619 
Nawashin’s mixture or picric acetic acid. The latter treatment was particularly 
good for demonstrating the nuclei and the pit-connections between the cells. 
The peripheral part of the frond is formed of long thin filaments, usually 
3—4 w thick, sometimes only 2 « or still thinner. Where there is ample room the 
filaments may be long and straight and consist of very long cells; they may be 
simple or bear opposite or alternate branches rising rectangularly near the upper 
end of the cells. The branches are partly different from the long straight filaments 
by being composed of shorter more or less swollen, spindle-shaped or more irregular 
cells. (Comp. figs. 617—619). Later the filaments become very much branched and 
densely felted together, forming a more or less continuous layer within the surface 
of the shell. The cylindrical cells contain a small number of long ribbon-shaped 
or perhaps branched chromatophores and, in the middle of the cell, a nucleus 
which is often rather inconspicuous and feebly stained with hematoxylin. The 
transverse walls are very distinct but it is usually impossible to distinguish a central 
pit by application of dry lenses; but with high magnifying power a callus button 
deeply stained with hematoxylin on each side of the transverse wall could some- 
times be observed (fig. 617 A, E), a sure sign of the presence of a pit. The structure 
of the inflated cells is similar to that of the cylindrical cells; the nucleus was in 
some cases very distinct (fig. 618 B-F), the chromatophores shorter. The transverse 
walls of the inflated cells have the same small diameter as those of the long cells. 
The contents of the inflated cells is at first not very rich, but later it often becomes 
dense and rich in granular matter, probably floridean starch, as it takes a red-brown 
colour with iodine, and the inflations may then resemble the chlamydospores of 
Fungi (fig. 618 J). 
A new kind of cell-rows, different from those hitherto mentioned, spring, usually 
from the inflated cells, but sometimes perhaps directly from the long thin cells. The 
cells of these rows are different from the others by greater breadth and denser con- 
tents. Besides the cell organs they contain much floridean starch which gives the 
cell a granular, untransparent character. The cells are broader, not only in the 
middle but also at the transverse walls, which have a much larger diameter than 
the other cells. The cells of these filaments are cylindrical or a little inflated, almost 
of the same length as breadth, and they are uniform. These cell-rows are always 
branched but have a limited growth. The branches are usually more or less curved. 
As it must be supposed that they normally produce monospores, they will be 
called fertile cell-rows. In material from different shells they showed, how- 
ever, certain differences; as I feel doubtful whether these variations are due to 
differences in the external conditions offered by the various tests in which the 
plants grow or if they may possibly be expressions of genotypical differences, 
specimens from various shells will be mentioned separately, and the drawings 
illustrating them are arranged so that each of the three groups of drawings 
(figs. 617—619) originates from plants growing in the same shell or, as to those 
growing in Spirorbis, at least from specimens of this Serpulid seated on the same 
79* 
