We overcame this difficulty by using an auxiliary clock during the transmission.
The seconds beats of this clock were displaced with respect to those of the transmitting clock sufficiently to avoid this interference.
The following records were obtained : I. Transmitting clock VS. auxiliary clock, five minutes prior to time signals. 2. Radio time signals VUS. auxiliary clock. 3. Transmitting clock zs. auxiliary clock, five minutes after time signals.
To go into the results would exceed the limits of this paper. The values obtained, however, show that there are real variations in the lag which are of the order of .OI second, the lag itself being roughly .06 second.
The observed variability indicates that in connection with precise longitude work the lags should be measured for each signal used in the determination.
Our work in this connection consisted simply in the experimental determination of the lag. We supplied the results to the U. S. Naval Observatory which was handling all other matters in connection with the longitude determinations.
In conclusion it is a pleasure to acknowledge the valuable contributions of my associates, Dr. J. C. Karcher and Mr. M. Keiser, who, no less than the writer, are responsible for the development of the apparatus and for its successful use in the solution of many interesting problems.