Electric Bells and All About Them Part 5

_Conductors._ _Insulators._ Silver. Paraffin Wax.

Copper. Guttapercha.

Iron. Indiarubber.

Bra.s.s. Sh.e.l.lac.

All Other Metals. Varnishes.

Metallic Solutions. Sealing Wax.

Metallic Salts. Silk and Cotton.

Wet Stone. Dry Clothing.

Wet Wood. Dry Wood.

Oil, Dirt and Rust.

See also the more extended list given at -- 5 for a more complete and exact cla.s.sification.

It will be seen, on reference to the above, that copper is a good conductor, being excelled by silver alone in this respect; and that silk and cotton are bad conductors. When, therefore, a copper wire is bound round with silk or with cotton, even if two or more strands of such a covered wire be superimposed, since these are electrically separated by the non-conducting covering, no escape of electricity from one strand to the other can take place, and the strands are said to be insulated. If the copper wire had been coiled _naked_ round a bobbin, each convolution touching its neighbour, the current would not have circled round the whole length of the coils of wire, but would have leapt across from one coil to the other, and thus the desired effect would not have been obtained. A similar result, differing only in degree, would occur if a badly insulating wire were used, say one in which the covering had been worn in places, or had been badly wound, so as to expose patches of bare copper wire. If the insulation of a wire be suspected, it should be immersed in hot melted paraffin wax, and then hung up to drain and cool.

The size of wire to be used on a 2-1/2 in. bell should be No. 24 B. W.

G., the size falling two numbers for each 1/2 in. increase in the diameter of the bell. In these wires the higher the number, the finer the size, No. 6 being 1/5 and No. 40 being 1/200 of an inch in diameter.

Silk-covered wire has an advantage over cotton-covered wire, inasmuch as the insulating material occupies less s.p.a.ce, hence the convolutions of wire lie closer together. This is important, as the current has less effect on the iron if removed further from it, the decrease being as the _square_ of the distance that the current is removed from the wire.

Magnets coiled with silk-covered wire admit also of better finish, but for most purposes cotton-covered wire will give satisfaction, especially if well paraffined. This wire must be wound on the bobbins, from end to end regularly, with the coils side by side, as a reel of cotton is wound. This may be done on a lathe, but a little practice will be necessary before the inexperienced hand can guide the wire in a regular manner. If, however, the spool of wire have a metal rod pa.s.sed up its centre, and this be held in the hand at a distance of a foot or more from the bobbin on the lathe, the wire will almost guide itself on, providing the guiding hand be allowed to follow its course. With a little care, the wire for these little magnets may be wound entirely by hand. Before commencing to wind on the bobbins, just measure off 8 in.

of the wire (not cutting it off) and coil this length around a pencil, to form a small coil or helix. The pencil may then be withdrawn from the helix thus formed, which serves to connect the wire with one of the points of contact. This free end is to be fastened outside the bobbin by a nick in the head; or the 1/8 in. length, before being formed into a helix, may be pushed through a small hole made on the head of the bobbin, so that 8 in. project _outside_ the bobbin, which projecting piece may be coiled into a helix as above described. The wire should now be wound exactly as a reel of cotton is wound, in close coils from end to end, and then back again, until three layers of wire have been laid on, so that the coiling finishes at the opposite end to that at which it began. To prevent this uncoiling, it should be fastened by tying down tightly with a turn or two of strong silk. The wire should now be cut from the hank, leaving about 2 in. of free wire projecting at the finis.h.i.+ng end of each bobbin. In cases where many bobbins have to be wound, either for bells, for relays, or for indicator coils, a device similar to that ill.u.s.trated at Fig. 21 A may be employed. This _electric bobbin winder_ consists in a table which can be stood on a lathe or near any other driving wheel. Two carriers, C C, somewhat similar to the back centre and poppet head of a lathe, hollow inside, and furnished with a spring and sliding piston spindle, stand one at each end of this table.

The sliding spindle of the one carries at its extremity a pulley, A, by means of which motion can be transmitted from the band of the driving wheel. The sliding spindles, B B, are fitted with recesses and screws, H H H H, by means of which the temporary wooden cores, or the permanent iron cores, of the bobbins can be held while the bobbins are being wound. The bobbin is placed as shown at D; a flat piece of metal, E, hinged at G, presses against the bobbin, owing to the spring F. The centre figure shows details of the carrier, C, in section. At the bottom is shown the spool of wire on a standard L. The wire pa.s.ses from this spot between the two indiarubber rollers, M M, on to the bobbin D.

[Ill.u.s.tration: Fig. 21 A.]

When the bobbins have been wound, they may be slipped over the magnet cores. They should fit pretty tightly; if they do not, a roll of paper may be put round the magnet cores, to ensure their not slipping when the bell is at work. The helix ends of the bobbins should stand uppermost, as shown at Fig. 22 A. A short length of the lower free ends of wire (near the base or yoke) should now be bared of their covering, cleaned with emery paper, twisted together tightly, as shown at Fig. 22 B, soldered together, and any excess of wire cut off with a sharp pair of pliers. To prevent any chance electrical leakage between this bared portion of the wire and the iron, it should be carefully coated with a little melted guttapercha, or Prout's electric glue.

[Ill.u.s.tration: Fig. 22.]

Of course, if the operator has any skill at winding, he may wind both bobbins with one continuous length of wire, thus avoiding joins, taking care that the direction of the winding in the finished coils be as shown at Fig. 22 B; that is to say, that the wire from the _under_ side of one bobbin, should pa.s.s _over_ to the next in the same way as the curls of the letter [rotated S].

[Ill.u.s.tration: Fig. 23.]

[Ill.u.s.tration: Fig. 24.]

[Ill.u.s.tration: Fig. 25.]

[Ill.u.s.tration: Fig. 26.]

[Ill.u.s.tration: Fig. 27.]

The part that next claims our consideration is the _armature_, with its fittings. The armature is made out of 5/16 square bar iron, of the best quality, soft, and well annealed, and filed up smooth and true. The proportionate length is shown at Figs. 23 and 24; and the size of the iron for other bells is regulated in the same ratio as that of the cores. Two methods of making and attaching the springs and hammers are shown. Fig. 24 shows the section of an armature fitted with back spring and contact spring in one piece. This is cut out of hard sheet-bra.s.s, as wide as the armature, filed or hammered down to the desired degree of springiness, then filed up true on the edges. It may be attached to the iron of the armature, either by soldering, by rivetting, or by means of two small screws. Rivetting is, perhaps, the best mode, as it is not liable to shake loose by the vibration of the hammer. The spring at its shank end may be screwed or rivetted to the bracket. Mr. Edwinson considers this the better form of contact spring. The other form is made in two pieces, as shown at Fig. 23, where two strips of hard bra.s.s are cut off, of the width of the armature, and the edges filed. A slot is then cut in the back end of the armature to receive the two bra.s.s strips, and these are soldered into it. The top strip is then bent back over the armature to form the contact-spring, the other strip being soldered or rivetted to a small bracket of angle bra.s.s. In either case a short rod of stout hard bra.s.s wire is rivetted or screwed into the free end of the armature, and to the end of this rod is screwed or soldered the metal bead, or bob, which forms the hammer or "clapper" of the bell.

The next portion to be made is the contact pillar, or bracket, with its screw, as shown at Fig. 25. This may either be a short stout pillar of 1/4 in. bra.s.s rod, about 1 in. high, tapped on one side to receive the screw, which should be fitted with a back nut; or it may, as shown in the figure, be made out of a stout piece of angle bra.s.s. The exact size and length of the screw is immaterial; it must, however, be long enough to reach (when put in its place behind the contact spring) the spring itself, and still have a few threads behind the back nut to spare. The screw should be nicely fitted to the pillar, and the lock nut should clench it well, as when once the adjustment of the parts is found which gives good ringing, it is advisable that no motion should take place, lest the perfection of ringing be interfered with. Some makers use a "set screw" at the side of the pillar wherewith to hold the contact screw; others split the pillar and "spring" it against the contact screw; but, all things considered, the back nut gives the greatest satisfaction. When the bell is in action, a tiny spark is produced at every make and break of contact between the contact spring and this screw. This spark soon corrodes the end of the screw and the back of the spring if bra.s.s alone is used, as this latter rusts under the influence of the spark. To prevent this, a piece of platinum must be soldered or rivetted to the spring, at the point where the screw touches, as shown at Fig. 26, and also at the extremity of the contact screw itself. It is better to rivet the platinum than to solder it, as the platinum is very apt to absorb the solder, in which case it rusts quickly, and the goodness of the contact is soon spoiled, when the bell ceases to ring.

To rivet the platinum piece on to the spring, as shown at Fig. 26, it is only needful to procure a short length of No. 16 platinum wire, say 1/8 in., then, having drilled a corresponding hole at the desired spot in the contact spring, put the platinum wire half way through the hole, and give it one or two sharp blows on an anvil, with a smooth (pened) hammer.

[Ill.u.s.tration: Fig. 28.]

[Ill.u.s.tration: Fig. 29.]

This will at once rivet it in its place, and spread it sufficiently to make a good surface for contact. The screw must likewise be tipped with platinum, by having a small hole bored in the centre of its extremity, of the same diameter as the platinum wire, which must then be pushed in, and rivetted by hammering the end, and burring the sides of the screw.

Whichever method be adopted, care must be taken that the platinum tip on the screw and the speck on the contact spring are adjusted so as to touch exactly in their centres. It will be hardly worth while for the amateur to cast or even turn up his own bells (which are generally of the cla.s.s known as clock gongs), as these can now be procured so cheaply already nickelled (see Fig. 28). The bell must be adjusted on its pillar (see Fig. 29^A), which is itself screwed into a hole in the base-plate, where it is held by a nut. The adjustment of the bell is effected by placing it over the shoulder of the pillar, and then clenching it down by s.c.r.e.w.i.n.g over it one or other of the nuts shown at Fig. 29. The bell should clear the base, and should be at such a height as to be struck on its edge by the hammer or clapper attached to the armature, Figs. 23 and 24. We still need, to complete our bell, two binding screws, which may take either of the forms shown at Fig. 27; and an insulating washer, or collar, made of ebonite or boxwood, soaked in melted paraffin, to prevent the contact pillar (Fig. 25) making electrical contact with the metal base. The best shape to be given to these washers is shown at Fig.

30. They consist in two thin circlets of wood or ebonite, that will just not meet when dropped, one on the one side, and one on the other of the hole through which the shank of the contact pillar pa.s.ses when set up on the base-plate. If a wooden base be used below the metal base-plate, then only one washer, or collar, need be used--that is, the one _above_--since the screw of the pillar will pa.s.s into the wood, and this is not a conductor. If the metal base alone be used, both washers must be employed, and a small nut (not so large as the washer) used to tighten up and hold the pillar firm and immovable in its place opposite the contact spring.

[Ill.u.s.tration: Fig. 30.]

Having now all the parts at hand, we can proceed to fit them together, which is done as follows:--The bell pillar, with its bell attached, is fastened by its shank into the hole shown near B, Fig. 17, where it is screwed up tight by the square nut shown at Fig. 29 _c_. In the same manner, we must fasten the contact pillar, or bracket, shown at Fig. 24 A. Whichever form be used, we must take great care that it be insulated from metallic contact with the metal base-plate by washers, as shown at Fig. 30 (similar washers must be used for the two binding screws if the _whole_ base-plate be made in metal). This being done, the metal frame, Fig. 18, is put in position on the wooden base, as shown at Fig. 17, and screwed down thereto by the screws indicated at _s s s_. The magnet may then be screwed down to the metal frame as shown. The small bracket of angle bra.s.s marked B, in Figs. 23 and 24, is next screwed into its place; that is, in such a position that the armature stands squarely facing the poles of the electro-magnet, but not quite touching them (say 1/16 of an inch for a 2-1/2 in. bell). In setting up this and the contact pillar, the greatest care must be taken that the platinum tip of the contact screw, Fig. 25, should touch lightly the centre of the platinum speck at the back of the spring, Figs. 23 and 24, shown full size at Fig. 26.

The free ends of the helically coiled electro-magnet wires should now be inserted into short lengths of small indiarubber tubing (same as used for feeding bottles), the extremities being drawn through and 1 in. of the copper wire bared of its covering for the purpose of making good metallic contact with the connections. One of these ends is to be soldered, or otherwise metallically connected, to the angle bra.s.s carrying the armature, spring and clapper, the other being similarly connected with the left-hand binding-screw, shown at Fig. 17. Another short length of wire (also enclosed in rubber tubing) must be arranged to connect the contact screw pillar Fig. 17, with the right-hand binding-screw. When this has been done, we may proceed to test the working of the bell by connecting up the binding screws with the wires proceeding from a freshly-charged Leclanche cell. If all have been properly done, and the connections duly made, the armature should begin to vibrate at once, causing the "bob," or hammer, to strike the bell rapidly; that is, provided the platinum tipped screw touches the platinum speck on the contact spring. Should this not be the case, the screw must be turned until the platinum tip touches the platinum speck.

The armature will now begin to vibrate. It may be that the clapper runs too near the bell, so that it gives a harsh, thuddy buzz instead of a clear, ringing sound; or, possibly, the clapper is "set" too far from the bell to strike it. In either case a little bending of the bra.s.s wire carrying the clapper (either from or towards the bell, as the case may dictate) will remedy the defect. It is also possible that the armature itself may have been set too near, or too far from the electro-magnet.

In the latter case, the clapper will not vibrate strongly enough, in the former the vibration will be too short, and the clapper may even stick to the poles of the electros, especially if these have not been carefully annealed. A little bending of the spring, to or from the magnets, will remedy these deficiencies, unless the distance be very much too great, in which case the bending of the spring would take the platinum tip out of the centre of the platinum speck.

[Ill.u.s.tration: Fig. 31.]

-- 43. Having thus constructed an efficient electric bell we may proceed to study its action and notice some of the defects to which it may be subject. In the first place, if we connect up the bell with the battery as shown in Fig. 17, viz., the left-hand binding-screw with the wire proceeding from the carbon of the Leclanche, and the right-hand screw with the wire from the zinc, then, if the platinum tipped screw touches the platinum speck, at the back of the contact spring, a current of electricity flows from the left-hand binding-screw all round the coils of the electro-magnets, pa.s.ses along the contact spring and platinum speck, thence to the platinum tipped screw along the short length of wire to the right-hand binding-screw, whence it returns to the zinc element of the battery, thus completing the circuit. The current, in thus pa.s.sing around the electro-magnet cores, converts them, _pro tem._, into a powerful magnet (see -- 13); consequently, the armature, with its contact spring and hammer, is pulled towards the electro-magnets and at the same time gives a blow to the bell. Now, if instead of having the platinum speck attached to a flexible spring, it had been attached bodily to the rigid iron armature, directly the electro-magnets felt the influence of the current, the platinum speck would have also been pulled out of contact with the platinum screw, therefore the electro-magnet cores would have _immediately_ lost their magnetism (see -- 13, last five lines). This would have been disadvantageous, for two reasons: 1st, because the _stroke_ of the hammer would have been very short, and consequently the ring of the bell very weak; and, 2nd, because, as even the softest iron requires some appreciable time for the electric current to flow round it to magnetise it to its full capacity, it would need a much greater battery power to produce a given stroke, if the contact were so very short. The use of an elastic contact spring is, therefore, just to lengthen the time of contact. But the electro-magnets, even when the flexible spring is used, do actually pull the platinum speck out of contact with the platinum screw. When this takes place, the circuit is broken, and no more current can flow round the electro-magnets, the spring rea.s.serts its power, and the contact is again made between the contact screw and contact spring, to be again rapidly broken, each break and make contact being accompanied by a correspondingly rapid vibration of the armature, with its attendant clapper, which thus sets up that characteristic rapid ringing which has earned for these bells the name of trembling, chattering, or vibrating bells.

-- 44. From a careful consideration of the last two sections it will be evident that the possible defects of electric bells may be cla.s.sed under four heads: viz., 1st, Bad contacts; 2nd, Bad adjustment of the parts; 3rd, Defective insulation; 4th, Warpage or shrinkage of base. We will consider these in the above order. Firstly, then, as to bad contacts.

Many operators are content with simply turning the terminal wires round the base of the binding-screws. Unless the binding-screws are firmly held down on to the wires by means of a back nut, a great loss is sure to occur at these points, as the wires may have been put on with sweaty hands, when a film of oxide soon forms, which greatly lowers the conductivity of the junction. Again, at the junction points of the wires with the contact angle bra.s.s and contact pillar, some workmen solder the junctions, using "killed spirits" as a flux. A soldered contact is certainly the best, electrically speaking, but "killed spirits," or chloride of zinc, should never be used as a flux in any apparatus or at any point that cannot be washed in abundance of water, as chloride of zinc is very _deliquescent_ (runs to water), rottens the wire, and spoils the insulation of the adjacent parts. If solder be used at any parts, let _resin_ be used as a flux. Even if any excess of resin remain on the work, it does no harm and does not destroy the insulation of any of the other portions. Another point where bad contact may arise is at the platinum contacts. Platinum is a metal which does not rust easily, even under the influence of the electric spark given at the point of contact. Therefore, it is preferred to every other metal (except, perhaps, iridium) for contact breakers. Platinum is an expensive metal, the retail price being about 30s. an ounce, and as it is nearly twice as heavy as lead (Lead 11. Platinum 215) very little goes to an ounce. For cheap bells, therefore, there is a great temptation to use some other white metal, such as silver, german silver, platinoid, etc.

The tip of the platinum screw may be tested for its being veritably platinum in the following mode: Touch the tip with the stopper of a bottle containing aquafortis, so as to leave a tiny drop on the extreme point of the suspected platinum. If it boils up green, or turns black, it is _not_ platinum; if it remains unaltered, it may be silver or platinum. After it has stood on the tip for a minute, draw it along a piece of white paper, so as to produce a streak of the acid. Expose the paper for a few minutes to sunlight. If the streak turns violet or pinky violet, the metal is _silver_; if the paper simply shows a slightly yellowish streak, the metal is platinum. The tip of the platinum screw must be carefully dried and cleaned after this trial before being replaced.

Secondly, as to bad adjustment. It is evident that the magnets and the armature must stand at a certain distance apart to give the best effects with a given battery power. The distance varies from 1/24 in. in the very smallest, to 1/8 in. in large bells. Sometimes (but only in very badly made instruments) the armature adheres to the poles of the electro-magnet. This is due to _residual_ _magnetism_ (see -- 14), and points to hard or unannealed iron in the cores or armature. As a make-s.h.i.+ft, this defect may be partially remedied by pasting a thin piece of paper over that surface of the armature which faces the poles of the electro-magnets. Another bad adjustment is when the platinum screw does not touch fairly on the centre of the platinum speck, but touches the spring or the solder. Rust is then sure to form, which destroys the goodness of the contact. To adjust the contact spring at the right distance from the platinum screw, hold the hammer against the bell or gong. The armature should now _just not touch_ the poles of the electro-magnet. Now screw up the platinum screw until it _clears_ the contact spring by about the thickness of a sheet of brown paper (say 1/50 of an inch). Let the hammer go, and notice whether the contact spring makes good contact with the platinum screw. This may be tried by the Leclanche cell as well, so as to make sure of the character of the _ringing_. When this has been satisfactorily adjusted the back-nut or set screw may be tightened, to insure that the vibration of the hammer shall not alter the adjustment. It sometimes happens that the spring that bears the armature is itself either too strong (or set back too far) or too weak. In the former case, the electro-magnet cannot pull the armature with sufficient force to give a good blow; in the latter, the spring cannot return the armature, with its attendant contact spring, back to its place against the platinum screw. To ascertain which of these two defects obtains, it is only necessary, while the bell is in action, to press the spring lightly with a bit of wire, first _towards_ and then _away_ from the electro-magnets. If the ringing is improved in the first case, the spring is too strong; if improvement takes place in the latter case, the spring is too weak. The third source of inefficient action, defective insulation, is not likely to occur in a newly-made bell, except by gross carelessness. Still, it may be well to point out where electrical leakage is likely to occur, and how its presence may be ascertained, localized, and remedied. If the wire used to wind the electro-magnet be old, badly covered, or bared in several places in winding, it probably will allow the current to "short circuit," instead of traversing the whole length of the coils. If this be the case, the magnet will be very weak: the magnet of a 2-1/2-in. bell should be able to sustain easily a 1 lb. weight attached by a piece of string to a smooth piece of 1/2-in. square iron placed across its poles, when energized by a single pint Leclanche cell. If it will not do this, the insulation may be suspected. If the wire has been wound on the bare cores (without bobbins), as is sometimes done, bared places in the wire may be touching the iron. This may be ascertained by connecting one pole of a bottle b.i.+.c.hromate, or other powerful battery, with one of the wires of the electro-magnet coils, and drawing the other pole of the battery across the clean iron faces of the electro-magnet poles. If there is any leakage, sparks will appear on making and breaking contact. Nothing but unwinding and rewinding with a well covered wire can remedy these defects. The other points where the insulation may be defective are between the binding screws and the base, if this be all of metal; or between the contact spring block and the base, and the contact pillar.

It is also probable (if the connecting wires have not been covered with indiarubber tubing, as recommended) that leakage may be taking place between these wires and some portion of the metal work of the base or frame. This must be carefully examined, and if any point of contact be observed, a little piece of Prout's elastic glue, previously heated, must be inserted at the suspected places. With regard to the binding screws, if they stand on the wooden base, their insulation (unless the base be very damp indeed) will be sufficiently good; but if the base is entirely metallic, then ebonite or boxwood washers must be used to insulate them from contact with the base-plate. With regard to the contact spring block and the platinum screw pillar, it is _permissible_ that one or the other should not be insulated from the base or frame; but one or the other _must_ be insulated by means of ebonite or other insulating washers. Personally, I prefer to insulate both; but in many really good bells only the platinum screw pillar is thus insulated. Any such leakage can be immediately detected by holding one pole of a powerful battery against the suspected binding-screw, or block, or pillar, and while in this position, drawing the other pole across some bare iron portion of the frame or metal base. Sparks will appear if there is any leakage.

The fourth defect--that is, warpage or shrinkage of the base--can only occur in badly-made bells, in which the entire base is of wood. A cursory examination will show whether the board is warped or swollen, or whether it has shrunk. Warping or swelling will throw the electro-magnet too far from the armature, or "set" the pillar out of place; shrinkage, on the contrary, will bring the parts too close together and jamb the magnets, the armature, and the contact pillar into an unworkable position.

-- 45. Before quitting the subject of the defects of bells, it may not be out of place to mention that no bell that is set to do real work should be fitted up without a cover or case. The dust which is sure to acc.u.mulate, not to speak of damp and fumes, etc., will certainly militate against good contacts and good action if this important point be neglected. The cover or case generally takes the form of a shallow box, as shown at Fig. 32, and may be made from 1/4-in. teak, mahogany, or walnut, dovetailed together and well polished. It is fastened to the base in the same manner as the sides of a Dutch clock, by means of studs, hooks and eyes. At the bottom of the box is cut a slot, of sufficient width and length to admit the play of the hammer shank.

[Ill.u.s.tration: Fig. 32.]

In the annexed table is given a general idea of the proportion which should be observed in the construction of bells of different sizes. It must be noted that if the bells are to be used at long distances from the battery, rather more of a finer gauge of wire must be employed to wind the magnets than that herein recommended, unless, indeed, _relays_ be used in conjunction with the bells.

-- 46.--

TABLE

Showing proportions to be observed in the different parts of electric bells.

---------+---------+----------+--------+---------+---------- Diameter Length of Diameter Length Diameter B. W. G.

of Magnet of Magnet of of Bobbin of Wire Bell. Cores. Cores. Bobbin. Head. on Bobbin.

---------+---------+----------+--------+---------+---------- 2-1/2" 2" 5/16" 1-3/4" 3/4" 24 3 2-1/4 3/8 2 7/8 24 3-1/2 2-1/2 7/16 2-1/4 1 22 4 2-3/4 1/2 2-1/2 1-1/8 22 4-1/2 3 9/16 2-3/4 1-1/4 20 5 3-1/4 5/8 3 1-3/8 18 5-1/2 3-1/2 11/16 3-1/4 1-1/2 16 6 3-3/4 3/4 3-1/2 1-5/8 16 6-1/2 4 13/16 3-3/4 1-3/4 16 7 4-1/4 7/8 4 1-7/8 16 7-1/2 4-1/2 15/16 4-1/4 2 14 8 4-3/4 1 4-1/2 2-1/8 14 8-1/2 5 1-1/16 4-3/4 2-1/4 14 9 5-1/4 1-1/8 5 2-3/8 14 9-1/2 5-1/2 1-3/16 5-1/4 2-1/2 14 10 5-3/4 1-1/4 5-1/2 2-5/8 14 10-1/2 6 1-5/16 5-3/4 2-3/4 12 11 6-1/4 1-3/8 6 2-7/8 12 11-1/2 6-1/2 1-7/16 6-1/4 3 10 12 6-3/4 1-1/2 6-1/2 3-1/8 10 ---------+---------+----------+--------+---------+----------

[Ill.u.s.tration: Fig. 33 A.]

[Ill.u.s.tration: Fig. 33 B.]

[Ill.u.s.tration: Fig. 34.]