Evans .— Vegetative Reproduction in Metzgeria. 299 
Among the thallose Jungermanniales the closest relative of Metzgeria 
is apparently Riccardia (Aneura ), and this is the only genus in which the 
gemmae are at all similar. Even here the resemblance is much less 
striking than in the leafy forms just described, largely because the gemmae 
are still very rudimentary structures at the time of their separation. They 
are wholly undifferentiated and are in the form of minute oval bodies, each 
consisting usually of only two cells. Each gemma arises from one of the 
surface cells of the thallus, which becomes at once the mother-cell of the 
gemma without undergoing a preliminary division. The contents of 
the cell separate from the wall, surround themselves by a new wall of 
cellulose, and then divide. The free outer wall of the mother-cell is then 
ruptured and the gemma escapes. According to Goebel (’ 82 , p. 338), the 
escape is effected by a swelling of the inner layers of the wall of the mother¬ 
cell. This process is comparable with the deposition of the gelatinous 
layer in the projecting cell of Metzgeria which is to give rise to a gemma, 
except that in the latter case the gelatinous substance does not completely 
enclose the protoplast of the cell, but only that portion of it which would 
have been exposed upon the rupture of the outer wall. In Riccardia 
further development is deferred until after the gemma is set free, while in 
Metzgeria the development is continued until a multicellular gemma is 
formed. In other respects the conditions are much the same in the two 
genera, as Goebel (’ 98 #, p. 275) has already emphasized. 
The large and complicated gemmae in the Marchantiales, known only 
in Marcha7itia and Lunularia , are totally distinct from the gemmae of 
Metzgeria, while the gemmae of the Anthocerotales are still too incompletely 
known to make comparison profitable. The gemmae in the Bryales, which 
have been so thoroughly studied by Correns (’ 99 ), are almost always set 
free by a rhexolytic process, in which a specialized stalk-cell, or tmema, 
is torn across. Among those in which the separation is schizolytic the only 
ones which at all resemble the gemmae of Metzgeria are found in Tortida 
papillosa and T. latifolia , both of which grow on the trunks of trees. In 
these two species the gemmae are irregular oval bodies, each consisting of 
about twelve cells and showing only a slight differentiation. A single leaf¬ 
cell gives rise to a number of these gemmae in succession, and when they 
germinate they first develop a branched protonema as in all other mosses. 
Conditions under which Gemmae are produced. 
The gemmiparous species of Metzgeria do not produce gemmae under 
all circumstances, and the same is true of other Bryophytes. Apparently 
the conditions which induce the formation of gemmae are similar to those 
which induce regeneration. As Goebel (cf. ’ 98 #, p. 277) has pointed out 
more than once almost any liverwort cell has the power of regenerating, 
that is, of giving rise to an entire new plant. In doing this, if it has 
