274 
Proceedings of the Boyal Society of Edinburgh. [Sess. 
bridge was carefully calibrated beforehand against a standard box of coils, 
as also were the comparison resistances. The latter took the form of three 
coils of 100, 200, and 400 ohms resistance respectively. In every case three 
different combinations of these coils were made — by means of a mercury 
key — and the three measurements so obtained were never found to differ by 
more than a unit in the third significant figure. The connections between 
bridge, coils, and U tube were made with heavy copper wires whose 
resistances had been found to be negligible on measurement. To ensure 
correctness of temperature in the case of the solution, the final readings 
were not made until the U tube had stood for half an hour or so in the 
bath, and had thereafter given identical readings on two occasions 
separated by a considerable interval of time. The resistance constant of 
the cell was measured by means of a N/50 KC1 solution. The electrical 
measurements gave the following results : — 
Concentration of the 
Molecular 
Molybdate present in the 
Conductivity at 
Solution terms of N/3. 
25° C. 
0 
92-2 
100 
i 
i 
110 
117 
Ti 
¥ 
134 
These are plotted in fig. 1. 
3. j Rotation Measurements. 
The instrument used for the rotation measurements was a special form 
of spectro-polarimeter invented by myself, the cost of which was defrayed 
by a grant from the Carnegie Trustees. The solutions were contained in a 
jacketed glass tube 49‘86 cm. long, through the jacket of which a constant 
stream of water was passed, that had previously been made to traverse long 
coils of tubing immersed in a special thermostat bath. The temperature 
regulation of the tube was not quite so satisfactory as could have been 
wished. A difference of about 0 o, 4 C. existed between the two ends of the 
polarimeter tube, as measured by thermometers let down branch tubes 
into the solution. It was also found difficult to keep the mean tempera- 
ture exactly the same in successive experiments, as will be seen from the 
figures attached to each set of measurements. 
The dispersion prism of the instrument was calibrated by the employ- 
