

in the bolster and step, and locking pins in the case both in an adjusted position when the proper fit was

This form of spindle was the standard till the spring of 1888, when it was decided to lengthen the spindle and its bolster, to give a greater length of bearing.

No substantial changes have been made since that time, and the operation of the spindle is so near absolute perfection that it would seem as though none were required.

I have omitted to mention another invention which has been used from the first on the Rabbeth and Sherman spindles, made by William T. Carroll of Worcester. On tapering spindles, prior to his invention, the bobbin was driven by the frictional contact of its bore with the outer surface of the spindle. With this arrangement the bobbins always stood at varying heights on the spindle, and were liable to be split. Carroll applied a cup to the spindle, which embraced the bottom of the bobbin, and drove it by external rather than by internal friction, thus obviating the difficulties mentioned.

Going back again, and referring to the "Sherman spindle," a modification was made in it about 1879, experimentally, by Mr. G. E. Taft, superintendent of the Whitin Machine Works, at Whitinsville, Mass. This modification was perfected and introduced in 1883, and has since been largely sold under the name of the "Whitin gravity" spindle. The main difference between this and the Sherman is as follows: Sherman's bolster is loosely fitted in the supporting tube, while Whitin's has, or is supposed to have, a sliding fit in its supportWHITIN SPINDLE. ing tube, without lateral motion, opposite the bolster bearing. There are some other differences in detail between the two structures.