MEMOIR OF DANIEL TREADWELL. 499 



Let US now suppose a roving of hemp to be passed into the tunnel l, and the pullc}- b to be 

 turned in the direction of the arrow ; the bobbins lying Ijetween the rows of hatchel-teeth must be carried 

 along with the hatchel-belt ; and as some of the bobbins, by their connection, in forming the bobbin-belt, 

 constantly pass out of the hatchcl-belt on the riglit, so others constantly pass into it on the left. When, 

 therefore, the roving of hemp, passing into the tunnel l, meets the hatchcl-points, it is pressed between 

 them by the bobbins, and passes along in the hatchel-belt until it arrives near the end f' of the rails 

 ff'. Now, as the rails pass beyond the pulley ii', to the riglit, and as the clearcrs (see Fig. 6), as has 

 been before described, pass on tiie top of the rails, it is evident that they will be raised upwards, upon the 

 hatchel-points ; and as the roving of hemp is always above them, this also must be lifted out of those 

 points. Suppose that the, end of the roving be tlien passed through the gatherer x, and that it be made 

 to enter between the drawing-rollers. If the drawing-rollers arc now turned faster than tlie hatchel-belt 

 carrying the roving, the roving must be drawn out through the hatchel-points by the rollers, forming a 

 filament as much smaller than the original roving as the rollers move- faster than the hatchel-belt. If, 

 however, the relative motions be constant, then tliese proportions will be true onl^- when the mean size 

 of the filament is compared with the mean size of the roving ; for the roving is not supposed to be of 

 equal size tliroughout, and, moreover, the ends of the fil)res of hemp cannot be distributed at equal dis- 

 tances in it ; hence more of these ends will be taken into the rollers together at some times than at 

 others. Now when manj' ends meet the rollers together, many fibres must be drawn out of the hatchel 

 together and the filament be made larger than at other places, where the ends of fibres, by being distrib- 

 uted in the roving at greater distances, meet the rollers less frequently. To prevent as much as possi- 

 ble this inequality in tlic filament, I use the apparatus now to be described, which I call the regulator 

 and tlie comb. 



A small cast-iron frame, eee (Fig. .3), is fixed upon the wooden frame a a. The shaft//" is made 

 to run in proper centres fixed in the frame ee, and placed upon it are the wiieels g, g' [the shaft, with 

 its connections, is drawn separate from the frame in Fig. 10]. These wheels are bored with a round 

 hole, and turn freelj- upon the shaft /y, except as hereinafter described, but are prevented b^' proper 

 collars from slipping in the direction of the length of the shaft. A second shaft, /* A, is made to run in 

 centres upon the same frame, ee, parallel toff. Tlie wheels i, i', are fixed firmly on the shaft h/t. The 

 ■wheels g and i' have fift3- teeth each, the wheels g' and i have seventy' teeth each, and the diameters of 

 all the wheels are in proportion to the number of teeth. [In the figure, g' is drawn too large, and i too 

 small.] The teeth of r^ p\ay into those of i, and tlie teeth ot g' into those of*'. 



There are two locking-pins, or catches, k,k' (Fig. 3), fixed upon the wheels (7, </'. Z, l' (Fig. 10) are 

 two clutch-boxes placed upon the shaft ff. These clutch-boxes are free to slide in the direction of the 

 axis of the shaft; but in the shaft are fixed pins which project into a groove cut lengthwise on the inside 

 of each clutch-box so that the clutch-box must always turn with the shaft. Several parallel grooves are 

 turned in the body of each clutch-box. A slide-bar, 00, is placed on the frame ee, parallel with the 

 shafts hh and ff and is capable of being moved in the direction of its length, and in that direction only. 

 Two arras, jOjo, are screwed upon the top of this slide-bar, and pass from it towards the shaft _/y, and ter- 

 minate directly over the clutch-boxes II, that is, one arm over each clutch-box. There are fixed to the 

 under side of each avm,p,p, met.al blocks or boxes, qq, which pass down to the clutch-boxes. The 

 under sides of q q are formed in the sliape of a hollow cylinder, and arc grooved so as to embrace the 

 clutch-boxes upon the parallel grooves formed upon them through about one third of their circumference. 

 [Figure 11 shows the slide-bar o o and axms p p, separate from the other works. The figure represents 

 the under side of these parts.] It will be seen tliat, when the shaft y/" is in motion, the clutch-boxes, b^' 

 reason of the pin in the shaft and the longitudinal groove in the chiteh-boxes, must move al.so, l)iit the 

 wheels g and g' , lieing loose on the shaft /|/", do not move. On eacli clutch-box is fixed a locking-pin, 

 »w and n, and also on each of the wheels g and g' a locking-pin, k and k' ; now if these locking-|)ins on 

 the clutcii-ljox and those on the wheels g and g' are l)rought in contact, these wheels must move witli the 

 shaft yy, and as they gear with the wheels i, i', fixed on the shaft /( A, tiic shaft must move also. Now 

 the elutch-boxes are connected by tlie two aruT; pp with the slidi'-bar n <>, and are moved with it to the 



