Some of the deductions from the observations with kites at Blue Hill follow. Plate VIII. is a facsimile of the record of the baro-thermo-hygrograph during two flights on October 8, 1896, when for the first time the height of a mile and a half was attained. The record-sheet, it may be said, is wrapped around a cylinder that turns on its axis in twelve hours, and the curved lines in each of the three horizontal sections divide them into quarter hours. The lower section contains the trace of the barometer, the horizontal lines being the heights in metres and feet that correspond to the barometric pressure with a temperature of 32∞ Fahrenheit; in the middle section is the trace of the hygrometer on a scale of relative humidity in percentages, and in the upper section is the trace of the thermometer on a scale of temperatures in Fahrenheit and Centigrade degrees. It will be observed that the record of the barometer is reversed, _i.e._ the trace rises for falling pressure, and in the second flight when the unexpected height of 8697 feet above Blue Hill was reached, the limit of the altitude scale was exceeded. [Illustration: PLATE VIII.--METEOROGRAM FROM THE KITE-FLIGHT OF OCT. 8, 1896, AT BLUE HILL.] In order to study the changes of these elements with height during the higher flight, in Plate IX., Figs. 4 and 5, the temperature and humidity of the automatic record are plotted as abscissÊ, with the heights above sea-level in metres as ordinates. For those not familiar with this unit of length, it may be said that 100 metres are about 330 feet, and that 1600 metres equal one mile approximately. When the meteorograph was ascending, dots indicate the recorded temperatures and humidities, which are each connected respectively by continuous lines; when the meteorograph was descending, crosses indicate the observations, which are connected by broken lines. Lines inclining upwards to the left indicate decreasing temperature and humidity with increase of height, and lines inclining to the right increasing temperature and humidity with height. The straight dotted lines show the adiabatic decrease of temperature for ascending dry air. The ascent was made during the warmest part of the day, and the descent for the most part after sunset. The two branches of the temperature-lines typify the temperature change with height which usually occurs in fair weather during the day and the night respectively. The continuous line, representing the day observations, shows a uniform fall of temperature at the adiabatic rate to the cloud level. During the night, the lower part of the broken line bends decidedly to the left, showing a body of relatively cold air near the ground, caused by radiation. There is a rise of temperature with increasing altitude above the ground up to a certain height, and afterwards a comparatively uniform fall as high as the clouds, if they exist; but the rate of fall with increasing altitude, shown by the