A third motion was also necessary to insure the cf r tainty of the hook always catching the loop when draw n hack ; and as the working needle required a rotatory motion on its axis, it became necessary that a corres ponding motion should be given to the feuding needle, so that one should be always as nearly as possible in the directly opposite part of the circle to the other. The means by which all these different motion:; were effect ed, will be best understood by describing; the particular construction of the machinery behind; and this is more clearly represented in I'latc CXXXVI. XVI. rig. 6. than in any other part of the Plate. The frame w Lich contains all this part of the machinery, is r•prest used at the let ters A 13, and a small additional frame is placed behind at I', for a certain portion of the wheels. The is hole of this frame is to be lixed firmly upon the centre of the cross rail behind, and at right angles to the rail, so that the oblique point 0 may project towards the back part of the cloth, at a few inches (listance. The oblique projecting of the spindle, or axis at 0, is designed to answer the purpose of any common revoking crank, by moving round a given point, and can ing w irh it the bar iu which the feeding needles arc driven ; and this ruction is very simply effected by the perpendicular rack I) %rot king into the wheel C, which is loose upon the axis. The rack 1) is fixed to an upright or vertical rod of polished iron, sliding freely in tw o guides fixed to the rail or beam which carries the whole apparatus. The lower part of this polished rod is worked by the middle treddle, so that when the treddle is forced down by the wiper, the rack descending gives rather more than a full revolution to the wheel C. Attached to this wheel is a circle of brass larger in diameter than the wheel C, and furnished with a spring catch G, which rests in the wheel F, which is a ratchet. This, when C is pulled down the spring G, acting upon the ratchet F, which is fast upon its axis, carries it round, and as the spring does not work in the cont•ary direction, the rack returns for the next stroke, without bringing back the spindle ; which would undo what had been done. Thus the whole revolution of the spindle is uniformly one way without returning. This is all that is necessary for the rotatory motion.
That motion, by the feeding need:c is always placed in the opposite part of the circle to the we rking needle, is effected as follows, by the whc els in the small frame behind. On the extremity of the spindle is fixed the pinion I, which revolves along with it, giving motion to the wheel with which it is connected, and which is loose upon its axis. To this wheel is fixed the ratchet K, containing only two notches which bisect the wheel, one being at each extremity of its diameter. The spring catch 1, pressing on the circumfercnce of the wheel, catches each of these notches in succession ; the effect of w hich will be immediately seen. The wheel I. is fast upon the axi.. and is wrought by the rack 'AI, which is screwed to a cross bar deriving its motion from the largest traverse wheel. Upon the large spindle the pully at H carries a small band, passing two or three times round it, and from each end of which is suspended a weight. That represented at N, Plate CXXXVI. Fig. 6. is much heavier than the one suspended from the oppo site side, as its intention is to pull in a direction contrary to that of the impelling power. This excess of weight must be sufficient to pullback the band by friction, after the pulley has been stopped by the spring-catch I laying hold of one of the notches in the ratchet K. The effect of this apparatus is as follows. When the main spindle O has performed a revolution, or somewhat more, by the descent of the rack D, any excess is instantly cured by the joint operation of the balance-weight N, and the friction of the rack D in reascending ; and the position of the tooth in the ratchet K becomes the stop, by which the relative positions of the feeding and working needles are preserved. Merely for the sake of convcniency, the
calculation between the front rack for the needles and the back rack for moving the regulator K, was taken as two to one. Both derive their motion from the inverted levers KK, (Fig. 1.); and hence the range of one will always be proportional to that of the other. When the inverted levers arc shifted either way, the needles before, and the regulating ratchet behind, change positions by the same power, and at the same instant. The perpendicular rack gives always more than a full revolution to the feeding crank, but the excess is instantly corrected by the action of the counterpoise N, and thus the relative position of the two frames is uniformly and correctly preserved. When the motion of shift is in the contrary direction, the spindle 0 is carried with the rest of the apparatus, to which the moving wheel C, being loose, presents no obstruction.
Of the operation of that part which produces the se cond, or reciprocating motion of the feeding-bar, the most correct idea will be formed by referring to Figures 3. and 4. In Fig. 3, a profile elevation of it is given, in which almost every part may be distinctly seen. Let A represent a cross section of the beam which carries the whole feeding apparatus. From this beam descends two projecting arms ; one of which appears at B, and the other is in every respect similar. Their horizontal ap pearance appears behind 11, in Fig. 4. Between these aims is a cross shaft, placed so as to vibrate freely in the centres ; and at each end the perpendicular arms are joined be the cross shaft, the arms being fast upon the shaft. From the middle of the shaft, the lever E pro jects at right angles to the arms, and at the extremity is 11% ed a weight, in order to give it a tendency to de scend ; by which descent the perpendicular arms are im pelled forward towards the cloth. From the upper part of the perpendicular arms, a connection is formed with the back part of the feeding bar, by means of two wires, with a hook or eye-joint at each end, to prevent them fmm impeding the circular motion of the bar. When the lever E is at liberty, the weight descends, and the feeders spring forward close to the cloth. When the perpendicular shaft which carries the rack of the feeding inatlincry (1), Plate CXXXV I. Fig. 6.) reascends, a jointt.•I catch, represented at F, catches the projecting yin E, lifts it up, and produces the retrograde motion of the needles, in order to pull the threads close into the baths of the working needles. When the work ing needles are shut, and leaving the cloth, a small hammer, fixed to the back wooden roller, by striking a crank, forces back the catch ; and the projecting arm E descending, brings the feeding needles again close to the cloth. Thus the reciprocating motion is commu nicated to the feeding- bar. The it'll)) of the crank at 0 is useful for this motion, because it tends, when the bar recedes, to bring- back the threads much more fully into the barbs of the working needles, than a rectangular crank, where much of the obliquity would still be pre served, could effect. In the first experiments, cranks of the common kind, and similar in appearance to what is represented by the lower dotted lines, were employ ed, and afterwards abandoned for this special reason, and after their inefficiency had become apparent from actual experience.