ON THE BOILING POINTS AND VAPOUK TENSION OF MEKCUBY, ETC. 317 



On the Boiling Points and Valour Tension of Mercury, of Sulphury 

 and of some Compounds of Carbon, determined by means of the 

 Hydrogen Thermo'ineter. By Professor J. M. Crafts. 



[A communication ordered by the General Committee to be printed in extenso among- 



the Reports.] 



A HYDROGEN thermometer, in preference to an air thermometer, was 

 adopted, because the more rapid flow of hydrogen through a capillary 

 tube permits the use of one of smaller dimensions than when air is taken. 

 A new form of constant-volume thermometer was used, in which an 

 electric contact between the mercury in the manometer and a platinum 

 point causes a current to excite a magnet, and close a cock to arrest 

 the flow of mercury into the manometer tube at the moment that the 

 gas attains a fixed volume, determined by the surface of the mercury 

 touching the platinum point. 



The advantages secured by the system are the reduction of the obser- 

 vations of the mercury level to one, instead of two ; the possibility of 

 diminishing the size of thermometer bulb from 200—500 cc. to 1-10 cc. j 

 and the rapidity of action due to the automatic movement of the 

 instrument. 



The calculation of results has been facilitated by the compilation of 

 an extended series of tables. 



The boiling point of mercury was redetermined with an apparatus- 

 similar to that used for taking the boiling point of water ; and the chief 

 causes of error, which Regnault noticed regarding his own experiments, 

 have been removed. It is proposed to use boiling mercury, like water, to 

 fix an additional fundamental point in thermometry. 



Mercury boils at 357-0 under a bai^ometric pressure of 760 millimetres. 



Sulphur boils at about one degree lower than the temperature given 

 by Regnault, and can be used to fix temperatures near 444°. 



Two other substances were found well adapted to obtain constant 

 temperatures near 200 and 800 Centigrade, and the means of purifying 

 them were carefully studied. Naphthaline, when pure enough to melt at 

 79'6-79-9, does not vary more than 0-1° from the boiling point 218'06, 

 bar. 760' mm. Benzophenone prepared by the action of oxychloride of 

 carbon upon pure benzine in the presence of chloride of aluminium, when, 

 pure enough to melt at 47°7-48°-0, boils at 306°-l, bar. 760 mm. 



The boiling points of these two substances were determined under 

 pressures varying from 87 to 2,300 mm., giving a series of temperatures 

 that can be easily established and maintained for any length of time. 

 The range of temperature is from 140° to 350°. 



It is very probable that benzine may also be easily obtained with a 

 perfectly fixed boiling point, and it is intended to study the properties of 

 benzine purified by fractional crystallisations on a large scale. When per- 

 fectly pure fragments of solidified benzine are allowed to melt with the 

 precautions usual in a zero-point determination, a delicate thermometer 

 only sinks from 5-17 to 5°-lG during the fusion. The product studied by 

 Regnault melted at 4'45, and was impure. 



The boiling points of the following carbon compounds are grouped to- 

 show that some of these experiments point to the conclusion that succes- 

 sive, similar, additions to the molecular weight do not cause the boilino- 



