HORO 
Fig. 94 reprefents Mr. Earnfhaw’s balance, which, 
like Mr. Arnold’s, has only two compenfation-pieces, and 
thole much (hotter than his, being little more than a 
quadrant each: a a are the fcrews of adjuftment for rate, 
fcrewed in to make the chronometer go falter, and out to 
make it go flower. The Hiding-pieces bb have each a cir¬ 
cular groove turned in a lathe or turning-frame, deep 
enough to form a bed for the expanfion-bars, in order 
that the interior fide lbrews cc may prefs againft the edges 
of the expanfion-pieces, and retain the Hiding-pieces in 
any given fituation. The pivots, Mr. E. fays, Ihould be 
conical, except very near the ends, which Ihould be cy¬ 
lindrical; and Ihould run in a “ jewel-hole "as lhallow as 
poliible, fo as not to endanger cutting the pivot; and the 
part of aftion of the hole Ihould be made quick back, 
with only a very lhallow chamfer behind to retain the oil.” 
The Hiding pieces bb are the weights of adjuftment for 
temperature; in making thefe weights, a brafs ring is firft 
made in the lathe, and is then cut through into fourteen 
equal parts in a clock-engine, by an operation fimilar to 
that of cutting a wheel into fourteen teeth, fo that each 
piece is the fourteenth part of a circle, after deducting the 
thicknefs of the cutter, which is not mentioned. One of 
the weights, in a box-chronometer, is about twenty grains 
generally. The expanlion-rims are turned out into a ring 
from a circular compound plate, which is made by filling 
the brafs in a crucible, containing the Iteel circle, held in 
a horizontal pofition during the time of fulion. The con- 
ftru&ion is'very Ample; but, as it appears to us, if the 
adjuftments for pofition were made by means of the fcrews 
for rate and weights for temperature only, one adjuftment 
would derange another, and that for polition would be 
very difficult to make nicely, as there is no tangent-fcrew, 
or other contrivance, to move the weights bb by gradual 
flow motions. As though to get. rid of this objeftion, 
which was too obvious to be overlooked, Mr. E. conceives 
that the caufes of a difference in the rate in different po- 
flfions, are large balance-pivots in part, but principally 
the badnefs of the balance-fpring ; his words relpefting 
the latter caufe, are thefe: “ Much difficulty has fallen to 
the lot of watch-makers in the endeavour to make time¬ 
keepers go nearly the fame in the different pofitions : I 
have, had my fnare of this, but it is now over; by far the 
greateft part of this difficulty arifes from the balance-fpring 
not being properly made. But, if the lpring is made as I 
lhall defcribe hereafter, you have only to make the balance 
of equal weight, and it will go within a few feconds per 
day in all pofitions alike; and, if it vibrates not more than 
one circle and a quarter,, by applying a Jmall matter of 
roright to that part of the balance which is downward, 
when in the polition that it lofes moll, (it) will correft it 
with great accuracy.” 
The board of longitude, not knowing what this “ fmall 
matter of weight” implied, nor how it is to be attached 
to the balance, w r ere induced to put the following quef- 
tion, viz. “ When the weight is wanted to adjuft the watch 
in the pofitions of III and IX, by what means do you ob¬ 
tain that weight, in the manner your balance appears to 
be made? If 5011 know any thing more that is material 
concerning the making time-keepers go nearly the fame 
iii different pofitions, communicate it.”— Anf. “To adjuft 
the watch in the pofitions of III and IX, (fays Mr. E.) 
I fix on to one of the compenfation-weights that is down¬ 
wards, when, in that pofition that it lofes molt, a fmall 
piece of brafs, not larger in diameter than a common pin¬ 
head, and nearly as thin as foolfcap paper. I fix it on 
with a very fmall particle of bees-zuax, not larger than the 
common dot of an i: that is, if the watch is gaining on 
mean time. But, if the watch is loling, I then take out 
the balance, and, with a drill, drill out a linall matter from 
that compenCation-weight that is uppermoft when in the 
pofition that the watch lofes molt; this I have found to 
correct it without fo many fcrews and fans.as I have feen 
in fome time-keepers.”—Sticking on brafs with bees-wax 
%a time-keeper is certainly a very curious proceeding!,. 
LOGY. 35') 
The rule for adjufting the balance for temperature is 
this: “ Put the watch into about 85 or 90 degrees of heat 
by the common thermometer, and mark down exaftly how 
much it gains or lofes in 1 2 hours; then put it into as 
fevere a cold as you can get, for 12 hours; and, if it gains 
one minute more in 12 hours in cold than in heat, move 
the compenfation-weights farther from the arm of the ba¬ 
lance about ^ of an inch; but, if it gains one minute 
more in 12 hours in heat than in cold, move the weights 
g of an inch nearer to the arm of the balance, and fo on 
in the like proportion, trying it again and again, till you 
find the watch go the fame in whatever change of heat 
and cold you put it.” Of courfe this adjuftment precedes 
the adjuftment for pofition, fo that moil’ probably the 
pofition, whatever it is, during thefe trials, is always the 
fame, a circumltance not fpecified. The total vibration 
at firft is confined to a circle and a quarter, which quan¬ 
tity, it is faid, will produce the moft fteady performance; 
but we are not informed what the banking is that limits 
the greateft vibration, nor indeed whether there is any 
banking at all, which we have’ been informed there is not 
in general. 
Mr. Earn (haw's greateft difficulty in the progrefs of his 
labours fieems to have been to find out what he calls the 
invi/ble properties of the balance-fpring, and to render it, 
not exaftly ifochronical, but nearly fo, that its deviation 
from ifochronifm may compenlate for a relaxation in the 
fpring, real or imaginary, ariling from conftant ufe, which 
is fuppofed. to atteft the permanency of the rate. After 
complaining bitterly of much difappointment in his firft 
labours, our author fays, “ I found, in the courfe of rea- 
foning on bodies, that watch-fprings relax and tire like 
the human frame, when kept conftantly in motion ; and 
this may be proved by the following experiment: Let a 
watch, that has been going a few months, go down, let it 
be down for a week or two, or more, then fet it going, 
and if it be a good time-keeper, fo as not to be affefted 
by the weather, it will go-fome feconds per day fafter than 
it did when it was let down, but it will again lofe its 
quicknefs in a gradual manner, gaining lefs and lefs, till 
it.comes to its former rate. Therefore, finding that ifo- 
chronal fprings would not do, and likewife having made 
fprings of fuch (hape as would render long and (hort vi¬ 
brations equal in time, (which) conftantly lofe (loft) the- 
longer the watch went; I then made them of fuch (hape 
as to gain in the fnort vibrations about five or fix feconds 
per day more than the long ones ; this quantity could only 
be found by long experience; and the way I proved this, 
was to try the rate of the watch with the balance vibrating 
about of a circle; I then try its rate when vibrating a 
circle and a quarter; and, if the (hort, vibrations go flower 
than the long ones, that watch will lofe on its rate, and 
if they are equal, it will likewife lofe, but that only from 
relaxation ; and, if it gains in the (hort vibrations more 
than fi^e or fix feconds in 24 hours, it will in the long 
run gain on its rate; but, if not more than that quantity,, 
and if the time-keeper is perfect in heat and cold, and 
every other part, the above properties will render it de- 
ferving of the name of a perfect time-keeper. This is a 
principal caufe of my time-keepers excelling all others ; 
and this is the principal caufe of fome of my time-keepers 
going better than others; the fprings of them being made 
to accord more exaftlv to the above proportions ; and this 
is the caufe which has enabled me to foretel what my 
time-keepers would do, which Dr. Mafkelyne, Mr. Crol- 
ley, and others, can teftify. The above effect is produced 
as follows : I find the common relaxation of balance-fpring^ 
to be about five or fix feconds per day on their rates in 
the courfe of a year; therefore, if the (hort vibrations are 
made, by the (hape of-the fpring, to go about that quan¬ 
tity fafter than the long ones, and as the fpring relaxes in¬ 
going,by time, fo the watch accumulates in dirt, (dirt accu¬ 
mulates in the watch more properly,) and thickening of 
the oil (takes place), which (hortens the vibrations, the 
(hort ones then being quicker compenfated (compenlates), 
