diate addition of argentic nitrate, and sometimes I used a 
mixture of argentic nitrate and nitric acid as the solvent. 
The argentic chloride was dissolved in ammonic hydrate, 
filtered and thrown down by nitric acid and eventually 
weighed on a tared filter. The very highest result that 
I could find given for chlorine in the ash of this plant was 
a little over eight per cent., and the usual quantity given 
was about two per cent, or less ; and still finding from five 
to six and even eight per cent., although I knew that chlo¬ 
rine had been lost in preparing the ash, I concluded that I 
was in error and resolved to determine the chlorine in the 
plant or dried sample. To this end from ten to thirty 
grams was taken, from two and one-half to eight grams of 
puresodic carbonate was dissolved in water and made up 
to a volume sufficient to wet the sample thoroughly. The 
sample was then placed over a free flame and thoroughly 
charred, the mass being extracted with hot water; the fil¬ 
trate was usually slightly colored, especially in cases where 
a larger portion had been taken, but when the charring had 
been successfully done, the solution was colorless. & The 
carbon was washed free from chlorine and then burned 
until the ash was white. As the organic matter is an un¬ 
pleasant material to have in the solution, I evaporated the 
same to dryness and ignited it to complete the carboniza¬ 
tion. This was easily effected at a temperature which 
would produce no voladzationof the sodic chloride from the 
mixed salts. The solution was filtered off and added to the 
nitric acid solution of that portion of the ash insoluble in 
water. As a matter of course, care was taken to avoid 
loss from effervescence, escape of hydrochloric acid, etc. 
From this point on the two processes were the same, but 
the results were much higher than before. This induced 
me to treat the whole series in this manner. It has been 
stated that a loss of chlorine can scarcely be avoided in in¬ 
cinerating the plants ; but either the Joss has been consid¬ 
ered insignificant or the determination of chlorine has been 
deemed of so little importance to our study of the plant’s 
requirements that the determination of chlorine in the ash 
been accepted, perhaps with good reason, as quite 
sufficient for our purposes. Be this as it may, we have 
made the series of determinations in which our results ap¬ 
proximate closely to the chlorine in^the plants at the time 
they were cut ; and, while the loss is very varying, it is ob¬ 
servable in every case and shows that -the amount of care 
- bestowed upon the preparation of the ash was not sufficient 
to give us more than a relative idea of the amount of this 
element in the plant. The results are expressed in the per- 
