596 Mr Lamplougli, On the determination of the rate of 
A was thus always in contact with pure carbon monoxide. By 
this method the solution nickel carbonyl could be kept for a 
month or more without decomposition. Two similar pieces of 
apparatus were used in which to store the solution of iodine 
and the pure solvent (alcohol) so that all the liquids used should 
be saturated with carbon monoxide, the gas given off through the 
reaction. 
In order to compare accurately the strengths of the solutions 
of nickel carbonyl and iodine, the solutions were titrated together, 
the tint imparted to the solution by any free iodine present after 
long standing being used to indicate the presence of excess of that 
constituent. 
In order to determine whether the products formed during the 
reaction in any way influenced the rate of the reaction, an ex- 
periment was performed in which solutions of nickel carbonyl 
and iodine of strengths in the ratio of 1 : T33 were used. 25 c.c. 
of the former, and 6 c.c. of the latter were made up to 110 c.c. 
and the rate of reaction measured. At the end of the reaction, 
14 c.c. of liquid were abstracted, 8 c.c. of nickel carbonyl and 
6 c.c. of iodine solution being added. The concentration of the 
reacting substances was therefore the same, but the solution 
contained in addition the products of the previous reaction. On 
measurement of the rate of reaction, however, it was found that 
there was no difference between the velocities of reaction in the 
two cases. 
To determine the part which each constituent plays in the 
reaction, experiments were performed in which there was a large 
excess, either of one constituent or the other. 
The rate of reaction was so great, especially when nickel 
carbonyl was present in excess, that it was found necessary to 
reduce the temperature to 0° before the velocity could be con- 
veniently measured. To maintain this temperature without 
variation proved to be a difficult problem. After many attempts 
a fair measure of success was obtained. The bath containing a 
mixture of ice and water was well stirred, and by means of a 
spiral circulator revolving in a cylinder, the water was made to 
flow through the jacket of the burette and back again into the bath, 
so that both the reaction vessel and the burette were maintained 
at 0°. 
The experimental results however did not appear to be capable 
of any simple interpretation. Straightforward results were not 
obtained when either constituent was in excess. The most remark- 
able observations were those made when the nickel carbonyl was in 
excess. As illustrated by the curves in Fig. 6 the action proceeded 
at a fairly uniform rate till a point was reached at which the 
reaction came to an almost abrupt stop, the colour of the solution 
