I note that in recently published results of the magnetic observatories under the direction of the United States Coast and Geodetic Survey the procedure has been adopted to tabulate the mean ordinate for Greenwich hourly intervals, centering around the half hour, the same as at Potsdam. The non‐cyclic effect may be determined from the adopted procedure just as satisfactorily as from that advanced by Dr. Chree, namely, by obtaining the value at the beginning and end of the day from the two adjacent hourly means. There is no ground for assuming that it is necessary for this purpose to adopt simply the mean of one hourly interval.As to the adoption of the Greenwich day for the daily tabulations, I put less weight upon this matter. To prefer not to use the Greenwich day but instead some day corresponding to the nearest 15‐degree meridian time, is natural from practical considerations. If anyone prefers to use, nevertheless, the Greenwich day (not simply the Greenwich hours which is the matter of chief importance), this may always be done; it will only be necessary to divide up the hourly series at some other point. However, I would suggest that at the observatories where the Greenwich day is not generally adopted, those data for which strict simultaneity over the Earth is actually requisite be also computed for the Greenwich day and published along with the other results; this suggestion would entail only a small amount of additional work. Such desired data are: