SEEDLING STRUCTURE IN THE LEGUMINOS. 91 
We find that in many cases the transition from root-structure to stem- 
structure 1s localised in the portion of the axis between the collet and 
the base of the cotyledons—. e. the region which is here, for descriptive 
purposes, termed the hypocotyl. In numerous instances, however, the 
transition occurs below the collet: occasionally intermediate structure is 
found above the base of the cotyledons (e. р. Cwsalpinia sepiaria). That is, 
there is no essential connection between the hypocotyl and the transition 
phenomena ; and we shall show later that the level of transition is primarily 
under the control of a factor of size. Moreover, the hypocotyl may be non- 
existent ; in many hypogeal seedlings the collet coincides with the base of 
the cotyledons ; and occasionally the change of surface-layer even occurs on 
the petioles of the cotyledons, some distance above the node. 
We are justified, therefore, in considering that the hypocotyl, when 
existent, owes its prominence and. peculiar features to its position in relation 
to stem and root, to the ground-level, and to its connection with the 
seed-leaves. It seems to be best regarded, not as a stem- nor as a root- 
structure, nor as a compound organ partly composed of stem partly of root, 
nor as a plant-member sui generis ; but as a specialised region of a primitively 
undifferentiated axis. 
HYPOGEAL AND EPIGEAL GERMINATION. 
The distribution of these two types of germination through the Leguminosce 
is of interest. Of the two most primitive tribes the Sophorez appear to 
contain a majority of hypogeal species, while the Podalyriez are, so far as 
known, exclusively epigeal. The Viciew are hypogeal, with the exception of 
Abrus precatorius (? other Abrus spp.), an anomalous member of this tribe. 
The Genistee, Galegew, Lotew, Trifoliew, and Dalbergies are almost 
uniformly epigeal, and so are the Hedysarew, with the partial exception of 
Arachis hypogea. The only other tribe of Papilionate, viz. Phaseolez, 
shows both types in abundance. Among the sub-orders Mimosoidew and 
Cresalpinioidex the epigeal type prevails, but hypogeal germination occurs in 
some cases, especially in the tribes Ingeæ and Eu-Cxesalpiniez ; while in the 
Bauhinie: several species seem to be intermediate in type. 
In several instances hypo- and epigeal germinations occur within the 
limits of a single genus, as in Pithecolobium, Cwsalpinia, Sophora, Phaseolus, 
Erythrina, &c. 
It thus appears that the distinction between epi- and hypogeal germination 
is easily overstepped, and that the change has occurred repeatedly in the 
evolution of the Leguminose in particular and of the Phanerogams in 
general. The difference between the two types is apparently not dependent 
on the absolute size of the cotyledons ; for certain species of Sophora with 
