682276-1881-Observatory-Hongkong- — Page 5

THE HONGKONG GOVERNMENT GAZETTE, 3RD SEPTEMBER, 1881.

803

It follows that the Observatory determinations of time should be made and published with scientific precision, and be such as to command public confidence. They should supply the mariner with data in which, whether bound on a short run through the seas of these regions, or on a long ocean voyage, he may implicitly trust. In short, the working of the establishment ought to be made worthy of the grave interests involved, and of the requirements of one of the busiest shipping ports in the world.

In order to satisfy the above conditions, it will be necessary to provide the following instruments and apparatus:—

(1). A Transit Instrument, of not less than three inches' clear aperture and about three feet focal length, constructed for fixing upon stone axis-piers, supplied with two setting-circles, and complete with the best modern fittings and auxiliary apparatus, spare webs, &c. The transit-threads in this instrument should be suited for chronographic observations. There is no necessity at present, in my opinion, to go to the expense of purchasing a transit-circle. Axis-piers, of Portland stone, may be furnished with the instrument.

(2). A Sidereal Standard Clock, geared to break an electrical circuit at every second, excepting one

second in each minute and two consecutive seconds at every fifth minute.

(3). A Chronographic register, on Bond's principle, with a cylinder about twelve inches long and six or seven inches in diameter, driven by clock-work, and fitted for registering (1) sidereal seconds from the standard clock, (2) transit-signals from the observer at the instrument: and complete with all electrical and other apparatus and gear for these purposes, including a signal-key for the observer, glass scale for reading the record, and spare springs for the spring-governor. (4). A Mean Solar Clock, fitted with contact-springs for discharging a time-ball automatically at a fixed instant by the clock each day. I recommend that this instant be made that of Local Mean Noon. Noon is a much more convenient time than one o'clock for ship-captains ge- nerally, and it is the hour to which all navigators refer their calculations. This clock should also be fitted with springs for sending a signal at ten o'clock daily. The necessity for such a signal is very likely to arise in the course of time.

The automatic dropping of a time-ball, or sending of a time-signal, on the arrival of the hands of a clock at a certain position necessitates the condition that the clock shall be right at that time. If it be not right, the time-ball or signal will obviously exhibit an error equal to that of the clock. The clock, therefore, must be fitted with apparatus by which minute errors, down to fractions of a second, may be corrected, and the clock set perfectly right beforehand. This can be effected by sending an electric current through a galvanic coil placed beneath the pendulum, which, by its action on a slender bar-magnet carried on the pendulum, imparts either an impulse or a check, according to the nature of the current sent through the coil, and thus accelerates or retards the stroke of the pendulum. This is done for such time (ascertainable by experiment) as may be needed in order to effect the required correction. apparatus serves for sending the proper current through the coil. Another method, the A commutating invention of Mr. SANG, C.E., consists of a mechanical arrangement by which the effective length of the pendulum may be diminished or increased by a minute amount, and the stroke thereby slightly quickened or retarded, for such interval as may be necessary in order to correct the error. One or the other of these contrivances should be fitted to the mean solar clock.

When the error of the standard sidereal clock has been computed from observations of stars, it is necessary to compare that clock with the mean-time clock, and then to calculate the error of the latter on mean solar time, in order that it may be set right as above explained. If the clocks are in different rooms, and cannot be compared directly with one another, it is necessary to employ hack chronometers. Or else, which is simpler and more accurate, to have in the same room with the solar clock a chronometer regulated electrically by the sidereal clock, and beating in unison with it. At Greenwich, there are two chronometers on a desk in one of the offices, one regulated by the solar clock, the other by the sidereal clock; and, as the commutating apparatus for correcting the mean solar clock is on the same desk, the whole operation of computing, comparing and correcting is done on one spot, by a single individual, in the shortest time possible. I think, however, that, by putting our chro- nograph into the same room with the mean solar clock, we may dispense with all of these devices. It will then only be necessary to place the chronograph in circuit with the sidereal clock, when the click of the armature or movements of the pencil on the former, taken in connexion with the beats of the mean solar clock, will furnish means for sufficiently accurate comparison.

(5). A Time-Ball and apparatus, complete. with electro-magnetic mechanism for dropping the ball at the instant of noon by the mean solar clock, and sending a return signal to the Obser- vatory: also the usual air-apparatus or springs for checking its fall. Provision must be made for connecting the clock and time-ball together by an underground telegraph line, the con- ducting wire of which should be of large gauge. This cable may be laid in earthenware pipes, and the line must be efficiently guarded by "lightning-protectors."