Astronomical Society. 4'45 



The periods of these two stars may be stated respectively at 55 and 

 43 years ; that of g Urssb at 56^, which last determination agrees 

 nearly with that of M. Savary, who has assigned 58^ years for this 

 element. The period of 70 Ophiuchi has been determined by Pro- 

 fessor Encke at 73-862 years. 



Lastly, Mr Herschel subjoins some practical remarks on the 

 management and action of telescopes, and takes occasion to intro- 

 duce a set of measures of the positions of Saturn s satellites, taken 

 under remarkably favourable circumstances, and a so a single ob- 

 servation of the place of the comet of 1830, the only one he could 

 succeed in procuring. ,, ,t iii, hj 



In a subsequent letter received from Mr. Hersche , he alludes to 

 a new method of taking the angles of position, viz., by reflected as 

 well as bv direct vision. He conceives that this method possesses 

 two capital advantages: Ist, That the quantity directly measured 

 is double the quantity sought, by which, of course, the error of de- 

 termination is halved; and, 2ndly, That the error of the zero is 

 destroyed altogether, the double arc being given by the difference 

 of readings in the limb of one and the same circle, maintained in one 

 and the same position. Mr. Herschel also adds, that subsequent 

 and very satisfactory measures of ^ Corons fully confirm the con- 

 elusion above mentioned with regard to the revolution of that star. 

 IV. On the dependence of a clock's rate on the height ot the 

 barometer, by the Rev. Dr. Robinson. 



The rate of a clock of good workmanship may be assumed to 

 depend on three things; first, the rate at a given temperature and 

 barometric pressure: secondly, on the variation of temperature, as 

 shown by the thermometer; and thirdly, upon the atmospheric 

 pressure, measured by the barometer. The variation of temperature 

 not only affects the materials of the pendulum, but, along with the 

 atmospheric pressure, greatly modifies the action of the air in the 

 way of buoyancy and inertia, &c. Possibly, the irregular action of 

 the wheelwork, the diminution of arc from the thickening of the 

 oil on the pallets, and even the hygrometric state of the air, may 

 sensibly alter the rate of a clock; but the present memoir is confined 

 to the consideration of the first-mentioned sources ot error. 



As the changes are minute. Dr. Robinson formed equations of 

 condition, where the quantities sought were the rate of the clock at 

 49'^ of the thermomer, and 295 inches of the barometer and the 

 retardations corresponding to a rise of 1° and 1 inch in each of these 

 instruments respectively. The absolute gain or loss of the clock 

 was determined with the utmost care. . ,. .u 



«A thermometer placed in the clock-case, with its bulb three 

 inches before the pendulum rod, and on a k-vel with the cover ot 

 the jar, and a barometer with its cistern at the same level, were 

 noted morning and evening at half-past nine; a time chosen not 

 merely as likely to give mean results tor the whole day, but also 

 because it nearly bisects the average period of evenmg observations 

 The thermometer was read tiirough a s(iuare of plate glass, cemented 



