F. Cavers. 
176 
arisen by reduction, correlated with adaptation to a moist and 
shaded habitat, from a Targionia Alke type, while the anomalous 
characters of Monoclea may also he explained as due to still more 
complete adaptation to a hygrophilous or almost aquatic mode of 
life. These three genera occupy a somewhat isolated position in 
the Marchantiales, and there is much to he said for the view that 
they represent an independent line derived from a BoscJiia-Uke type. 
It is hardly an exaggeration to say that, beyond the possession 
of a stalked carpocephalum, the Marchantiacese are not distinguished 
from the four lower families of Marchantiales by a single character 
of importance. The Astroporze form a fairly natural group, marked 
off from the Corsiniaceae by the stalked carpocephalum and the 
well-developed elaters, and from the Operculatae by the fibrous 
thickenings in the cells forming the capsule-wall. The Operculatae 
and Composite, however, form much less natural divisions. The 
apical thickening of the capsule-wall and the dehiscence of the 
capsule by the throwing off of a lid are characters found in the 
Targioniacese and in practically every genus of Marchantiaceae, 
though apparently it is only in the Operculate genera Grimaldia 
and Neesiella that the operculum is definitely delimited by an 
annulus of small cells. The structure of the thallus is remarkably 
uniform throughout the Operculatae, and a second common 
character is the absence of fibres from the capsule wall, but in all 
the genera that have been investigated, except PlagiocJiciswa, we 
find an approach to the definitely “Composite” type of carpoce¬ 
phalum. On the other hand, Fegatella, which we can hardly place 
elsewhere than in the Composite, resembles the typical Operculate 
in having as a rule a single archegonium in each involucre; while 
Bucegia and Lunulnria show “Operculate” characters in the air- 
chamber layer and the capsule-wall respectively. It is hardly 
necessary to emphasize the contention that the organisation of the 
carpocephalum affords the best clue to the phylogeny of the 
Marchantiaceae. For instance, Marchaiitia and Lunularia alone 
among Marchantiales possess gemmae, and it is hardly likely, even 
in a group so rich in examples of homoplasy as the Hepaticae, that 
precisely the same peculiar type of gemma should be evolved in 
widely separated types. Yet the two genera differ widely in 
apparently important characters; Marchantia has barrel-like thallus 
pores, a stalked male receptacle with several growing-points, a 
perianth around each archegonium, and ring-fibres in the capsule- 
wall, while Lunularia has simple thallus pores, a sessile male 
