100 SCIENCE.
lished some important conclusions. Generally speaking, the duration of an earthquake varies directly as the magnitude of the disturbed area, and inversely as the distance of the observing station from the place of origin. The average duration of the vertical component is about four- fifths that of the horizontal component. The period of the maximum movement, both horizontal and vertical, ranges between 0*53 and 1*7 seconds for slow undulations, and between 0*12 and 0*15 second for ripples. The range of the vertical motion is always less than that of the horizontal motion.
Professor Milne remarks that the average velocity with which waves pass through the earth varies with the square root of the average depth of the path they follow. It appears that the elasticity which governs the transmission of the precursors of the real earthquake augments at the rate of about 1 per cent, for every mile of descent.
Balloons have been used with increased success for exploring the atmosphere. The loftiest theoretical height which can be attained by such means is twelve miles, or with the aid of the sun's heat, fourteen miles. The greatest altitude actually reached is eleven miles by an unmanned balloon, and five miles and a half by an aeronaut.
In the case of the descent of a balloon, it has been shown by Dr. Hergesell that the velocity of the fall is not accelerated, as is often stated, but, on the contrary, decreases in proportion to the greater height ; so that the higher the point at which the descent begins, the less necessary it is to throw out ballast.
Since April, 1898, more than a hundred unmanned balloons have been sent up from Trappes by M. de Bort. His most important observation is that the air is subject to an annual variation of temperature, even up to six and a half miles, the maximum being towards the end of summer, and the minimum towards the end of winter.
Dr. Tuma has made several ascents for the purpose of investigating atmospheric electricity. He finds that the positive potential decreases with the greater height, so that positive charges must be accumulated in the lower regions of the air. But there is no evidence that the balloon was electrically charged, or that there was any danger of the ignition of hydrogen from such a cause.
At the Blue Hill Observatory, Massachusetts, on February 21, an altitude of 12,440 feet was reached by a recording instrument attached to a string of tandem kites. The temperature was found to be 12° Fahr., whilst that at the surface was 40°.
Mr. Pilcher, who had made great progress in the construction of an aerial machine which should soar as well as fly, attempted a flight on the last day of September. • He had risen to a height of about sixty feet, when, a sharp gust of wind snapping the tail of his apparatus, he was precipitated to the earth and was mortally injured.
In striking an average it is the custom, when the number of readings is small, to omit any single one that differs widely from the others* This is not fair, since if the reading is a high one a lower mean is obtained than if it had been less high. M. Vallier advances the correct method. With a limited number of readings that includes a widely divergent value for which there is no intrinsic improbability, take
VjUUv IL,
