ATMOSPHERIC AIR AS THE FOOD OF PLANTS. 79 
glass, porcelain, silver, etc.,) so that water would evapo- 
rate rapidly from its surface. The purest water was then 
dropped into the warm dish in small quantities at a time, 
each portion being allowed to evaporate away before the 
next was added. Over the vapor thus generated was held 
the mouth of a cold bottle until a portion of the vapor 
was condensed in the latter. 
The water thus collected gave the reactions for nitrous 
acid and ammonia, sometimes quite intensely, again faint- 
ly, and sometimes not at all. 
By simply exposing a piece of filter-paper for a suffi- 
cient time to the vapors arising from pure water heated 
to boiling, and pouring a few drops of acidified iodide-of- 
potassium-starch-paste upon it, the reaction of nitrous acid 
was obtained. When paper which had been impregnated 
with dilute solution of pure potash was hung in the va- 
pors that arose from water heated in an open dish to 100° 
F., it shortly acquired so much nitrite of potash as to re- 
act with the above named test. 
Lastly, nitrous acid and ammonia appeared when a 
sheet of filter-paper, or a piece of linen cloth, which had 
been moistened with the purest water, was allowed to dry 
at ordinary temperatures, in the open air or in a closed 
vessel. (Jour. fir Prakt. Chem., \xvi, 131.) These ex- 
periments of Schénbein are open to criticism, and do not 
furnish perfectly satisfactory evidence that nitrous acid 
and ammonia are generated under the circumstances men- 
tioned. Bohlig has objected that these bodies might be 
gathered from the atmosphere, where they certainly exist, 
though in extremely minute quantity. 
Zabelin, in the paper before referred to (Ann. Ch. Phy 
Cxxx, p. 76), commtnicates some experimental results 
which, in the writer’s opinion, serve to clear up the mat- 
ter satisfactorily. j 
Zabelin ascertained in the first place that the atmos- 
pheric air contained too little ammonia to influence Ness- 
