Evans .— Vegetative Reproduction in Metzgeria. 279 
undifferentiated for a long distance. It may even show a branching, by 
dichotomy, while it still consists of a single layer of cells, a condition which 
has not been observed in either of the preceding species (Fig. 5, b). It 
may also give rise to new gemmae at a very early stage. Except for the 
difference in position the distinction between gemmae and adventive 
branches is not always clearly marked in M. quadriseriata. In typical 
cases the branches show a definite costa from the beginning, but they are 
sometimes in the form of narrow band-like structures only one cell thick 
throughout. Such a branch takes its origin from a single cortical cell of 
the costa, and is perhaps to be considered an example of regeneration rather 
than a true adventive branch. The close connexion between regeneration 
and the production of gemmae will be discussed later. 
Metzgeria myriopoda , Lindb. 
This species, although first described in 1874, is still incompletely 
known. It has been collected in several of the Southern States and also 
in Brazil and Argentina. The gemmiparous 
material studied came from Sanford, Florida 
(Rapp, No. 8), and from Opelousas, Louisiana 
(Langlois); it agrees closely with the speci¬ 
mens distributed by Drummond and by 
Sullivant, both of which are quoted by 
Lindberg. The branches which bear the 
gemmae are a little narrower than normal 
branches, but are otherwise scarcely modified. 
The gemmae are irregularly scattered and 
are sometimes abundant, but their develop¬ 
ment does not seem to affect the growth of 
the branch in any marked degree. The early 
stages in the development of a gemma are 
similar to those described for M. fur cat a ^ the original projecting cell 
becoming at once the mother-cell of the gemma. In most cases the 
gemmae maintain a width of six cells (about 0*12 mm.), but they may 
be still narrower or even a little broader (Fig. 6 , A). Their length 
varies greatly and sometimes attains 1*5 mm. before separation takes 
place. The gemmae are usually but one cell thick throughout, but in rare 
cases a few of the cells at the base undergo a division parallel to the surface, 
thus forming an exceedingly rudimentary costa, which may or may not 
bear a very few postical hairs. This condition, however, apparently never 
persists, and the gemma as it continues its growth quickly becomes reduced 
to a simple cell-plate. Marginal hairs make their appearance very early, 
but usually remain short. Although some of them are truly marginal, the 
majority are slightly displaced to one surface, thus showing the first 
Fig. 6. M. myriopoda. a. A young 
gemma. x 80. b. Portion of a 
longer gemma, still attached to the 
gemmiparous branch, but showing 
a secondary gemma at the left, 
x 80. 
