Tubiflorce. 149 
vulatae. This has doubtless arisen from the presence of but a 
single ovule in each ovary-cell, in contrast with the multiovulate 
ovary of the typical “ personalian ” flower. But the single ovule 
is pendulous from the apex of the cell, while in the schizocarps of 
Diovulatae the ovule is usually erect and basal. There is in 
Selagineae no indication of what we have named “ schizocarpy ” in 
the last chapter, and to which the uniovulate condition of the 
ovary-cell in Diovulatae is due: each loculus corresponds to a 
carpel, which has therefore one ovule, not two, associated with 
it. In habit and general facies, again, the Selagineae appear to be 
more or less closely linked with some of the smaller-flowered 
members of Scrophulariaceae ; and it may be urged that this tribe, 
with its commonly occurring dense clusters and spikes of small 
flowers, represents a tendency to acquire conspicuousness by the 
aggregation of florets. Indeed, this tendency finds definite reali¬ 
zation in the closely allied family Globulariaceae, in which the 
inflorescence is capitate, delimited by an involucre of bracts. This 
family differs further from the typical Selagineae in the possession 
of a unilocular ovary, the solitary ovule being pendulous from 
the apex. This condition is, however, foreshadowed in Selagineae, 
for here the two free achenes are commonly unequal in the fruit, 
and one is sometimes sterile or obsolete; the ovary in Globularia¬ 
ceae is thus probably the result of suppression of one of the 
carpels. 1 
On the whole the writer is inclined to associate Selagineae 
with Globulariaceae in one family, Selaginaceae, in agreement with 
the system of Bentham and Hooker, but in close connection with 
Scrophulariaceae,’as in Engler’s system. 
The fruit of Scrophulariaceae is nearly always a two-celled 
capsule, revealing no tendency to schizocarpy; and neither fruit nor 
seeds display any special adaptations for dispersal comparable, at 
any rate, with those found in certain of the allied families with 
which we shall have shortly to deal. 
The large majority of Scrophulariaceae are herbaceous in habit, 
the shrubby and arborescent forms being very few in number ; this 
is, of course, consistent with the very general trend of advance in 
the higher Sympetalae. An interesting special tendency in relation 
to habit is seen in the heterotropbic mode of existence adopted by 
many of the Rhinanthoideae ( Euphrasia , Pedicularis, etc.), which 
1 See Lc Maout and Decaisne. A General System oj Botany, 
Descriptive and Analytical. Mrs. Hooker’s translation (1873), 
p. 619. 
