532 
MK. T. J. BAKEK ON THE TlIEBMO-CIIEMlSTEY 
Since this, in itself, mnst introduce a difference in the heat of dissolution according 
as the metals are mixed or alloyed, the use of nitric acid is unsuited to the purpose 
in view. 
Dr. Galt’s'’^ experiments, mentioned on the first page, were made with this solvent, 
and the conclusions drawn from them are, therefore, open to grave douht. 
After very numerous experiments, two solvents were found which gave very satis¬ 
factory results and were not open to the ohjections urged against nitric acid. They 
are, respectively, mixtures of solutions of ferric chloride and ammonium chloride, and 
of cupric chloride and ammonium chloride. 
Both attack the metals and their alloys rapidly, and no gas is evolved. The 
reactions involved are simple and give rise to a moderate evolution of heat only. 
The latter point is one of considerable importance, because the rather small differ¬ 
ences between the heats of dissolution of the metals when merely mixed and when 
alloyed thereby assume greater relative values, and are consequently less affected by 
experimental errors. 
This matter is further referred to on p. 544. 
A2yparatiis. 
The calorimeter consisted of a thin glass beaker of rather more than 500 cub. 
centims. cajDacity. 
This was supported within a highly-polished nickeled vessel by a stout rubber ring 
of square section, which encircled the beaker just below its rim and rested on a gallery 
soldered inside the metal vessel. 
The latter was screened from external fluctuations of temperature by a surround¬ 
ing copper water-jacket furnished with a stirrer, the air-space between the nickeled 
vessel and the water-jacket being covered in with a copper lid. 
The beaker itself was covered with a thin ebonite slieet pierced with holes for the 
stirrer and thermometer. 
The stirrer consisted of a thin sheet of mica perforated with several holes and 
carried by a thin glass tube, which could be actuated from a distance by a cord pass¬ 
ing over a pulley. 
The thermometer, by Geissler, of Bonn, was divided into ’01° C., and it was easy to 
read to '001° C. 
The position of the mercury thread was so adjusted that at the beginning of an 
experiment it always stood as nearly as possible at the same division, and since the 
)-ise in temperature during an experiment was usually about 1° C., the same portion 
of the scale was always used. 
This 1'^ interval was found sensibly correct when compared with a standarilised 
thermometer. 
* ‘ Eep. Brit. Assoc.,’ 1898 and 1899. 
