695 
Costaria , Undaria , and Laminaria . 
geometrical figure (Fig. 18) is diagrammatized from this series. From the 
diagrammatic figure it may be easily understood that the growth of the 
embryonal lamina of Costaria Turneri results from the division of the two 
initial cells situated side by side beneath the apical cap. This mode of 
cellular arrangement is rather disturbed in Saccorhiza bulbosa , judging from 
the figure given by Thuret, 1 and in Undaria and Laminaria as described 
below. 
The cells which constitute the monosiphonous stipe elongate as the 
lamina extends, to a length much exceeding the diameter. At the same time 
they divide longitudinally and transversely, quite irregularly, to form a 
cylindrical stipe. The resulting cells are poor in chromatophores, and 
those in the lower part are entirely devoid of them. From the basal cells of 
the stipe, a few rhizoidal filaments are given off, in a manner reminding us 
of the root-hairs of a land plant (Fig. 5). Two forms answering to such 
a stage have been illustrated by Kiitzing 2 in Laminaria saccharina ; in them, 
the terminal cells are shown in regular transverse zones, but their longitudinal 
arrangement is not in any apparent order. 
When a plant has attained about 2 mm. in length, the monostromatic 
lamina becomes distromatic. The process is first carried on at the point 
which corresponds to the trasitional region in an adult frond. The division 
of the cells is by a simple plane passing through the middle points between 
the two surfaces of the lamina. Hence, in the cross-section of a lamina at 
this stage, the sister cells of the two layers are exactly facing one another. 
At this stage the cells at the transitional region undoubtedly play the 
part of a meristem, although the apical cells multiply as well. The activity 
of the latter, however, is gradually retarded as the blade is extended by the 
so-called stipo-frondal growth. Most former writers, if not all, seem to have 
omitted to recognize the apical growth of the sporelings of Laminariaceae, 
and hence the crucial stage in the mode of growth has not been hitherto 
considered. So far as I am aware, Reinke 3 is the only one who, arguing 
from the figures given by Thuret, has remarked that the embryonal blades 
of Laminariaceae extend at first by cell-division in all parts of their area, 
and that afterwards the intercalary growth is localized in the transitional 
region. 
While the lamina becomes two-layered after the monostromatic stage, 
the stipe becomes gradually thicker. Cross-sections show many polygonal 
cells compactly but irregularly arranged without any remarkable distinction 
in their size. The boundary between the distromatic lamina and the 
polysiphonous stipe lies, of course, at the transitional point. 
When the stipe has become cylindrical the rhizoidal filaments are 
copiously given off from its basal point (Fig. 14). Each filament is simply 
1 Thuret: 1 . c. 2 Kiitzing : 1 . c., p. 345, Tab. XXV, I, Fig. 5. 
3 Reinke : Studien zur vergleichenden Entwicklungsgeschichte der Laminariaceen, 1903, p. 8. 
