404 Dr. R. Schafer—On Phillipsastrea, d’ Orb. 
the Zetracorallia. The reason of this may be that the arrangement 
above mentioned is shown usually only in young Corals. Kunth* 
himself could not distinguish the bilaterally-symmetrical arrangement 
of the septa in the genus Phillipsastrea, but could distinguish it 
clearly enough in the closely-related genus Acervularia. Under 
these circumstances I must refrain from indulging in any speculation 
as to the position of the “ Haupiseptum,” the “ Gegenseptum” and the 
«« Nebensepta,” and confine myself solely to giving a description of 
those relations which I really observed. 
All the corallites in the sections show two kinds of septa; when 
regularly and normally developed, 12 longer septa and 12 shorter 
alternating seemed to be the number. Such regularity in the 
development I could not, however, observe in every case; in most 
cases more than 24 septa were present, and here and there two of 
the shorter septa came between two of the longer, and often some 
of the longer septa were shortened and could hardly be distinguished 
from the shorter. The sections (see Pl. XII. Figs. 3-7) show clearly 
that in most cases two or more opposite septa unite and form a con- 
tinuous line, which traverses the whole corallite. In the central part 
this line is often somewhat thicker ; it isin some cases joined by others 
of the longer septa. No columella is seen in any of the corallites, 
either in the horizontal or in the vertical sections. 
The vertical section of two neighbouring corallites of the hest 
preserved specimen, Pl. XII. Fig. 4, cuts exactly the vertical axis of the 
latter, only the upper part of the section (owing to an accident of 
growth) is slightly excentric, that is to say, intersects the imaginary 
axis of the corallite at a very acute angle. Accordingly the upper 
part of the figure shows the section of two neighbouring septa, but 
further down, where the section runs through the central axis, it 
intersects the point where two neighbouring septa join, and the 
lower part of the figure shows but one septum line; here these two 
septa are joined together. The four other vertical sections which 
I examined agree with this in every particular of importance— 
a columella is nowhere to be observed. 
Wall.—On the upper surface of the coral no wall can be perceived, 
but the distal ends of the septa of the neighbouring corallites appear 
to run into one another. Nevertheless some indication of a wall 
may be perceived here and there in that the septa in some places 
are not altogether confluent, but the line of apparent junction is 
broken, and so suggests the place where a wall might be. Neither 
is any wall perceptible in the horizontal sections of Phillipsastrea 
radiata. When more highly magnified however, one often sees 
that the septa of neighbouring corallites are not really confluent, 
but that they overlap each other and are separated by a small space. 
Sometimes indeed, but very rarely, the septa do run together in a 
continuous line. 
It is worthy of notice that a so-called interior wall unquestionably 
does not exist. It was formerly generally supposed that in all 
1 A. Kunth, Das Wachsthumsgesetz der Zoantharia rugosa, etc., Zeitschrift der 
Deutschen geologischen Gesellschaft, vol. xxi, p. 659, Berlin, 1869. 
