154- TRANSACTIONS OF THE [FEB. 15, 



pendulum by the same amount. But after these many forms of 

 compensated pendulums were found not to thoroughly compen- 

 sate, attention was turned to the question of how far the tem- 

 perature of a room could be kept at a certain fixed point. This 

 question has been solved by placing the clock, provided with 

 compensated pendulum, inside of two or three cases, the whole 

 l)eing established in a room under the ground, which room is 

 kept at a fixed or very near to a fixed temperature. Such ar- 

 rangements were found to give very much better clock rates 

 than any other; but it Avas observed that the clock rate was still 

 not constant. Investigation showed tliat the changes in the rate 

 were undoubtedly due in part to changes in the density of the 

 air through which the pendulum bob swung.' Sir Ocorge Airy 

 devised an apparatus which is still in use at the Greenwich Ob- 

 servatory, England, for the purpose of automatically making the 

 correction rendered necessary by the changes in the barometric 

 pressure.^ But here again it was found that although the appa- 

 ratus did largely correct the changes of rate due to variations of 

 density of the air, yet there was a residual correction necessary. 

 'No one has yet succeeded in showing how to eliminate this resid- 

 ual error, although attemptshave been made; and in some recent 

 apparatus there is promise of a closer a])proach to complete suc- 

 cess. Some years ago, Mr. Ballon, of Connecticut, invented an 

 electric winding apparatus Avhicli he adapted to several astro- 

 nomical clocks. This apparatus was so arranged that the wind- 

 ing was automatic, and in a test made at Cambridge, Mast-'., 

 during a six weeks' trial of one of these clocks, "it Avas found 

 to run better than the Harvard College Observatory's best clock." 

 In a clock furnished by Mr. Ballon to parties in Washington, 

 there was adopted a form of air-tiglit case which completely sur- 

 rounded the apparatus, and therefore kept the density of the air 

 inside of the case always at the same point. In this apparatus 

 of Ballou's there Avas considerable complication, and it Avas quite 

 expensive. 



In the form Avhich we describe to the Academy to-night, 

 Aveights are done aAvay with entirely, and in their place a spring 

 is employed. This spring is coiled on the arbor of the hour 

 Avheel; in uuAvinding it delivers its poAver direct to the hour hand. 

 During one hour the revolution of the hour wheel brings a 

 " toggle" piece to a position Avhere an electric contact is made, 

 when a current from a single Leclanche cell passes through a 

 small electric motor attached to the lower j)art of the clock 



' A decrease of one inch in the barometer reading increased the daily- 

 gaining rate of the GreenAvich clock by about three-tenths of a second. 



- See Nature for April 1, 1875. Lockyer's Star-Gaziiuj , p. 194. The 

 Observatory for November, 1885, p. 3C6. 



