SEEDLING STRUCTURE IN THE LEGUMINOSA. 99 
correlation is due to a tendeney on the part of trees to produce large seeds and 
consequently large seedlings, and for the other types of plants to produce 
small seeds and seedlings (p. 109). 
THE LEVEL oF TRANSITION AND PHYLOGENY, 
The view is held in certain quarters that the level of transition is probably 
of value in phylogenetic discussion. Tansley & Thomas (1906, p. 763) state 
that “It seems that a high transition, typical of Gymnosperms and very 
many Dicotyledons, is probably primitive. ... . Stem structure of the hypo- 
cotyl has apparently been acquired... . . ' 
The results of the previous sections, however, throw great doubt on this 
theory, for though they only apply strictly to the Leguminosie, yet there is 
much probability that similar principles are in operation outside this family. 
It is necessary to eliminate the influence of size before the transition-level 
can become a phylogenetic criterion. Unfortunately, records of the size of 
the seedlings hitherto studied are deficient, and so comparisons are hampered. 
It is noteworthy, however, that in the Coniferm with high transitions the 
diameter of the hypocotyl is often small, at least at the summit, as shown in 
Hill & de Fraine’s figures (1908, I. pl. 35); while Libocedrus decurrens, 
with an intermediate transition (as gathered from the description), is shown 
with a hypocotyl 2:0 mm. in diameter. It is clear that the ascription of a 
high transition to the Cycads, in which the seedling is usually large, rests on 
a misconception ; for though “ the hypocotyl throughout the greater part of 
its length shows a root-structure ” (Hill & de Fraine, 1909, III. p. 455), yet 
“it is obvious from the strongly mesarch nature of the cotyledonary bundles 
n that only a slight amount of re-arrangement is necessary in order to 
bring the protoxylem into its exarch position " (loc. cit. p. 441) *. In Arau- 
caria. Cunninghamii it is stated that “the actual transition takes place very 
slowly ” (Hill & de Fraine, 1909, II. р. 214), and the same is the case in 
A. imbricata and A. Bidwillii (Seward & Ford, 1906; Shaw, 1909). In the 
Gnetales * the transition to root-structure takes place in the lower region of 
the hypocotyl” (Hill & de Fraine, 1910, IV. p. 331). 
Thus it appears that the Gymnosperms should not be quoted as ex- 
hibiting characteristically a high transition comparable to that found in the 
Dicotyledons. 
Since it has been shown that cenogenetic preponderate over palingenetic 
influences in determining the level of transition of the Leguminose, and since 
the basis among the Gymnosperms of the theory of the phylogenetic value of 
* It must be noted that these authors trace the structure downwards, from cotyledons te 
root. 
H2 
