4 AMERICAN MACHINIST 
keep it hot, and there is no danger of igniting the turpentine, as there would be over a flame. Finally I will mentioP  , ol„ process, which, for pati,rns that have been finished to a nice even surface to start with, is superior to others, in my opinion. Bayberry tallow is a vegetable product, found as a film or blush on the outside of bayberries which ripen on low bushes in the autumn, and is separated from them by boil ing, they tallow then rising to the surface. When cold it is much harder than beeswax and somewhat lighter in color, having an aromatic odor and feeling very slippery to the touch. uses are quite varied, and it can usually be f, and in drug stores. To prepare it for use on patterns, melt it, adding enough turpentine so that when cold again it will be about the consistency of butter. The pattern should be first rusted or else covered with the copper solution previously mentioned, and then rubbed over thoroughly with the tallow. This leaves an extremely smooth, shiny surface, and it is clean to handle and has the advantage that it is not necessary to heat the mixture after being once made, in order to apply it, or the pattern either, without one chooses; thus a pattern can be kept in fine condition contin-ually if a molder will provide himself with a can of the preparation, and keep it on his bench and occasionally rub over the pattern with it before going home at night, when the next day it will be found hard and ready for use, as the turpentine will then be evaporated. While I may not have offered anything very original, still I believe that some points will be helpful to many. As these are probably not all the ways nor the only ones in use, I shall not go so far as to claim any one of them to be the absolutely best method which can be devised. 
Pouring, Flowing-off and Feeding Cast-ings—II. 

BY S. BOLLAND. 
dle of the swiftly rotating mass, and just how this may be accomplished is shown at Fig. 18; where a number of castings, directly con-- —,;ted with a central ball, may be fed with comparatively pure iron, with no possibility of dirt other than may be gathered within 

Eig .23. A B pipes or‘columns at the flanges, the gates A A being intended when it is desired to run down through the cope, and the one at B to be used when it is thought that the former plan would be too hard on the mold at that point. 

Ely. 19. far metal will reach in such wor method of running long columns one end co, ,...icts with the first prince filling molds with iron of equal temp ture, it is very evident that all long castings should be filled from both ends. The wis-dom of the above always strikes us the more forcibly when we see any viola-tion of these principles result in a cracked casting. Keep all risers away from brackets ; for should there be but a very slight commotion in the mold the bracket is sure to suffer if the disturbance finds a vent at that point. When the flask will admit of running round columns at the end it is by all means the best plan to adopt, and the best kind of runner for this purpose is shown at Fig. 22, where main runners A A are seen to connect with a circular runner B, cut round the bearing and entering the casting at one or more gates ample to run the column safely. Round columns will run 18 feet inch thick from one end, providing all other things are favorable ; but the remarks on square columns apply with equal force to round ones, and risks should not be taken. Sometimes it is found advisable to change the location of the runner ; if this must be done, choose some flange or collar into which the gates can be directed either on the side, as seen at C, or dropped down at D. Fig. 23 shows how to run a large wheel through the hub core. The cen-ter dry sand core, with a hole large enough to fill the mold at a proper rate, rests on another dry sand core in which the requisite gates have been prepared. To save making the bottom core, holes for gates may be made at A, indicated by broken lines, but this plan is some-what risky, if the noses against which the iron beats are not made in dry sand, as seen at B. It is plain that a wheel filled after this manner is preferable to any other, as it makes what is otherwise a critical job a 

Fig. 18. 
AinAmerican Machinist _Fig. 21. 

Pouring, Flowing-off and Feeding Castings—II. 
rt 

BY S. BOLLAND. 
Fig. 16 (September 29) illustrates a system of dams set before the castings when it is de-sired to produce clean work from a spray. When best results are looked for, all such gates should be connected with the patterns on a match board, so as to insure a good hard surface for the molten iron to pass over. Gates cut with tools are, as before stated, un-trustworthy, on account of the soft, broken surface yielding to the extreme heat to which they are subjected, and thus forming slag that invariably finds its way into the casting. - The form of the leader, in this instance, is a noteworthy feature; the iron entering at A travels rapidly along the smooth, round sur-face of the leader, passing the gates and out at C, carrying a large proportion of the dirt along with it; whatever portion remains is held on the upper surface of the leader whilst the casting is being fed from the bottom. The dams D, as seen, are formed with cores, and make "assurance doubly sure" by check-ing any inflow of dirt, should the pouring from any cause be lacking in force. There is no other casting that has helped the science of running as much as the gover-nor ball. During the early part of my ap-prenticeship this job was held back for some particular man who alone could be trusted with such an important job, and not unfre-quently have I known the best men to fail time after time to produce a casting that was clean all over when turned. The manner of running a ball which must be turned bright is shown at Fig. 17, where it is seen that the metal passes down A into the ball at B; the direction given the metal by this form of ingate causes the metal to re-volve rapidly in the mold, and this causes the lighter substances which gather on the sur-face to collect towards the center, as indicated by the arrows in the plan, to be ultimately ejected at the riser C. The principle involved to keep the ball clean must naturally suggest the propriety of filling other molds from such a ball whilst the dirt is being held a prisoner in the mid-

Fig. vo. _Fig. 22.  
_Fig. 25. 
Pig. 26 Fig. 28. POURING, FLOWING-OFF AND FEEDING CASTINGS. 
their own limits. Unless in large castings, there need be no riser on the ball when it is used for running purposes. Fig. 19 shows how the principle may be made general and used for almost any class of work. Fig. 20 illustrates two methods of running 
The regulation method of running square columns is shown at Fig. 21. All such col-umns will run from one end, under ordinary head pressure, 17 feet at one inch thick if the iron is in good condition ; beyond this it is unwise to go, especially if the column should be less than one inch. I speak now of how 
The ball shown at Fig. 25 is supposed to be 12 inches in diameter, and sup-pose that such a ball was cast (without riser) with hot iron, and left to cool in the same position it was cast; it is cer-tain that the upper surface would have fallen in, in proportion to the amount of shrinkage which would have taken place before the crust was firm enough to sustain itself; the amount of shrinkage would, of course, be according to the nature of the iron it was cast with. Now if this ball, when cold, was split in two, it would be found that the upper hemisphere would show a sponginess similar to that seen in the sectional illustration at Fig. 26. The figure 
by broken lines, but this plan is some-what risky, if the noses against which the iron beats are not made in dry sand, as seen at B. It is plain that a wheel filled after this manner is preferable to any other, as it makes what is otherwise a critical job a very safe one, and insures a good cast-ing every time, at least so far as the running is concerned. Fig. 24 is a plan and elevation of a spiral drum, or at least as much of it as will serve the purpose of showing how to arrange for a system of bottom gates when such gates must be made in the pit around a brick mold. Where the pits are damp it is absolutely necessary to have the gates protected from the moisture. The method shown needs no explana-B tion other than can be discovered by a careful examination of the figure. A A are the gates prepared in the mold, against which are set cores B B, these again being surmounted by other cores, as seen, until the top is reached. If the mold is unusually long, and there is danger of the metal becoming too sluggish to fill the upper parts of the mold correctly, then apply the gates on the top, after the manner as fully ex-plained in " The Iron Founder," page 163. The utility of feeding castings is questioned by some, but a little reflec-tion will, I am sure, lead all who deny its efficacy to see the erroneousness of their conclusions. 

 

pg-5  gouge, chisel and mallet are used to remove the bulk of the wood. As there is risk of the concussion of the blows splitting out the short grain, their direction should be rather diago-nally, and away from the rim outwards in the direction of the line A A, Fig. 62, and as more and more is removed, inclining to parallel. When the bulk of the stuff has been removed the curves of faces and flanks are set in a lit-tle way with paring gouge and chisel, and the edges whitened with chalk or reddened with red lead in oil. The flanks and faces are now worked straight across from one side to the other with paring gouge and chisel, and the setting in and coloring of the curves at the ends prevents risk of the lines being cut un der with the gouge and chisel. A small, thin, narrow straight-edge is used with coloring matter, to test the straightness of the flanks as the work proceeds—a quite essential pre-caution when the cutting is done with gouge and chisel, tools 'which do not possess the initial faculty of guidance possess( d by the plane. The plane, as we see, can only be used for the extreme faces, Fig. 63 A. When the teeth are worked accurately nothing more is done to them except to give them a good oiling with linseed oil with a brush or rag. This imparts a slight hard-ness and a good appearance to the surface. The marking out of bevel wheel cogs necessarily proceeds on different methods. Still there is no more actual difficulty, but only more time required with the latter. Since a bevel wheel is a frustum of a cone the tooth centers, roots, points, flanks, and faces all run in exact planes to the apex A of the cone, Fig. 64. The shapes of the teeth are marked out accordingly on the drawing from which the wheel is Made, and are transferred to the ends of the cogs. I need not go into the details of the operation, but the drawing, Fig. 64, will convey the whole , meaning of the process to the mind of a pat-tern maker. At B there is the sectional form of the tooth, and at C D the forms and di-mensions of their outlines on the large and small diameters respectively are seen pro-jected. The radii are not taken along the semi diameters a b c d, but along the project-ed radii e b, f 4, on lines at right angles with A e. These are then drawn at C D. The thickness of the cogs on the-small diameter is 
The small grooved pulley on the arm over which the slack side of the rope passes holds the rope up and causes it to encircle nearly the whole of the driving pulley, making it next to impossible for the rope to slip. It is manufactured by Albert L. Colburn, New Haven, Conn. 
Two-spindle Boring and Mortising Machine. 

This machine will bore two holes at once, 
between centers. Weight, abo-ut 500 pounds. Either of these machines can be furnished with hand-lever and a treadle. The manu-facturer is P. Pryibil, 498-510 W. 41st street, New York City. 
Two firemen who were on the steamer "Alice Blanchard " on a trip to the Yukon River brought back two novel Aleut canoes. The boats are made of rawhide supported by horizontal ribs The only opening is a circular hole in the center large enough for 
LETTERS FROM PRACTICAL MEN. 
Fly-wheels. Editor American Machinist : In the matter of fly-wheel diameter and weight, I like to apply Matthew Baldwin's old rule in locomotive construction : " Make it about right, and then throw in another ton of iron." With this it is suitable to remember J. A. Fay & Co.'s rule for horse-power to drive wood-working machinery: " A little too much belt power is about enough." ROBERT GRIMSHAW. 

SMALL PORTABLE DRILLING MACHINE. 
and the line joining the center of the two holes may be at any desired angle to the table, from horizontal, which is the position shown in the engraving, to the vertical, or where one spindle would be vertically over the other. The table can be set 10" below the centers of spindles. It slides .forward and back 
a man's body so that he can sit down. A flap of rawhide surrounds the hole, and when the occupant plants himself in the bot-tom of the craft he draws this flap up, fastening it tightly around his body under the arms. This makes the boat practically water-tight, and he can paddle along in the roughest kind of weather without danger of drown-
Why Are Gears Noisy ? Editor American Machinist : In reading " Gascon's Discourses Upon Some Shortcomings in Text-books," by Jarno, I am led to attempt to put a few thoughts of my own on the same subject be-fore the readers of the AMERICAN MACHINIST.

Take any of the text-books that are now published treating upon " Involute Gearing "—all give a theory of constructing the gear 8ooth, and gears so constructed fail to give good practical results. In fact, some will not mesh at all. What the average ma-chinist wants is this gear question discussed in plain, every-day talk. Most writers, in explaining the theory of gearing, use technical terms, which, while probably correct, are understood by but few of the mechanics having, anything to do with gearing. The writer is one of that class who repre-sent the practical side of gearing, and what I shall say will be the result of a long and varied practice. First, I will ask the professors a few ques-tions. Can they tell us what amount of interference there is in the true involute in a pair of gears, say 3 P 19 and 69 T, and can they tell how to draw the form of teeth for these gears so they will run smoothly and quietly?. Why do gears make a noise ? How should the teeth bear upon each other to prevent a noise ? Which gear is the best and smoothest run-ning—the 144-°, 18°, 20°, or 24° pressure angle, and why ? T