A series of experiments was carried out at Saint Louis using the Macelwane‐Sprengnether horizontal component electromagnetic seismometers working in synchronism with type H‐S, L and N galvanometers having a period of twenty seconds; both the seismometer and the galvanometer were critically damped. With this equipment some more or less regular long period disturbances were recorded. These disturbances were finally eliminated by operating the seismographs in a seismic vault in which both the temperature and the humidity were controlled. While no physical explanation for this can be given at the present time, it appears worth while to point out that these results are similar to those found by Milne in Tokyo, the United States Coast and Geodetic Survey in Honolulu, and by Shaw in Engl