( $95 ) 



the observation and which I could only remove partially by boiling 

 long with water and benzene, 



2. thai the substance be boiled under nitrogen, as pure benzidine 

 is soon contaminated by the oxygen from the air, 



3. thai superheating, which causes the liquid benzidine at times 

 to rise l 1 /,, decimeter in the jacket, be prevented by a good quantity 

 of glass-wool. 



When benzidine was used, a cooler was not wanted, the vapour 

 condensing in the nan-owed par! of the vapour-jacket. 



The benzidine being heated under nitrogen, 1 required a nitrogen 

 reservoir. For this purpose a IOO L. flask was used, which was 

 tilled with nitrogen obtained from a saturate solution of equal units 

 of' weighl of potassium nitrite and ammonium chloride. When the 

 reservoir had once been tilled, the nitrogen could be expelled from 

 the flask by the admission of water, and be conducted through ;i 

 drying apparatus to the vapour-jacket. The pressure in the reservoir 

 was measured by a manometer. 



The experiments proper began with the filling of the nitrogen 

 reservoir. The phenol and water were weighed in the required ratio, 

 and heated with a spirit lamp. The temperatures at which water 

 and phenol do not mix appeared then to be exceeded, ami a homo- 

 geneous liquid was formed. A thick-walled quartz tube closed on 

 one side, and drawn out capillarh on the other side in the voltaic 

 arc was now heated in the free flame, and the air expelled under 

 the phenol water solution: the liquid now rose in the tube which 

 was tilled for more than ' ,. and then fused together in the voltaic 

 arc. The quartz appeared then to be a very suitable material. It is 

 as clear and transparent a- glass, but, when hot. may be cooled in 

 cold water without bursting; it may be heated in a i'veo flame, 

 and is proof to high pressures, as the critical pressure of water, 

 viz. '200 aim. Protective measures proved to have been most likeh 

 superfluous. 



Thermometer and quartz tube were fastened by means of copper 

 wires in a hole of the glass tube which pierces the airtight rubber 

 stopper of the vapour- jacket. The apparatus were filled with nitrogen, 

 the pressure in the boiling vessel was reduced to about 30 cm., the 

 gas-burner lighted, and every five minutes a lamp was inserted of the 

 resistance, by the aid of which the current required to check radia- 

 tion was regulated. .After an hour the thermometer indicated about 

 350°. As soon as this temperature had become constant, which also 

 appeared from the fact that the boiling phenomena Stopped m the 



