Stubbs. — Coiid iirtivit ij of Ar/ueniis Soliitwns of Carbon -dioxide 15 



another the pressure was given as 16 atmospheres ; even the composition was 

 in doubt, being given as having 6, 8, or 9 molecules of water of hydration). 

 It was anticipated that the formation of a hydrate should cause a very notice- 

 able variation in the conductivity, and this anticipation was fully realized. 



The work to be attempted consisted in — 



(1.) Investigating the conductivities, at 18° C, of saturated solutions of 

 CO, in water, under pressures ranging from 1 to 30 atmospheres. 



(2.) The same, if possible, at 0° — i.e., if formation of a solid hydrate did 

 not interfere with the obtaining of a complete result. 



(3.) Investigating, by means of the conductivity, the formation and proper- 

 ties of a CO2 hydrate at various low temperatures and at various pressures. 



(4.) Determining the variation in conductivity of COg solutions due to 

 a variation in temperature. (This could only be done to a limited extent 

 in the available time.) 



(5.) One or two other points of interest in connection with CO, solutions 

 came up during the course of work, and will be referred to. 



Apparatus and Method. 

 The ordinary Kohlrausch method of determining the electrical con- 

 ductivity of solutions was adopted, using induction coil and telephone. 

 Bridge-wire and known resistances were calibrated. 



Conductivity-vessel. 



A special form of closed conductivity-vessel had to be used in order 

 to stand the high pressures of the experiment. It was made of gun-metal, 

 as it was not considered that glass would stand the pressure used. The 

 only disadvantage attending this was that in the later stages of the experi- 

 ment the interesting changes taking place within the vessel were not visible 

 to the eye. The diagram on next page shows the details of construction. 



The casting consisted of two parts — the vessel B, and the head A, which 

 screwed on to B by the thread C. At D, fitting on to B was a broad lead 

 washer, by screwing the head tight down jon to which the apparatus was 

 rendered absolutely gastight, even at the highest pressure. The head was 

 bored with two slightly coned holes E to receive the electrodes F, G, H, L, 

 the tops of which were coned at the same angle. Over the coned tops F 

 of the electrodes pieces of the best rubber tubing were tightly stretched, 

 the electrodes being then pulled home into their sockets ; this both made 

 the electrodes perfectly steady and gastight, and also effectively insulated 

 them. The electrodes, which were stout brass rods, tapered off from the 

 shoulders J, and terminated in the platinum electrodes L. The exposed 

 brass parts were encased in hard Jena-glass tubes which fitted tightly on 

 to the shoulder at J. Into the ends H were soldered with silver-solder stout 

 platinum wires, shielded by Jena-glass tubes sealed at K, and terminating in 

 the electrodes L. The latter were made of heavy platinum sheet, measured 

 each 20 mm. by 22 mm., and were placed from 8 mm. to 10 mm. apart, being 

 closer at the bottom than at the top. They could be adjusted to a suitable 

 distance by bending the platinum wires K, L. The ends of the electrodes 

 protruding from the head were threaded to receive the nuts N and P. 

 M was a thick fibre washer, acting as an insulator ; N a nut which was screwed 

 down tight, thus pulling the electrodes fast into their sockets. P was 

 another nut, which with N formed a binding screw for the lead from the 

 electrode. The whole arrangement of electrodes was found to be quite 

 steady and satisfactory from a conductivity point of view. 



