The zodiacal light is subject to considerable variations in brightness, and also varies in extent, the apex of its cone varying in distance from the Sun's place, from 40 to 90 degrees. These variations cannot be attributed to atmospheric causes alone, some of them being due to real changes in the zodiacal light itself, whose light and dimensions increase or decrease under the action of causes at present unknown. From the discussion of a series of observations on the zodiacal light made at Paris and Geneva, it appears certain that its light varies from year to year, and sometimes even from day to day, independently of atmospheric causes. Some of my own observations agree with these results, and one of them, at least, seems to indicate changes even more rapid. On December 18th, 1875, I observed the zodiacal light in a clear sky free from any vapors, at six o'clock in the evening. At that time, the point of its cone was a little to the north of the ecliptic, at a distance of about 90 degrees from the Sun's place. Ten minutes later, its summit had sunk down 35 degrees, the cone then being reduced to nearly one-half of its original dimensions. Ten minutes later, it had risen 25 degrees, and was then 80 degrees from the Sun's place, where it remained all the evening. On March 22d, 1878, the sky was very clear and the zodiacal light was bright when I observed it, at eight o'clock. At that moment the apex of the cone of light was a little to the south of the Pleiades, but this cone presented an unusual appearance never noticed by me before, its northern border appearing much brighter and sharper than usual, while at the same time its axis of greatest brightness appeared to be much nearer to this northern border than it was to the southern. After a few minutes of observation it became evident that the northern border was extending itself, as stars which were at some distance from it became gradually involved in its light. At the same time that this border spread northward, it seemed to diffuse itself, and after a time the cone presented its usual appearance, having its southern border brighter and better defined than the other. It would have been impossible to attribute this sudden change to an atmospheric cause, since only one of the borders of the cone participated in it, and since some very faint stars near this northern border were not affected in the least while the phenomenon occurred. Besides these observations, Cassini, Mairan, Humboldt, and many other competent observers have seen pulsations, coruscations and bickerings in the light of the cone, which they thought could not be attributed to atmospheric causes. It has also been observed that at certain periods the zodiacal light has shone with unusual intensity for months together.