Steil. — Apogamy in Nephr odium hirtipes , Hk. 119 
17-21, PI. V). Sometimes the leaf is developed from the surface of the pro- 
thallium (Fig. 19, PI. V), and in a few instances it has been found that leaves 
are developed from both surfaces. Occasionally roots were observed to 
make their appearance on both surfaces of the prothallium. Secondary 
prothallia produce embryos in all essential respects similar to tho^se formed 
by the primary prothallia. Large irregular prothallia whose form results 
from changes in the intensity of illumination may produce more than one 
embryo each (Fig. 24, PI. V, a and b ), and these embryos may develop into 
normal sporophytes. Fig. 23, PL V, represents a portion of a prothallium 
of irregular form with two embryos, each with a leaf and a root. On another 
portion of the same prothallium a third embryo had been produced. 
A careful study was made to determine whether nuclear migrations and 
fusions similar to those described by Farmer and Digby ( 1907 ) occur among 
the vegetative cells of the prothallium previous to the development of the 
embryo. For this purpose both stained prothallia mounted whole and 
stained sections were studied. In not a single instance were nuclear 
migrations, nuclear fusions, or binucleate cells observed. 
Sporogenesis. After it was discovered that in the sporangium of 
N ephr odium hirtipes the chromosome number is doubled in consequence of 
incomplete nuclear divisions in the manner described below, it appeared 
desirable to make as thorough a study as possible of the whole subject of 
sporogenesis. The observations on the sporogenous cells are based on 
prepared slides of thousands of sporangia in different stages of development. 
Most of the work was checked by a microscopical examination of living 
material. The presence of numerous nuclear division figures in the sporo- 
genous cells belonging to the generations previous to the spore mother-cells, 
in tapetal cells, and in the wall cells of the sporangium, facilitated the counting 
of the chromosomes. In the dividing nuclei of cells of all these classes 
between sixty and sixty-five chromosomes were counted. The same 
approximate number of chromosomes was counted in dividing nuclei of the 
gametophyte and the young sporophyte. 
The nuclear and cell divisions in the sporangium, preceding the stage 
at which eight sporogenous cells are present, are normal in every respect. 
At this stage the sporogenous cells are nearly isodiametric, although there 
are frequent exceptions, some of them being at times much elongated. The 
resting nucleus is large, although not so large as the new diploid nucleus 
formed at a later period. The resting nucleus passes through the normal 
prophase stages. At this time the tapetum is composed of two layers of 
cells. Normal division figures are not uncommon in the tapetal cells and 
the sporangial wall cells even, at a later stage. 
The metaphases following these prophases are at first normal in 
appearance, when observed either from the side or in polar view (Fig. 24, 
PL VI). Instead, however, of the metaphases being followed by the 
