74 GERM-CELL CYCLE IN ANIMALS 
chromatin aggregated into large masses. Only 
from the latter do the sex cells arise. These two 
kinds of cells could be distinguished in larval stages 
and the early separation of germ cells from somatic 
cells was pointed out. Maas, however, does not 
insist that there is here a demonstrated continuity 
of germ cells, since the cells which become sex-cells 
are separated from the egg by a long series of genera- 
tions. 
The recent investigations of Jérgensen (1910) 
on Sycon raphanus and S. setosa have added consider- 
ably to our knowledge of the origin, structure, and 
early history of the germ cells of sponges. Jérgensen 
does not agree with Maas (1893) regarding the early 
segregation of the germ cells from somatic cells, 
but finds no particular difference between so-called 
mesoderm cells and wandering or egg cells. It is 
worthy of note, however, that the youngest recog- 
nizable odgonia were found to contain several distinct 
bodies in their cytoplasm (Fig. 27, B). 
The method of formation of the gemmules has 
engaged the attention of many investigators, but 
several important points concerning it are still in 
doubt. Gemmule formation is of particular interest 
since the cells (amebocytes), which by most authori- 
ties are said to give rise to the germ cells, are also 
considered the cells which form the reproductive 
portion of the gemmules. At least four views have 
been held concerning the origin of the gemmule 
cells: (1) Carter (1849) believed that the gemmule 
is derived from a single cell, the “ovi-bearing cell’’; 
