The spark produced by a coil in good condition should be blue-white with a small pinkish flame surrounding it, when the gap is ¼ of an inch or less. The sparks should pass in a continuous stream with this length of gap without irregular stopping and starting of the vibrator. Coils giving a sputtering, weak discharge that causes sparks to fly in all directions are broken down and should be remedied. The secondary windings of coils are often punctured or broken down by operating the coil with the high tension circuit open, or by trying to cause long sparks by increasing the spark gap over ⅜ of an inch in the open air. Coils are also broken down by allowing excessive currents to flow in the primary coil. Never cause a spark to jump over ⅜ of an inch. High compression in the cylinder shortens the jumping distance of a high tension spark. Coils that will cause a stream of sparks to flow across a gap of ½ an inch in the open air are often unable to cause a single spark to jump a gap of 1/32 of an inch under a compression of 80 pounds per square inch in the cylinder. Remember that a hot spark causes rapid combustion, and will fire a greater range of mixtures and “leaner” charges, than a straggling, thin, weak spark. Spark coils that give poor results with a long spark gap under high compression are often benefited by the shortening of the spark gap. Shortening the gap will increase the heat of the spark, and will insure the passing of a spark each time that the timer makes contact. A good coil should have no difficulty in igniting a piece of paper inserted between the wires forming the spark gap in the open air. [Illustration: Fig. 96. Kingston Dash Coil. ] The adjusting screw affords a means of increasing or decreasing the tension of the vibrator spring, and the amount of battery or magneto current flowing through the primary coil. Increasing the tension of the spring requires stronger magnetization of the core to break the circuit of the contact points. This in turn calls for more current from the battery; hence in order to lessen the demand for current on the battery, the tension should be as little as possible to obtain the necessary spark. An increased tension produces more spark as the magnetization of the core is increased, but for the sake of your batteries decrease the tension as much as possible with a satisfactory spark. Almost all operators have a tendency to run with too stiff a vibrator, and hence use too much current. An efficient coil should develop a satisfactory spark with ¼ to ½ of an ampere of current in the primary coil. I have often found coils that would work well with ½ ampere, that were screwed up so tight that the coils were consuming 4 to 5 amperes or 8 to 10 times as much as they should. A battery ammeter used for testing the current consumed by coil will save its cost many times over in batteries and burnt points if used at frequent intervals in the primary circuit.