ME C H x 
prevent the inTuing of the fleam as much as poffible, let 
the wooden bar be again brought forth, applied to the in- 
Itrument, and the degrees of the micrometer and thermo¬ 
meter read off as before. The difference of degrees of the 
micrometer correfponding to the difference of degrees of 
the thermometer will exprefs the expaniion of the bafis 
between thofe degrees of heat; that is, upon the fuppoii- 
tion that the wooden bar was of the fame length at the 
time of taking the fecond meafure as at the firff. In¬ 
deed, a meafure can hardly be taken without fome lofs of 
time, as the whole of the inffrument, when the hot mea¬ 
fure is to be taken, is confiderably hotter than the wooden 
bar; and, in the cafe of boiling-water, the (team being 
very repellent and active, the bar is liable to be fenfibly 
aft'e&ed in its length, before the meafure can be taken, 
both by heatand moifture, which both tend to expand the 
bar; but, as the quantity is fmall, and capable of being 
nearly afcertained, a wooden bar thus applied will anfwer 
the fame end as if it were unalterable by heat or moilture. 
In order, therefore, to know the quantity of this altera¬ 
tion, let the time elapfed between the firff approach of the 
bar to the inffrument and the taking of the meafure, be 
obferved by a fecond-watch, orotherwife; after another 
equal interval of time, le* a fecond meafure be taken ; and,\ 
aftera third interval, a third and a fourth. The three differ¬ 
ences of thefe four meafures will be found nearly to tally 
with three terms of a geometrical progreffion, from which 
the preceding terms may be known, and will be the cor¬ 
rection ; which, if applied to the meafure firff taken, re¬ 
duces it to what it would have been if the wooden bar 
had not expanded during the taking thereof. From a 
fetv obfervations of this kind carefully repeated, the ex- 
panfion of the bafis may be fettled ; and, this once done, 
the making experiments upon other bars will become very 
cafy and compendious. 
The bar of brafs which compofes the bafis is an inch 
broad by half an inch thick, and (lands edgewife upwards ; 
one end is continued of the fame piece at right angles, to 
the height of three inches and a half, and makes a firm 
lupport for the end or the bar to be experimented ; and 
the other end aifts upon the middle of a lever of the (e- 
cond kind, whole (uicrum is in the bafis ; therefore, the 
motion of the extremity of the lever is double the differ¬ 
ence between the expansion of the bar and the bafis. 
This upper-part of the lever riles above the lid ot the cif- 
tern, lo that it and the micrometer-fcrew are at all times 
clear of the water; the top of the lever is furnifhed with 
an appendage called the fcekr : it is the extremity of this 
piece which comes in contact with the micrometer-fcrew. 
The conffrudiion and application hereof will better appear 
from the figure than from many words: it hence appears 
that, having the length of the lever from its fulcrum to 
the point of fufpenfion of the feeler, the diffance between 
the fulcrum and the point of contact with the bar, the 
inches and parts that correfpond to a certain number of 
threads of the micrometer, and the number of divifions 
in the circumference ot the index-plate, the fraction of an 
inch expreffed by one divifion of the plate may be tie-' 
duced ; thofe meafures are as follows : 
Inches. 
From the fulcrum of the lever to the feeler - 5 875, 
From the fulcrum to the plate of contact - - 2'8y5 
Length of 70 threads of the (crew - - 2‘455 
Divifions in the circumference ot the index-plate 100 
Hence the value of one divifion will be the part of 
an inch ; but, if the (crew be altered one-fourtn ot one of 
thefe divifions, when the contadt between the fcrew and 
feeler is well adjufted, the difference of contact will be very 
perceivable to the (lighted obferver; and, confequentiy, 
--tj—th part of an inch is perceivable in this inffrument. 
One thing (till remains to be fpoken of; and that is, 
the verification of the micrometer-fcrew, which is the only 
part of this inffrument that requires exadtnefs in the exe¬ 
cution ; and how difficult thefe are to make perfectly good 
is well known to every perfon of experience in thefe mat¬ 
ters ; that is, that the threads of the fcrew ntay not only be 
1 N I C S, ' 735 
equidiftant in different places, but that the threads fhall be 
equally inclined to the axis in every part of the circum-, 
fere nee. As nearly the lame part of the fcrew is made 
ufe of in thefe experiments, the latter circumffance is 
what principally needs inquiry ; for this purpofe, let a 
thin (tip of lteel or other metal be prepared, tvhofe thick- 
nefs is about one-eighth of the diffance of the threads; let 
the edges of this thin plate be cut into fuch a fltape as ex¬ 
actly to fit into the fixed notch in which one end of the 
bar is laid ; let a fcrew pafs through the ltandard of brafs 
on which that notch is fupported, in fuch a manner that 
the end of the bar to be meafured, that is fartheff from 
the lever, may take its bearing againff the point, or rather 
the fmall hemif'pherical end, of this fcrew. Let one of 
the brafs bars ufed in the other experiments be applied to 
the inffrument, and a meafure taken : then let the thin 
plate be put in between the end of the bar and the point 
of the fcrew laft-mentioned, and again take the meafure ; 
but firff obferve that the plate is put down to the notch, 
fo that the fame place of the plate may always agree with 
the point of the fcrew ; and confequentiy no error may 
arife from a different thicknefs in different places of the 
plate; obferve alfo that the whole comes to a true bear¬ 
ing; then advance the fame fcrew till the micrometer- 
fcrew is pufhed backwards one-fourth of a revolution ; 
again repeat the meafure with and without the thin 
plate ; again advance the former fcrew, fo as to make that 
of the micrometer recede another quarter of a turn, and 
repeat the meafures with and without the thin plate. 
This method being purfued as far as neceffary, it is evi¬ 
dent, that, the thicknefs of the plate being always the 
fame, if the difference of meafures take 1 with and without 
it are not always the fame in the different parts of a revolu¬ 
tion of the micrometer-fcrew, that this fcrew is not equi¬ 
angular; but from the differences of the meafures corref¬ 
ponding to the thicknefs of the fame plates in the differ¬ 
ent parts of a revolution, the errors thereof may be nearly 
afligned. For greater certainty in this examination, left 
the heat of the obferver’s body (hould aftedt the bar or in- 
ftrument during the obfervation, let the whole be im- 
inerfed in the cittern of water, which ought to (tandaluf- 
ficient time before the obfervation is begun to acquire the 
fame temperature as the air, which alio ought to be in a 
fettled ffate. 
A B C D, fig. 43, is the main bar or bafis of the inftru- 
ment. EF is the bar to be meafured, lying in two notches ;. 
one fixed to the upright ftandard'AB, the ether to the 
principal level, H I. The end E of the bar bears againft 
the point of G, a fcrew of ufe in examining the micro¬ 
meter-fcrew. The other end of the bar, F, bears againft 
a fmall fpherically-protuberant bit of hard metal, fixed, at 
the fame height as G, in the principal lever, HI. K is 
an arbor fixed in the bafis which receives at each end the 
points of the (crews H, L, upon which the lever H I turns, 
and ferves as a fulcrum thereto. O is a (lender fpring to 
keep the lever in a bearing (fate againft the-liar; and R is 
a check, to prevent the lever from falling forward, when 
the bar is taken out. N is the feeler, lomething in the 
ffiape of a T inverted, x* and moveable up and down 
upon the points of the ferews I, M, which, as well as L, H, 
are fo well adjufted as to leave the motion free, but with¬ 
out fhaking. R is the handle of the feeler, moveable upon 
a Joofe joint at R; fo that, laying hold of the knob, the 
feeler is moved up and down, without being affefted by 
the irregular preffure of the hand. The extremity, S, of 
the feeler is alfo furniffied with a piece of protuberant hard 
metal, to render its contact with the point of the inicro- 
meter-ferew the more perfect. T, the micrometer-fcrew. 
V, the divided index plate; and W, a knob or handle. 
The micrometer-fcrew paffes through two (olid fcrewed 
holes at D and Y. The piece Y Z, is a little fpringy, and 
a£ts upon the fcrew backwards from the hole D, keeping 
the micrometer-fcrew conltantly bearing againft its threads 
the fame way, thereby rendering the motion thereof fteady 
and gentle. X, the index, having divifions anfvvering to 
the turn of the fcrew. This points out the divifions of 
