190 
of the dead parts. When, however, the decay of the soft parts 
spreads over a greater portion of the corallum and especially when 
the mouth also looses its living parts, a great many buds originate 
from isolated patches of living tissue, especially in the marginal 
region. A very fine instance illustrating this phenomenon is repre¬ 
sented in figs. 1 and 2. In this specimen the corallum has died 
off, but from the remainder of the living parts of the margin 94 
buds have developed, a great many of which show a broadening of the 
anthocaulus under the present anthocyathus, indicating that already 
a former anthocyathus has detached from it. In the largest buds 
still found on this dead specimen three complete cycles of septa 
and a few members of the fourth cycle are visible, their longer 
diameter amounting to 3 mm. The largest of these buds have already 
assumed a hexagonal form (fig. 3). 
Two specimens in the collection furnish a very curions example 
of regeneration. In each of them the central part with a greater 
diameter of 25 mm has been more or less separated from the 
marginal part by a suture. These irregularities have come about 
in the following way: When the corallum reached a size of about 
25 mm part of it lost its living tissue, which brought about the 
formation of buds in the neighbouring portion. In one specimen 
three buds are present, in the other ten. Apparently the growth 
of the corallum stopped for some time after part of the tissues 
died off. Afterwards a new part of the theca arising from the centre 
of the corallum formed itself and grew out over the dead part. 
In most parts of the margin of the old corallum this new theca 
has fused with the old one, but the difference in thickness of the 
two parts is still clearly noticeable. At first sight these coralla 
seem to be the outcome of a fusion of two individuals. That the 
only true interpretation is the one given above is borne out by 
the comparison with other specimens in which also this kind of 
regeneration has taken place, but at a less advanced stage. 
Fig. 10 represents a specimen showing the first traces of such 
a regeneration. The living part next to the defunct portion has 
formed a kind of ridge. In the course of further development a 
new theca may grow out from this ridge overlapping the dead 
portion. When at last the new-formed parts reach the margin of 
the dead portion, the corallum seems to consist of two discs that 
