The outside ends of 1 and 4 serve as terminals for the coil.
This method of connection leaves all the turns so joined that the current circulates in the same direction through them all, as will be seen by an examination of the enlarged diagram, Fig. 5.
Sprague, in his "Electricity: Its Theory, Sources, and Application," recommends that the turns of wire in the secondary coil shall gradually increase in number until the middle of the spool is reached, and then decrease to the spool end, in order that the greatest number of turns be in the strongest part of the magnetic field (see Fig. 6). D D D are section dividers, S secondary windings, P primary coil. The selection of the size of wire to be used depends on the requirements as to the spark. If a short thick spark be desired, use a thick wire, say No. 34 B. & S.; if a long thin one, use No. 36 to No. 40 B. & S.
Fig. 6.
Although it is impossible to lay down rules for determining the exact amount of wire to be used to obtain a certain sized spark, yet a fair average is to allow 1¼ pounds No. 36 B. & S. per inch spark for small coils and slightly less for large ones.
The most satisfactory and perhaps the easiest way for large coils is to wind the secondary in separate coils, made in a manner similar to that employed in winding coils for the Thompson reflecting galvanometer. This method, first described by Mr. F. C. Alsop in his treatise on "Induction Coils," is somewhat as follows:
A special piece of apparatus (Figs. 7 and 8) is necessary, but presents no great difficulty in manufacture. A metal disk, D, one-sixth of an inch thick and 7 inches in diameter, is mounted on the shaft S. A second disk is provided with a collar and set screw, A, in order that it may be adjusted on the shaft at any desired distance from the stationary one. When the diameter of the coil to be wound has been decided upon, a wooden collar, W, with a bevelled surface is slipped on the shaft, it corresponding in diameter with the desired diameter of the hole through the centre of the secondary coil. As these coils are going to be made as flat rings and slipped on over the insulating tube, a remark here becomes necessary on this diameter. Reference to Fig. 9 will show that it is intended that the coils near the reel ends shall fit very loosely on the tube T (Fig. 1)—in fact, that there shall be a clearance of possibly one-half inch in the extreme end, diminishing gradually to a fifteenth of an inch in the centre coils. Therefore it becomes necessary to provide a number of wooden rings equal to the desired diameter of the central hole in the coil. The thickness of the wood determining the width of the individual coil depends on the selection of the operator; but the rule may be laid down that the narrower the coils the better the insulation of the complete coil will be on completion.