347. When the Moon goes through the center of the Earth’s shadow she is directly opposite to the Sun: yet the Moon has been often seen totally eclipsed in the Horizon when the Sun was also visible in the opposite part of it: for, the horizontal refraction being almost 34 minutes of a degree § 181, and the diameter of the Sun and Moon being each at a mean state but 32 minutes, the refraction causes both Luminaries to appear above the Horizon when they are really below it § 179. [Sidenote: Fig. V. Duration of central Eclipses of the Moon.] 348. When the Moon is Full at 12 degrees from either of her Nodes, she just touches the Earth’s shadow but enters not into it. Let _GH_ be the Ecliptic, _ef_ the Moon’s Orbit where she is 12 degrees from the Node at her Full; _cd_ her Orbit where she is 6 degrees from the Node, _ab_ her Orbit where she is Full in the Node, _AB_ the Earth’s shadow, and _M_ the Moon. When the Moon describes the line _ef_ she just touches the shadow but does not enter into it; when she describes the line _cd_ she is totally though not centrally immersed in the shadow; and when she describes the line _ab_ she passes by the Node at _M_ in the center of the shadow, and takes the longest line possible, which is a diameter, through it: and such an Eclipse being both total and central is of the longest duration, namely, 3 hours 57 minutes 6 seconds from the beginning to the end, if the Moon be at her greatest distance from the Earth: and 3 hours 37 minutes 26 seconds, if she be at her least distance. The reason of this difference is, that when the Moon is farthest from the Earth she moves slowest; and when nearest to it, quickest. [Sidenote: Digits.] 349. The Moon’s diameter, as well as the Sun’s, is supposed to be divided into twelve equal parts called _Digits_; and so many of these parts as are darkened by the Earth’s shadow, so many Digits is the Moon eclipsed. All that the Moon is eclipsed above 12 Digits, shew how far the shadow of the Earth is over the body of the Moon, on that edge to which she is nearest at the middle of the Eclipse. [Sidenote: Why the beginning and end of a lunar Eclipse is so difficult to be determined by observation.] 350. It is difficult to observe exactly either the beginning or ending of a lunar Eclipse, even with a good Telescope; because the Earth’s shadow is so faint, and ill defined about the edges, that when the Moon is either just touching or leaving it, the obscuration of her limb is scarce sensible; and therefore the nicest observers can hardly be certain to four or five seconds of time. But both the beginning and ending of solar Eclipses are visibly instantaneous; for the moment that the edge of the Moon’s Disc touches the Sun’s, his roundness seems a little broke on that part; and the moment she leaves it he appears perfectly round again. [Sidenote: The use of Eclipses in Astronomy, Geography, and Chronology.]