the moon may frequently be observed. Eclipses of Jupiter's satellites may also be seen; and (if the navigator have a telescope somewhat better than is usually carried in ships, and steadily mounted) the passage of Jupiter's satellites, either behind the planet or in front of the planet, may be seen, and the times at which the centres of the satellites just touch the edge of the planet may be observed. All these observations will be useful the observations must be recorded with the same fulness which has been mentioned before.

4. It may chance that the navigator is in some climates where the air is much more damp, and in others where it is much more dry, than in Europe. It is possible that in these places he may be able to make observations which will throw some light upon the influence of moisture in atmospheric refraction. It is recommended that repeated observations of the altitude of the sun's upper and lower limb be taken when the sun is very near the horizon. It will be necessary that the time at the ship and the latitude be very well known. The thermometer must be read, as also some hygrometrical instrument, and the barometer, if there is one on board, during the observations. The observations of every kind must be recorded with the utmost fulness.

5. It is certain that some of the stars of the southern hemisphere are variable in magnitude; the most remarkable of these is n Argûs. It is desirable that, on favourable nights, the magnitude of this star should be observed and recorded. The best way of doing it will be, not to state that it looks like a star of the 2nd magnitude, or of the 3rd magnitude, &c., but to compare its brightness with that of some of the stars near

it. Thus it will be easy to say that it appears pretty exactly as bright as one star, certainly brighter than a second, and certainly not so bright as a third.*

6. Much attention has been excited by the appearance, in several years, of meteors in great numbers, on or about the 9th of August and the 12th of November. It is probable that these appearances may be seen by persons at sea, when, either from the hour at which they occur, or from other causes not yet understood, they cannot be seen in Europe. It is impossible to observe them with accuracy; but very valuable information will be given by counting repeatedly how many can be seen in some fixed interval of time, as five minutes; and by remarking whether they all come from, or go to, one part of the heavens; what is that part of the heavens; whether they usually leave trains behind them; what is their usual brightness (as compared with that of known stars); and by any other remarks which may be suggested by their appearance.

7. Many opportunities will occur of observing the zodiacal light; more especially when the observer is near the equator, where probably it can be seen at all seasons, before sunrise and after sunset; or, if in northern latitudes, after sunset in February and March, and before sunrise in September and October;

See a list of variable stars, and some suggestions for observations of brightness of stars, in the Appendix.-(Ed.)

† Humboldt (Kosmos, i. 387) enumerates the following epochs as especially fertile in meteors, viz. April 22-25; July 17-26; Aug. 9 —11; Nov. 12—14 and 27-29; Dec. 6-12. Of all these epochs, that of August has hitherto proved to be the most regular. The star B Camelopardali has for several years been their point of divergence about that epoch. Any change in this respect should be recorded. (ED.)

if in southern latitudes, before sunrise in March and April, and after sunset in August and September. The zodiacal light consists of a pyramid of faint light, whose base is somewhere near the place of the sun, and whose point is at a distance of perhaps 30° from the sun; the axis of the pyramid being usually inclined to the horizon, following nearly the direction of the ecliptic. Although it presents to the eye a considerable body of light, yet the light of any portion of it is so feeble, and the definition of its outline is so imperfect, that it cannot be observed with a telescope. The observer, therefore, should only attempt to observe it with the naked eye, when the sky is very clear, and when the sun is so far below the horizon that no twilight is visible. He should then endeavour, with the assistance of a chart of the stars, to define as accurately as possible its boundary with reference to the stars; remarking especially the place of the point of the pyramid, the width where it rises from the horizon, whether its sides are curved, and in what parts the light is brightest. It will be found that these observations are made most accurately by occasionally turning the eye a little obliquely from the zodiacal light. registering the observation, in addition to the particulars to be recorded as prescribed above, there should be a statement of the latitude of the ship, the day, the time at the ship (or the Greenwich time and the longitude of the ship), the state of clearness of the sky, and the state of the weather for the day preceding the observation.

Improvement of Nautical Astronomy.

In

8. So much attention has been given to every detail

of Nautical Astronomy, that it is very difficult to fix upon any part of it to which the attention of navigators should be specially directed with a view to its improvement. Perhaps the principal deficiency at the present time is in the want of well-understood methods of observing (with the sextant) the altitudes of stars at night, and of observing the altitudes of the sun and moon when the horizon is ill-defined. Every endeavour ought to be made to become familiar with the use of Becher's horizon, or some equivalent instrument, and to acquire a correct estimate of the degree of confidence which can be placed in the use of it.

9. It is likewise desirable that efforts should be made to facilitate the observation of occultations of stars by the moon, and the observation of eclipses of Jupiter's satellites at sea. Occultations occur rarely, but the result which they give for longitude is usually so much more accurate than that given by lunar distances, that, in long voyages where little dependence can be placed on the chronometer, the observation of an occultation must be extremely valuable. The eclipses of Jupiter's satellites afford less accurate determinations of longitude, but they occur very much more frequently, and may be very useful where chronometers cannot be trusted.*

* Attempts may laudably be made to devise some available mode of suspending a chair, so as to afford a steady seat to the observer. Hitherto such attempts have failed of practical success, from setting out with the principle of perfectly free suspension, a principle which tends to prolong and perpetuate oscillations once impressed. It remains to be seen what stiff suspension, as for example by a rigid rope or cable, or by a hook's joint, purposely made to work stiffly (and that more or less at pleasure), by tightening collars-as also deadening and shortening oscillations, by lateral cords passing through rings to create friction-and other similar contrivances

Astronomical Geography.

10. The intelligent navigator, on arriving at any port which has not before been visited, or whose position is not very well settled, ought to consider it his first duty to determine with all the accuracy in his power the latitude and longitude of the port. Supposing him to have determined by the usual nautical methods the approximate latitude, longitude, and error of chronometer, the best method of determining the latitude will be to find the chronometer-time at which the sun or any bright stars of the Nautical Almanack list will pass the meridian, and to observe the double altitude of any such object by reflection in a mercurial horizon, several times, as near as possible to the time of the meridian passage. If the place is in the northern hemisphere, the observation of the double altitude of the pole-star may be made at any time when it is visible: convenient tables for the reduction are given in the Nautical Almanack. For these and other observations the navigator ought to be provided with a proper trough and a store of mercury. For determining the longitude, there is probably no method superior to that of lunar distances (the exactness of which will be increased if the sextant or reflecting circle be mounted on a stand), unless the stay at the port is so long that transits of the moon can be observed. In any case, if there be a transit-instrument in the ship, it ought to be mounted on shore as soon as possible. The instrument ought, on the first evening, to be got very nearly into a meridional position, and

may do. In the suspension of a cot, at least, I have found this principle signally available.-(ED.)