214 
as much as 2.5 mm in L. geniculata. The frustule of L. genicul- 
ata figured in the present paper (fig. 2) has a length of 0.87 mm; in 
my material of L. longissima I have seen frustules of 0.43—0.48 
mm in length. 
The frustule sticks to any solid object and creeps forwards over 
the support, secreting new chitin and leaving behind itself an ex- 
tremely delicate, empty tube, open behind. After some time (some 
minutes, according to Billard) this movement is ceased, and there 
is a resting period during which a vertical chitinous wall is devel- 
oped from the hind part of the coenosarc, separating a closed 
chamber containing the coenosarc from the open, empty tube. Some 
of Allman's figures (7 and 8) might indicate, that he has observ- 
ed this limitation, though it is not clearly shown in the figures 
and not mentioned in the text. Billard, in any case, has failed 
to see this wall; according to his figures the stolon is continually 
open behind. The transversal wall is shown in figs. 4, 5 and 6. 
At the front end the chamber is obliquely rounded. During the 
resting period the walls ef the chamber are consolidated all over. 
When examining the polyps on the above-mentioned glass-plate I saw 
in some cases, that the fine empty tube had disappeared, leaving 
on the glass a track as two very fine parallel lines, while the 
chamber which is more capable of resistance remains, limited be- 
hind by the vertical wall. 
On the upper side of this primary stolon a bud is developed, 
growing up to form a pedicel and developing a hydranth with its 
hydrotheca (fig. 4). The developing hydranth consumes a great 
deal of the substance contained in the frustule, as no nutrition 
can take place, until the first hydranth is fully developed. The 
coenosarc of the stolon, therefore, shrinks and becomes thinner 
and shorter, releasing the walls of the chamber, and is partly drawn 
up into the pedicel of the hydranth (fig. 5). According to Billard 
the stolon and the lowest part of the pedicel may even be quité 
emptied. 
When the development of the primary hydranth is completed, 
the new little colony is able to take nourishment and make further 
progresses in its development. — From the primary hydranth new 
hydranths are formed by budding. Among my small colonies which 
are 4 or 5 days old, I have found one or two with two hydranths 
