fore,) but in a ratio constantly decreasing. It is therefore evident that, ascend as high as we may, we cannot, literally speaking, arrive at a limit beyond which no atmosphere is to be found. It must exist, I argued; although it may exist in a state of infinite rarefaction.
On the other hand, I was aware that arguments have not been wanting to prove the existence of a real and definite limit to the atmosphere, beyond which there is absolutely no air whatsoever. But a circumstance which has been left out of view by those who contend for such a limit, seemed to me, although no positive refutation of their creed, still a point worthy very serious investigation. On comparing the intervals between the successive arrivals of Encke's comet at its perihelion, after giving credit, in the most exact manner, for all the disturbances due to the attractions of the planets, it appears that the periods are gradually diminishing; that is to say, the major axis of the comet's ellipse is growing shorter, in a low but perfectly regular decrease. Now, this is precisely what ought to be the case, if we suppose a resistance experienced from the comet from an extremely rare ethereal medium pervading the regions of its orbit. For it is evident that such a medium must, in retarding the comet's velocity, increase its centripetal, by weakening its centrifugal force. In other words, the sun's attraction would be constantly attaining greater power, and the comet would. be drawn nearer at every revolution. Indeed, there is no other way of accounting for the variation in question. But again:—The real diameter of the same comet's nebulosity, is observed to contract rapidly as it approaches the sun, and dilate with equal rapidity in its departure toward its aphelion. Was I not justifiable in supposing, with M. Valz, that this apparent condensation of volume has its origin in the compression of the same ethereal medium I have spoken of before, and which is dense in proportion to its vicinity to the sun? The lenticular-shaped phenomenon, also, called the zodiacal light, was a matter worthy of attention. This radiance, so apparent in the tropics, and which cannot be mistaken for any meteoric lustre, extends from the horizon obliquely upwards, and follows generally the direction of the sun's equator. It appeared to me evidently in the nature of a rare atmosphere extending from the sun out-
