582 Mr Lcimplough, On the determination of the rate of 
very simple, for it will be shown later that no appreciable period 
of induction is observed if the liquid is vigorously agitated, whilst, 
if the liquid is not stirred, supersaturation of the gas takes place, 
in some cases to the extent of 100 times the normal saturation 
amount. The time taken to effect this supersaturation depends on 
the rate of the reaction and the nature of the gas. If however 
the solutions are vigorously stirred, supersaturation of the gas is 
avoided, and the rate of evolution of gas becomes a reliable method 
for the investigation of the course of the reaction. 
The starting-point of this work was an attempt to elucidate 
the reaction between nickel carbonyl and iodine by measuring 
the rate of evolution of the carbon monoxide. During this in- 
vestigation it was noticed that when the reacting liquid was 
allowed to remain at rest, a large amount of gas became stored 
up in the liquid in a state of supersaturation, a considerable 
volume of gas being expelled immediately the liquid was agitated. 
Hence the measurement of the rate of evolution of gas from the 
liquid at rest could by no means be taken as the measure of the 
velocity of the reaction. Experiments were therefore undertaken 
to determine whether the rate of evolution of gas from a solution 
thoroughly agitated by stirring was a true measure of the rate of 
reaction. For this purpose the decomposition of hydrogen per- 
oxide in the presence of colloidal platinum was studied, because 
in this case the progress of the reaction could also be followed by 
means of titration. 
Experimental method. 
The apparatus used was the result of experience gained by 
many trials, and not a few failures. Fig. 1 represents the vessel 
in which the reaction was carried out. It consists essentially of a 
flask R, to contain the reacting liquid, fitted with a delivery tube D, 
connected with the burette in which the evolved gas was measured. 
Violent agitation of the liquid was produced by means of a platinum 
paddle P, rotated by the attached pulley A. (The bottom of the 
vessel was closed after the paddle had been inserted.) The paddle 
was generally rotated at 800 to 1200 revs, per min.; a higher 
velocity was used in the case of the decomposition of formic acid 
in sulphuric acid, the gas produced being held more tenaciously 
by this solvent. It was found that the temperature of 50 c.c. of 
alcohol was not raised as much as T° by vigorous stirring for 
5 minutes. The shaft B carrying the paddle rotated freely in a 
tube T ; a rubber collar G was stretched over this tube, but was 
allowed to project a few millimetres so that the shaft B rotated 
stiffly in the collar. The rubber was lubricated with vaseline or 
glycerine and water, and formed an excellent airtight joint between 
