the inner 19, and bringing the two ends of the cables thus prepared together, so that the inner 19's butt up to each other. Interlace these outer wires so that the respective pairs also butt though alternately on either side of the centre. Then bind this

first joint of the inner cable with four strips of tinned copper binding wire, each of some ten turns, and the strips equally spaced over the joint, so that the three sets of butts come in between the four strips of binding wire. Lastly, by means of two

FIG. 128.

ladles, keep pouring melted solder over the joint, catching it in the ladle held underneath and using it over again and again until the joint takes the solder and becomes tinned and soldered into a solid mass. Fine powdered resin must be used as a flux in just sufficient quantity.

When cool enough, taper the end of the first insulating covering, when the joint should then present the form shown in Fig. 129. Now wrap on tightly in the usual way, first pure rubber strip, with the application of rubber solution between layers, and then prepared tape up to a thickness slightly exceeding the other insulation.

Joint in Outer Main.-Wrap a sheet or sleeve of copper plate which has been previously cleaned on the outside carefully with emery cloth, and which is of such a thickness as to be comfortably pliable. This sleeve must be the length of the outer conductors, and make one turn or wrap fitting the insulated cable closely. Now cut off half of every alternate wire of each outer main and interlace, after cleaning and straightening each as before. Then bend them closely over the sleeve of copper by tinned binding wire in, say, four strips about " wide, the butts of the conductors being between the two middle strips.

Solder as before with the ladles. appearance shown in Fig. 130.

The joint now has the

When cool enough insulate up in the usual way, tapering the ends of the old insulation of the cable beforehand.

Protections to insulated joint.-Wrap a piece of sheet lead over the last insulation, so as to form a sleeve of just one complete turn and overlapping the ends of the ordinary lead sheathing of the cable. Then solder wipe the ends and down the seam so as to make a good water and air-tight joint.

Note. Any part of the lead may be previously painted if desired so as to prevent the lead solder wiping from taking to that part. Next re-serve the inner yarn as far as it will go over the lead joint, adding more to complete the serving.

Then repeat this operation with the strip armouring, and lastly with the outer yarn, when the joint may finally be well tarred over and is then complete for laying in.

APPENDIX

Deviation of the deflections of Reflecting Galvanometers from the direct proportional law.

IT has been stated on p. 6 that the scale deflections of a D'Arsonval galvanometer are directly proportional to the currents producing them. Though this is not rigorously correct, it is sufficiently true for most practical purposes in the usual forms of instruments belonging to this class. For very accurate work, however, it is necessary to apply a correction, usually amounting to a small fraction of 1% for deviation from this law, and this we now proceed to indicate.

Q

D

A

d

a

P

FIG. 131.

Let O be the zero of the scale PQ, presumably in the centre, though, for convenience, only onehalf of the scale is shown, and let Od and OD be the scale deflections of the spot of light on PQ from zero for currents C1 and C2 through the galvanometer coil.

2

If B is the centre of the needle ns, then OB is the incident ray of light from some source at ( and Bd, BD the reflected rays for

the two positions of the mirror and its attached needle ns.

By drawing the normals to the mirror in each position we get

Ba and BA respectively, and it can at once be shown that the angle dBD = 2aBA, or that the angular motion of the mirror is half that of the reflected ray.

Now since ns is contained by the plane of its coil for no current and is parallel to PQ, we shall always have (for the small angular motion of ns usually obtained in mirror galvanometers)——

or if Od and OD are not very different and are small, then—C1: C2Od: OD (approximately).

=

(Fall of Potential

Measurement of the Internal Resistance of
Secondary Cells.
Method.)

Proof of Formula.-Referring to p. 73, let E=the total E.M.F. of the cell or battery, and V the potential difference at its terminals, when sending a current A.

If then B is the internal resistance of the cell and R the resistance of the external circuit, we have by Ohm's Law--Fall of Potential round external circuit = AR=V,

But the Fall of Potential in the cell itself is also = E - V,

.. E-V=AB,

-