166 
Tubiflorce. 
those common among plantains. The diagonal placing of the 
calyx-segments has led to the assumption that a posterior calyx- 
segment has been lost in Veronica and Plantago in the course of 
descent. But in the allied Littorella the calyx-segments are not 
arranged diagonally (see Eichler, Bliithendiagramme, I, fig. 27). 
It has been assumed further that the tetramerous condition in both 
has resulted from fusion of two corolla-lobes. This may not 
impossibly be the case in Veronica with its unequal corolla-lobes 
and its affinities with pentamerous forms of the same family ; but 
to make the same assumption in the case of Plantago with its 
perfectly regular corolla seems to be quite unwarranted. In neither 
Veronica nor Plantago do the processes of floral development favour 
the assumptions in question, in regard to either calyx or corolla. 1 
The androecium of Plantago is thus in every sense isomerous 
with the corolla, and this militates further against the idea of its 
origin from a stock which is essentially anisostemonous. 
We conclude, therefore, that the ancestry of Plantaginales 
must be sought more remotely than among the descendants of 
Multiovulatae, and we are led to regard them as derived from the 
apocynal plexus on an evolutionary branch contemporary with the 
branches of the Transitional Group; the original tendency deter¬ 
mining the branch was aggregation of florets; but reduction 
supervened, the result, perhaps, of the adoption of a geophilous 
habit. 
1 Payer, loc. cit., pp. 543, 607. 
