It was now that the balloon, showing a little waywardness, slightly upset a portion of the experiment, for, instead of getting to the neighbourhood of earth just at the moment of sunset, the travellers found themselves at that epoch 600 feet above the ground, and over the ridge of a hill, on passing which the balloon became sucked down with a down draught, necessitating a liberal discharge of sand to prevent contact with the ground. This circumstance, slight in itself, caused the lowest point of the descent to be reached some minutes late, and, still more unfortunate, occasioned the ascent which immediately followed to be a rapid one, too rapid, doubtless, to give the registering instruments a fair chance; but one principal record aimed at was obtained at least with sufficient truth, namely, that at the culminating point, which again was 6,200 feet, the temperature read 35 degrees, or about 6 degrees warmer than when the balloon was at the same altitude a little more than an hour before. This comparatively warm temperature was practically maintained for a considerable portion of the descent. We may summarise the principal of Mr. Glaisher's generalisations thus, using as nearly as possible his own words:-- "The decrease of temperature, with increase of elevation, has a diurnal range, and depends upon the hour of the day, the changes being the greatest at mid-day and the early part of the afternoon, and decreasing to about sunset, when, with a clear sky, there is little or no change of temperature for several hundred feet from the earth; whilst, with a cloudy sky, the change decreases from the mid-day hours at a less rapid rate to about sunset, when the decrease is nearly uniform and at the rate of 1 degree in 2,000 feet. "Air currents differing in direction are almost always to be met with. The thicknesses of these were found to vary greatly. The direction of the wind on the earth was sometimes that of the whole mass of air up to 20,000 feet nearly, whilst at other times the direction changed within 500 feet of the earth Sometimes directly opposite currents were met with." With regard to the velocity of upper currents, as shown by the travel of balloons, when the distances between the places of ascent and descent are measured, it was always found that these distances were very much greater than the horizontal movement of the air, as measured by anemometers near the ground. CHAPTER XVI. SOME FAMOUS FRENCH AERONAUTS.