Oct. 1 8, 1 883 J 



NATURE 



593 



balance serving to determine densities ; and here, again, 

 details of operations have been perfected up to the last 

 limits compatible with the actual state of science. The 

 water which is to be used for hydrostatic weighings, 

 having been once distilled by an ordinary still, is re- 

 distilled by means of an apparatus of platinum, and 

 then re-collected in a platinum vase. The latter is placed 

 under the balance employed for the weighings, and by a 

 series of ingenious apparatus the weights can be plunged 

 in the water and all the manipulations performed, — 

 manipulations often very delicate, but indispensable if all 

 possible chances of error are to be reduced to a minimum. 

 The weighing section contains, besides, a beautiful 

 collection of weights, — in platinum, iridium, and in quartz 

 for weights of the first class, and in gilt brass for weights 

 of the second class. 



Irrespective of the fundamental apparatus we have 

 just mentioned, the bureau possesses a large number of 

 different instruments ; some intended for certain special 

 labours, others necessary for accessory processes in close 

 connection with operations of comparisons or of weighings. 

 Among the former, one of the most remarkable, by reason 

 of the exceeding delicacy of the method it sets in opera- 

 tion, is the Fizeau apparatus, bymeans of which expansions 

 in small standards, or fragments of some millimetres in 

 thickness, are measured by applying an optical process 

 founded on observation of the phenomenon of the inter- 

 ference of light. This apparatus enables variations of 

 distance between two points to be determined and mea- 

 sured down to some millionths of a millimetre. 



The accessory instruments are cathetometers, sphero- 

 meters, meteorological instruments, barometers, thermo- 

 meters, hygrometers, &c. Fig. 4, for example, shows the 

 normal barometer of the section of weighings, a splendid 

 instrument combining all the most perfect contrivances 

 for the measurement of atmospheric pressures with the 

 utmost possible degree of accuracy. 



The measurement of temperature plays an essential 

 part in all operations which have to be performed with 

 standards either of length or of weight ; the studies in 

 connection with the thermometer have likewise a place so 

 important that they may be regarded as constituting a 

 section by themselves, with instruments peculiarly their 

 own. 



The measurement of temperatures does not form a 

 separate section in the International Bureau, but the 

 importance of the operations connected with it and the 

 precision of the apparatus employed for this end entitle 

 it to a special description. 



The air thermometer depends on a remarkable principle, 

 the knowledge of which science owes to the classical ex- 

 periments of Regnault on the expansion of gas. The 

 illustrious physicist has in effect demonstrated that the 

 increase of tension which gas suffers when heated while 

 its volume is kept constant is sensibly proportional to the 

 temperature. 



It is easy to conceive that Regnault utilised this property 

 for the measurement of temperatures. The first air ther- 

 mometer of precision constructed by him consisted essen- 

 tially of a glass globe filled with air, connected by a 

 capillary tube with one of the arms of a mercurial mano- 

 meter. Special contrivances allowed the mercury to be 

 constantly maintained at the same height in this arm, 

 while the globe was exposed to different temperatures. 

 The tension of air at each temperature is then measured 

 by the height of the mercurial column balancing it. 



This instrument, generally employed by experts, has 

 subsequently undergone numerous modifications, the 

 most of which aim at imparting greater precision to the 

 measurement of pressures. This measurement, which 

 consists in determining the difference of the height of 

 the mercurial level in the two arms of the manometer, 

 really presents great difficulties, seeing that tubes of large 

 dimensions are employed for the manometer, a condition 



necessary to avoid the capillary depression of the mercury. 

 The surface of the mercury in the large tubes presents a 

 plane superficies, so smooth that it is impossible to dis- 

 tinguish the level of the mercury when it is viewed hori- 

 zontally. To attain this end, one makes use of movable 

 points, bringing them gradually close to the surface. 

 Observing, then, in a telescope the level of the mercury, 

 the point is seen drawing nearer and nearer its image ; 

 the instant of contact between the movable point and its 

 image indicating exactly the level of the mercury. 



The apparatus represented in Fig. 4 is the one which 

 has been constructed at the observatory of the International 

 Bureau. The arrangement we have just spoken of has 

 been adopted as much for the sake of the readings of the 

 manometer as of the measurement of the atmospheric 

 pressure in the normal barometer, e b. In the part to the 

 left of the diagram, one distinguishes in a the arm of 

 the manometer which is connected by a capillary tube with 

 the globe placed in the interior of the warming apparatus. 

 The latter has been placed in an adjoining position, in 

 order not to expose the measuring apparatus to variations 

 of temperature. 



Having determined the normal positions of the instru- 

 ments, one can proceed to the comparison of the mercurial 

 thermometers, / /, the reservoirs of which are placed in 

 the warming apparatus in proximity to the globe, A, of the 

 air thermometer. The observations consist in reading on 

 one side the temperature indicated by the mercurial ther- 

 mometers, and on the other in measuring the difference 

 of the mercurial level in the two arms of the mano- 

 meter; the open arm of the latter experiences the atmo- 

 spheric pressure, the manometer merely indicating the 

 difference between the atmospheric pressure and the ten- 

 sion of air inclosed in the globe. To get the total 

 pressure balancing the tension of the air in the globe, 

 the barometric pressure must be added to that indicated 

 by the manometer. The measurement of the pressures is 

 effected by means of three horizontal telescopes, movable 

 vertically through the length of the upright fixed to the 

 pillar shown on the left side of the diagram. This upright 

 is able to turn on its axis. Having adjusted the level of the 

 mercury, one can turn the telescopes, without deranging 

 them, round this vertical axis in such a manner as to 

 read the graduations on the scale attached to the adjacent 

 manometer. The telescopes fixed on the upright form to 

 some extent, a pair of beam compasses, as they allow the 

 difference of the mercurial level in the arms of the 

 manometer to be transferred to the scale serving to 

 measure it. 



These instruments are easily adequate to the measure- 

 ment of a hundredth of a millimetre. In the ordinary 

 conditions of experiments three hundredths of a milli- 

 metre correspond with a variation of temperature of one 

 hundredth of a degree Centigrade. To maintain the 

 temperature constant within these limits, the steam of 

 different liquids, such as water, ether, methylic alcohol, 

 ordinary alcohol, the ebullition of which takes place at 

 continuous temperatures, is used with advantage. The 

 regular ebullition of these liquids being one of the con- 

 ditions essential to the constancy of the temperature, the 

 arrangement indicated in the diagram has been decided 

 on. A vessel, a, placed in a water-bath, c, contains the 

 liquid, the steam of which escaping by the tube x x, pene- 

 trates into the double cased heating apparatus. Having 

 traversed all the parts of the apparatus, including the 

 glass tubes in which the mercurial thermometers 1 1 are 

 placed, the steam issues by the tubes y y to liquefy in the 

 condenser R, whence the liquid returns to the vessel by 

 the tubes a. Other tubes serve equally to return to the 

 vessel so much of the steam as becomes condensed on 

 the way or in the apparatus, so that the same quantity of 

 liquid will serve for a sufficiently long time. The water- 

 bath c, which supplies the heat necessary for keeping up 

 the boiling of the liquid employed, is itself heated by 



