191 
have fused together. The phenomena in the marginal parts of the 
Cycloseris-\ovm of Fungia distorta, as described by Dbderlein 
(1902), lead to a form of corallum similar to these specimens of 
F. hexagonalis. With the latter species it is always obvious that the 
apparent doubling of the margin was caused by regeneration. 
In many cases the regeneration of the defunct parts by means 
of a new theca over the old causes irregularities, as for instance 
in the case of the specimen of fig. 7. Parts of the margin of this 
specimen have died off and the neighbouring portions in their 
further growth spread towards these dead regions, the septa and 
costae assuming a curved course. The corallum now shows some 
resemblance with a Diaseris-hrm, in which also the septa in each 
lobe strengly diverge. Here, however, the lobes are connected by 
outgrowths of the corallum above the dead parts, so that no clearly 
marked divisions between the lobes exist as in the case of the 
Diaseris-forms. 
Sometimes these lobes in regenerated coralla have the septal 
arrangement so characteristic in Diaseris-spQcimQns. One of my 
specimens (fig. 6) has regenerated from one half of a broken corallum, 
as a ridge at the lower surface indicates. The greater part of the 
disc has regenerated to a sector that completes the larger portion 
of the remaining half; in a smaller section, however, the septa 
strongly diverge whilst they are curved, having assumed in this 
way quite a Diaseris-WkQ appearance. 
Phenomena of regeneration often lead to the formation of a 
number of secondary mouths. One specimen illustrating this faet 
(fig. 11) has very irregularly regenerated. Nearly half of the corallum 
broke off as it appears and was afterwards partly restored. The 
septa and costae of this portion are arranged more or less at right 
angles to those of the rest of the corallum. At least six mouths 
are present, arranged on every side of an abnormal protuberance 
from the disc. It is obvious, especially at the lower surface, that 
the further regeneration of this specimen would lead again to an 
approximately circular corallum and not result into a Diaseris-form. 
Among the material are a few specimens that possess two 
mouths of equal size, apparently the results of fission of the mouth 
into two halves. The septa around each new mouth are partly the 
regularly arranged old ones, partly new ones formed between the 
