496 
(comp. fig. 454 4, B). Floridean starch-grains are often very abundant, in particular 
in the inner large cells of the globular thallus. The statement of SrurcH that there 
Fig. 455. 
Harveyella mirabilis. Cells 
from the cushion of a male 
plant connected with pri- 
mary and secondary pits 
(May). 350 : 1. 
The antheridia 
are no secondary connections between the cells of the gameto- 
phyte does not agree with my observations as I have repeatedly 
met with secondary pit-connections between the cells situated 
a little within the antheridia-producing cells (fig. 455). 
According to Scumirz (1889, p. 29), the cells of the intra- 
matrical filaments are here and there connected with the cells 
of the host-plant through secondary pits. An immigration of a 
nucleus from the parasite into the host-plant or vice versa 
probably takes place, as in the ordinary formation of secondary 
pits, but I have not ascertained this with full certainty, as the 
nuclei were usually not distinct in the specimens examined 
when treated after NAWASCHIN. In some cases a nucleus seemed 
to have penelrated from the parasite into the host-plant. 
As emphasized by SturcH (1924, p. 39), “the cystocarps, 
antheridia, and tetraspores are invariably developed on separate 
individuals”. The three kinds of individuals may attain about 
the same size, in the Danish waters scarcely exceeding 1 mm. 
arise by oblique alternate divisions of elongated cells composing 
the outer layer of the external cushion of the male plants (Sturcn 1924, fig. 20). 
Numerous procarps arise in the female specimens, but 
one cystocarp only is produced, which occupies most of the 
cushion. As to the details of the development of the cystocarp, 
reference may be made to Sturcu’s papers. It shall only be stated 
that I found a four-celled carpogonial branch in accordance with 
STURCH. According to STURCH, the cystocarpial wall is composed 
of branched cell-filaments consisting of fairly short cells. 
In a cystocarp collected by LIEBMAN December 1838 I found the 
filaments feebly branched and consisting of long cells (fig. 456). 
The tetrasporangia are cruciately divided, as shown by 
Kuckuck and STURCH. In specimens preserved in spirit the 
nuclei in the dividing sporangia sometimes showed irregular 
lobed features that may perhaps be due to the imperfect 
state of preservation (fig. 457). This picture shows, moreover, 
that the division of the tetrasporangium does not consist 
in two consecutive bipartitions but that the transverse and 
the longitudinal divisions proceed almost simultaneously 
from the periphery towards the centre of the sporangium. 
Fig. 456. 
Harveyella mirabilis. Hel- 
singer, December. Section 
through the cystocarpial 
wall; below, cells of the 
gonimoblast. 350 : 1. 
The relationship of the genus Harveyella is doubtful. Schmitz (Engler a. 
Prantl. I. 2, 1897 p. 344) classed it among the Gelidiacew, whilst Srurcx referred it 
to the Gigartinacee. As emphasized by Kyrın (1923, p. 124) it is not in accordance 
