M E C II A N I G S 
■with one flame in the middle ; then with two flames; and 
f'uccefiiveiy to one with three, four, and five, flames. Bur, 
previous to this trial, he took care to cool them equally, 
by expofing them (bme time upon the fame Itone, when 
it began to freeze, and Fahrenheit’s thermometer was at 
32 degrees. The effects of thefe experiments are digefted 
in the following Table, where the degrees'of expanfion 
are marked in parts equal to the T 2 ^g -h part of an inch. 
Exparjion 'of 
Iron 
Steel 
Copper 
Bra/s 
Tin 
Lead 
By one flame 
80 
8J 
89 
I 10 
153 
r SS 
[By two flames placed 
dole together 
117 
123 
220 
274 - 
[By two flames t.\ inches 
diltant 
IO9 
94 - 
92 
Z41 
219 
2o 3 
By three flames placed 
dole together 
142 
168 
193 
2 7 5 
By four flames placed 
clofe together 
21 I 
270 
270 
361 
By five flames 
230 
3 10 
310 
377 
5 
It is to be obferved of tin, that it will ealily melt when 
heated by two flames placed together. Lead commonly 
melts with three flames placed together, efpecially if they 
burn long. 
From thefe experiments, fo far as they are correct, it ap¬ 
pears, at firft view, that iron is the lealt rarefied of any of 
thefe metals, whether it be heated by oneor more flames.; 
and therefore is molt proper for making machines or in- 
ftruments which we would have free from any alterations 
by heat or cold, as the rods of pendulums for clocks, &c. 
So likewife the meafures of yards or feet fnould, if of me¬ 
tal, be made of iron, that their length may be as nearly as 
poflible the fame in fummer and in winter. The expan¬ 
lion of lead and tin, by only one flame, is nearly the fame; 
that is, alnioft double of the expanfion of iron. It is like- 
tv ife obfervable, that the flames placed together caufe a 
greater rarefaction than w hen they have a fenfible interval 
between them ; iron, in the former cafe, being expanded 
317 degrees, and only 109 in the latter; the reafon of 
•which difference is obvious. By comparing the expan- 
fions of the fame metal, produced by one, two, three, or 
more, flames, it appears, that two flames do not caufe dou¬ 
ble the expanfion of one ; nor three flames three times 
that expanfion, but always lefs ; and thefe expanlions differ 
fo much the more from the ratio of the number of flames, 
as there are more flames a fling at the flame time. It is 
a!fo obfervable, that metals are not expanded equally at 
the time of their melting, but fome more, fome lefs. Thus 
tin began to run when rarefied 219 degrees; whereas brafs 
was expanded 377 degrees, and yet was far from melting. 
The great difference between the expanfions of iron and 
brafs has been applied with good fuccefs to remedy the 
irregularities in pendulums ariling from heat. See the ar¬ 
ticle PIOROLOGY, vol. X. p. 308, 9. 
Mr. Graham ufed to meafure the minute alterations, in 
length, of metal bars, by advancing the point of a micro- 
meter-fcrew, till it fenfibly flopped againlt the end of the 
bar to be meafured : this fcrew, being fmall and very 
lightly hung, was capable of sdjuftment within the three 
or four-thoufandth part of an inch. On this principle 
alfo Mr. Smeaton contrived his pyrometer, in which the 
meafures are determined by the contact of a piece of me¬ 
tal with the point of a inicrometer-fcrew. This pyrome¬ 
ter (Smeaton’s) is preferred by Mr. Adams to all others : 
lie has given the following delcription of it. 
It is founded on the following principles: i. The 
quantity of expanfion being in proportion to the length 
of the bar, the longer the bar, the more fenfible the ex¬ 
panfion. 2. The fcale on which the alterations are mea¬ 
sured ought to be fo large, as that the fmalieft alteration 
may be vilible. 3. The materials of which the ineafuring- 
parts of the inftrument are made (hould fuffer no expan¬ 
fion during the experiment, or the degree or expanfion 
produced in them llfould be known and accounted for j 
be caufe the expanfion of the inftrument, fuppofing the 
bar to be meafured does not expand, will produce the fame 
appearance as the expanfion of the bar, fuppofing the in- 
firument not to expand. 4. As all bodies continue to ex¬ 
pand in proportion as the heat applied to them is increafed, 
it is neceffary to afcertain the degree of heat applied, in 
order to determine the comparative expanfion of different 
metals. 3. The meafuring-parts of the inftrument ought 
to be fo large, as that the quantities of the meafured ex- 
panfion may be known in real meafure. 
This inftrument, which is represented at fig. 43, 44, is 
fo con ft rutted as to receive a bar two feet four inches long, 
and might be made capable of receiving bars of a much 
greater length of fome kinds of materials, but not of others, 
on account of their flexibility, even with a degree of heat 
not exceeding boiling-water. The meafures are deter¬ 
mined by the conta< 5 ! of a piece of metal with the point 
of a micromef’er-ferew. The obfervation is belt judged 
by the hearing, rather than that of the fight or feeling : 
by this method it has been found pratticable to repeat the 
fame meafurement feveral times, without differing from 
itfelf above the 20000th part of an inch. The degree of 
fenfibility attained by this method is fuperior to any tiling 
that can be done by fight or feeling. 
As no lubltance has hitherto been difcoverecl that is 
perfectly free from expanfion by heat, the bar which makes 
the baiis of the inftrument fuffers the fame degree of heat 
as tlie bar to be meafured ; confequently, the meafures 
taken by the micrometer are the differences of their expan¬ 
fion. The expanfion then of the bafis between two given 
degrees of heat being once found, the abfolu.te expanfion 
ofany other body, by adding or fubtracting the difference 
to or from the expanfion of the bafis, according as the 
body to be meafured expands more or lefs than the balls, 
will be determined. 
When the inftrument is made ufe of. it is immerfed, to¬ 
gether with the bar to be meafured, in a ciilern of water: 
which water, by means of lamps underneath, is made to 
receive a certain degree of heat, not exceeding that of 
boiling; and thereby communicates the fame, degree of 
heat to the inftrument, to the bar, and to the mercurial 
thermometer immerfed therein, for the purpofe of alcer- 
taining the degree. 
All that remained, was to find the abfolute expanfion 
of the bafis between any two given degrees of heat, not 
greater than that of boiling-water; which is thus effected. 
Let there be prepared a bar of ftraight-grained white deal, 
or of cedar, which it is well known are much lefs expan- 
five by heat than any metal hitherto difcovered ; let th« 
bar be adapted to the inftrument, in like manner as the 
ether bars intended to be meafured ; but, that the foft.- 
nefs of the wood may not hinder the juftnefs of its bear¬ 
ings, let its ends be guarded with a bit of brafs let into 
the wood at the points of contafl, to prevent as much as 
may be the moifture or (team of the water from affecting 
the wood ; let it flrft be well varniihed, and then, being 
wrapped round with coarfe flax from end to end, this whi 
in a great meafure imbibe the vapour before it arrives at 
the wood ; let the ciftern alfo be fo contrived, that, the 
inftrument being fupported at a proper height therein, the 
bar to be meafured may, upon occaiion, be above the co¬ 
ver, while the bafis remains in the water; thus will the 
cover alfo be a defence againft the moifture. Lift the wa¬ 
ter in the ciftern be now brought to its lower degree of 
heat, fuppofe at or near the freezing point, the bails hav¬ 
ing continued long enough in the water to receive the 
fame degree of heat, and the wooden bar having been 
previouily kept in an adjacent room, notfubjed to iudden 
alterations of temperature by fire, or other caules ; let the 
bar be applied to the inftrument, and the degrees of the 
micrometer and the thermometer read off and let down. 
Let the wooden bar be then reftored to its former place, 
till the water is heated to the greater degree intended, fup- 
pole at or near that of boiling-water ; the lid being now 
(hut down, and the chinks flopped with coarfe flax, to 
4 ' . present 
