1

theory of the modern velocipede than those who study lectures like those delivered by Mr. Boys, or are influenced by the personal experience of riders. The simple fact is that there is no such thing as a perfect machine in the market. There are certain designs which are so far perfect in their way that they are the best for certain purposes, and for the rest the personal likes and dislikes of the intending purchaser must decide the matter; but anyone who expects to find a machine capable of all sorts of work and free from-we will not say defects, but-disadvantages, runs great risk of being disappointed; although at the same time the velocipede of his choice may be a very excellent one when confined to the work it is constructed to accomplish. It is not the theory that is at fault, but the practice, which expects what is next to the impossible.

PRACTICAL NOTES ON PLUMBING.– XCVII.*

By P. J. DAVIES, H.M.A.S.P., &c.

(Continued from p. 25.)

HOULD you require to cut out an old pipe, and have no such tools at hand, then, with a cold chisel and hammer, cut off, or rather split, the socket. Two hammers sometimes will do this, that is by placing one hammer at the back of the socket and hammering away at the front of the socket, and occasionally using the cold chisel at the end, trying your best to split it. You are sure to succeed if you try. Having cut your pipe, next begin to prepare it for screwing. First wipe the face free from oil, to prevent the file getting clogged, and with a 14in. or 16in. bastard cut, flat taper (or a three-cornered file may be used) file down the end of the pipe until it is true and smooth all round. Now take the stocks and

dies, Fig. 449. These solid dies are the best

straight threaded pipe; though, of late, the oldfashioned dies are fitted with this guide. Having cut the pipe and thread on the same, next fit together one piece of lin. barrel with a the outgo of the Tee to be fitted on the end of bend at one end, and an equal Tee at the other, the pipe, the extreme length to be 5ft. 3in.; the one end of the straight way of the Tee to be plugged, the other to be fitted with a nipple and elbow, and all as shown at Fig. 451. Now, it

B

IR

F

Fig. 451

may happen that this Tee will not screw on to your pipe, nor can you, perhaps, get the plug to enter the Tee socket or elbow. If this is the case, the thread of the pipe may be burred, which is done sometimes by thoughtless persons dabbing down the ends of these pipes and sockets on hard

carefully selected, or had from the same maker as your pipes. B, Fig. 448 K, is the plug tap for making running sockets, &c. Fig. 448Q is a wrench for working the taps and reamer. Now, your pipes or fittings in as straight a line as in order to screw this properly together, keep possible, so that the threads may properly enter each other, and without getting on the cross, which, if proceeded with, will totally destroy the threads, and thereby will be sure to cause leakage. For screwing together lin. barrel, &c., two pairs of tongs are required. These tongs should be lin. and 1in.; the lin. to grip the barrel R, Fig. 451, the 14in. to grip the socket CA, and the Tee at 4 and F. These tongs are illustrated at Figs. 452 and 453, which shows them placed for screwing up the socket and pipe. They are also shown at A B, Fig. 446. B is put on to unscrew the pipe, and A to unscrew the socket. If the socket is required to be screwed up, reverse the position of the tongs by turning them upside down.

sort, as they generally cut quickly to one gauge, and with a taper thread. Now place plenty of oil in the threads of the dies, and also upon the pipe; place the dies on the pipe and push hard and evenly, so that they may catch hold of the pipe, then gradually work them up the pipe by turning them forward and backwards, occasionally oiling the dies, so that they cannot become dry or too hot, to soften them, &c. Keep at this until the thread is of the desired length, which should be the length of the thickness of the dies. The thread should be, as before remarked, a little tapering, so that it may tighten itself as it is screwed up the pipe. This stock, if what is known as in. stock, may be fitted with fin., in., in., in., dies, sometimes only in.,

FIG. 450.

B

Fig. 453.

H

fin., in., and at other times in., in., in., according as ordered. The next size is in., in. lin.; next, lin., 1in., 1in., or 1in., 1in., and 2in. My sets run from in., fin., Jin., in. The next size is in., in., lin. The next lin., 1in., 2in., made by the inventor Cowley. I have also a set of the old pattern (Fig. 450), but they do not get much used, as the men prefer the solid dies, which do not require half the labour nor time to work them; besides, the solid dies have a guide C, which insures their being put on parallel with the pipe, and therefore gets a

• From the Building News, All rights reserved.

illustrates a very handy wrench, which will take in almost all sizes. It is especially handy for unscrewing nuts, and also useful in pipe work.

Parrot-Nose Tongs.

These tongs for their action also depend upon the same principle as the universal tongs, the difference being that the jaw J works through through the jaw A in a more circular direction, as may be seen at J K. These parrot-nose tongs are handy for taking out to do jobbing work, as they will fit many different-sized pipes. There are scores of others of this class; but for good hard work I prefer the old pattern tongs (Fig. 452). Sometimes it will so happen that you have not tongs with you, and that you want to unscrew a pipe, but you have a good screw wrench, as shown at Fig. 455 and 455A, which is the constant companion of the jobbing plumber. (I say jobbing plumber not out of any disrespect to him, for I consider a good jobbing hand to be the most practical amongst us that is, if he knows his trade; but unfortunately many jobbing plumbers are to be found who are incapable of doing good work when called upon, and who are the means of sadly disgracing our trade generally.) You will no doubt say that the wrench is useless for such work; but this is not so—it is of great use, and I will show you how to make it answer very well on a pinch. Screw the wrench to fit the pipe loosely, now get a round pin of iron or steel, such as a small piece of rat-tail file, or a piece of in. or in. iron wire, &c., and place it by the jaws of the screw wrench and side of the pipe, as shown at the round black speck B, Fig. 455, in just such a manner that this pin will firmly grip between the jaw of the wrench and pipe, and in such a manner that the more you pull the wrench round (in the direction indicated by the arrow M), the tighter it will become, and you will find no difficulty in unscrewing the pipes. You should be careful at this work, otherwise, if you go incautiously to work, it may slip and hurt you.

supposition should be preferred to others, as it is struction of the double o.g., therefore, appears not in any way connected with any necessary to me excellent. We have, accordingly : quality of a good telescope, but at most facilitates the calculations, which cannot be considered an advantage in investigations of this kind.

=

1.04394)

r1 == 3.29651

=

=

radii, crown lens radii, flint lens. It would lead too far in the present place, even Focal length B′ = 3.702231, aperture 0.09973 B'. only historically, to enumerate all the different § 9. In concluding this subject I will adduce suppositions which I have tried in order to attain my object. Suffice it, therefore, to say that the the example which has given rise to this inquiry. final result of my investigations hitherto gone An artist of this place brought to me two pieces = μ 1.53, 1.60, into, leads me to think that it is the most ad- of glass, for which I found ·004, du' 008; therefore, λ= 0·5. vantageous to use the liberty which we have ac-du cording to § 4, in the choice of the two first radii Although one had a number of veins and waves, in such a manner as to obtain the greatest he yet resolved on using them for a double o.g., quantity of light in the telescope; that is to say, and requested me to give him the dimensions of to make the aperture of the o.g. as great as it. I could have wished that he would wait for possible, and I will, therefore, shortly explain a better piece of glass, which in the end he will the process which is to be followed for this probably do. However, I communicated to him for these two glasses the following results, derived purpose. from the formulæ contained in 8, which appear to deviate very much from those hitherto used, especially with regard to the last radius r1, and the very large aperture. 10°,

As it is necessary in order to obtain the greatest possible aperture to make the first radii equal, we obtain by the well known optical formulæ—

With the above values of μ, μ', λ and a = With this value one may calculate the quan-01, I found the four radii, r1 = r1 = = 1.06, tities B and (B) by the seven equations of $2, r3 = = 1-04266, r1 = + 76-10095; so that the which, however, may be taken more conveniently first lens is bi-convex, the second concavefrom the following small table. Here the first convex. For this construction I found the angle of incidence is supposed 10°, and the focal length :thickness of the crown lens t = 01, its focal length being unity.

1.52

B

=

0.94380

0 94261

0.94141

0.94019

11 22 14'9

1.53 1'54

1.55 1'56

10 56 39 1 11 9 26.5

The following equations will now give M and that one might still desire would be a ready means B':

Coddington's formulæ supply this want in as easy and concise a manner as can possibly be desired. By them, the proper ratio of radii for either lens may be found, which shall approximately balance any ratio selected for the other lens.

I must remark that Littrow has not shown. how the heterogeneous marginal chromatism may be most effectually eradicated. It seems to me that the error arising from this cause must be more obtrusive than that developed by the central rays. For just as the spheric error becomes more violent as the aperture is increased, so also must the dispersive error be more apparent, since it is only an extended case of spheric error. And, although Littrow has conclusively shown how the mean central, and marginal rays may be made 3702231 for mar- to unite, it is certainly not by any means safe to ginal rays. infer that the same condition exists amongst 3.702292 for central the coloured rays, except those incident very near to the axis. The construction of o.g. which Littrow prefers, has the merit of giving a convenient shape to the crown lens; but, beyond this I fail to see any superior advantage. No doubt the plan is a good one; but that it is better than others often used, is not very clear. The other two constructions mentioned by him must have been found satisfactory in practice, as they are both found in the best work of the present day. Bohnenberger's ratio of 2 to 3 agrees with that used in what I have termed the English form of object-glass. Klugel's choice of 1 to 3 may be used with advantage with some species of glasses; but in selecting a form for the crown lens, it must not be forgotten that its influence on the shape of its companion flint lens must also be taken into account. If not,

Difference under-correct ⚫000061

be brought away from these cores; so to ensure a communication with the outer air we will make the lightening cores unite with the bar cores by the round holes indicated in section, Fig. 159; therefore one core-box will fill up print, take out bar and lightening cores at once, and all the air will be drawn away at once into the print.

Loose Pieces.

Take a plate of the annexed shape (Fig. 165). For some cause or another, generally for the purpose of securing a sound face, as in a surface the proper way is to leave the deep ribs loosely plate, this must be cast face downwards. Then dowelled on. We do so in order that they may be lifted up with the top sand, to be properly rapped and drawn out afterwards, rather than that the sand should be torn away from the ribs and entail a lot of mending up. If these ribs were shallow the necessity for leaving them loose would not exist; but, in the case of deep top lifts, there should be no exception to this rule. Somewhat akin to this is the case of a pipe or cylinder. We know the rule is almost invariably to make the patterns of these in halves, jointed longitudinally. The top half of the pattern is then lifted in the top box and rapped and drawn away from the sand. Occasionally, however, we see a pattern made solid, either from ignorance, or from a false idea of economy, and then the evil results of such a method are apparent in broken edges in the sand, and lapping joints in the casting, the result of imperfect mending up. The evil is seen at its maximum where a loam pattern is used. Here a joint is not practicable, and the rough loam tears the top mould so much that mending-up pieces are always necessary.

Further, if we take a common trolly wheel, though the back boss is almost as often fastened on as left loose, it is better to leave it loose, with a stud in the centre. For, although, owing to the abundant taper given to such bosses, the box will often be lifted away without causing fracture of the sand, yet a little unsteadiness in lifting will certainly break the mould, and then the boss must be unscrewed from the pattern and embedded in the broken mould for the purpose of mending it up. The advantage becomes still more obvious if there are deep arms (Fig. 166) in addition to the boss in the top, so that in bevel wheels, deep spur wheels with flat arms in the centre of the rim, and similar instances, the portion in the top should be dowelled on: boss and arms usually being fastened together and lifted out as one (Fig. 167.)

Stopping Off.

This is a term we have had occasion to use several times in the course of these articles. It is a term in common use in the workshop, and means broadly the modification of the shape of a mould, so that the casting shall be, to the extent

of such modification, unlike the pattern from

which it is made.

From these illustrations it is apparent that the principle is capable of almost indefinite application. There is, in fact, scarcely a limit to its made from them, either deeper or shallower, adaptability. Pattern wheels may have castings without altering the depth of the pattern itself. In the first instance the wheel is drawn; in the second, strickled. That is, supposing the pattern were 3in. deep, and the casting were wanted 4in., the sand would be rammed level with the top faces of the teeth, then the pattern would be drawn up as in delivery, but just lin., and the sand rammed level again. The pattern then finally drawn would leave the mould 3in. + lin. =4in. deep. Or, in the second case, the pattern being 3in., we want the casting 2in. deep. Then the sand, as before, being rammed level with the top faces, a strickle shouldered lin. deep is worked around and between the teeth, making a new joint 2in. from the bottom of the pattern (Fig. 171). of side frames, cheeks, and are altered as occasion arises by this simple method of stopping off. Thus a wall bracket like Fig. 172, A would easily be cast like Fig. 172, B in the manner shown, the shaded parts representing the new pieces which indicate the shorter length required, and the piece C the stopping-off templet for the use of the moulder. In the economy of the trade this principle of stoppingoff is of the utmost importance.

Even the outlines brackets

Skeleton Patterns.

Loam patterns may be considered as makeshifts; but, apart from loam work, there are methods of making some pieces of casting without a complete pattern. Thus, if we wanted a ring somewhat large in diameter, and moreover for a temporary job only, we should make a sweep (Fig. 173) of the same section as the ring, and a foot or so in length. To this we should attach a radius bar B, in which a hole would be bored corresponding in position with the centre of the ring. A stake Ĉ driven into the sand would form a fixed centre, around which this would swivel on a bit of wire or hard wood pin d, while the moulder rammed the sweep up a sufficient number of times in succession to complete the circle. Or we may omit the bar, and simply mark two circular trammel lines on a true sand bed, by which to set the sweep.

As a rule, a whole pattern would not be made for a large rectangular casting such as a balance

weight for a crane, or a boilermaker's or bith's levelling block. In these cases, a frame only Much can be done in this way by the moulder, would be screwed together (Fig. 174), rough inand considerable expense saved thereby. A com- side, but corresponding outside with the block, in paratively small amount of labour will often length, width, and depth. A level bed would be save the expense of a new pattern, or that of ex-made in the sand, then the frame laid upon it, tensive alterations to one already in stock. Take, the sand rammed round flush with its top edge, for example, the simplest case we can instance. and strickled off level. The pattern would then Say we have a solid plunger pattern, 12ft. long, 6in. be withdrawn, cores put in by measurement, and diameter, and we want a casting of the same a plain top box, rammed on the foundry floor and diameter, but only 10ft. long. If we cut the struck off level, would cover the mould over. pattern off, we spoil it for the greater length if These are simple illustrations, but the principle required at any future time; but neither should is often extended. The cast-iron centres for we make a new 10ft. length. So we simply mark the wrought-iron truck frames of cranes are made a length of 10ft. in the joint of the pattern, and in this way (Fig. 175.) The curb ring facing, boss, provide the moulder with a semicircular piece of and bed are strickled by a board working round the same diameter as the plunger and two or a core bar, the skeleton pattern frame is rammed three inches long. Before the pattern is drawn around on this, and the top and bottom arms, out, he carries the 10ft. mark across the sand and middle ribs and boss are formed by cores. joint both in top and bottom, then withdrawing Neither in making a plated casting is a whole the pattern, he lays the face of the stopping-off pattern necessary. The familiar form of a tank piece level with the mark, and fills up the mould plate will illustrate this. A frame is made beyond that face with sand (Fig. 168.) corresponding in outside dimensions and in thickInstead of a solid cylinder, let it be one hollowness with those of the plate, and to this frame the in section, say a pipe 5in. in bore by 6in. outside flanges for bolting are screwed. When the pattern diameter. The stopping-off piece will then take is rammed up and turned over, its position is that rethe form of Fig. 169. The half-circle, 6in. presented in Fig. 176. The sand within the frame is diameter, will, as in the last illustration, fit the strickled over level with its upper face. The top mould of the pipe body, but it will have a 5in. box is then placed over it, and upon the level half print screwed upon it. This, then, laid in the and hard-rammed surface of the frame the top mould and rammed round with sand will give sand is rammed (parting sand of course interboth the length of the pipe and the shape and vening). The top is then removed, and the sand size of its print also. within the frame is scraped out with a strickle, shouldered to the same depth as the thickness of the plate, after which the pattern is drawn out. This method is also applied in modified forms to core-boxes as well as to patterns. In making

If, instead of a plain pipe, we have one double flanged, we should proceed in this way. A pipe, say as before, 6in. diameter by 12ft. long, has to be be made 8ft. long, and at each end of the Sft. length

FLEXIBLE PHOTOGRAPHIC PLATES. HOTOGRAPHERS, both amateur and proPfessional, have long wanted some thoroughly supply the want with more or less of success. The efficient substitute for glass as a support for dryplate films, and a few attempts have been made to following method has been recently patented in this country on behalf of Messrs. Fickeissen and Becker, of Villingen, Baden. The plates or surfaces can be prepared from paper, cloth, or other suitable fabric or material, but by preference from white paper containing very little size and not much grain. This paper is first extended on a frame, or other arrangement, according to the size of the plate or surface which is desired. After it is dry the surface varnish or composition, such as copal varnish, for is covered in any convenient manner with a fine the purpose of rendering the fabric transparent; it is then dried, and after it is quite dry the surface is rendered smooth by the application of powdered pumice stone or other suitable material, or it may be smoothed by suitable machinery. This process of smoothing may be repeated, if necessary, two or three times until the surface or plate is smooth or so prepared is then transparent. The surface covered on one or both sides with a solution of gelatine, isinglass, or other substance possessing similar properties, and allowed to dry. The surface so prepared may, if desired, be further treated with a preparation of ox-gall from which the fatty matter has been extracted by acetate of alumina or similarly acting agents which will precipitate the fat of the gall, the resulting preparation being then passed through a filter, whereby a clear solution will be obtained with which the plate or surface may be covered, so as to secure the safe reception of the emulsion for photographic or other use. Instead of ox-gall any similarly acting substance or material may be used. The plates or sheets prepared as above may be used with great advantage in reproducing photographs from nature in lines or stipples for calico and other printing, as the stipples or lines can be printed first on the material before it is made transparent. Any photographic design or drawing can be put on the transparent surface in the usual way, and by using the film as a negative or positive in photographing from nature or from a drawing half-tones will be reproduced in lines and stipples available for any kind of printing. also be used as surfaces upon which can be printed As these plates or sheets are waterproof they can or produced all kinds of ornamental and useful work.

PLATING WITH NICKEL.-I.* ICKEL-PLATING is an American industry, N in the sense that it was first successfully practised on the commercial scale in the United States, and here received that practical demonstration of its usefulness that has since made it the most successful and most widely practised branch prominence and popularity about ten years ago, it of the art of electro-plating. Coming first into has since that time rapidly grown, until to-day it has developed into an industry of great magnitude. The almost silvery whiteness and admirable brilliancy of electro-deposited nickel; its cheapness as compared with silver; the hardness of the electrodeposited metal, which gives the coating great power to resist wear and abrasion; the fact that it is not blackened by the action of sulphurous circumstance that it exhibits but little tendency to vapours which rapidly tarnish silver; and the oxidise even in the presence of moisture, are sufficient to explain the great popularity which nickel-plating enjoys.

The industrial development of the art, however, which has been surprising both in respect to its rapidity and extent, may be attributed in a large measure to certain favouring circumstances, quite independent of the excellent adaptability of the metal for electro-plating purposes. These circummade within the period above named in the prostances are: first, the great advances that have been duction of nickel on the commercial scale, by which the cost of the metal has been greatly reduced and its purity greatly increased, for which we are indebted largely to the American Nickel Works of Camden, N. J., under the scientific management of Mr. Joseph Wharton; and, second, the introduction and great improvement within this period of the dynamo-electric machine, which placed at the disposal of electro-platers a constant, powerful, and cheap source of electricity, in the place of the un

By WILLIAM H. WAHL. A paper read before the Chemical Eection of the Franklin Institute.