68 BULLETIN OF THE BUSSEY INSTITUTION. 
colleague, Professor Farlow, pronounced to be a species of Pziza, 
appeared upon the surface of the soil in every one of the jars that 
contained roasted sole leather; while the most careful scrutiny failed 
to detect any trace of the fungus upon either of the remaining twenty- 
five jars in the series. It so happened that the position of these 
roasted-leather jars was at the very middle of the collection of jars 
charged with loam and sand. The loam had been calcined and the 
leather roasted just before they were put in the jars, and neither the 
jars nor the materials had been exposed to any other air than that of 
the glass-house. There was no escape from the conviction that the 
fungus had found something to feed upon in the roasted leather 
which was absent from all the other jars. But this edible substance 
could have been no other than some form of nitrogen, since the loam 
contained all the other elements of plant food. It may here be men- 
tioned that only negative results were obtained on testing some of the 
freshly roasted leather for ammonia, and for nitrites and nitrates. 20 
grms. of the roasted leather were ground to powder in a drug mill, 
and then boiled in half a litre of pure water, the first 50 cc. of dis- 
tillate from which were tested with Nessler’s reagent. A slight 
greenish yellow coloration was obtained, together with milkimess, but 
no trace of the color that is produced by ammonia. 200 ec. of the 
dark-colored liquor in the flask was decanted and boiled with metallic 
cadmium in an atmosphere of hydrogen, and, after the removal of the 
cadmium, distilled with acetic acid, and the distillate tested for nitrous 
acid, with iodo-starch ; but no reaction was obtained. For the sake of 
controlling this test, a minute quantity of nitrate of potash was added 
to another portion of the liquor, and the same method was pursued as 
above: the reaction for nitrous acid was obtained without the least 
difficulty, in this case. 
Several different kinds of sands or soils were used in these experi- 
ments, in order, as has been stated already, that the mechanical condi- 
tion of some one at least of the sets of mixtures might prove favorable 
for the growth of the plants. The importance of this consideration 
had been enforced, with perhaps undue emphasis, in the course of a 
series of experiments with New Jersey green sand, where the texture 
of the sand seemed to be decidedly unfavorable for the prosperity of 
the crops. For these experiments nine glass preserve jars were 
charged as follows: No. I., with 1250 grammes “ West Jersey green 
