from some errors of his observations; but, if such were the case, it might be expected that the result would betray that kind of irregularity which is always the character of such errors. Thus it would be expected that the predicted time would sometimes be later, and sometimes earlier, than the observed time, and that it would be later and earlier to an irregular extent. On the contrary, it was observed, that while the earth moved from E to E"", the observed time was continually later than the predicted time, and, moreover, that the interval by which it was later continually and regularly increased. This was an effect, then, too regular and consistent to be supposed to arise from the casual errors of observation; it must have its origin in some physical cause of a regular kind.

The attention of Roemer being thus attracted to the question, he determined to pursue the investigation by continuing to observe the eclipses. Time accordingly rolled on, and the earth, transporting the astronomer with it, moved from E" to E'.

It was now found, that though the time observed was later than the computed time, it was not so much so as at E'; and as the earth again approached opposition, the difference became less and less, until on arriving at E, the position of opposition, the observed eclipse agreed in time exactly with the computation.

From this course of observation it became apparent that the lateness of the eclipse depended altogether on the increased distance of the earth from Jupiter. The greater that distance, the later was the occurrence of the eclipse as apparent to the observers, and on calculating the change of distance, it was found that the delay of the eclipse was exactly proportional to the increase of the earth's distance from the place where the eclipse occurred. Thus when the earth was at E'", the eclipse was observed sixteen minutes, or about 1000 seconds later than when the earth was at E. The diameter of the orbit of the earth, E E'", measuring about two hundred millions of miles, it appeared that that distance produced a delay of a thousand seconds, which was at the rate of two hundred thousand miles per second. It appeared, then, that for every two hundred thousand miles that the earth's distance from Jupiter was increased, the observation of the eclipse was delayed one second.

Such were the facts which presented themselves to Roemer. How were they to be explained? It would be absurd to suppose that the actual occurrence of the eclipse was delayed by the increased distance of the earth from Jupiter. These phenomena depend only on the motion of the satellite and the position of Jupiter's shadow, and have nothing to do with, and can have no dependence on, the position or motion of the earth, yet unquestionably the time they

appear to occur to an observer upon the earth, has a dependence on the distance of the earth from Jupiter.

To solve this difficulty, the happy idea occurred to Roemer that the moment at which we see the extinction of the satellite by its entrance into the shadow is not, in any case, the very moment at which that event takes place, but sometime afterward, viz. such an interval as is sufficient for the light which left the satellite just before its extinction to reach the eye. Viewing the matter thus, it will be apparent that the more distant the earth is from the satellite, the longer will be the interval between the extinction of the satellite and the arrival of the last portion of light which left it at the earth; but the moment of the extinction of the satellite is that of the commencement of the eclipse, and the moment of the arrival of the light at the earth is the moment the commencement of the eclipse is observed.

Thus Roemer, with the greatest felicity and success, explained the discrepancy between the calculated and the observed times of the eclipses; but he saw that these circumstances placed a great discovery at his hand. In short, it was apparent that light is propagated through space with a certain definite speed, and that the circumstances we have just explained supply the means of measuring that velocity.

We have shown that the eclipse of the satellite is delayed one second more for every two hundred thousand miles that the earth's distance from Jupiter is increased, the reason of which obviously is, that light takes one second to move over that space; hence it is apparent that the velocity of light is at the rate, in round numbers, of two hundred thousand miles per second.

By more exact observation and calculation the velocity is found to be 184,000 miles per second, the time taken in crossing the earth's orbit being 16m 356.

543. Eclipses of Saturn's satellites not observable. Owing to the obliquity of the orbits of the Saturnian satellites to that of the primary, eclipses only take place at or near the equinoxes of the planet, the satellites revolving nearly in the common plane of the equator and the ring. When they do take place, these eclipses are so difficult of observation as to be practically useless for the determination of longitudes, and have, consequently, received but little attention.

IV. TRANSITS OF THE INFERIOR PLANETS.

544. Conditions which determine a transit. When an inferior planet, being in inferior conjunction, has a less latitude or distance from the ecliptic than the sun's semi-diameter, it will be less distant from the sun's centre than such semi-diameter, and will

therefore be within the sun's disk. In this case, the planet being between the earth and sun, its dark hemisphere being turned towards the earth, it will appear projected upon the sun's disk as an intensely black round spot. The apparent motion of the planet being then retrograde, it will appear to move across the disk of the sun from east to west in a line sensibly parallel to the ecliptic.

Such a phenomenon is called a TRANSIT, and as it can only take place with planets which pass between the earth and sun, it is limited to the inferior planets.

545. Intervals of the occurrence of transits. - -The transits of Mercury and Venus are phenomena of rare occurrence, especially those of Venus, and they are separated by very unequal intervals. The following are the dates of the successive transits of Mercury from 1845 to the end of the present century:

Those of Venus occur only at intervals of 8, 122, 8, 105, 8, 122, &c. years; the last took place on December 8, 1874. The next transit will occur on December 6, 1882, and the two succeeding ones on June 7, 2004, and June 5, 2012.

546. The sun's distance determined by the transit of Venus. The transits of Venus have acquired immense interest and importance, from the circumstance of their supplying data by which the sun's distance from the earth can be determined with far greater precision than by any other known method. The transits of Mercury would supply like data, but owing to the greater distance of that planet from the earth when in inferior conjunction, the conditions affecting the data are not nearly so favourable as those supplied by Venus.

The transit of Venus on the 3rd of June, 1769, was considered of sufficient importance for sending out an astronomical expedition to Otaheite, in the Pacific Ocean, for the purpose of obtaining observations of this rare phenomenon at a distant part of the globe, which would supply the necessary data, in conjunction with those found from observations in other localities, for ascertaining the amount of the sun's parallax. This expedition was under the command of the celebrated Captain Cook. The French, Russian, and other governments also fitted out expeditions in the most liberal manner, which were sent to various parts of the globe.

On a comparison of all the observations, it was found that they gave 8" 5776 as the value of the sun's horizontal parallax, or 171552 as the angle which the earth's diameter subtends at the (See Appendix, 807.)

svin.

Extensive preparations were made by astronomers of all nations for the observation of the transit of Venus on December 8, 1874, the preliminary arrangements for the British stations being undertaken by Sir George Airy, the Astronomer Royal. The transit was very successfully observed at some of the most important stations, both by the eye and by photographic registration.

V. OCCULTATIONS.

547. Occultation defined. When any celestial object, the sun excepted, is concealed by the interposition of another, it is said to be OCCULTED, and the phenomenon is called an OCCULTATION.

Strictly speaking, a solar eclipse is an occultation of the sun by the moon, but usage has given to it, by exception, the name of an eclipse.

548. Occultations by the moon. The phenomena of this class which possess greatest astronomical interest are those of stars and planets by the moon. That body, measuring about half a degree in diameter, moves in her monthly course so as to occult every object on the firmament which is included in a zone extending to a quarter of a degree at each side of the apparent path of her centre. All the stars whose places lie in this zone are successively occulted, and disappearances and reappearances of the more conspicuous ones, as well as those of the planets which may be found within the limits of the same zone, present some of the most striking effects which are witnessed by observers.

The astronomical amateur will find in the Nautical Almanac a table in which all the principal occultations, both of stars and planets, are predicted.

The disappearance takes place always at the limb of the moon, which is presented in the direction of its motion.

From the epoch of full moon to that of new moon the moon moves with the enlightened edge foremost, and from new moon to full moon with the dark edge foremost. During the former interval, therefore, the objects occulted disappear at the enlightened edge, and reappear at the dark edge, and during the latter period they disappear at the dark, and reappear at the enlightened edge.

The disappearances and reappearances when the moon is a crescent are especially remarkable. If the disappearance take place at the convex edge, notice of its approach is given by the visible proximity of the star, which, at the moment of contact, is suddenly extinguished. Its reappearance is more startling, for it seems to be suddenly lighted up at a point of the firmament nearly half a degree from the concave edge of the crescent. If the disappearance take place at the dark edge it is much more striking, the star appearing to " go out" of itself at a point of the sky where nothing interferes

with it.

When stars, however, are of less magnitude than the fifth, the overpowering light of the moon makes the star invisible before the occultation takes place at the enlightened edge. At the reappearance the same effect is produced, the star before becoming visible is frequently some distance from the bright limb of the

moon.

The moon's horizontal parallax amounting to nearly twice its diameter, the part of the firmament on which it is projected and which is its apparent place, differs at different parts of the earth. In different latitudes the moon, therefore, in the course of the month appears to traverse different zones of the firmament, and consequently to occult different stars. Stars which are occulted in certain latitudes are not occulted at all at others, and of those which are occulted, the durations of the occultation and the moments and places of disappearance and reappearance are different.

To render this more intelligible, let N 8, fig. 92, represent the earth, N being its north, and s its south pole. Let m m' represent the moon, and m* and m'* the direction of a star which is occulted

N

m

Fig. 92.

m

by it. It must be observed that the distance of the star being practically infinite compared with the diameter of the moon, the lines m* and m' are parallel. Let these lines be supposed to be continued to meet the earth at land l'. Let similar lines, parallel to these, be imagined to be drawn through all points of a section of the moon made by a plane at right angles to the direction of the star passing through the moon's centre. Such lines would form a cylindrical surface, the base of which would be the section of the moon, and it would be intersected by the surface of the earth, a portion of which would be included within it, one-half of which is represented by the darkly shaded part of the earth between and '. It is clear that the star will be occulted by the moon to all observers situated within this space.

While this cylindrical space is carried by the moon's orbital motion from west to east, the surface of the earth included between the parallels of latitude l n and l'n', is also carried from west to east, but much more rapidly, by the diurnal rotation,