there is a total eclipse; the breadth of this space is not large, averaging only 140 miles. Beyond this umbra there is a lighter shadow, penumbra (pene, almost--umbra, a shadow), where only a portion of the sun's disk is obscured. Within this region there is a partial eclipse. To those persons living north of the equator and of the umbra, the eclipse passes over the lower limb of the sun; to those south of the umbra, it passes over the upper limb.* When the eclipse occurs exactly at the node, it is said to be central. If the eclipse takes place when the moon is at apogee, or furthest from the earth, her apparent diameter is less than that of the sun; as a consequence, her disk does not cover the disk of the sun, and the visible portions of that luminary appear in the form of a ring (annulus); hence there is an annular eclipse in all those places comprised within the limits of the cone of shadow prolonged to the earth.

General facts concerning a solar eclipse.—The following data may perhaps guide in better understanding the phenomena of solar eclipses.

(1.) The moon must be new.

(2.) She must be at or near a node.

(3.) When her distance from the earth is less than the length of her shadow, the eclipse will be total or partial.

(4.) When her distance is greater than the length of her shadow, the eclipse will be annular or partial. (5.) There can be no eclipse at those places where the sun himself is invisible during the time.

*South of the equator the reverse of these phenomena would happen.

(6.) An eclipse is not visible over the whole illumined side of the earth. As the moon's diameter is so much less than that of the earth, her cone of shadow is too small to enshroud the entire globe, so that the region in which it is total cannot exceed 180 miles in breadth. As, however, the earth is con- stantly revolving on its axis during the duration of the eclipse, the shadow may travel over a large surface of territory.

(7.) If the moon's shadow fall upon the earth when she is just nearing her ascending node, it will

Fig. 49

SOLAR ECLIPTIC LIMIT (17°).

only sweep across the south polar regions: if when nearing her descending node, it will graze the earth near the north pole. The nearer a node the conjunction occurs, the nearer the equatorial regions the shadow will strike.

(8.) At the equator, the longest possible duration of a total solar eclipse is only about eight minutes, and of an annular, twelve minutes. One reason of the greater length of the latter is, that then the moon is in apogee, when it always moves slower than when in perigee. The duration of total obscuration is greatest when the moon is in perigee and the sun in apogee; for then the apparent size of the moon is greatest and that of the sun least. We see from

this that the relative situation of the moon and sun decides the length and kind of the eclipse.

(9.) There cannot be more than five nor less than two solar eclipses per year. A total or an annular eclipse is exceedingly rare. For instance, there has not been a total eclipse visible at London since 1715, and previous to that, there had been none visible for five and a half centuries.

(10.) A solar eclipse comes on the western limb or edge of the sun and passes off on the eastern.

(11.) The disk of the sun and moon is divided into twelve digits, and the amount of the eclipse is estimated by the number of digits which it covers. Thus an eclipse of six digits is one in which half the diameter of the disk is concealed.

Fig. 50.

Curious phenomena.-Various singular appearances always attend a total eclipse. Around the sun is seen a beautiful corona or halo of light, like that which painters give to the head of the Virgin Mary. Flames of a blood-red color play around the disk of the moon, and when only a mere crescent of the sun is

visible, it seems to resolve itself into bright spots interspersed with dark spaces, having the appear

ance of a string of bright beads

(Baily's Beads.)

Fig. 51.

Attendant circumstances of a total eclipse. These are of a peculiarly impressive character. The dark

ness is so intense

BEADS.

that the brighter stars and planets are seen, birds cease their songs and fly to their nests, flowers close, and the face of nature assumes an unearthly cadaverous hue, while a sudden fall of the temperature causes the air to feel damp, and the grass wet as if from excessive dew. Orange, yellow, and copper tints give every object a strange appearance, and startle even the most indifferent. The ancients regarded a total eclipse with feelings of indescribable terror, as an indication of the anger of an offended Deity, or the presage of some impending calamity. Even now, when the causes are fully understood, and the time of the eclipse can be predicted within the fraction of a second, the change from broad daylight to in

stantaneous gloom is overwhelming, and inspires with awe even the most careless observer.

Curious custom among the Hindoos.-Among the Hindoos a singular custom is said to exist. When, during a solar eclipse, the black disk of our satellite begins slowly to advance over the sun, the natives believe that some terrific monster is gradually devouring it. Thereupon they beat gongs, and rend the air with most discordant screams of terror and shouts of vengeance. For a time their frantic efforts seem futile and the eclipse still progresses. At length, however, the increasing uproar reaches the voracious monster; he appears to pause, and then, like a fish rejecting a nearly swallowed bait, gradually disgorges the fiery mouthful. When the sun is quite clear of the great dragon's mouth, a shout of joy is raised, and the poor natives disperse, extremely self-satisfied on account of having so successfully relieved their deity from his late peril.

THE SAROS.-The nodes of the moon's orbit are constantly moving backward. They complete a revolution around the ecliptic in about eighteen and a half years. Now the moon makes 223 synodic revolutions in 18 yr. 10 da.; the sun makes 19 revolutions with regard to the lunar nodes in about the same time. Hence, in that period the sun and moon and the nodes will be in nearly the same relative position. If, then, we reckon 18 yr. 10 da. from any eclipse, we shall find the time of its repetition. This method was discovered, it is said, by the Chal