14 WITTROCK, SPORES OF THE MESOCARPEZA. 
of spore-formation as purely neutral. Then a conjugation 
would not even have been designed, and the spores would be 
pure agamospores. In this case Gonatonema ventricosum ought 
to be regarded as a primitive neutral form which has not yet 
attained the higher degree of development, when an act of 
fecundation is necessary to the formation of the propagative 
cells. Or we may — knowing that in two Mesocarpemw, in 
which normally an act of conjugation precedes the formation 
of spores, spores may nevertheless in rare and exceptional 
cases be formed by the instrumentality of a single cell — assume, 
that the plant of which I am speaking is a form where this 
exception has become a general rule, and the spores would 
then be regarded as parthenospores and not as agamospores. 
At present I think it scarcely possible to decide with ab- 
solute certainty which of these interpretations is the right 
one. But the reasons which speak in favour of the first in- 
terpretation seem to me decidedly stronger. One of these 
reasons — and I think a very powerful one — is the circum- 
stance, so characteristic in Gonatonema, that the cells in the 
same filament bend, when forming spores, in a zig zag line 
alternately in two opposite directions. Ifithad been du- 
ring some earlier period in the nature of the species, asitis in 
that of other Mesocarpee, to conjugate, the bending of the 
cells ought to take place (as in these) in the same direction. 
Otherwise only part of the cells (in G. ventricosum every other 
one) would find a possibility of meeting with its conjugation- 
outgrowth the corresponding cell in another filament placed 
near it. At the formation of parthenospores in Mougeotia 
calearea it always occurs, that all the cells in the same fila- 
ment bend (or send forth conjugation-outgrowths) towards the 
same side; see figs. 17 and 18. — Another circumstance which 
favours the view that the spores of G. ventricosum are agamo- 
spores and not parthenospores is, that the spores are here of 
a comparatively considerable size, and that they are fully devel- 
oped both as to their contents and their membrane. The par- 
thenospores of Mougeotia calcarea are comparatively much 
smaller (compare figs. 17 and 18) and have, as far as I have 
had opportunity to observe, a membrane of only one or two 
layers instead of, as normal spores, of three. 1!) On account 
1) In the well-known parthenogenetic form of Zygnemer, Zygnema mirabile 
Hass. (Spirogyra longata (Vauch.) CLEVE, 1. c. pag. 20. pl. 4) the spores 
