1848.] 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



26r 



at 0", to determine immediately the weifjht of the litre of air. 

 Now, according to tlie principle upon which the French system of 

 measures was estahlished, the kilogramme is the weight of a litre 

 of distilled water, freed from air, at the temperature of its maxi- 

 mum density, which is about i° (39'2 Fahr.) ; it will suffice then to 

 determine the weight of water at 4° which fills the capacity which 

 the globe presents at 0°. 



To do this M. Regnault operated in the following way : — 

 The open globe was weighed upon a good balance ; its weight 

 was found to be 12S8-S5 gr., the surrounding temperatui-e being 

 4-2% and the height of the barometer reduced to 0", 757'89 mil. 



A small quantity of water was introduced into the globe, and 

 the globe exhausted by means of the air-pump, and at tlie same 

 time heated. In tliis way the atmospheric air was completely ex- 

 pelled by means of the vapour of water whicli was constantly 

 developed. The stop-cock of the globe was then closed. 



On the other hand, perfectly pure distilled water was boiled in a 

 large globe to free it completely from tlie air which it always holds 

 in solution at ordinary temperatures. Upon the tubulure of the 

 first globe was fixed by caoutchouc a glass tube, twice bent, one of 

 whose branches descended to the bottom of the vessel in which the 

 water was kept boiling. On opening the stop-cock of the globe, 

 the boiling water entered it slowly, without coming in contact with 

 air ; it was consequently perfectly free from that gas. 



The globe being completely filled, the recurved tube was removed, 

 and replaced by a tube having a bulb %vhich was kept filled with 

 the boiling water, and furnished the quantity of water necessary to 

 keep the globe filled as its temperature lowered. 



Wlien the globe, filled with water, had come down to the sur- 

 rounding temperature, it was placed in a zinc vessel, and com- 

 pletely surrounded with melting ice, care being taken to pack the 

 ice in proportion as it melted, upon the walls of the globe. 



The globe was left in the ice for a time varying from 6 to 18 

 hours ; the stop-cock was then closed, the bulbed tube detached, 

 and the tubulure above the stop-cock carefully wiped. 



The globe was placed in a large vessel filled with water at a 

 temperature a little above that of the chamber in which the balance 

 was ; it was left for two hours, so that it should take nearly the 

 temperature of the chamber. As the water contracts in proportion 

 as its temperature rises from 0° (to 4°), the globe could be kept 

 closed without danger of breaking. When the globe had acquired 

 the temperature of the chamber, it was weighed, and this weighing 

 (the temperature of the room and the height of the barometer 

 being noted) gives the means of calculating the weight of water at 

 4°, which fills the capacity which the globe presents at 0°. 



According to the experiments of M. Pierre {Aiinales de Chimie et 

 de Physique, Zd serie, tome xv., p. 348), if the density of water at 0° 



be taken as 1, at 4° it is . 



' 0-999881 



Whence we can calculate the weight of the water at 4° (its 

 maximum density), which fills the capacity which the globe pre- 

 sents at 0°. Three experiments give the following results : — 

 I. 9881-060 grammes. 

 II. 9881-113 „ 

 III. 9881-299 „ 

 The third weighing gave a number probably a little too high, 

 because the globe was intentionally left but a little time in the ice, 

 in order to see what influence this circumstance would have upon 

 the result. On this account, M. Regnault adopts the mean of the 

 former experiments, viz. : 9881-086. 



Desiring to ascertain whether the correction made to reduce the 

 weiglit of water from 0° to 4° was sufficiently exact, M. Regnault 

 made two direct experiments, which gave a mean differing only 



0-1 52 or ■ from the result of the calculation. The capacity 



of the globe at 0" was therefore 9-881086 lit., and since {see 

 Second Memoir) the weight of air which filled it at 0°, under a 

 pressure of 760 mil., was 12-7781 gr. ; the weight of the litre of air, 



under these normal circumstances is sr. =: 1-293187 er. 



9-881086 ^ ^ 



a value notably less than that which was heretofore admitted from 



the experiments of MM. Biot and Arago* (1-299541). 



* M. Regnault remarks, in a note, that all the numerical corrections made by MM 

 Blot and Arago, for the purpose of reducing the weight of air to 0^, and to absolute dry 

 ness, contributed to render the number which they adopted too high. Another circum- 

 ■ tance may ha?e produced a similar effect. These experimenters exhausted the globe 

 several times with a very good air-pump, and they supposed that the slight tension which 

 remained in their globe was produced by the vapour of water which the walls of the 

 globe abandoned in vacuo, which they re-condensed when the air entered again. It is, in 

 fact, probable that this.was the case ; but it seems to be also very probable that when the 

 globe was filled with sir yery nearly aaturated with moisture, it gave a new portion of 



From this and the numbers obtained in the preceding memoir 

 for the densities of the gases we deduce, that at Paris 



The litre of Atmospheric Air weighs 1-293187 grammes. 

 „ „ Nitrogen „ 1-256167 „ 



)) » Oxygen „ 1-429803 „ 



„ „ Hydrogen „ 0-089578 „ 



„ „ Carbonic Acid „ 1-977414 „ 

 Strictly considered, these values are only correct for the locality 

 in which the experiments were made— that is, for a latitude of 

 48° 50' 14", and a height of about 60 metres above the level of 

 the sea. 



M. Regnault finds the weight of the litre of air, under the 

 parallel of latitude 45°, and at the same distance from the centre 

 of the earth as that at wliich his experiments were tried, = 1-292697. 

 And assuming this as the standard number, he deduces for any 

 other latitude, and any other distance from the centre of the earth, 

 the formula 



w= 1-292697 gr. (1-00001885) 1 -f 2A (1 -0-002837, cos 2,\). 



R 

 Or, more simply 1 



w = 1-292673 gr. 1 + 2/i (1 —0-002837, cos 2\) ; in which w is 

 R 

 the weight of the litre of air (the litre is 61-09908 cubic inches) ; 

 R, the mean radius of tlie earth = 0,366,198 metres ; h, the height 

 of the place of observation above this mean radius, expressed in 

 metres ; and A, the latitude of the place. 



Applying this formula to the level of the sea, in the latitude of 

 Philadelphia (39° 56' 51-5"), and assuming the radius of the earth at 

 this point 6,367,653 metres : 



The weight of the litre of air will be 1-2914392 grammes. 

 And assuming the litre as 61-09908 cubic inches, and the gramme 

 as 15-433159 grains troy: 



The weiglit of a cubic inch of air wUl be 0-32621 grains troy. 

 Or, (assuming IMr. Hassler's determination of the weight of a 

 cubic inch of water, 252-6934 gr.) water is 774-63 times heavier 

 than air. 



Density of Mercury. 



The density of mercury has been determined several times by 

 M. Regnault, and with the greatest care ; as he wished to satisfy 

 himself whether tills liquid, purified by the means employed ordi- 

 narily in the laboratories, presented a constant density. 



A glass globe, of a capacity of from 250 to 300 cubic centimetres, 

 was filled with mercury. The globe terminated in a capillary tube 

 of about 2 mil. diameter, upon which a mark was made, and this 

 tube was surrounded by a larger one which was used as a funnel. 

 Tlie funnel could be hermetically closed by a ground glass stopper. 

 The globe being filled with mercury, this liquid was boiled, and 

 suffered to cool. The globe was then placed in ice for several 

 hours, and the level of the mercury brought exactly to the mark. 

 As soon as it was satisfactorily ascertained that the level of the 

 mercury did not change, tlie mercury was suffered to take the tem- 

 perature of the air, and its weiglit determined. The same globe 

 was then filled with distilled ivater, first boiled to deprive it of air. 

 It was suffered to cool, the funnel being kept full of boiled water, 

 and closed with its stopper. The globe was then surrounded with 

 ice, and when the water had taken exactly the temperature 0', the 

 water level was brought to the mark, and the sides of the funnej 

 wiped with filtering paper. The closed globe was then placed in 

 water having nearly the temperature of the surrounding air, so as 

 to bring it more quickly to the temperature of the air in which it 

 was to be weighetl. 



The three determinations of the density of mercury, which are 

 reported, were made at very different times, upon specimens from 

 different sources, and in three different globes : — 



I. The first specimen was mercury designed for the construction 

 of a standard barometer for the obserxatory of Paris. This mercury 

 came directly from the mine ; it had been twice distilled in an iron 

 vessel. It was then suffered to stand for several days under weak 

 nitric acid, to dissolve the oxide of mercury which always forms 

 during distillation. The metal was then washed with much water, 

 and dried in the air-pump. The density of the mercury at 0°, 

 compared with that of water at 4°, was 13-59599. 



II. In the second experiment, the mercury employed was that 

 used by M. Regnault, in the construction of his manometer. This 

 mercury was distilled several years ago, in an iron retort, and has 



water to the glass. This portion, which was not taken into account, was considered as 

 making a part of the weight of the air, and necessarily made that weight too great,— (See 

 Blot's " Traits de Physique," tome 1, p. 367.) 



S5» 



