ME. J. J. MANLEY ON THE APPARENT 
or;9 
^ tJ ^ 
were obtained between the hours of 10 a.m. and 8.30 p.m. 19 independent R.P. values. 
The weighings were continued during the three succeeding days. The results obtained 
during the four days over which this experiment extended are given in a very 
condensed form in the following table :— 
Table VII. 
Date. 
R.Ps. reduced for 16° C. 
Number of 
determinations. 
Maximum. 
Minimum. 
Differences. 
Means. 
October 11, 1911 . . 
110-0 
91-6 
18-4 
102-8 
19 
„ 12 „ . . 
109-7 
108-4 
1-3 
108-9 
45 
,, 13 . 
109-8 
109-1 
■7 
109-6 
18 
„ 14 „ . . 
112-2 
109-6 
2-6 
111-1 
12 
Tlie results obtained during October 11-14 inclusive are represented graphically and 
in a more detailed manner by Curve No. 2 (fig. 3). Curve No. 1 exhibits the 
small variations observed in the relative weights of X and Y before the contents of 
either were mixed, and at the same time indicates not only the degree of accuracy 
attained under the improved conditions introduced for weighing glass vessels, but also 
serves the useful purpose of a standard with which the other curves may be compared. 
The several curves are formed by plotting the differences in the apparent weights of 
X and Y against the times when those differences were measured. 
Curve No. 2 presents two interesting features, the most arresting of these being the 
well defined undulations in the first portion. The slope of the axis of the undulations, 
though at first steep, becomes with lapse of time, less pronounced ; finally the 
undulations die away and the further apparent-changes in weight are from that time 
onward best represented as an aj^proximately straight line curve having a direction 
which tends to become horizontal. The mean path of the whole graph suggested to 
our minds a very strong resemblance to a cooling curve ; we were thus led to suspect 
the existence of some simple relationship between the decreasing temperature of the 
recently mixed contents of X and the changes in the apparent weight of the same. 
We therefore decided to measure the rate of cooling for X and its contents under 
conditions similar to those which obtained during the weighings. With this object in 
view, another vessel, similar in all essential respects to X, was charged with solutions 
of silver nitrate and ferrous sulphate, the quantities used being the same as those 
employed in our Landolt experiment. A thermometer graduated to 0°‘l C. was 
introduced and read; the two solutions were then mixed, and the thermometer again 
read ; the increase in the temperature equalled 4°'8 C. and this was therefore the 
range of temj^erature for which the rate of cooling had to be determined. In carrying 
