510 Osborn. — Some Observations on the Tuber of Phy llog loss um. 
In section it is said to have an almost circular outline. 1 This is true of the 
leaf for the greater part of its length, but towards the base it is usually 
decidedly elliptical and sometimes becomes, by compression against the 
other leaves, segment-shaped. The usual result of this is that the leaf, when 
detached, lies either on the abaxial or adaxial surface, generally the latter, 
aud seldom on its side. Further, the simple leaf bundle, which is said to 
lie in ‘ the centre of the transverse section \ only has this position in the 
upper portion of the leaf. Towards the base it lies on the minor axis of 
the ellipse but displaced from the median line, being nearer to the adaxial 
surface (Text-fig. 40 and Photos 2 and 6). More than ‘ four or five xylem 
elements ’ have been found in transverse section ; the number is usually ten 
or more near the base. This eccentric position of the leaf bundle within 
the somewhat flattened leaf has the effect of producing curvature when 
turgidity is reduced, and it is the abaxial surface which becomes convex 
under such circumstances. The inference, therefore, is that the adventitious 
growth occurs from the abaxial surface. This conclusion is supported by 
the transverse sections seen in Text-figs. 37, 38, and 40 and Photos 2 and 6. 
It is less easy to judge from longitudinal sections, in which plane most of 
the leaves were cut, but Photo 7 supports this conclusion. 
The amount of adventitious growth made by the leaves of the various 
series was markedly different. The difference can be correlated with the 
period of the growing season at which the leaves were detached. The series 
may be conveniently placed in three groups : A, started early in the growing 
season (June 13), a little after the tips of the largest leaves were noticed 
above the soil; B, C, and D, started in July, when the plants had been 
growing for at least six weeks in the field ; and E, started two months later 
(September 13). Series B and C were very successful. 2 Of the 18 leaves 
involved, 12 matured tubers and 3 others would probably have done so had 
they not been removed for sectioning before the tubers were fully developed. 
The remainder all showed some growth. Series E, on the other hand, was 
a comparative failure. Unlike the leaves of the other series, which remained 
green for four months or longer, many of the leaves of this series quickly 
rotted away. Four only formed any growth, and in the case of one of 
these it was very slight. The remaining three leaves only developed as far 
as the cell mass ; this, however, became white and starchy and in two 
instances formed a ‘ leaflet \ During the two months, July 21 to September 
13, which elapsed between starting Series C and E, considerable develop- 
ment occurred in the new tubers formed by plants growing in the field. In 
July the new tuber merely showed a stalk in process of elongation with 
a small unswollen apex. By September the stalk had reached its full 
1 Bower, F. O. (1885) : loc. cit., p. 673 . The other quotations in this paragraph are from 
the same source. 
2 It seems legitimate to neglect Series D, since its failure may be attributed to other reasons 
than the age of the leaves. 
