ON THE ATOMIC WEIGHT OF CHLORINE. 189 



7. Method of Carrying Chit the Combustion. 



The weighings of the palladium bulb and the chlorine condensation bulb completed, 

 the next step was to set up the combustion apparatus (fig. 5). This consisted of a 

 stout glass globe A, the " combustion globe " made of Jena glass. Its capacity was 

 about 750 cub. centims., and it was provided with three ground-glass tubulures. In 

 order to ignite the hydrogen at the jet, two platinum-iridium wires* (totally enclosed, 

 save for their extreme tips, in glass covers) were fused into the combustion globe on 

 each side of the hydrogen tubulure. By the passage of electric sparks between their 

 tips, the jet of hydrogen was easily ignited. 



Into the combustion globe was run sufficient water to absorb all the hydrochloric 

 acid gas formed during the combustion, and to leave dilute acid of a not greater 

 strength than one-seventh concentrated. Then two very thin glass bulbs (capacity of 

 each about 6 cub. centims.), which had been previously filled with a hot, concentrated 

 solution of potassium iodide and sealed, were cautiously slid into this water through 

 one of the tubulures. The palladium bulb B, the chlorine condensation bulb 0, and 

 the three-way tap D were then, respectively, fitted to the tubulures E, F, and G, care 

 being taken that none of the lubricant (phosphoric acid) was squeezed into the 

 combustion globe through -the interstices of the ground-glass joints. To one limb of 

 the three-way tap D, a generator of carbonic acid in an atmosphere of which the 

 subsequent titration of residual iodine was carried out, was attached by a short length 

 of thick-walled indiarubber tubing ; to the third limb was fused the apparatus H, 

 through which any residual gases from the combustion were drawn. It consisted of a 

 wash-bottle which could be taken to pieces by means of the ground-glass joint J. 

 The tap K controlled the passage of the gases through the liquid, an alkaline solution 

 of sodium thiosulphate, contained in the wash bottle. 



The tube L was attached to the mercury pump by a short piece of thick-walled 

 indiarubber tubing. 



These two short lengths of indiarubber tubing were employed so as to enable us to 

 give a jerking motion to the combustion globe and the bulbs when fitted together : 

 (i.) to break the drawn-out cul-de-sac of the chlorine bulb, and (ii.) to break the 

 potassium iodide bulbs after the combustion. The only danger arose from a possible 

 in-leakage of air through the tube connecting the wash-bottle with the pump, by 

 which the residual oxygen, nitrogen, and hydrogen were withdrawn from the globe. 

 This tube was wired on to the glass when hot, and was well " drowned" before being 

 used to evacuate the globe. We found that no readable volume of air had leaked 

 through into the highest vacuum attainable during three days. 



The different parts of the combustion apparatus having been fitted together, the 

 strength of a neutral solution of sodium thiosulphate was determined by titrating a 



* The position of these wires is shown by the dotted lines P and PI ; they lie in a plane at right angles 

 to the vertical section shown in fig. 5. 



