320 
K.J. McNamara 
A 
Test Length (in mm) 
C 
Figure 5 Plots of pore pairs in, A, ambulacrum I, B, 
ambulacrum II and, C, ambulacrum III 
aborally against test length for Schizaster 
(Schizaster) compactus. 
On the adoral surface significant ontogenetic 
changes occur to the relative size and shape of the 
peristome and the periproct (Figure 6). The 
peristome becomes relatively smaller with growth, 
decreasing in width, from 20%TL to 16%TL (Figure 
6A). Likewise, the peristome length (Figure 6B) 
decreases from 9%TL to a little over 4%TL. This 
relatively greater reduction in length, compared 
with width, arises, in part, from a relative increase 
in anterior growth of the labrum, which projects 
more across the peristome in larger specimens than 
in smaller ones (Figure 1). Unlike S. (Ova) 
myorensis, where the peristome becomes positioned 
relatively closer to the anterior ambitus through 
ontogeny, the peristome position in S. ( Schizaster) 
compactus is highly variable (see below) and shows 
no such anterior trend. However, like S. (Schizaster) 
compactus, the peristome in S. (Ova) myorensis also 
becomes relatively smaller during ontogeny 
(McNamara and Philip 1980a). 
The periproct likewise becomes relatively smaller 
during ontogeny. It is particularly wide in the 
juvenile specimen, being almost 15%TL, reducing 
during adult ontogeny from 12.5%TL to about 
11%TL (Figure 6C). The slightly greater degree of 
reduction in relative length (Figure 6D), from 
19%TL to 15%TL, compared with relative width, is 
reflected in a change in periproct shape, from 
subcircular to distinctly elongate. S. (Ova) myorensis 
similarly shows a decrease in periproct size during 
ontogeny. Other characters to undergo ontogenetic 
change are the anterior notch, which becomes 
relatively deeper and narrower (Figure 1), and the 
periplastronal area which becomes relatively 
narrower (Figure 1). 
Intraspecific variation 
The most notably variable character in S. 
(Schizaster) compactus is the number of gonopores, 
which vary between two and four (Figure 3). Of 
the 27 specimens used in this study that had fully 
opened gonopores, 8 had two, 4 had three, and 15 
had four. Thus, while Koehler (1914) defined the 
species, in part, on its possession of four 
gonopores, the examination of a larger population 
size than the two specimens Koehler used in his 
description, shows that only 55.6% of specimens 
had, in fact, four gonopores. A further 29.6% had 
three and 14.8% two (Figure 8). It would be 
interesting if sufficient numbers of specimens from 
the type locality in the Bay of Bengal ever became 
available, to see if similar variation exists in 
topotype material. The relative percentages of 
specimens with two, three and four gonopores 
show quite a close correspondence with the 
situation in S. (Ova) myorensis (Figure 8), where 
14.3% had two gonopores, 11.5% had three and 
74.2% four. 
There is no correlation between number of 
