abstract Observations of earth strains, covering a wide spectrum of periods, have been made with a three component installation located 1800 feet below the surface in a deep mine at Ogdensburg, New Jersey. This paper is a description of the strain meters and their environment and is a report of several studies based largely on strain observations. The instruments, modeled after those of Benioff, employ quartz tubes as their standard of length, use variable capacitor transducers, and record simultaneously tidal data at low gain and seismic data at a sensitivity about 40 times that of the tidal signal. A significant reduction of the noise level in the seismic spectrum has been achieved by sealing the tube rooms against air pressure fluctuations. The present secular strain rate is less than 1.7 × 10−8 per month, almost an order of magnitude less than that observed by Benioff in California, and almost two orders of magnitude less than that observed by Takada in Japan. A transient strain with a duration of about 100 hours was associated with the storm tide which damaged the eastern coast of the United States on March 5-8, 1962. The strain corresponds to that produced by an anomalous high water mass whose seaward extent was limited to 100 ± 30 km in the azimuth southeast from Sandy Hook. Direct observation of the seaward extent of such storm tides, for areas off coastlines bordering the open ocean, has not been possible in the past because of the lack of deep water tide recorders. Love's number h and Shida's number l are determined to be 0.53 and 0.037, respectively, from observations of strains associated with the M2 tidal constituent, on the basis of the assumption that the ocean load generated strain has the same phase on both horizontal instruments. The uncertainty in h and l is large, as it is in other measurements of these numbers, because of the effect of ocean load. This uncertainty is examined in a quantitative manner by mapping h and l as functions of ocean load amplitude and phase in the vicinity of the solution. Strain seismograms of the Kenai Peninsula earthquake of September 5, 1961, and the Prince Edward Island earthquake of December 29, 1961, are unique. The exceptional long period response of the strain seismometers has made it possible to record, on the first pass, waves associated with the group velocity maximum of the Rayleigh wave dispersion curve over a purely continental path across North American and an almost pure oceanic path across the Atlantic. Such data have never before been reported. The shield area of North America is characterized by higher than average Rayleigh wave velocities. The Atlantic data, when compared with data from the Pacific, do not indicate any appreciable difference between the mantle structure below the Atlantic and that below the Pacific.