]\gt ; VII .
of the of a Cubic of Distilled H. J. CHANEY .
Cornmunicated by Sir G. GABRIEL STOKES , Bart. , F.R. Received February 4 , \mdash ; Read June 19 , 1890 , \mdash ; Revised January 16 , 1892 .
THE evaluation of the mass of a cubic inch of distilled , as at present accepted , was based on weighings made in 1798 by Sir G. SHUCKBURGH Phil. Trans 1798 , p. 133 ) ; and on measurements made in 1821 by Captain KATER ( ' Phil. Trans 1821 , pp. 316 and 326 ) .
Subsequent researches however , particularly those in relation to the mass of a cubic decimetre , show that it is desirable to -determine the mass of the cubic inch of distilled water .
The result of SHUCKBURGH'S was that the cubic inch of distilled water at the temperature of Fahr. , the barometer being at inches , weighed grains , and this value as corrected by Captain KATER , became grains at the temperature of Fahr. , the barometer being at 30 inches ; or in vacno Fahr. the cubic iuch of distilled water weighed grains .
This corrected value , grains , has been adopted in various legislative enactments ; \mdash ; for instance the Weights and Measures Act , 5 Geo. IV .
, , section 5 ( 1824 ) , declared that " " a cubic inch of distilled water , weighed in air by brass weights , at the temperature of degrees of FAHRENHEIT'S thermometer , the barometer being at thirty inches , is equal to two hundred and fifty-two graius and four hundred and fifty-eight thousandtl ] parts of a grain , of which the imperial standal 'd Troy pound contains five thousand seven hundred and sixty Owing to doubt as to the true mass of a cubic inch of distilled water , the above section of the Act of 1824 was led in 1878 , and has not been -enacted .
No -determination of the mass of a cubic inch of distilled water in terms of the present imperial pound has yet been made ; the above value , grains , being based on the old Troy pound of 1758 .
Dr. WILD'S investigations appear indeed to show that later results obtained in the evaluation of a definite volume of distilled water , differ appreciably from those obtained in terms of the old Troy pound .
If , for instance , the mass of a cubic decimetre of distilled water at C. , as originally determined by and TRALLES in 1799 , is taken as grms. * Bericht uber die Arbeiten sir Reform der Schweizerischen Urmaasse , ' Zuricl ] , 1868 .
MR. J. CHANEY'S fION OF Then the equivalent value of the cubic decimetre , as deduced from the SHUCKBURGH and KATER value of the cubic inch , would be . . . . . . . . . . . . . . . .
grms. Other investigations have given:\mdash ; ( 1825 ) BERZELIUS / ( 1S34 ) STAMPFER , , ( lS41 ) KUPFFER The arithmetical mean of these results is grms. , and there is a diffel.once of .
between he lowest result .
The difference between the results appears to mainly from inaccuracy in the measurem of temperature , and uncertainty as to the condition of the water used .
Dr. BROOCH reyised a report by TRALLES on original work , and at conclusion that probably the true of a cubic decimetre of distilled water at its maximum , and in a vacuum , is from 90 to 120 mgrms .
less than 1000 grms. Methods arpd Instruments .
It has been the object of the present investigation to ascertain the weight of water displaced by a body or gravimeter , whose masses , in air and in , and linear dimensions , had been carefully ascertained , rather thar1 to determine the theoretical mass of a cubic inch of water .
It has been considered that the mass of a definite volume of distilled water cannot well be ascertained by weighing the water contained in a vessel of a given capacity ; or by the use of gravimeters , made only of one material ( as brass ) , and of one particular form ( as a sphere ) .
Three gravimeters ( or hydrometers ) therefore , of the following forms were now adopted:\mdash ; Two gravimeters , and , of cylindrical form , one hollow and the other solid , the third gravimeter being hollow , and of spherical form .
The hollow gun-metal circular cylinder ( fig. 1 , page 340 ) was nearly nine inches in height and diameter , and was adopted as being a body the weight and dimensions of which might give the least probable error in the several operations of weighing in air and in water .
The gravimeter was protected form oxidation by platinising , it having been first made airtight .
For the purpose of linear measurement , there were traced on the gravimeter a series oflines , as indicated in fig. 1 .
These lines were cut rather deeper than was desirable , but the space taken up by them on the surface of the cylinder has been considered .
* ' Proces-Verbaux\mdash ; Commission Inter.nationaledu Metre .
Reunions des Membres Franqais .
1873 .
' THE MASS OF A CUBIC INCH OF DISTILLED WATLR .
The cylinder was adopted as a solid gravimeter because the density and expansion of quartz are well known , as well as on account of the hardness of quartz , its capability of receiving a high polish and absence of hygroscopic properties .
is a pure crystal , and was originally adjusted under the directions of Dr. VOIT , its coefficient of linear expansion for 1o C. being taken as in a direction parallel to the principal axis of the crystal .
The third gravimeter , is a hollow brass sphere , having on its surface two engrayed lines for the purpose of linear measurement , and its approximate diameter is six inches .
Expansion of Water.\mdash ; As the weighings of the gravimeters in water could not be actually made at the normal temperature of Fahr. , corrections were necessary for the density of water at various temperatures .
There has not yet been made any determination of the rate of expansion of water which might alone be accepted , and we have therefore to adopt the mean result of several selected determinations .
ROSSETII has stated the result of his experiments ( ' Atti del Istituto Veneto , ' 1867-8 ) , as well as those of KOPP ( 1847 ) , PIERRE , DESPRETZ ( 1839 ) , HAGEN ( 1855 ) , and MATTHIESSEN ( 1866 ) , in a table which gives as the ratio of the density of water at Fahr. to its maximum density at C. It does not appear whether ROSSETTI corrected the earlier results by more recent determinations of the rates of expansion of mercury and of glass ; which corrections at some temperatures would affect the last place of decimals in the above expression .
The differences between the above determinations would also affect the present investigation to grain in the mass of the cubic inch of water .
In 1870 Dr. FOERSTER , after a critical examination of the results above referred to , as well as of those of JOLLY ( 1864 ) and W. H. MILLER ( 1856 ) , adopted the mean results of MILLER and ROSSETTI ( after SCHIAPARELLI , 1868 ) ; which give as the ratio at Fahr. ; or if the maximum density of water is taken the temperature of C. , then at Fahr. th density of water may be expressed by ; and these mean results have , therefore , been followed in the present investigation .
* ' Metronomische ' No. 1 , Berlin , 1870 .
'Phil .
Trans 1857 ( Part III .
for 1856 ) .
MR. H. J. CHANh'Y'S RE-DETERblINATION OF TABLE I.\mdash ; Density of Water .
Logarithms of Ratios of the Maximum Density of Water to its Density at Condition of the Water.\mdash ; The water was twice distilled in pure tin and glass stills , was found to be free from any impurities likely to affect the weighings .
It was deprived of air by boiling , and no correction for the absorption of air was therefore made .
It was , however , found that at the temperature of Fahr. ( B. inches ) a cubic foot of distilled water , freed from air , weighed bout 3 ] .
grains more than when nearly " " saturated\ldquo ; with air .
of Length Standards of Length and .\mdash ; For the purpose of measuring the external dimensions of the gavimeters in inches of the present imperial yard , four 9-inch end standards of length were used ; as well.as a steel 6-inch and a stee13-inch standard ( 1884 ) .
At Fahr. , the true mean of the four 9-inch standards was inches inch ; the 6-inch and 3-inch standards having a true length of and 299975 inches respectively .
For tervals between and inch there was also used a subdivided inch " " the accuracy of the subdivisions of which had been measured by the Standards Commission in 1868 inch .
For the comparison of the dimensions of the gravimeters with the standards of length , there were used two comparators of well-known forms ; one a Whitworth contact comparator , other a micrometer-microscope comparator designed by Mr. J. SIMMS .
By means of the former comparator measurements by touch might be made to inch ; and optically to inch , by means of the microscopes .
THE MASS OF A CUBIC INCH OF DISTILLED WATER .
When two stan dards or measures of extension are to be compared by means of the Whitworth comparatol , first one standard and then the other is placed in a horizontal position between two contact points ; one of which is a fixed point , the other being point or the termination of a micrometer-screw .
Any difference , in parts of an inch , between the two standards can then be measured by means of the screw , the value in parts of an inch of one ] ution or whole turn of the screw being known .
In this comparator , the appreciation of the moment of contact is ascertained by means of a " " grayity-piece.\ldquo ; When the gravimeter was placed in position , one of its sides was in contact with the fixed point , its other side being in contact with the gravity-piece , which was interposed between the gravimeter and the screw , and the moment of contact between the gravimeter and the screw was then ascertained by gradually turning or releasing the screw until the gravity-piece fell by its own weight .
In the general form of the ordinary micrometer-microscope comparator , or optical beam-compass , there are two fixed microscopes , the distance between which is determined by comparison with a standard of length .
micrometer heads of the microscopes are divided into 100 divisions , the mean value of one division being equal to nearly inch ; and the linear Lqtandards or measures of extension are compared by being alternately placed under the microscopes , a microscope being fixed over either end of the standard under observation .
As the comparators were only required for the purpose of measuring small differences between the standards and the gravimeters , it was unnecessary to verify the whole of the micrometer screws ; but the particular parts of the screws used were verified by comparison with the subdivided inch D. There were no defining lines marked on the gravimeter by which its dimensions could be measured , although certain lines , as previously mentioned , were engraved on and , for the purpose of indicating generally the particular parts of the surface of the gravimeter brought under measurement .
Therefore , in the microscopic comparisons of , it became necessary to develop defining lines , visible through the microscope , and this was done after the method of FIZEAU and CORNU .
When the grayimeter was placed position under the fixed microscopes , two pencil points , made of polished silver , were brought nearly into contact with the gravimeter , within inch of the gravimeter at each end .
The " " points\ldquo ; did not therefore actually touch the sides of the gravimeter , but were so reflected as to form an apparent line of contact , visible through the microscope , on either side of the gravimeter .
The actual distance between the two apparent lines formed by the " " points\ldquo ; and their reflections , as measured by the fixed microscopes , was then compared with a standard of length which was placed under the microscopes after the removal of the gravimeter .
Rates of of the Gravimeters by .\mdash ; The rates of expansion by heat of the three gravimeters were not ascertained by actual experiment , as the obable error which would arise in ascertaining the rite of expansion of a body having the particular form of either one of the gravimeters would be greater the probable error of experiments on the dilatation of a body of the simple form of a rod or bar .
gravimeter for instance , is made gun-metal alloy , the coefficient of linear expansion of which alloy was found by SHEEPSHANKS in 1848 to be for 1o .
It was subsequently found by CLARKE ( 1860 ) to be ; FIZEAU , also , has shown that SHEEPSHANKS ' coefficient was too low , owing possibly to the form of the mercurial thermometer adopted in 1848 .
For the gravimeter , CLARKE'S coefficient of linear expansion has been therefore taken .
the gTavimeterS the coefficient of linear expansion for 1o Fahr. , , has been taken ; and for the cubic expansion for 1o Thermometers .
Six standard mercurial thermometers were used\mdash ; Centigrade , Fahrenheit , , 4,517 4 , 518 4 , .
20,065 The errors of these thermometers were originally determined in relation to the hydrogen thermometer at the Bureau International des Poids et Mesures , Paris , and were re-determined after the experiments .
Thermometers 4517 and 4518 were made of a hard glass ( including silica , lime , soda , alumina , sulphuric acid , and potasb ) , and were highly sensitive .
Thermometer No. 4517 .
Divided into ] C. from to .
Distance from of reservoir to C. is 53 mm. Total length of thermometer is 702 mm. of a degree is mm. OBRECTION for Calibration .
Divisions equidistant .
Probable error of correction C. THE MASS OF A CUBIC INCH OF WATER .
Coefficient of pressure ( or pressure on reservoir ) for one millimetre of mercury , as applied when thermometer was used in a vertical position is Boiling-point in a horizontal position is .
The value of a division in " " normal\ldquo ; degrees being Error of zero-point , when the thermometer was placed in a horizontal position , was C. Thermometer No. 4518 , is a Centigrade thermometer of similar form to 4517 .
Boiling-point in a horizontal position is 10 0665 .
The value of a division being Zero-point ; horizontal position , Thermometer No. 4575.\mdash ; Divided into C. from to .
Distance from middle of reservoir to C. is 64 mm. Total length of thermometer , 482 mm. Divisions not equidistant , but by comparison with a standard thermometer the following corrections were obtained:\mdash ; Coefficient of pressure is Zero-point ; horizontal position , Thermometer ; Kew , 430.\mdash ; Divided into 1o Fahr. from to Fahr. Nos. 12,765 and 20,065 .
Divided into tenths of a from to Fahr. 20,065 .
mm. It was found that the construction of the above thermometers , Kew , 12,765 , and 20,065 , did not permit of exact calibration , and the following corrections were determined in a vertical position , by comparison therefore with a standard thermometer .
Estimated probable error of correction is Fahr. The coefficients of interior pressure at 10 C. for one millimetre of mercury is , for No. 430 ; and for the other thermometers the coefficients of pressure were stated 12,765 . . . .
20,065 . . .
Zero-points:\mdash ; Kew , 430 .
Fahr. 12,765 .
20,065 .
Weighings .
Weighings.\mdash ; For the weighings three ances were used , by means of which differences of , and of a grain might be ascertained respectively ; for weighings in water it was possible , however , to weigh only to of a grain .
The weighings were made by BORDA'S method of counterpoise , the position of equilibrium of the beam , and the weight of water displaced by the cylinder being calculated after the methods given by MILLER and BROOCH .
The gravimeter was suspended in a glass vessel by a platinum wire hook , and was surrounded by at least two inches of water , the depth of the water being regulated so that it always rose to the same height the vessel , whether the gravimeter was suspended therein or the wire hook only .
The wire hook was kept polished , any water found to adhere to its upper surface being either wiped off or dried off by a blowpipe before the weighings were taken .
THE MASS OF A CUBIC INCH OF DISTILLED WATER .
In such weighings in water minute bubbles of are found to be carried down by the gravimeter or by the wire by which it is suspended , but by repeated immersions , and examination with a small telescope and a glow-lamp , it is possible to avoid appreciable error in this direction .
When not in use the gravimeters were kept in air of similar temperature to that at which the weighings were made .
It is , of course , difficult , during such weighings , to find the precise temperature of a comparatively large volume of water , and as any uncertainty of Fahr. in the reading of the thermometer during the weighing of the cylinder in water would amount to nearly three rains , it was desirable not only to use sensititive thermometers , but to place their bulbs at proper depths , and to read them quickly by means of cathetometers during the periods of weighing .
All weighings in air were reduced to " " normal air or air which at the temperature of Fahr. , the barometer being at 30 inches at } Fahr. , reduced to latitude and at sea level , contains four volumes of carbonic anhydride in every 10,000 volumes of air , and also contains two-thirds of the amount of aqueous vapour contained in saturated air .
This is the average condition of the air at Westminster ( latitude at 16 feet above sea level ) , where the weighings were made .
A litre of such air , if dry , but containing the above proportion of carbonic anhydride , would weigh grammes ; and , if dry , but containing no carbonic anhydride , would weigh grammes .
MR. H. J. -DETERMINATION OF Details of and Weighings .
ing details of Jrem e and weigbings of the three gravinleters , and ; together with statements of the results obtained .
Fig. 1 .
CYLINDER UPPER FACE ' '/ LOW ER FACE THE MASS OF A CUBIC INCH OF DISTILLED At Fahr. the mean height of the cylinder was therefore taken at inches .
* Ten consecutive micrometer readings on November 1 , gave a result by one observer as follows : \mdash ; .
7 THE MASS OF A CUBIC INCH OF DISTILLED WATER .
@E . . . . .
; co THE MASS OF A CUBIC INCH OF DISTILLED IL . . . . . . . . .
.34 MR. H. J. RE-DETERMINATION OF The lines traced on have the effect of diminishing its cubic capacity ; these lines are -shaped , and have a depth of nearly -inch , with an average breadth at the top of -inch .
There are 24 vertical lines , each nearly 9 inches long ; three circular lines , each nearly 28 inches long ; besides two circular lines nearly 14 inches long .
The edges of the cylinder were not quite sharp .
If , therefore , the diameter of is taken at inches , then the cubic contents of may be taken at ) cubic inches .
WEIGHING of Cylinder in Air .
Date , 1888 .
Description oused ( Before immersion .
in .
19 After immersion .
10 , grains grai ainC sgrains .
10 : 5grains The following are the weights of air displaced by and respectively .
The density of the weights used ( w ) was , but their values were always corrected to the density of , or to the density of the brass representative of the Imperial ' Pound adopted in this country .
displaces displaces 1888 , Noy .
16 177.03 grains grains .
, , \ldquo ; , In normal air , displaces , and grains , whence the weigt , in grains , of the , linder was taken as follows:\mdash ; In vacuo .
In air .
grains grains .
, Reference Avoirdupois Standards ; having small in relation to the present Imperial Pound , which errors were allowed for .
THht MASS OH ' A CUBIC INCH OF DISTlLL ] As the weighings of the gravimeters were made against standard weights , which were parts and multiples of the Imperial Standard Pound , it necessary , in order to find the masses of the gl.avimeters , to reduce the weigbings to vacuum , as the Imperial Standard Pound itself has its true mass stated WEIGHING of in water .
Date , 1888 .
Nov. 20 . .
Date , 1888 .
Nov. 20 . .
Weight of counter- Mean .
Mcan .
B. .
56 .
The erences in the above weighings of on the several days were considered to be owing partly to thermometric variations , and partly to the presence of minute bubbles of air carried down by this large cylindrical body .
Results , however , could not be rejected merely because they did not always closely agree .
At Fahr. , to which temperature densities are reduced : or 1040169 .
The of the volume of in grains at Fahr. , may be essed i follows:\mdash ; vol. 5.1598148 or weight of wat , er at As the cubic contents of cubic inches , the mass of one cubic inch of distilled water at Fahr. , B. inches , has been taken as grains .
DIAMETER and Height of Q. diameter of was measured at 2.5 different points , each measurement being the mean of three readings .
The method of FIZEAU was followed , the true distance between the " " points\ldquo ; being ascertained after the diameter of had been measured . .
Jan. ( 21 ) ( 22 ) ( 2 : ; ) ( 24 ) 3.083546 3.083921 3.083924 3.083910 '3.084124 At the true distance been was found to be inches by the 3-inch steel measure ( 2999749 inch ) .
Hence at Fahr. the diameter of was taken at inches .
THE MASS OF A CUBJC INCH OF DISTILLED HEIGHT of as measured at 25 different points by Whitworth lSIicrometer .
61.7 For the standard 3-inch measure , the mean reading was found to be divisions , hence the height of was taken as inches .
By the above measurements of the height of it may be seen that the quartz gravimeter is not of true cylindrical form , and that its height could only be approximately ascertained .
After allowing for bevels on the upper and lower edges of the cylinder , as measured by the microscopic comparator , its cubic contents were taken as cubic inches .
WEIGHING of in Air .
The wing are the weights of air displaced by and 84924 respectively : \mdash ; Q. 7 .
7 .
2.2943 In normal air displaces .
and displaces , whence , in vacuo and in normal air respectively , the weight in grains of the sphere would vacuo .
In normal air .
WEIGHING of in Water .
bgrains at Fahr. , inches , vol. in graUns , or grains .
The density of the quartz cylinder being:\mdash ; At Therefore , as the cubic contents of are equal to cubic inches , the mass of one cubic inch at Fahr. ( B. 30 inches ) may .
by these experiments with , be taken as grains .
THE MASS OF A CUBIC INCH OF DISTILLED WATER .
MR. H. J. CHANEY'S RE-DETERMINATION OF After for the difference between the rates of expansion of the steel standard 6-inch and the brass sphere , and the error of the steel standard in relation to the yard , the diameter of the sphere is 5992444 inches .
On the sphere , however , two lines had been traced , for which a correction is to be made .
These lines triangular in section , nearly inch wide and inch deep , therefore the actual contents of would be at Fahr. cubic inches .
WEIGHING of Sphere in Air .
DeSCri trains 3 gilt l. Nos. ; , , 12 gilt l. Nos. ; .
gilt grains 300 , 100 , 10 .
, , 17 . . .
gilt lbs. ; gilt g , 100 , 10 .
After imIl .
, I8 . . .
gilt l. Nos. ; gilt grains 300 , 100 , 10 .
The following are the weights of air displaced by and respectively:\mdash ; displaces displaces grains .
35 .
, , 4.245 , , , , . . .
, , . . . . .
\ldquo ; In normal air displaces and grains respectively ; whence in , and in normal air , the weights in grains of the sphere are as follows:\mdash ; THE MASS OF A CUBIC INCH OF DISTILLED In vactlo .
In normal air 28 , .
28 , ains .
, , . . . . .
, , . . . . .
, , . . . . .
grains .
Or the logarithms of the weights in grains In vacuo .
In air .
4 .
4.4534759 WEIGHING of in Water .
B. 30.37 just floats in water and was suspended and kept in position by a " " sinker The weights of water displaced by the sinker , with and hout the gravimeter attached to the sinker , were ascertained by separate equipoises , the smull grain weights added each day representing the differences of such equipoises .
The density of the weights used was ; hence the density of is\mdash ; 0.999265 or A.s the weight of in air is grains , the volume of in grains at Fahr. will be grains .
As cubic contents of have been ta.ken as cubic inches , the mass of one cubic inch of water would be grains .
ON THL ' MASS A CUBIC INCH DISTILLED WATER .
Result .
The ements of the sphere appear to afford more accurate results than the measurements of the cylinders .
The measurements of the larger cylinder afford a more result than those of the cylinder ; although , on the other hand , there appears to be less uncertainty as to the volume of the solid quartz cylinder at the various temperatures .
If all the weighings and measurements of the three gravimeters had been of equal value , then a mean of the three results:\mdash ; grains . .
might have been taken as the final result of the present experiments ; but , for the cons..derations above stated , it was thought desirable to take as the final result , or grains grain , as the mass of the cubic inch of distilled water freed from air , weighed in air at the temperature of Fahr. , the barometer being reduced to 30 inches , against brass weights of the density of A cubic foot of such vater under the above conditions would weigh grains , or lbs. avoirdupois .

