In the first series of measurements the source was one No. 6 blasting cap, while in the second series one pound of 50 percent nitroglycerine stick dynamite was used as source. A telephone carbon button microphone was the receiver and was held fixed in location while shots were fired successively at 5, 8, 12.5, 25, 35, 50, 100 and 600 meter distances. A two‐element string oscillograph was used for timing, one element recording the instant of firing, the other element recording the arrival of the wave at the microphone. The work was done at time of no perceptible wind; air temperatures were measured carefully; no humidity measurements were made. Travel times could be read reliably to 10−4 second and distance measurements were at least correspondingly good. Instantaneous speeds were obtained by plotting time computed minus time observed against distance and measuring slopes of the resulting curve. Since this work was incidental to seismic prospecting, the observations were not quite as numerous as those of von Angerer and Ladenburg. However, the results are similar as regards abnormally high speeds near the source. Also, the use of instantaneous speeds appears to show abnormally low speeds a little farther from the source, perhaps masked in the work of von Angerer and Ladenburg due to the use by them of average speeds instead of instantaneous speeds.