Liagoropsis — Doty and ABBOTT 
449 
Perhaps the nature of the carpogonial branch 
itself is of systematic value. The straight carpo- 
gonial branch of Liagoropsis with a relatively 
large and somewhat variable number of not 
highly differentiated cells can be considered 
simpler than the curved carpogonial branches 
consisting of fewer, more highly differentiated 
and numerically and morphologically more uni- 
form cells, for example, of Helminth ora. In this 
respect Liagoropsis, Nemalion, Trichogloea, and 
Trichogloeopsis are similar. 
Kylin (1956:108) distinguished Liagoropsis 
from Nemalion on the basis of the carpogonial 
branches, which in Liagoropsis were said to be 
accessory lateral 8 branches on the cortical fila- 
ments, whereas in Nemalion the carpogonial 
branches were said to correspond to young 
branches of the cortical filaments. While this 
distinction may be useful for separating Liagora, 
Helminth or a, and H elminth o cladia from Ne- 
malion or Trichogloea, it is not a valid one for 
distinguishing Liagoropsis, for this genus shows 
both derivations of the carpogonial branches 
but, particularly frequently, the more terminal, 
less specialized Nemalion type. 
When the carpogonial clusters containing 
two or three carpogonial branches in a vegeta- 
tive branch system are considered, the attach- 
ment and derivation of the different branches 
appear to be so different that some may be in- 
terpreted as replacing vegetative branches ter- 
minally (Figs. 12, 13), and others, either in the 
same (Fig. 15) or separate (Figs. 10, 11) fer- 
tile clusters, appear to have arisen laterally on, 
respectively, a carpogonial or a vegetative fila- 
ment. In these small clusters the carpogonial 
branches also may seem to be attached to each 
other (Fig. 10), or in larger clusters they may 
appear to be (Figs. 15-18) parts of a cymose 
system. Thus the position or attachment of the 
carpogonial branches is not a very useful sys- 
tematic criterion in this genus. 
The first and subsequent divisions of the fer- 
tilized carpogonium do not show regularity and 
thus the sequences described by Papenfuss 
(1946) for various of the Helminthocladiaceae 
are not applicable to Liagoropsis. Some oblique 
8 However, further on in his key and descriptions 
Kylin says "terminal or subterminal,” but still places 
the genus near Liagora rather than near Nemalion. 
and some transverse divisions have been seen 
in the zygote contents as well as many cases 
where gonimoblast initials seem to have been 
produced without any division at all. No con- 
sistency was found. This is quite different from 
the situation in Trichogloea or Helminth ora. 
This could be considered as primitive variabil- 
ity, whatever the unknown nuclear cytology. 
The gonimoblast is small when compared to 
the size of the carpogonial branch. Its size is 
comparable to the cystocarp of Nemalion (Ky- 
lin, 1916), but in most other genera in the Hel- 
minthocladiaceae sensu Kylin the mature cysto- 
carp is large and conspicuous and almost always 
associated with a large number of sterile fila- 
ments. The gonimoblast of Liagoropsis consists 
of but a few fertile filaments developing termi- 
nal carpospores on the more centrally formed 
filaments first. Subsequently carpospores are 
produced, also terminally, on the surrounding 
and later formed gonimoblast filaments. This 
sequence is not that of either Liagora or Ne- 
malion. It would appear that the gonimoblast 
grows by producing young branches radially, 
which in age produce terminal carposporangia 
successively more distant from those first pro- 
duced. This cymose development is not precisely 
the indeterminate growth of the gonimoblast 
initials described for the classical Nemalion el- 
minthoides (Veil.) Batters (1902); nor is it that 
of Trichogloeopsis, where the primary gonimo- 
blast filaments terminate in rhizoids. Neither 
does it, in its development, lead to the mixture 
of young, mature, and empty carposporangia 
characteristic of most Liagora species. In other 
words, it seems to be intermediate of, or per- 
haps more primitive than, these other two types. 
In Trichogloeopsis (Abbott and Doty, I960) 
and in Nemalion, as classically illustrated by 
Kylin (1956: figs. 57, 58A), the zygote divides 
transversely, and the primary gonimoblast fila- 
ments protrude radially and develop most at 
90° or more from the trichogyne. With further 
growth the gonimoblast filaments curve toward 
the base of the carpogonial branch, and they 
may become parallel and close to it. The carpo- 
spores are produced terminally on secondary 
gonimoblast filaments which radiate or curve 
toward the outside of the thallus and which are 
branches of the primary gonimoblast filaments. 
Though differing in detail, the gonimoblast in 
