of the corona; and a very large number of photographs of the corona (including many in polarized light on several different plans), of its spectrum, and of the spectrum of the chromosphere, were obtained by the various parties, which will afford copious material for discussion. Newall also obtained a polarized spectrum of the corona. Altogether no less than eighty stations were occupied. There were English, American, Russian and German observers in Egypt; English and French in Algeria and Tunisia; English in Majorca; observers of almost all nationalities in Spain; and English and American in Labrador. In Egypt the weather was bright, though the sun was low; in Majorca and Spain there were local clouds. Consequently many observations, in addition to those in Labrador, were lost, notably the special spectroscopic observations undertaken by Evershed on the northern limit of totality, and the observations of radiation undertaken by H. L. Callendar. A search for intra-Mercurial planets was conducted on an elaborate plan, with similar batteries of telescopes, in Egypt, Spain and Labrador, by three parties from the Lick Observatory, but the examination of the plates showed nothing noteworthy. Pending discussion of the greater part of the material, some interesting preliminary results were published in 1906 by the French observers. C. E. H. Bourget and Montangerand conclude that there is a marked division of the chromosphere into two regions or shells, a lower or “reversing-layer,” extending only 1″ from the limb, and a chromospheric layer extending to 3″ or 4″; and that the coronal light contains less blue and violet, but more green and yellow, than sunlight; while Fabry, by visual methods, obtained measures of the total and intrinsic intensity of the light from the corona closely confirming recent photographic observations, finding the total brightness about equal to that of the full moon, and the intrinsic brightness at 5′ from the limb about one quarter of that of the full moon. (H. H. T.)
II. Eclipses of the Moon.
The physical phenomena attending eclipses of the moon are no longer of a high order of interest either to the layman or scientific observer. A brief statement of them and their causes will therefore be sufficient. An observer watching such an eclipse from the moon would see the earth, which has nearly four times the apparent diameter of the sun, impinging on the sun’s disk and slowly hiding it. The phenomenon would be quite similar to that of an eclipse of the sun seen from the earth, until the sun was completely covered. During the progress of this partial eclipse the moon would be passing into the earth’s penumbra. As the moment of total obscuration approached, a red band of light would rapidly form in the neighbourhood of the disappearing limb of the sun, and gradually extend around the earth. This would arise from the refraction of the sun’s light by the earth’s atmosphere, and the absorption of its blue rays. When the light of the sun was completely hidden, a reddish ring of great brilliancy would, owing to this cause, surround the entire dark body of the earth during the period of the total eclipse.
The aspect of the moon, as seen from the earth, corresponds to this view from the moon. The fading of the moon’s light, due to its entrance into the penumbra, is scarcely noticeable without direct photometric determination until near the beginning of the total phase. Then, as the limb of the moon approaches the earth’s shadow, it begins to darken. When only a small portion has entered into the shadow, that portion is completely hidden. But, as the total phase approaches, the part of the moon’s disk immersed in the penumbra becomes visible by a reddish coppery light—that of the sun refracted through the lower parts of the earth’s atmosphere. The brightness of this illumination is different in different eclipses, a circumstance which may be attributed to the greater or less degree of cloudiness in those regions of the earth’s atmosphere through which the light of the sun passes in order to reach the moon. Its colour is due to absorption in passing through the earth’s atmosphere.
III. Laws and Cycles of Recurrences of Eclipses of the Sun and Moon.
It has been known since remote antiquity that eclipses occur in cycles. These cycles are known now to be determined principally by the motion of the moon’s node and the relations between the revolutions of the earth round the sun and the moon round the earth.
 |
| Fig. 4. |
Owing to the inclination of the moon’s orbit to the plane of the ecliptic, an eclipse of the sun can occur only when the conjunction of the sun and moon takes place within about 16° of one of the nodes of the moon’s orbit. The eclipse can be total only within about 11° of the node. An eclipse of the moon can occur only when the Eclipse seasons.line sun-moon-earth makes an angle less than about 11° with the line of nodes; and the eclipse can be total only within about 8° of the node, the average limiting distances varying 1° or 2° according to the circumstances. These conditions being understood, the cycles of recurrence of eclipses of either kind can be worked out geometrically from the mean motions of the sun, moon, node and perigee by the aid of geometric conceptions shown in their simplest form in fig. 4. Here E is the earth, at the centre of a circle representing the mean orbit of the moon around it. MN is the line of nodes which is moving in the retrograde direction from N towards S1, at a rate of about 19·3° in a year, making a complete revolution in 18·6 years. Let the sun at the moment of some new moon be in the line ES1, continued. If the angle NES1 is less than 16° there will probably be an eclipse of the sun, which may be central if the angle is less than 11°. Let the next new moon take place in the line ES2 a month later. The mean value of the angle S1ES2 is about 29°; but as the node N has moved towards S1 about 1·4° during the interval, the sum of the angles NES1 and NES2 will be somewhat greater than S1ES2 by about 1·6°. The result is that if these two angles are nearly equal there may be two small partial eclipses of the sun, after which no more can occur until, by the annual revolution of the earth, the direction of the sun approaches the opposite line of nodes EM, nearly six months later. The result is that there are in the course of any one year two “eclipse seasons” each of about one month in duration, in which at least one eclipse of the sun, or possibly two small partial eclipses, may occur. One eclipse of the moon will generally, but not always, occur during a season.
Owing to the retrograde motion of the node the direction ES of the sun returns to the node at the end of about 347 days, so that a third eclipse season may commence before the end of a year. In this way there is a possible but very rare maximum of five eclipses of the sun in a year. Owing to the motion of the line of nodes each eclipse season occurs about 19 days earlier in the year than it did the year before. Another conclusion from the greater eclipse limit for the sun than for the moon is that in the long run eclipses of the sun, as regards the earth generally, occur oftener than those of the moon. But as any eclipse of the sun is visible only from a limited region of the earth’s surface, while one of the moon may be seen from an entire hemisphere, more eclipses of the moon are visible at any one place than of the sun.
If, starting with a conjunction along some line ES1, we mark by radial lines from E the successive conjunctions year after year, we shall find that at the end of 18 years and about 11 days the 223rd conjunction will fall once more very near the line ES1, the angle NES1 being about 24′ greater than before. Successive eclipses will then occur very nearly in the same order as they did 18 years and 11 days before. This period of recurrence has been known from remote antiquity and is called the Saros. What is most remarkable in this period is that in addition to the distance from the node being nearly the same as before, the longitude of the sun increases by only 11° and the distance of the moon from its perigee has changed less than 3°. The result of this approach to coincidence is that the recurring eclipse will generally be of the same kind—total, annular or partial—through a number of successive periods.
