Francis J. Lewis. 
256 
In order to show that an increase of sodium chloride takes 
place in the tissues of the leaf during immersion in sea-water it 
was first necessary to determine the average amount of sodium 
chloride in the leaves of experimental plants. For this purpose 
leaf sections were tested with silver nitrate in the following way. 
Sections were cut dry and mounted directly in a T x ff gram mole¬ 
cular solution of silver nitrate and exposed 1-ft. from an electric 
lamp for twenty minutes. Leaves tested in this way after immer¬ 
sion in sea-water (the surface of the leaf being washed repeatedly 
with distilled water before the sections were cut) always showed 
a much greater sodium choride reaction than fresh control leaves 
of the same plant. The reaction was strongest in the cells of the 
lower and upper epidermis and the palisade layer. In all these tests 
the upper mesophyll cells showed the least reaction. This method 
does not, however, yield quantitative results, and was therefore 
only used as a preliminary test. A series of trials were carried out 
on extracts obtained by grinding up the leaves in a mortar, and 
making up the extract after filtration to 100 cc. with distilled 
water and titrating. This method was found to he unsuitable, 
owing to the strong colouration of the extract with chlorophyll. 
The following method was then tried and found to give 
reliable results. 
The leaves in which the amount of sodium chloride had to 
be determined were washed with distilled water, until the wash- 
water gave no opalescence with silver nitrate. The leaves were then 
dried in an oven at 100°C for one-and-a-half hours and weighed. 
They were then incinerated in an open porcelain dish at a low 
temperature, no redness being visible in ordinary daylight. In¬ 
cineration was continued until all the carbon had been driven off, 
leaving a grey or white ash. According to Strecker 1 no chloride of 
sodium volatilizes during this method of incineration. The 
resulting ash was then dissolved into 100 cc. distilled water. In 
most cases the resulting solution was slightly alkaline, and was 
neutralized with a trace of pure nitric acid, phenolphthalein being 
used as an indicator. The solution was then titrated in the usual 
way with potassium chromate and a ^ normal silver nitrate 
solution. 
This method was used in determining the sodium chloride 
content of the untreated leaves in Table I and in the treated leaves 
in Tables II, III, IV, V and VI. The same individual plants were 
1 Strecker. Ann. d. Chan. u. Phavm. 54, 353. 
