314 BULLETIN OF THE BUSSEY INSTITUTION. 
Note on the use of Rain- Water for experiments in growing plants. 
In the beginning, that is to say before any of the experiments 
above recorded had been made, I had many doubts as to the prac- 
ticability of using rain-water for the proposed assay; but these 
doubts were quickly dissipated by the results of the experiments. 
The results of every trial have shown most conclusively that the 
rain-water was extremely well-fitted for the use to which it was 
put, and I believe that in general pure rain-water, such as I used 
(compare page 295), will be found better than ordinary distilled 
water for experiments like the foregoing. ‘The constancy of the 
composition of rain-water, as taken from the great cistern which 
supplied me, is one very important consideration, since it relieves 
the experimenter from the risk of variations such as might readily 
arise through accidents, mistakes, or carelessness in the prepara- 
tion of distilled water during the long terms of the experiments. 
Boussingault showed long ago that in preparing distilled water in 
the ordinary way we concentrate in the distillate the ammonia 
which was previously disseminated throughout the entire bulk of 
water taken for the operation, whence it may readily happen, in 
case the amount of distillate prepared is small, that the ‘‘ puri- 
fied” water will be contaminated with a hurtful amount of ammo- 
nia. For the sake of the comparison, I have myself sometimes 
grown plants in several pairs of pots of sand, with and without 
the addition of inorganic chemicals, 7. e. ash ingredients, and have 
watered one set of the pots with rain-water and the other with 
distilled water, and have found that as a rule it was from the dis- 
tilled water rather than from the rain-water that the plants tended 
to get some traces of nitrogenous food. When the distilled water 
was carefully prepared, with the express view of excluding ammo- 
nia, it yielded to the plants no more nitrogen than the rain-water, 
that is to say it was as good as rain-water for my purposes; but 
unless these pains had been taken to exclude ammonia, the dis- 
tilled water, though such as would be used without hesitancy in 
the laboratory for most purposes, was no better than the rain- 
water but on the whole rather worse. Those chemists who are 
most familiar with the preparation of water free from ammonia, 
such as I have described in the ‘‘ American Journal of Science,” 
1876, (3.) 12. 182 note, will most keenly appreciate the advantage 
of being able to use rain-water in experiments with plants. 
In order to gain some idea of the character of the rain-water 
used in these experiments, I have had it tested from time to time 
with Nessler’s reagent, with the result that the water as taken 
from the cistern contains from 0.00048 to 0.00053 grm. of free 
ammonia (N H,) to the litre, and 0.00006 to 0.00008 grm. to the 
litre of the albuminoid ammonia of Wanklyn. Most commonly 
the tests gave 0.0005 grm. of free ammonia. No differences in 
the amount of ammonia were noticed on testing the cistern-water 
during or after heavy rains. But it appeared that the plants 
really received less ammonia from the rain-water than the above 
