524 PROCEEDINGS OF THE AMERICAN ACADEMY. 



The heating coil had a resistance of about 0. 7 ohm, and was supplied 

 ■with a suitably controlled current of from twelve to eighteen amperes 

 from eight large storage cells. The ends of the coil were sealed into 

 the ends of small glass tubes within which were stout copper wires, 

 contact being made by mercury. It is necessary that the copper wires 

 be heavy so that they may not become heated, and thus superheat the 

 vapor coming into contact with the glass tubes encasing them. 



The temperatures at which distillations took place were read from 

 small standardized Anschiitz thermometers, whose mercury threads were 

 entirely within the vapor, so that no correction for projecting mercury 

 thread was necessary. The bulb of the thermometer was placed oppo- 

 site the entrance to the hood, in order to measure the temperature of 

 the vapor actually admitted — a point especially emphasized by Lou- 

 guinine, who nevertheless merely inferred the temperature from the 

 barometric pressure and the coefficients expressing the dependence of 

 boiling point on pressure. 



Another point, usually neglected, is worthy of brief notice. Before 

 the liquid comes to the boiling point, an appreciable quantity of vapor 

 may pass over and be condensed. Obviously this may introduce error, 

 since the assumption is made, in calculating the result, that all the 

 vapor was at the boiling point of the liquid. Moreover, the heat of 

 vaporization is different at different temperatures. In the present 

 experiments this cause of error was eliminated by providing the vapor- 

 izer with an outlet and stopcock above, and by passing a very slow 

 current of dried air backwards through the condenser coil and vapor- 

 izer until the liquid was boiling at a lively rate, and the temperature 

 of the whole interior was quite at the boiling point of the liquid. This 

 air prevented the vapor from passing into the condenser, so that no 

 premature condensation was possible ; its initial temperature was kept 

 close to that of the calorimeter. 



Preliminary experiments were made to find whether or not the 

 proximity of the hot coil of platinum wire might superheat the vapor 

 passing in its downward course to the condenser. The wire was first 

 made into a coil about 3 cm. in diameter, and a thermometer sus- 

 pended therein, the bulb of the thermometer being at the center. The 

 coil and thermometer bulb were then immersed in distilled water in a 

 large open test tube, the current was connected, and the temperature 

 noted at which boiling occurred. The coil was then made about 2 cm. 

 in diameter and the process repeated ; finally the wire was coiled as 

 tightly as possible around the thermometer bulb, without actual con- 

 tact, and the temperature at which boiling took place was noted as be- 

 fore. It was found that there was no danger of superheating from this 



