[Fig. 6] gives the change in wind velocity at the different levels. There was an increase of wind at all the levels from the time of the passage of the anti-cyclone to the passage of the cyclone. The minimum of wind at 200 metres was in the anti-cyclone, with a secondary minimum during the passage of the centre of the cyclone.
Figs. [7] to [10] show the changes in height from day to day of the equal conditions at the different levels. [Fig. 7] shows the change in level of the isobars, which, although very small, is largest at the lower levels. The light broken lines in [Fig. 7] and subsequent figures indicate the axes of the anti-cyclone and cyclone. That the axis of the cyclone was inclined backward, and that the high pressure occurred later at high than at low levels, was confirmed by the wind observations on the twenty-first.
[Fig. 8] shows the heights at which the same temperatures were found on successive days. Since the isotherms rose until the twenty-third, the temperature of the air up to 3000 metres was higher on the day of the cyclone than on the day of the anti-cyclone. Previous high flights indicate that this is the normal condition in the moving cyclones and anti-cyclones of the eastern United States. As the light broken lines represent the axes of the anti-cyclone and cyclone up to 3000 metres, it is seen that at this level the temperature at the place of maximum pressure is probably higher than at the place of minimum pressure, although this is not true for a vertical column of air above the earth.
[Fig. 9] gives the positions of equal humidities on successive days, saturated and cloudy areas being indicated by crossed shading, and less humidity by single ruling. From the laws of thermo-dynamics the unshaded curves should represent descending currents, and the shaded portions ascending ones. In the first case, increased warmth and a lower relative humidity are produced in the descent to a lower altitude; in the last case, cooling, increasing relative humidity, and condensation are produced by expansion in the ascent to a higher altitude. Consequently, two regions of descending air are indicated, one in the centre of the anti-cyclone, the other in the centre of the cyclone.
[Fig. 10] shows the change in height of the lines of equal wind velocity. With ascending currents and precipitation, high wind velocities were found at low levels, because of increased barometric gradient, while with the descending currents in the anti-cyclone and centre of the cyclone, the high velocities were found only at great altitudes. The study of these data indicate that the cyclonic and anti-cyclonic circulations observed in this latitude do not embrace any air-movements at greater altitudes than 2000 metres, except in front of the cyclone, when the air appears to be carried upward to a great height. Above 2000 metres there are probably other weak cyclones and anti-cyclones, or secondary ones, with their centres at different places from those at the earth's surface and producing a different circulation of wind. The observations of the cirrus clouds at Blue Hill indicate that at their level exists a cyclonic circulation above the anti-cyclone apparent at the earth's surface. The shallowness of our anti-cyclones would be inferred from the great differences in speed of the general atmospheric drift, for since the velocity of the general drift from the west is more than thirty times greater at 10,000 metres than it is at 200 metres, a circulation of great depth could not endure long. Cyclones and anti-cyclones appear to be but secondary phenomena in the great waves of warm and cold air which sweep across the United States from periodic causes.
The origin of cyclones and anti-cyclones is perhaps the most important problem remaining for meteorological study. The theory that they are produced by differences of temperature in adjacent masses of air, or, as it is called, the convectional theory of the American meteorologists, Espy and Ferrel, is opposed by the observations on mountains in Europe which were collected by Dr. Hann of Vienna. If the question can be solved by the use of kites, as seems to be foreshadowed by the results just stated, another foundation-stone will be laid in the science of meteorology and the status of the kite established as an instrument of research. The kite fails when there is little or no wind at the ground, but it seems possible in such cases to lift the kite into the upper air, where there usually is wind, by attaching it to a small balloon that, after the kite can support itself, shall be detached automatically. While the height to which kites can rise is limited, and the limit is probably being approached, judging from the less gain of altitude in recent flights, yet it seems reasonable to expect that, with favourable conditions, a height of at least three miles will be reached.
Besides lifting the meteorological instruments described, kites can carry apparatus for other investigations in the free air, such as the measurement of atmospheric electricity, and the collection of samples of air, to be examined for cosmic dust and bacteria. Cameras have been lifted by kites, as already said, and for the purpose of photographing the upper surfaces of clouds there is being constructed for the Blue Hill Observatory a very light automatic camera, similar in principle to M. Cailletet's apparatus for photographing the ground from a balloon.
The use of the kite as an aeroplane can only be alluded to in this book, and it may be sufficient to say that if a motor attached to a kite can, by wings or screws, propel it against the wind, the sustaining string is unnecessary, and we shall have the flying machine which Professor Langley tells us will soon be realized. The surface of our globe has been tolerably well explored; the exploration of the atmosphere by balloons and kites will continue to make great progress during the last year of the century, and at the end of the twentieth century we may confidently expect that as the seas now are a medium for transportation, so the ocean of air will have been brought likewise into man's domain.