1900} STRUCTURE AND REPRODUCTION OF COMPSOPOGON 265 
of the cortex or of the younger uncorticated portions of the fila- 
ments may divide to form two cells, one of which is usually, 
however, far smaller than the other and often separated from it 
(fig. 10) by an oblique septum. These small cells, instead of 
being converted directly into sporangia, become pushed outward 
so as to occupy a more or less peripheral position, and soon begin 
to divide independently, at first by anticlinal septa only, the 
early divisions sometimes following one another at right angles 
with such regularity that the resultant cells may be in many cases 
so conspicuously arranged in groups of four as to suggest aggre- 
gations of tetraspores; a resemblance, however, which soon 
disappears, since the divisions follow one another with increasing 
irregularity, and may even become periclinal. The filaments 
thus become covered with scattered sori (figs. 8, 17) irregular 
in form as well as in the number of cells which compose them, 
and might well be mistaken for groups of parasitic alge having 
a habit somewhat resembling that of Xenococcus. 
The discharge of spores from the cells composing these sori 
was not actually observed, but preparations left over night in 
Van Tieghem cells showed that such a discharge undoubtedly 
takes place, the sori being surrounded by numerous aplanospores, 
which, like those of the normal type, were spherical and from 
their regular position evidently non-motile ; but were, on the 
other hand, far smaller (fig. 9) and paler from the presence of a 
relatively much smaller number of chloroplasts. That these 
microaplanospores are similar in nature to the normal spores, and 
that they germinate directly is altogether probable; but the 
exigencies of my departure made it impossible for me to deter- 
mine this as well as other points in connection with them. The 
appearance of the older sori suggests the conditions that have 
been figured in connection with the supposed formation of 
antherozoids in Bangia; but there was no indication that the 
ao thus separated in Compsopogon were sexual in func- 
ion, 
A careful examination of the coloring matter of these plants 
Would be a matter of no little interest, and it may be worthy of 
