18G5.] Astronomy. 75 



can be registered on the paper. During the time that the paper has 

 been unreeling by means of a clock-work driving apparatus, the normal 

 clock has recorded the seconds, and indicated the commencement of 

 each minute on a scale of seconds placed parallel with, and in close 

 juxtaposition to, the indications transmitted from the transit. The 

 traces are made with diamond points on mat-black paper, which be- 

 comes burnished under their pressure, and shows two bright lines, 

 which, the points being always on the paper, would be continuous, but 

 for the interruptions of the current, produced, on the one hand, by the 

 beats of the clock, and, on the other, by the operator's signals from 

 the transit. These interruptions cause a sort of battlement, like that 

 on a crenulated wall, to project from the lines, the commencement 

 being the exact second, or the precise moment of transit, as the case 

 may be. The reading off is perfectly easy, and is capable of great 

 exactness, either by estimation or by actual measurement. The bat- 

 teries used are Meidinger's modification of Daniell's, and give great 

 satisfaction, the elements being copper in contact with a solution of 

 sulphate of copper, and zinc in contact with sulphate of zinc, or sul- 

 phate of magnesium ; and the two fluids, neither of which is placed in 

 a porous cell, are kept in surface-contact. 



A very great improvement has been adopted in the arrangement 

 of the normal clock belonging to this observatory. Clocks vary owing 

 to three causes. The first being unavoidable errors of workmanship ; 

 the second, the variation of temperature causing the pendulum to 

 lengthen and contract ; and the third the alteration in the atmospheric 

 density, the pendulum meeting with less resistance while passing 

 through the air when the barometer is low than when it is high. 

 Horology has now been brought to such perfection, that it is believed 

 that if the second and third causes of error could be compensated for, 

 the first would be almost imperceptible. The temperature error is 

 generally corrected by the mercurial cistern, but the far greater bulk 

 of the latter and its contents in relation to surface as compared with 

 that of the pendulum rod renders it impossible for the mercury to 

 take up any change of temperature so rapidly as the rod ; hence the rod 

 varies in length far before the mercury can, by expanding or contract- 

 ing, prevent an inevitable shifting of the centre of oscillation in re- 

 ference to the axis of suspension. This serious defect in the mercurial 

 pendulum has been guarded against at Pulkowa, by removing the 

 clock which was formerly embedded in one of the piers of the central 

 hall under the great equatorial, and placing it in a subterranean cham- 

 ber beneath the hall, in a situation where changes of temperature occur 

 very slowly, and where the limits of change are very narrow from one 

 period of the year to another. The barometric error it is also intended 

 to remove by placing the clock in an air-tight case, in which a con- 

 stant pressure will be maintained by means of a pump to be brought 

 into use whenever it shall appear from the indications of a pressure 

 gauge to be connected with the clock case, that there is a variation in 

 the density of the air around the pendulum. 



Mr. Do la Rue also describes a very ingenious method of commu- 

 nicating time signals for controlling other clocks by the normal clock 



