49 8 Osborn. — Some Observatio?is on the Tuber of Phy Hog l os sum. 
was too rotten to remove for sketching, but, on examination, a white 
rounded body, barely 2 mm. in diameter, was found on the under side near 
the proximal end of the leaf. This was the only failure in the series, and 
it was not a completely negative result. 
By November 16 it was found necessary to stop the experiment, and 
the remaining six leaves were all sketched before fixing. At this date the 
original leaves were yellowed, and though they still retained their shape 
they were found to be mere hollow shells, the mesophyll being quite 
collapsed. The leaflets also were yellowing, but the cell mass appeared 
turgid and green. Text-fig. 18 D shows that the two tubers developed 
from the single leaf were without any rhizoids. These tubers were opaque 
and white, but smaller and less regular than those formed singly by the 
other leaves. The remaining five leaves (Text-figs. 16 D, 19 B, 20 B, 21, and 
22) all showed tubers, borne at the end of more or less elongate stalks, 
externally resembling the tubers normally produced by Phylloglossum 
plants, and having the usual abundant development of rhizoids that 
characterizes the later stages of tuber growth. In most cases (Text-figs. 
16 B, 20 B, 21, and 22) the tuber was sunken for less than 4 mm. below the 
surface of the soil. In Text-fig. 19 B the stalk is seen to be much longer. 
As will have been noticed from Text-fig. 19 A, this leaf showed quite early 
in the experiment a pronounced upturning of the ends, so that it rested 
upon the soil by a small portion of its surface only. The resistance offered 
by the soil to the downward growth of the stalk caused the proximal end 
to be still further elevated, thus the stalk became unusually elongated. 
(ii) Series C. 
On July 21 ten leaves were broken off plants of Phylloglossum g rowing 
in a sod of soil as before, and laid horizontally upon it. In this experiment 
the sod was placed in a large china developing dish and covered with 
apiece of plate-glass, in order to secure better illumination. Under these 
conditions the atmosphere was at a less constant humidity than in Series B, 
for the glass plate did not fit tightly. While care was taken to prevent the 
soil drying up, no precautions were taken to keep it constantly damp, for it 
was felt that by allowing drier periods to alternate with wet ones, field 
conditions would be more closely reproduced. 
It is unnecessary to describe the results in detail. The leaves, as in 
Series B, showed no visible change other than curving until after August 1 5. 
By August 31 several were noticed to be forming cell masses near the 
proximal end, and one was removed for microtoming. It then showed an 
irregular nodular swelling, yellowish in colour (Text-fig. 23). Three other 
leaves were removed early in October. Two of these showed irregular cell 
masses, from the bases of which white growing apices were projecting. 
