The purpose of the condenser as regaids the induction coil is two-fold, it being used primarily to absorb the spark caused by the breaking of the primary circuit at the vibrator contact points. The second and more important value of the condenser is its intensifying power whereby the extra discharge from the primary circuit is held in store and at the proper time released into the primary circuit again. The value of the condenser in this second case is readily appreciated by disconnecting the condenser from any coil, and noting the depreciation in the secondary spark discharge. As a matter of fact, a 1-in. spark coil will hardly spark at all, with the condenser removed.

The question now arises: With given primary core and winding specifications and a vibrator operating a stated number of times per minute, by what formula or method is the proper capacity for the condenser calculated? Here is an instance where experience counts more than mathematics. Coil manufacturers, as a rule, have stock specifications for use with all kinds of coils, but it is true that some one had to figure out all the details carefully in the first place.

Let us consider the iron core of the coil. There are all grades of iron and steel wire used in the coil manufacture. The best wire is Swedish iron wire. This comes in all wire gauges. The finer the wire used the better the operation of the coil. In moter and dynamo construction we learn the value of laminated armatures. We find that all solid masses of iron are avoided. The thinnest discs of soft iron are bolted together to form a laminated armature so that the losses by hysteresis may be at a minimum. In such a small affair as an induction coil, the difference in results between a steel wire core and a Swedish iron core would seem so trifling as not to warrant the extra cost of several cents per pound for the iron wire. In the cheaper grades of coils, particularly in automobile coils manufactured by the thousands at a very low wholesale price, the finer details of core construction are not considered. It is purely a problem of getting a minimum cost for all materials and turning out something that will do the work and beat the other fellow out of the market. In such coils we find single cotton covered wire primaries, with no insulation between layers, and No. 20 gauge soft steel wire cores. If such unscientific construction was put into a watch or dynamo the manufacturer would soon have to mend his ways or go out of business.

It will be found that the more perfect the core, the easier it will be to provide a condenser to work in harmony with the primary winding. The more turns of wire about the iron core for the primary, the greater the magnetic effect of the core; subject to certain limitations.

The more turns in the primary the greater the spark at the contact points before the condenser is installed. Sufficient turns have to be made on the primary to make the core powerful enough to attract the hammer head of the vibrator. In building a coil, therefore, where the spark is to be not over one inch in length in the open - for automobile purposes, for example - it is advisable to put just as few turns of No. 16 or 18 waxed double covered wire over the core for a primary as will magnetize that core, (when the proper amount of battery is used) and attract the vibrator hammer head. The larger its area exposed to the end of the core and the shorter the vibrator stroke, the faster and fatter will be the secondary discharge.

It was once the custom in small coil manufacture to-wind on four or more layers of No. 13 wire for the primary in the belief that the added resistance of 4 layers in preference to two of larger cross section would reduce the current consumption one-half and prolong the life of the batteries accordingly. Coil manufacturers have since found by experience what every electrician has long known to be a fact, that the proper way to get around the sparking at the platinum points was not through the increase of condenser capacity, but by reducing the number of turns on the primary as much as possible. It was found that a vibrator could not be constructed too speedy to fail to give a secondary discharge, and also that a fast vibrator did not require anwhere near the condenser capacity as a slower vibrator.

It was also found that a fast vibrator, working through a low inductive winding, was just as easy on the battery as the slower vibrator and the many-layered primary winding. Also, the wear on the platinum, points was less. Based on the cost of platinum, as in comparison with the cost of batteries, the fast vibrator and low winding would prove more economical even if the batteries did become exhausted more rapidly than in the previous case.

The cost of constructing a condenser for the latter coil is much less than that of the former, owing to the less amount of tinfoil and waxed paper required. Sumin in if up the cost in each case, we find that there is a saving in the primary wire of just one-half, (which permits the use of the best grade of core wire), a saving in cost of vibrator materials, and a saving in condenser cost.

With larger induction coils, giving two or more inches spark, we have much more wire on the secondary than in the one inch coils. This mass of copper wire is practically a shield of copper over and about the primary core and winding, and it is impossible to use a vibrator intended for a 1-inch coil on this larger coil and get maximum results.

The vibrator, if core-operated, must be much slower to allow a more complete saturation of the core, and the condenser will have to be correspondingly increased in capacity to handle the spark at the make and break.

The construction of a condenser for a coil will fur-nish the best lessons in the work - far better than any mathematical formulae to be had at the present time. Any one undertaking this task, should make provisions for increasing or decreasing the capacity as needed until the proper size is ascertained, by making the condenser in sections.