H. F. Wernham. 
Taking the natural order Solanacese as it stands in our modernly 
accepted systems of classification, we find that only about 6% have 
obviously zygomorphic flowers. These, comprised in the tribe 
Salpiglossideae, were originally included by Hooker in the Scrophu¬ 
lariaceae, 1 and the constitution of Solanaceae as a truly natural order 
is still a matter for debate. These zygymorphic forms display 
undoubted affinities with the sub-regular forms of Solanaceae on 
the one hand, while on the other they pass almost insensibly into 
Scrophulariaceae. The latter order, again, includes many sub-regular 
forms with isomerous androecium (Verbascum). It is apparent, then, 
that the line between the two orders in question is extremely 
difficult of description ; as a result, they appear in the same sub¬ 
cohort in Engler’s system. There is small doubt that research 
would reveal the necessity for no little rearrangement upon the 
border-line between the two orders ; but our concern here is with 
the bulk of the species typical of each order. On one side we have 
the Solanaceae, the large majority of which are more or less 
manifestly closely related inter se, and have regular or sub-regular 
flowers with isomerous androecium. On the other side we have 
the Scrophulariaceae, the large majority of which have strongly 
zygomorphic flowers with oligomerous androecium. The claim 
of the small minority of Solanaceae which have flowers with 
the two last-named characters to remain in this natural order 
must rest upon other critical features, and upon the strength of 
their relationship with the typical members of this order estimated 
in the light of those features; this, in any case, is beyond our 
present scope. Sufficient has been said, perhaps, to justify our 
inclusion of the Solanaceae in the Transitional Group, in agreement 
with the system of Bentham and Hooker; and this order is especially 
important in the history of the group, in so far as it seems to lie, 
from the evolutionary aspect, very near to the direct line leading 
from an apocynal ancestry to the multiovulate zygomorphic 
Tubiflorae. The general question of the origin and affinities of this 
Transitional Group must now be approached forthwith. 
In the first place, the bond between the Convolvulaceae and the 
descendants of the apocynal stock appears to be remarkably close. 
Many Convolvulaceae recall climbing, suffrutescent apocynaceous 
forms strikingly in habit and general appearance, while the essential 
floral characters are very similar. The former order is, it is true, 
1 See J. D. Hooker’s Appendix to Le Maout & Decaisne, A General 
System of Botany, Descriptive and Analytical, Mrs. Hooker’s 
translation (1873), p. 586. 
