350 
HOROLOGY. 
every minute by the maintaining-power, or main-fpring, 
and which actuates the contrate and balance wheels, and 
confequently impels the pallets with a conftant impulfe, 
independently of the maintaining-power, the latter being 
employed for no other purpofe but for winding up the 
former, as before fpecified. This {lender fpring, which is 
ten inches long, though it weighs only 3J grains, is coiled 
in a'fpring-barrel, apparently concentric with the contrate- 
wheel, and has its outer end attached to a hook in the 
barrel, with its inner end attached to another hook on the 
contrate-wheel. In order to make this delicate fpring 
anfwer its purpofe perpetually, two wheels, and as many 
pinions, in addition to the foregoing train, called the 
fourth and fifth wheel, and fourth and fifth pinion, are 
introduced in connection with a fly; as is likewife a de¬ 
tent with five arms, refembling a ftar, turning on the pi¬ 
vots of a common arbor. The numbers of thefe wheels 
and pinions appear to be as follow: of the fourth wheel 
112, which ftands concentrically over the contrate-wheel 
and fpring-barrel, of its pinion 14, of the fifth wheel 104, 
and of its pinion 12; and it is to the arbor of this lad 
pinion, we prefiume, that the fly is attached, though it 
does not appear evident either from the drawings or de- 
fcription. The mode of applying the three effective arms 
of the detent bears fome refemblance to the aCtion of the 
detents in the driking-part of a clock, in which the count- 
wheel is ufed inftead of a fnail, and where the locking and 
unlocking are alternately effected at meafured intervals of 
time. The detached fcapements, however, have now ren¬ 
dered this complex mechanifm fuperfiuous; and indeed 
it is manifeft, that the propofed object, of having a com¬ 
pletely detached power to act alike at all times on the 
pallets, is not thus perfectly effected; for the locking and 
unlocking eight times in every minute, to be produced 
by the {lender fpring in queftion, mull be luppofed to in¬ 
terfere in a certain degree with its regularity of action on 
the pallets; and, if it fnould be contended, that the eight 
deductions from the force of the fpring are regular and 
periodical in every minute, and that therefore they pro¬ 
duce an equable effeCl, yet the fame argument may be ufed 
in favour of a well-made train, in which the irregularities in 
the tranfmiflion of force from the main-fpring may like¬ 
wife be periodical, the wheels and pinions not being com- 
pofed of prime numbers. 
But it was not enough for Harrifon, that nearly an 
equable force was applied to maintain the motion of the 
balance, while the balance itfelf and alfo the balance-fpring 
were fubjeCt to alterations in their dimenfions by changes 
of temperature; he well knew that the fpirai fpring ufed 
in watches had more power when contracted, and lefs 
when elongated, than a mean power, and alfo that an en¬ 
larged balance has a greater momentum than a diminifhed 
one, and vice verfa ; the mode alfo of effecting an adjult- 
ment to counteract the confequent lofs or gain in the rate 
of going, was well known, not only to him, but to all 
watch-makers, and mechanically applied with fuccefs, as 
at the prefent day, by a moveable fluid, (called the regula¬ 
tor ■,) to limit the effective length ot the fpring); but this 
adjuftment was not a Jef-aSivg one ; an index was re¬ 
quired to be moved by a manual operation, which index 
was connected with the moveable fluid, and pointed out 
the quantity of the adjuftment; it remained for Harrifon 
to devife a mode of action, entirely dependent on the ftate 
of the atmofphere at any moment, which of itfelf would 
produce the requilite adjuftment; this he did, as we have 
already faid, by riveting together a flip of brafs and a flip 
of fteel,- which two metals are of unequal expanfibilities, 
to the remote end of which compound bar he attached his 
clip to hold the exterior thread of his fpring, after the 
outer end of it had paffed through, and been pinned to, a 
fluid of brafs attached to the upper plate of the frame ; the 
confequence proved, what none but a real genius would 
have forefeen, that the brafs elongating and contracting 
alternately in oppofite temperatures more than the fteel, 
produced a curvature in the compound piece, the concave 
fide of which was always occupied by the metal leaft 
elongated, that is, by the fteel in hot, and by the brafs 
in cold, weather; hence the compound piece, which car¬ 
ried the ftud backwards and forwards, he called a kirb, a 
Lincolnfnire word for curb, which contrivance is deftined 
to curb or command the effective length of the regulating 
fpring. The only material objection that experience has 
pointed out againft this felf-compenfating mechanifm, is, 
as we have before obferved, that {’mail pieces of metal and 
pieces in motion do not alter their temperature at the fame 
time with large pieces and pieces not in motion; nor yet 
fteel fo foon as brafs, even in fimilar circumftances; which 
confideration conftitutes an objection to the ufe of the 
thermometrical curb; an objection firft raifed by Harri¬ 
fon liimfelf, when it was greatly his intereft to have fup- 
preffed fuch a fuggeftion, as he was a candidate for the 
parliamentary reward. 
Harrifon had befides remarked, that in an ordinary 
watch the power which the main-fpring has over the ba¬ 
lance, through the medium of the train, compared with 
the power that the regulating fpring has over the fame, is 
as one to three generally; this power from the main-fpring, 
he obferved, being fulficient to put the watch in motion 
from a ftate of quiefceace, mult be too imperious for the 
balance, fmall and light as it was, to control it; accord¬ 
ingly he propofed, reafoning thus a priori, to give an ad¬ 
ditional momentum to his balance, compared with his 
maintaining-power; but momentum was to be attained in 
three different ways, namely, by additional weight given 
to a balance of the ufual diameter, by enlarging the dia¬ 
meter without increafing the weight, or laftly, by increaf- 
ing both in a certain degree: the firft mode was objecti¬ 
onable on account of the friction likely to be produced 
on the balance-pivots by a heavy balance; the fecond was 
alfo objectionable on account of the refiftance of the air 
it was likely to experience; and therefore he fixed upon 
theTaft mode of gaining momentum, by partly enlarging 
the diameter, and partly increafing the weight, of the ba¬ 
lance in ufe; by which means he conftruCted a balance 
over which the force from the maintaining-power of the 
remontoir has not more than one-eightieth part of the do¬ 
minion that the balance-fpring has. Hence in the time¬ 
keeper there is not force enough in the maintaining-power 
to excite motion from a quiefeent ftate, though there is 
power enough to overcome all the obftacles to continued 
motion, and to keep the piece going, when put in motion. 
In this he feems to have been followed by Mudge, whofe 
time-keeper does not ftrike off of itfelf.. 
The balance of the time-keeper before us is deferibed to 
be of more than three times the weight of that of a large 
ordinary watch, and of three times its diameter; for, ac¬ 
cording to the notes taken by Dr. Mafkelyne at the time 
of its examination, its diameter was 2% inches, and that 
of the plate 3 T g; hence a point in its circumference will 
pafs through 24 inches, or about four times the fpace of 
an ordinary watch, in each fecond, as Harrifon calculates 
from the arc of its vibration ; an advantage which has 
not been loft fight of by fome of the modern makers of 
chronometers ; and it may be confidered as an axiom in 
chronometry, “ that the perfection of a balance, confi¬ 
dered Amply as a regulator, independently of its compen- 
fation mechanifm, confifts in its having the greateft pofii- 
ble quantum of momentum with the leaft force from the 
train, and fmalleft quantity of friCtion and refiftance from 
the ain” 
The fufee has 6jt turns ; the pivot-holes are all bulbed 
with rubies containing pieces of diamond at the bottom 
of each; and the pallets are of diamond. The fly at the 
fifth pinion is ufed to regulate the velocity with which 
the fpringat the contrate-wheel is wound up every eighth 
part of a minute by the main-fpring when unlocked. 
iiz 96 
The dial-work is —— X —=12, and the feconds are 
28 32 
concentric by means of a wheel of 104 attached to the 
contrate- 
