238 
NATURE 
| Fan. 30, 1873 : 
a eee 
defining lines of the metre will be cut at the distance of 
1 centimetre from each end of the bar. 
There appear to be many advantages in this new form 
of measuring bar, of a geometrical, mechanical, thermal, 
and economical character. Much importance is attached 
to the absolute uniformity of the bar throughout its whole 
length, as equalising its resistance and molecular action, 
and also to the adoption of a geometrical form as sym- 
metrical as possible. The absence of any acute angle 
was also dwelt upon as facilitating the mechanical dis- 
placement of the surplus metal; and it has since been 
practically ascertained that the planing can be executed 
with the utmost regularity and precision. It will also 
prove an excellent test of the soundness of the metal 
throughout the whole length of the bar. The great 
rigidity of this form of bar, combined with the advantage 
of the high elasticity of the platinum-iridium, was fully 
shown ; as compared with the rigidity of the Metre des 
Archives, it will be as 25°9to1, although its sectional area 
is only half as much more, The new form will also be 
highly favourable for equalising the temperature through- 
out its whole length and thickness, and for taking the 
temperature of the surrounding medium ; and it will 
afford a most convenient lodgement for mercurial ther- 
mometer tubes, thus enabling the actual temperature of 
the measuring axis of the bar to be readily and accurately 
determined. This measuring axis will be in one open 
and unbroken line, and quite unaffected in its dilatation 
by any contact with the support of the bar. Lastly, in 
an economical point of view, the form proposed will give 
the greatest possible strength with the least quantity of 
the costly material used. 
This form for the s¢re @ traits can be employed with 
merely a slight modification for any métres d bouts, or end- 
standards, that may be required. The form of the bars 
for the metres @ bouts will have a similar sectional figure, 
but symmetrical, the measure being defined by the sphe- 
rical end of two small cylinders, 3 millimetres in diameter, 
and projecting 1 millimetre from the middle of the ends 
of the bar, the radius of curvature being 1 metre. 
One other point may be noticed as to the mode of deter- 
mining the temperature and dilatation of the standards. 
The temperature at which the new metre will have its 
true length has been decided to be the same as that of 
the Metre des Archives, that is to say, o° C. All the 
necessary arrangements have been already made for 
making comparisons at this temperature by constructing 
a cold chamber expressly for the purpose, and surround- 
ing it with non-conducting materials. By a blast of cold 
air driven by a steam-engine in an adjoining room over a 
surface of ether and through pipes into the cold chamber, 
the temperature in it may be reduced in a few hours to 
the freezing point, and maintained constant there. From 
this adjoining room also the requisite light is conveyed 
into the cold chamber, ani is thrown by reflection on the 
bars and apparatus. There is an inner part of the 
chamber in which the standard metres and the comparing 
apparatus are placed, whilst the observer is enabled to 
make the adjustments and the comparisons through the 
microscopes from an outer part, and thus the heat of his 
body is prevented from exerting any disturbing influence 
on the bars and apparatus, 
Many comparisons of the metre will, however, be made 
at other temperatures, and_in all such cases, as well as for 
ascertaining the rate of expansion of the bars, the accurate 
determination of the temperature by thermometers will be 
requisite. The question of the amount of dependence to 
be placed on the indication of the temperature by mer- 
curial thermometers, which has recently been a good deal 
agitated in this country, was considered by the Commis- 
sion to be one of great importance. They found that in 
all mercurial thermometers, the dilatation of the glass 
envelope, which, so far as it is known, is only about one- 
seventh that of the quicksilver, renders the reading of the 
best calibrated thermometers liable to errors amounting to 
some tenths of a centesimal degree. The best authorities 
are also of opinion that implicit dependence cannot be 
placed on the constancy of mercurial thermometers, so 
far as they indicate the temperature, nor on the constancy 
of the dilatation of the glass envelope. It was thought, 
therefore, that for ascertaining the temperature with a 
degree of precision exceeding o”r C., recourse must be 
had to an air thermometer. 
On the other hand, the air thermometer is an instru- 
ment complicated in construction and difficult to use. It 
requires the greatest precautions and practised skill in its 
manipulation ; and the necessity of having recourse to an 
air thermometer on every occasion of making com- 
parisons with the primary standards would create very 
serious embarrassments. On these grounds it was decided 
that every one of the new metres should be accompanied 
by two detached mercurial thermometers, carefully com- 
pared with an air thermometer, and which should be re- 
verified with it from time to time. 
It was stated to the Commission by M. H. Saint-Claire 
Deville, as the result of twenty years’ use of an air ther- 
mometer, that no instrument could be more precise and 
convenient in reading, more easy and expeditious in use. 
He estimated that by employing an air thermometer 
according to a method suggested by him, the mean tem- 
perature of ‘a standard metre under comparison could be 
determined with precision to the zisth of a degree of 
the centigrade scale. : 
On the subject of dilatation, we can only briefly allude 
to M. Fizeau’s admirable method of accurately determin- 
ing the rate of expansion of solid bodies by heat, by em- 
ploying the length of a wave of monochromatic sodium 
light (a constant = 0'005888 millimetre, or 0’00002318 
inch), as his standard of measure. By means of an in- 
genious apparatus constructed by M. Soleil, the yellow 
ray is made to fall vertically through a piece of plate glass 
on a horizontal plane of the solid body, and is reflected 
in the under-surface of the glass. By counting the 
number of Newton’s rings passing a fixed point upon the 
glass, when they are set in motion from the expansion of 
the surface of the solid body by observed degrees of heat, 
its dilatation can be computed with the greatest precision, 
This method has been described in the proceedings of the 
Royal Society on November 30, 1866, when the Rumford 
gold medal was awarded to M. Fizeau for it. The Com- 
mission also hope to obtain a standard of dilatation by 
marking a measure of length of one or two decimetres 
upon the plane surface of a piece of Beryl in its axis of 
non-dilatation. M, Fizeau has shown that Beryl varies 
in its dimensions from heat less than almost any other 
body, and that it possesses this peculiarity, that whilst it 
expands by heat in the direction of its axis of crystallisa- 
tion, it contracts by heat in the direction perpendicular to 
