500 Osborn. — - Some Observations on the Tuber of Phylloglossum. 
(iii) Series D. 
This series was grown on a similar sod of soil to Series C and in the 
same dish. It was, however, much less successful. The leaves were often 
buried by the castings of small animals or overgrown by Bryophytes. Two 
leaves fixed early in October presented a similar appearance to that shown 
in Text-fig. 23. By the end of the month six of the remainder had damped 
off. Only one of these showed any adventitious growth, a whitish swelling 
about 1 mm. in diameter near the proximal end on the lower side. The 
last two leaves were fixed at the end of November, when one showed 
a dropper, the apex of which was hardly swollen. The other had a series 
of nodular swellings extending almost around the leaf near the cut surface ; 
several of these appeared white, like minute tubers (Text-fig. 25). 
It is difficult to account for the comparative failure of this series, unless 
it were due to competition with various organisms. 
(iv) Series E . 
This series was not started until September 13, when twelve leaves 
were laid on soil in a glass dish as above. Three of these collapsed in 
about a fortnight, and five others before the end of November. Of the 
remaining four, three (Text-figs. 26-28) each formed a small irregular cell 
mass near the distal end. Two of these (Text-figs. 26 and 28) produced 
• leaflets \ and all showed rhizoids, but no dropper was developed. The cell 
masses themselves appeared opaque white, like resting tubers. The fourth 
leaf (Text-fig. 29) only produced a minute white structure about 1 mm. 
diameter. This swelling was developed from the surface of the leaf remote 
from the soil. 
(v) Series A. 
The consideration of this series has been postponed until the other 
experiments had been described, because, of the eight leaves used, only one 
formed a single tuber (Text-fig. 30) in the manner described above. All 
the remainder produced abnormalities. 
As has been explained, the conditions of the experiment were slightly 
different, in that soil from the locality in which Phylloglossum was growing 
was placed in a Petri dish and lightly tamped down, instead of using an 
undisturbed sod. Water was sprayed over the soil from time to time to 
keep it moist. Though the experiment was begun on June J 3, nearly 
a month before any other series, no development occurred until the latter 
end of August, by which date all other series also showed growth. The 
only noticeable feature displayed by the leaves during the first two months 
was the remarkable curving exhibited by them. With a single exception, 
in which case the leaf lay flat along the soil throughout the five months the 
