108 MR. R. H. COMPTON : AN INVESTIGATION OF THE 
This is only a partial solution of the problem, however. Two alternatives 
still remain open, viz.: (i.) The first leaf trace in the triarch Viciez is a new 
structure superposed upon a diarch type. (ii.) The Vicie type has arisen 
by reduction from а tetrareh type, probably with opposite primordial leaves 
as in the Phaseolew, the intercotyledonary xylem retained not being the same 
one as is kept in the epigeal type, and being converted from a cotyledonary 
to a plumular function. Between these two alternatives it is difficult to 
decide ; but the existence of the tetrarch Abrus and the closely-allied 
Phaseolez tells in favour of the second. The evidence afforded by Pithe- 
colobium Unguis-cati and Сеѕаїріта sepiaria points in the same direction 
(cf. Vicia Faba); while the peculiar ease of Bauhinia racemosa is hardly 
evidence for the first alternative. On the other hand, the second explanation 
involves the conversion of a joint conducting channel composed of plumular 
secondary xylem superposed on cotyledonary primary xylem into a strand of 
plumular primary xylem; and this change may seem too radical to be probable; 
particularly as in the case of epigeal triarchy it has been shown that an 
active replacement of cotyledonary by plumular xylem does not seem to 
occur. The first explanation, moreover, is simpler and more attractive than 
the second ; though it conflicts with the usual rigidity of diarchy. On the 
whole the balance between the two alternatives appears to be nearly even, 
with a slight inclination towards the second, viz., the theory of the derivation 
of the triarch Viciew type from tetrarchy by specialisation in the opposite 
direction to that which has occurred in the epigeal Galegese, Hedysarez, Xc., 
combined and eonnected with the adoption of hypogeal germination and 
early plumular development. 
OTHER Types or SYMMETRY. 
In a few instances more than four protoxylems are present in the hypo- 
cotyledonary axis, The simplest case is that of Pithecolobium Saman, where 
four new proxytolems are added in the root, alternately with the existing 
four. This may be simply in connection with the large size of the root in 
this species ; though other species with equally large roots remain tetrarch. 
Another simple case is that of Bauhinia racemosa, where it is clear 
that the hexarch structure of some seedlings is derived from the tetrarch 
by the interposition of plumular metaxylems (p. 15). In these two cases 
the tetrarch type appears to be fundamental, the extra complications 
acquired. In other eases, however, this is not so obvious. In Voandzeia 
is too remote for serious consideration. 
It may be noted that in the hypogeal Sophores (see Lubbock’s figures of Sophora secundi- 
„Йога and Edwardsia chilensis, 1892, pp. 452, 453) the first leaf is borne on the same side 
as in the epigeal Hedysarew, &c., i, e.—on the reverse side to that on which it is borne 
in the hypogeal Viciem. 
