TEMPERATURE OF THE RATE OF A CHEMICAL CHANGE. 
189 
by-paths have to be left unexplored. For the purpose of these observations it was 
only necessary to make sure that the dark standard, made in the manner described, 
would not change appreciably, the tube being corked and kept in the dark except 
when in use, by exposure to diffused daylight. A set of observations only occupied 
five or six hours, and the same standard was in use for several weeks without any 
noticeable change. 
The pale standard made by diluting some of the dark standard was less constant, 
and differed in colour from the liquid in the reagent tube which had faded to a similar 
depth of colour. This difference of colour made the observations less easy, and therefore 
less exact, than they would otherwise have been. The cause was found to be that the 
constancy and precise tint of ferric sulphocyanide, which in dilute solutions varies 
between pink and orange, depends not only upon the relative proportions of iron, 
sulphocyanide, and acid, which dilution leaves unaltered, but also upon the concentra¬ 
tion of the latter two. In the reagent tube the concentration of these two influential 
bystanders is practically if not wholly unchanged. 
By several preliminary testings a degree of concentration of hydrogen chloride was 
found which would prevent the formation of a cloudiness by hydrolysis of stannic 
chloride without increasing unduly the rate of change. The margin is rather 
narrow. 
To maintain a constant temperature the reagent tube was placed on an iron tray 
covered by white paper, which was heated beneath at one end by a small gas flame, 
while the liquid was stirred by the passage of large bubbles of carbonic acid from an 
inverted thistle-funnel. By moving the tube nearer to or further from the heated end 
of the tray, or by moving the gas burner, the mean temperature of the liquid was 
regulated during intervals varying from five minutes to three-quarters of an hour 
within about 0°'05 C. Four or five bubbles a minute stirred the liquid sufficiently 
when the desired temperature was near that of the room, but for temperatures above 
20° C., and during the delivery of stannous chloride from a pipette, a more rapid 
stream was used. 
The colour on mixing was very deep, the excess of ferric chloride being such as to 
allow at least two or three minutes for adjustment of temperature before the depth 
of colour approached that of the dark standard. To make the comparison, the cork 
holding the thermometer and thistle-funnel was replaced by a plain cork, and the 
dark standard was placed close alongside the reagent tube. The observer looked at 
the two, watch in hand, closing his eyes at intervals for a few seconds to get a fresh 
impression, till the colours seemed undistinguishable, made a mental note of the time, 
and continued to watch till the contents of the reagent tube seemed the paler of the 
two. Then thermometer and thistle-funnel were quickly replaced, and any disturbance 
of temperature estimated as having prevailed after the time of agreement, which is the 
start of the race, was compensated by causing a similar divergence on the opposite- 
side of the thermometer line. 
