IN the last issue of the Bulletin of the St. Petersburg Academy of Sciences (vol. xxix. No. 2), M. Rykatcheff publishes a very interesting paper on the atmospheric waves produced by the Krakatoa eruption. General Strachey and Mr. R. H. Scott (NATURE, vol. xxix. p. 181) have already shown how the eruption must have produced an atmospheric wave which has been noticed by the barometers at many meteorological observatories. The wave was propagated in concentric circles, increasing in diameter until it reached the great circle; then, it contracted until reaching a point on the antipode of Krakatoa, whence the wave returned in the same way to its point of origin; then, gradually diminishing in intensity, it made for a second and third time its way around the earth. M. Rykatcheff now publishes the curve of the barograph of Pavlovsk for August 27 to 30, where the influence of the atmospheric wave is pretty well seen; and he discusses the results obtained from observations at thirty-one different stations (Pavlovsk, St. Petersburg, Berlin, Leipzig, Magdeburg, Brussels, Paris (I. and II.), Toulouse, Greenwich, Kew, Aberdeen, Stonyhurst, Liverpool, Glasgow, Falmouth, Armagh, Valentia, Georgia Island, Coimbra, and Toronto). It appears from these observations, when calculated according to Gen. Strachey's method, that is, by taking the time between two successive passages of the wave at the same station, that, for European stations, on the average the wave took 36h. 38m. to make its way around the earth when it was going from east to west, and 35h.54m. when going from west to east. The accordance of the figures for different observatories is striking (excepting Tolosa), the greatest deviation from the average being only + 33m. and - 38m. in the first case, + 27m. and - 39m. in the second. The average speed would thus be: for the first wave, 303.3 metres, and 316.1 metres for the second. The calculated time of the Krakatoa eruption would be between gh. 6m. and 9h. 42m. Krakatoa mean time; or, on the average, 9h. 23m. When the calculations are made on M. Wolf's method (which admits the same speed in both directions), the average speed of the wave is 334.3 metres, and the time of the eruption would be loh. 39m. Krakatoa mean time. Finally, M. Rykatcheff makes the calculations by deducing both speed and time of eruption from observations made at two stations next to Krakatoa (Pavlovsk and St. Petersburg), and then he calculates from equations made for all other stations the error of the two observations. He receives thus 321.4 metres for the speed of the wave, and l0h. 16m. for the time of the eruption at Krakatoa. These results are more in accordance, he says, with the result obtained by Herr Wolf's method, and, combining both, M. Rykatcheff takes as probable 327.9 metres for the speed, and l0h. 27m. for the time of the eruption. As to the amplitudes of the oscillations of the barometers at different stations, they vary from 0.9 to 1.7mm. and reach 2.5 mm. at Georgia Island.