158 William J. Hodgetts 
(obtained at the time the monthly samples were collected) over a 
steam-bath and weighing the dish with its residue 1 . 
It will be seen on reference to the lowest curve of Fig. 3, where 
the results, in parts per 100,000 parts of water (in the sequel termed 
“ degrees ”) 2 , are plotted, that the total dissolved matter in the water 
varied during the 3J 3^ears between rather wide limits, namely 
between 8 and 30 degrees. A comparison of this concentration-curve 
with the curve showing the fluctuations in the water-level, and also 
with the rainfall data (all given in Fig. 3), reveals several interesting 
and somewhat unexpected results. In the first place the lowest 
concentration in both 1919 and 1920 was reached in the month of 
April, a result of the dilution of the pond by the winter and early 
spring rains. As one would rather expect, the two curves of Fig. 3 to a 
great extent van? inversely, the periods of low water-level generally 
corresponding to high concentrations, and vice versa , this being very 
obvious in the latter half of 1919; but this is not always the case, and 
in the autumn months especially the water-level is often rising at 
the same time as the concentration of the water is increasing. 
Thus from September 21st to October 19th, 1918, the water rose 
6*2 cm. but the dissolved matter in the water between the same 
dates increased from 12 to 16 degrees; and notwithstanding a further 
rise in the water-level of 3-8 cm. between October 19th and November 
16th the concentration had increased to 18 degrees by the latter 
date. Again from October 18th to November 22nd, 1919, the level of 
the water rose 8 cm., but the dissolved matter of the water also 
increased from 28 to 30 degrees. During the autumn of 1920 the 
phenomenon of the concentration increasing with a rising water-level 
was not shown so well, since the water-level was fairly constant 
between September and November, but even so the concentration 
showed a decided increase between October and November. 
It is obvious, therefore, that a correct idea of the variation of 
the concentration of the water cannot be obtained merely from an 
examination of the fluctuations of the water-level of a pond, and 
1 The same dish was used throughout, and it was kept specially for this 
purpose, never being used for any other work. After evaporating the water to 
dryness the dish with the residue was allowed to cool in a desiccator, and was 
then carefully weighed, using a chemical balance sensitive to a milligram. It 
was found that the weighing process had to be carried out rather quickly 
since the deposit was very hygroscopic and rapidly absorbed moisture from 
the air, an exposure of a minute or so to the air being sufficient to alter its 
weight appreciably. 
2 It is convenient to term one part of dissolved matter in 100,000 parts of 
water a concentration of one “degree.” The term thus falls in with the 
chemist’s “degree of hardness.” 
