after inception, and therefore, even though tests may be made in which the circuit breaker is tripped before the short circuit is applied, it is as much the current 2 to 5 cycles after inception that determines the usefulness of such a generator as described by Mr. Kilbourne as the current in the first Figure 1. Highvoltage 5-position selector switch for transformers tops located in outdoor switchyard loop. Furthermore, in practice in the field, a circuit breaker faces a short circuit whose a-c component decays relatively little from inception to interruption at these times. Circuit breakers are proportioned accordingly, and therefore, while a small deviation from this condition in the test circuit can be tolerated without serious consequences, a wide variation will create conditions whiah are either unfairly severe in some respects or inadequate in others. For this reason, Mr. Kilbourne was asked to hold the a-c component of short-circuit current at 5 cycles as close as possible to 75 per cent of the value at 1/4 cycle. When this request was made, it was realized that most of the things that could be done to increase the short-circuit current from a machine to a given size had a strong tendency to increase the sharpness of the decrement curve. It is therefore with very high appreciation that we note how closely Mr. Kilbourne has approached this goal. The curve which he shows in Figure 4 makes it possible to test three phase at maximum short-circuit current with a decrement curve quite comparable with those to be found in our major generating stations and makes possible, within the power delivered, tests which are adequately severe in all respects but seriously unfair in none.Although deceleration as high as 6,000 rpm per second arises from alternating torques, this is associated with a cyclic exchange of energy back and forth between the mechanical kinetic form (W/wIc,2) and the electromagnetic form ('/2LI2), and the average deceleration over a number of cycles is very much less, as is indicated by Mr. Kilbourne's oscillograms. Maximum frequency loss in a 5-cycle test at 60 cycles is approximately 1 cycle. This is not sufficient to invalidate test data or to require de-energization of the driving motor at the time of a test.THE substantiated theories of frequency modulation' have been extended to the transmission of coded and other types of low-frequency signals. A very narrow-band, frequency-shift, powerline, carrier-current system has been developed. Theoretical work has been done, sample circuits have been built, and the simplest ones have been selected. Transmitters and receivers have been produced in sizable quantities and applied to power-line use. The first of these now has had several years of successful operation. As a result of field experience, additional development work, and new operating needs, an improved transmitter and receiver have been designed.In the frequency-shift transmitter, a Hartley oscillator is modified by the introduction of a network of two crystals in the feedback circuit to the gr...