Stubbs. — Couflricii tnty of Aqueous Soliifions of ('(trhon-dioxide . 17 



the vessel rose to 0-5° at times even when immersed in the ice ; but before 

 taking a final reading in any experiment the temperature was always kept' 

 (by constant stirring) exactly at 0° till the reading assumed a constant 

 value. The difference in solubility of CO, at the temperature slightly 

 above 0° would be small, and would make (especially at higher pressures) 

 a negligible difference in conductivity ; so that by the above precautions 

 it may be assumed that an accurate value of the conductivity was obtained. 

 For the experiments at low temperatures other than 0°, ranging from — 1^ to 

 9°, the observer's continual presence was required to keep the tempera- 

 ture of the bath constant. The thermometer used w^as graduated to tenths 

 of a degree, and the temperature was kept at the required point by the use 

 below 0° of a salt solution and ice, and above 0° by adding small pieces of 

 ice to the water, with constant stirring and w^atching the thermometer. In 

 every case a final reading was taken only after keeping the temperature 

 constant for a considerable time (at least thirty minutes), and making sure 

 that readings were no longer varying. 



• Stirring, etc. 



Owing to the form of the conductivity-vessel, the position of the elec- 

 trodes, and the high pressure employed, it w^as deemed inadvisable to 

 attempt to employ a stirrer in the solution ; nor could a thermometer be 

 introduced. The large mass of metal in the apparatus, however, insured 

 that the temperature inside was the same as that of the thermostat. The 

 absence of a stirrer was a drawback from a time point of view, as is was found 

 that saturation at any pressure of CO2, though aided by slight shaking of the 

 apparatus and by convection currents, took twenty-four hours to accomplish. 



Carbon-dioxide was obtained chemically pure from a bomb of liquid CO,. 



Pressure-gauges. 



Two pressure-gauges used to determine the pressure of CO 9. The dial 

 gauge was used for higher pressures ; it was a new one of 8 in. diameter, 

 and previous to use in the experiments was checked by a standard gauge. 



For pressures less than 5 atmospheres an open mercury gauge was used. 

 It could be connected at pleasure with the apparatus by means of a needle 

 valve, and by it pressures could be read off in millimetres of mercury with 

 great accuracy. By means of a screw below the needle valve the pressure 

 in the apparatus could at any time be relieved to the desired extent. 



Conductivity- WATER. 

 As the electrical conductivity even of fairly concentrated solutions of 

 CO2 is very small — according to Pfeiffer, " about one-twentieth of that of 

 spring w^ater under normal conditions"^ — it was necessary to obtain by dis- 

 tillation water of special purity. Walker and Cormack, whose work has 

 been referred to, obtained by three distillations a supply of water with 

 a conductivity at 18° of 0-7 x lO"'^ Siemens units, or about 0-75 x 10^" 

 reciprocal ohms. This appears to have been the best water used by any 

 of the workers who have been referred to. Walker and Cormack worked 

 with very dilute solutions where the conductivity was very small, and the 

 relative effect of impurity in the water consequently great ; it was there- 

 fore considered satisfactory for the purposes of the present experiments 

 w^hen a supply of water was obtained giving a conductivity of less than 

 1 X 10~^ ohms at 18° C. The supply was got by distilling in the open air 

 with a modification of the Bousfield still. The distillate condensed on hard 



