8 4 
F. Cavers. 
advisable, if these families must be removed from the Jungerman- 
niales, to place them in a special group—for which the name 
Sphcerocarpalcs is suggested—quite independent of the Marchan- 
tiales and the Jungermanniales, though showing affinities with 
both of these groups. 
Were we to take the further step of transferring the Ricciacese 
from the Marchantiales and uniting them with the Sphaerocarpales, 
we should have a synthetic generalised group which would 
practically realise our conception of the common ancestral stock 
from which were probably derived the Marchantiales and the 
Jungermanniales. But the affinities of the Ricciaceae with the 
higher Marchantiales are so close, especially when we take into 
account the transitional group Corsiniaceae, that such a step 
hardly seems justifiable. However, starting from a hypothetical 
ancestral form (Lotsy’s “ Sphaero-riccia ”) with the simple type of 
thallus found in Sphcerocarpus (but without the envelopes of the 
sexual organs) and the simple sporophyte of a Riccia, the evolution 
of the Hepaticae (excluding the Anthocerotales) may readily be 
regarded as having proceeded along two diverging lines. The 
Marchantia line is marked by elaboration in the internal structure 
of the thallus and by the aggregation of the sexual organs into 
receptacles of increasing complexity; while the Jxingermannia 
line shows external differentiation of the plant-body, leading to the 
typical leafy Acrogynae, the internal structure of the vegetative 
organs remaining in most cases simple, though the capsule attains 
greater specialisation in internal structure than in the Mar¬ 
chantiales. 
SPH.EROCARPACE/E. 
The Sphaerocarpales may be divided into two orders—the 
Sphaerocarpaceae with two genera ( Sphcerocarpus , Geothallus) and 
the Riellaceae with the single genus Riella. 
The most striking features of the Sphaerocarpales are the 
presence of a special envelope around each antheridium and arche- 
gonium ; the absence of typical elaters, which are represented 
only by sterile archesporial cells mixed with the spores ; the feeble 
development of the sporogonial seta ; and the absence of fibrous 
thickenings from the single-layered capsule wall, which merely 
decays to set the spores free. 
Most species of Sphcerocarpus have a very simple thallus, with 
notched margins, each notch marking the position of a growing 
