213 
on the sides of the jar. The attendant, Mr. Smith, told me, that 
this phenomenon was quite familiar to him; he had seen it sever- 
al times, just in the case of the same species, and it occurred 
when the hydroids were starving. — I left it all untouched for a 
few days more and then placed the glass-plate and the colonies 
in bottles with formalin. 
On the 27th of May 1915 I collected some colonies Laome- 
dea geniciilata in the harbour of Frederikshavn on the east coast 
of Jutland. When later on I examined those colonies in Copen- 
hagen, I saw on some of them a large number of frustules in vari- 
ous stages of development. Then I brought forward the plunger- 
plate and the colonies from Plymouth and undertook a doser 
examination of the whole question. — The following is an account 
of the formation, deliberation and further development of the frust¬ 
ules as far as I have been able to follow up the progression of 
the process by means of the examples in my possession. At the 
whole my observations are in accordance with the statements of 
Billard, but I may add a few facts unobserved by this author. 
A bud, founded in the usual way at the base of a hydranth, 
grows out as a thread, covered with a perisarc annulated near the 
base but elsewhere thin and smooth. At the beginning the tube is 
filled by the coenosarc, but when the thread has reached a certain 
length the coenosarc is seen to become thinner in its middle part, 
redrawing itself from the perisarc (fig. 1, a). Somewhat later the 
thread has the appearance of a proximal and a distal mass of coen¬ 
osarc connected by a thin string, all covered by the perisarc (fig. 
1, b and fig. 3). The two masses of coenosarc remove more and 
more from one another, at the same time as the growth is con- 
tinued at the distal end. At last the string is cut over, and the 
tissues are closed in the remaining proximal part as well as in the 
isolated portion, the frustule. At first this is provided with a point¬ 
ed “tail“ (fig. 1, c), but this tail is withdrawn, and before the 
frustule is deliberated it has the shape of a small cylinder, round- 
ed in both ends (fig. 2). By the power of the waves or the current 
in the water the thin chitinous tube is broken below the frustule, 
which swarms out into the surrounding water, still covered with a 
layer of perisarc. — According to Billard the length of the frust¬ 
ule at the time of liberation is 0.4—0.9 mm in L. longissima, 
