otherwife have done; the spherical atmosphere performing, in fome measure, the office of a large convex lens, or burning-glass. The more obliquely the rays fall, the greater is their deviation from their original course; and those rays that pass close to the Earth are found, by obfervations on the fetting Sun and other heavenly bodies, to fuffer a refraction of about 33 minutes of measure. The laws of optics, hereafter to be explained, require that they fhould undergo an equal refraction in paffing out through the oppofite part of the atL mofphere. Each exterior ray of the real fhadow will therefore pafs 66 minutes within the rays that would have formed the cone CKD, fig. 55. and confequently, the angle at the vertex of the cone will be 132 minutes, or 2°. 12. greater than it would have been; that is, it will be equal to the diameter of the Sun 32 (48, T), added to M2°. 12. which gives 2°. 44. Hence the axis of the cone, or length of the fhadow, is found to be no more than 42 femidiameters of the Earth; whereas the radius of the Moon's orbit, or mean diftance of the Moon, is about 60 femidiameters N of the Earth. In the fpace between the penumbra and the Earth's real fhadow it is much darker than the penumbra, though that space is illuminated by the rays of the Sun, which are variously refracted, according to the denfity of the air they pafs through. Many rays are reflected back, and the rays that go forward are fuch whofe nature does not admit of their being easily reflected. We are to fhew in

future

་

future that these are the red, orange, and yellow. Hence it is that the Moon in an eclipfe appears o red; and a spectator on the Moon would, after lofing fight of the Sun, behold the Earth environed with a narrow luminous edge of bright red light, fhaded off with yellow on the outfide.

Since the Earth, when beheld from the Moon, P must always appear in the part of the heavens immediately oppofite the Moon's apparent place as feen from the Earth, the enlightened fide of the Earth will have the fame figure, when feen from the Moon, as the dark fide of the Moon would exhibit if it could be feen at the fame inftant from the Earth. Thus, when the Moon is invifible, or near the conjunction, the Earth is in oppofition, and presents a full luminous face to the Moon; and on the contrary, when the Moon is at the full, or oppofite the Sun, it must be on the dark fide of the Earth, which confequently then becomes invifible. Near the beginning of the first or end of the last quarter, the dark fide of the Moon is rendered vifible by the full Earth fhining on it, but is. fcarcely fo luminous as the Moon when eclipsed. The Earth's difc, feen from the Moon, is about thirteen times that of the Moon seen from the Earth. If the Earth reflected as great a part of the light that falls on it as the Moon does, its light at the Moon would exceed the moon-light with us in that ratio. This, however, is not probable, though it may fairly be fuppofed that it is three or four times as great. But

we

we have already observed that the Earth's atmofphere, in a lunar eclipfe, illuminates the Moon Qrather more than this. Whence it follows, that

the narrow ring of light encircling the Earth, when feen from the Moon during an eclipfe, gives a light far exceeding our moon-light at the full, or or even that of the Earth's full face fhining on the Moon: and as the furface of the ring, by computation, can hardly equal the one hundredth part of the Earth's difc, or the one eighth part of the Moon's difc, its. brightnefs will be more than twenty-four times that of the Moon. It must confequently be very luminous and dazzling.

Notwithstanding this, there have been eclipfes of the Moon, when in that part of its orbit near the Earth, in which that luminary entirely disappeared. But these observations are very rare.

CHAP.