288 Evans .— Vegetative Reproduction in Metzgeria. 
branches, for example, these cells are arranged in from four to six rows, 
while on gemmiparous branches there are often only two rows present. An 
opposite tendency has already been noted in M. eras sip ills, where the 
number of rows is often increased from two to three or four. It is possible, 
however, that this difference between the two species is not of very great 
significance, because it may be due to factors which are not concerned in 
the production of the gemmae. The development of the gemmae in 
M. Liebmanniana is very much the same as in M. crassipilis. Their 
structure at the time of separation is also similar, except 
that they are usually less differentiated. A mature 
gemma is in the form of a flat or concave plate of 
cells, orbicular in form and measuring about 012 mm. 
across (Fig. 12, a). It consists throughout of a single 
layer of cells and does not have a stalk, although the 
two basal cells can usually be distinguished. Very 
young gemmae sometimes show two apical cells, but 
it is exceedingly rare for more than one to persist. 
Marginal hairs are sometimes absent altogether and 
sometimes appear, in very small numbers, as short 
rudiments. If the gemma is plane it shows absolutely 
no indications of dorsiventrality ; if it is not plane the 
concave surface represents what is morphologically 
the postical surface. Even here, however, the dorsi¬ 
ventrality is probably not stable at the beginning. 
In germination (Fig. 12, b) the rudimentary hairs 
about ready to separate, develop further and assume the function of rhizoids. 
gemma 8 ^ g g e J mlnatlng New marginal hairs also make their appearance and tend 
to be slightly displaced to one surface. In cases where 
the gemma is concave, the hairs are always displaced to the concave surface, 
thus emphasizing its postical nature. The apex of the gemma soon gives rise 
to a new thallus which consists at first of a strap-shaped plate of cells bearing 
a few marginal hairs. Since these are truly marginal, the young plant is 
quite without evidences of dorsiventrality, even if the gemma from which 
it grew was concave. It is only later, as the differentiation of the thallus 
proceeds, that the dorsiventrality again becomes apparent. The steps of 
the process are essentially the same as in previously described cases. In 
rare cases the young thallus forks while it still consists of a single layer 
of cells. 
Fig. 12. M. Lieb¬ 
manniana. A. A gemma 
Metzgeria dichotoma , (Swartz) Nees. 
The present species was originally collected in Jamaica by Swartz and 
is apparently not uncommon in the Blue Mountains. Lindberg records it 
also from Cuba, Mexico, and Brazil. The gemmiparous material described 
