In the following paper three arrangements for a rotating-beacon transmitter are discussed from the theoretical point of view. These three systems are the single coil, the double-spaced coil, and the Adcock aerial arrangement.In the case of the single-coil system it is shown that under night conditions, when wireless waves can be deflected from the upper ionized layer, the signal minimum as observed at the receiver can occur at such times during the rotation of the coil as to indicate an error in the observed bearing ranging up to 80 deg. It is shown that this error is of the same order as that observed on an ordinary direction-finder under similar conditions.The treatment of the arrangement of two co-planar spaced coils, as sometimes employed for directional reception, shows that, while the time of occurrence of the minimum giving the correct observed bearing remains unaltered during all conditions of night effect and downcoming waves, other secondary minima may be obtained under these conditions, the observed bearings on which may be in error by the same amounts as those obtainable with the single-coil transmitter. As the observer at the receiver will, in general, be unable to distinguish between the main and the secondary minima, this double-coil system would appear to possess nearly the same liability to error as the single-coil system. The possibility of the two minima occurring close together will also detract from the accuracy of observation of the main minimum, even supposing it could be distinguished from the other.The third case to be considered is that proposed by Mr. C. E. Horton, viz. the use of the Adcock aerial system as a rotating transmitter. With this arrangement it is shown that a signal minimum is produced when the plane containing the pair of vertical aerials, or their equivalent in space, is perpendicular to the direction of the receiver. This minimum remains unaltered in the presence of downcoming waves at night, and as no secondary minima occur it appears that this system will give correct bearings at all times when the signal is strong enough for observation.In searching for a type of rotating-beacon transmitter upon which the observed bearings are independent of night errors, it would appear that the Adcock arrangement is the most hopeful; and it is, therefore, proposed to consider this system in more detail from the practical standpoint.