147 
TubiflorcE. 
regular flower of Verbascum (including about 150 species), together 
with its short corolla-tube and isomerous androecium, links the 
Scrophulariacea* closely with the Transitional Group as represented 
by Solanacese ; while Verbascum is united with the more typical 
members of its family by way of the genus Celsin, generally similar, 
but with oligomerous androecium. The flower of Verbascum 
shews, again, distinct zygomorphy in its unequal and dissimilar 
filaments. The isomerous androecium occurs hut rarely in Scro- 
phulariaceae outside Verbascum ,—e.g., in the small genera Scoparia, 
Bacopa, and sometimes in Sibthorpia and Capraria ; and the total 
number of species in which this condition obtains represents 7—8% 
of the gross number contained in the family. 
In more than 60% of Scrophulariaceae the corolla-segments 
exceed the stamens in number by one— i.e., the latter number four, 
the posterior stamen being aborted or represented by a staminode 
(Pentstemon ); the four stamens are usually arranged didynamously. 
In the remaining 30% the androecium is composed of two fertile 
stamens only, the other members of a hypothetical pentamerous 
whorl being, in some cases, replaced by staminodes; these may 
amount to mere rudiments only, or may be altogether absent 
( Calceolaria , Veronica). 
It will be convenient to urge at this point with a certain confi¬ 
dence that the homology of the staminodes in the higher Tubiflorae 
is very probably with the fertile stamens, and that this conclusion 
is not—at least, in most cases—open to the same objections which 
we have admitted in the case of the Primulales (chapter III). We 
cannot discuss this question at length here ; but favourable evidence 
is not lacking, e.g., in respect of form, position and development, 
in the progressive degradation of the anthers in certain series of 
allied forms, and so on. 
Payer, in his TraiU d' Organogenie comparee de la Fleur (Paris, 
1857, p. 542), furnishes details of development in the flower of 
Lophospermum ( Maurandia) erubescens, a member of the Antirrhi- 
noideas-Antirrhineae section of Scrophulariaceae. The posterior 
stamen appears first, then two lateral stamens ; then an anterior 
pair. The posterior stamen, however, never bears a fertile anther, 
but persists as a staminode. In this case there seems to be little 
room for doubt that the staminode is the true homologue of a 
stamen. In the case of Veronica speciosa, however, Payer observed 
no sign of more than two stamens at any stage of development. 
In regard to adaptation for insect-visits, over and above the 
