152 



HOROLOGY. 



Comptnss- clock which had one, with that which had a simple pen- 

 tion (luluni, as shewn in a letter from Mr Bliss, at Oxford, 

 l'2th July 1752, to Mr Short, in London. " I find, 

 upon examining my book, that the greatest difference in 

 the going of the clock, between the coldest weather of 

 the two last winters, and the hottest weather of the 

 two last summers, is no more than one second per day ; 

 and this was occasioned by the levers being made too 

 short, of which I advised Mr Ellicott above a year 

 since: Whereas a clock with a simple pendulum and 

 brass rod, made by Mr Graham, and which belongs to 

 Dr Bradley, in the coldest weather lost above fifteen 

 seconds per day, and in the warmest gained above thir- 

 teen seconds per day, and went very near the equal time 

 in temperate weather." It is plain that Mr Bliss must 

 have mennt, gained in the coldest weather, and lost in 

 the hottest, otherwise there would be no analogy with 

 the effect of the temperature in summer and winter on 

 the brass pendulum rod. See Ellicott's pamphlet, al- 

 ready quoted. 



Graham's Mercurial Pendulum. 



Graham's The mercurial pendulum, invented by Graham, hav- 

 tnercurial j n g j, een the first that was applied to a clock for the 

 pcndujum. p Ur p Ose o f compensation, we shall begin with the de- 

 scription of it, taking the others nearly in the order of 

 their invention. This pendulum consists of a pendu- 

 lum rod, which carries a large glass jar filled with mer- 

 cury, so that the expansion or contraction of the rod 

 may be counteracted by the opposite expansion or con- 

 traction of the mercury. .To make this pendulum in 

 the way which has hitherto been adopted, is attend- 

 ed with considerable trouble. From the nature of the 

 material, the construction of such a pendulum must 

 always be troublesome, because any filling in, or taking 

 out of the mercury from the cylinder or glass jar, to bring 

 about the compensation, will cause a change of place in 

 the index point on the graduated arch or index plate, 

 if such a thing is used. A pendulum which will reme- 

 dy this evil will be afterwards described, so that we 

 shall now proceed to give a description of one made 

 in the common way. 



Dimensions The length of the pendulum over all, from the bot- 

 of a mercu. torn of the sole to the upper end of the pendulum 

 rial pcndu- spring, was 43.95 inches ; the inside bottom of the jar, 

 .6 of an inch from the bottom of the sole ; and the 

 height of the column of mercury in the jar, about 7-47 

 inches. From the upper end of the spring, take a length 

 of 39.2 inches on the pendulum downwards, then 

 43.95 inches 39.2 + C, will give that part of the co- 

 lumn of mercury below the centre of oscillation equal 

 to 4.15 inches, and that above the centre 3.32 inches. 

 The height of the jar outside, was 7.8 inches ; a wire 

 put down inside, measured 7.6 inches ; mean diameter 

 inside, 2.018 inches ; weight, 7.5 ounces. Although it 

 would be still better to have it of a less weight, yet it 

 is doubtful if it would be then strong enough to sup- 

 port such a column of quicksilver. The weight of the 

 stirrup or cylinder frame was 1 Ib. 6 ounces, and was 

 reduced 6 ounces; that is, from the sole was taken '.2.55 

 ounces, and from the top 3.45 ounces, both equal to 6 

 ounces. When the clock was set agoing after this al- 

 teration, with the pendulum the same length as before, 

 it went slow at the rate of 46 seconds in 24 hours, and 

 when shortened by touching up the regulating nut, it 

 was found to be about .15 of an inch less than the for- 

 mer length. 



The length of the stirrup bars outside, including sole 

 and part of the top, was 8.125 inches; the breadth of 

 the frame from outside of the bars, 3.25 inches ; the 



thickness of the pendulum rod and bars, 0.1 36 inch; Coinpnis*- 

 breadth of ditto, 0.384 inch ; thickness of sole outside, lion 

 0.515 of an inch; distance from bottom of the sole to 

 upper side of the jar cover, 8.1 87 inches ; jar sunk into n>enic 

 the sole about or near to 0.25 inch; distance from up- O f a'mer- 

 per side of jar cover to under side of stirrup top, 1.25 curial-pcn- 

 inch ; height of stirrup top for flat of pendulum rod, dulura. 

 1.75 inch; thickness of the flat, 0.220 inch ; diameter 

 of the regulating screw, 0.218 inch; ditto of the nut, 

 1 . 1 50 inch. The screw had 36.25 turns in an inch; and 

 the nut was divided into 30 prime divisions, each being 

 equal to a second in a day. The prime divisions were . 

 subdivided into four. 



Inches. 



Length of stirrup bars inside 8.05 



Thickness of sole outside 0.515 



Length of stirrup top 1.75 



From the stirrup top to the upper end of the 

 pendulum spring 33.485 



Length of the pendulum over all 43.800 



Inch. 



Length of pendulum spring 625 



Breadth of the double laminae, including 164 

 the space between theun,eachlamina being. 168 .500 

 Thickness of ditto 007 



Weight of quicksilver in the jar ... 



Ditto of stirrup frame 1 



Pendulum rod and spring, regulating nut, 



Jar cover, &c. 



Jar 



Ib. oz. 

 11 12 



drams< 

 5.35 

 



13 



7 



Total weight of Pendulum 14 013.35 



Before the pendulum was altered, the rate of the 

 clock shewed that the compensation was not sufficiently 

 effective, although the height of the column of mer- 

 cury was 7-5 inches nearly, and the jar being full, al- 

 lowed no more to be put in : By reducing the weight 

 of the jar frame or stirrup, the rod required to be short- 

 ened, as has been stated, .15 of an inch; whether or not 

 this will be enough, remains to be determined. The 

 daily rate for a month or six weeks was 0".l slow, when 

 the temperature was from 36 to 40" of Fahrenheit's ther- 

 mometer, and got gradually slower as the height of the 

 thermometer increased. When between 60 and 66, 

 the daily loss was from 0".37 to 0".45 ; for about two 

 weeks, when the weather was extremely cold, the ther- 

 mometer at the freezing point and under, the clock 

 showed a tendency of rather gaining. 



Of late years, the mercurial pendulum has been 

 adopted in some astronomical clocks; and it seems upon 

 the whole to answer very well. The author of the Ele- 

 ments of Clock and Watch-work thought, that it was not 

 fit for this purpose, being too quick in its operations of 

 expansion and contraction ; but if the cover is well 

 fitted to the top of the jar, we can venture to say, from 

 the resistance made by the glass to any change of tempe- 

 rature, that the operations will be too slow; and for this 

 reason, it is proposed to make such a pendulum with a 

 thicker rod and stirrup bars, that they may not take 

 heat and cold too hastily. A steel jar would perhaps 

 answer the end as well as any other contrivance ; but 

 some would object to this, on account of the danger of 

 magnetism. But even a jar of this kind, from its being 

 made thin, (for it would be heavy were it as thick as 

 the glass one,) would be easily affected by the changes 

 of temperature ; and mercury being still more suscep- 



