ZOOLOGY: R. GOLDSCHMIDT 
221 
not so much to internal changes inside the cell as to changes in the 
physical constitution of the surrounding medium, controlled by the fol- 
licle membrane, forcing the cell to react in that specific way. In order 
to test this hypothesis experiments have been made in influencing the 
osmotic conditions inside the follicular membrane. One of the clearest 
results was that hypertonic condition in the haemolymph forces all sex- 
cells from spermatogonia to spermatids to grow out in one direction, 
namely, against the follicle wall. The resistance of the latter forces 
the cell to grow along its inner surface so that finally the follicle appears 
as a whorl of thread-like cells. The amount of growing out depends 
upon two factors: the age of the cell and the hypertonicity of the me- 
dium. Spermatocytes during or before the maturation divisions form 
long spermatozoa-like threads but with inverse polarity, the axial fila- 
ment on the inner surface of the cell (present as it is well known in but- 
terflies from the spermatocyte stage on) not being involved at all. The 
whole process proved to be reversible without change of the medium, as 
were also the slight changes produced by a hypotonic medium, since 
after some days all cells returned to their normal form. This means 
that the follicle membrane has the faculty of active control of the os- 
motic conditions inside the folHcle. In addition to these experiments 
an effort was made to induce the cells of different age to grow in the 
normal direction along the axial filaments. Up to the present only the 
first steps have been obtained by treatment with diluted methyl alcohol. 
Another normal process that could be induced at a different point in the 
cell history was the formation of the axial filament. In these experi- 
ments interesting observations have been made upon the formation of 
flagella by the plasm of all sperm cells from the spermatogonia to ripe 
spermatozoa. They arise in Ringer's Fluid in a similar way as described 
for other cells by Merk, Kite, Oliver, Chambers, and they can be pro- 
duced and made to disappear at will by change of temperature. 
All these experiments together with the observations of the normal 
process point to the probabihty that the general processes of spermato- 
genesis are necessary reactions of the cells to a systematic regulation 
of the osmotic conditions on the part of the follicular membrane. The 
individual specific processes are caused by the specific properties of the 
reacting cells. This afi^ords us the possibility of an experimental attack 
upon other closely related problems, viz., the function of SertoH's cells, 
which may prove to be the same as the function of the follicle cells in 
the present case, and the problem of apyrene spermatozoa, which, ac- 
cording to some of my observations, may prove nothing else than a 
Itisus naturae, an abnormality produced by slight changes in the physi- 
