No. 426. MADREPORARIAN CORALS. 6 
493 
a monticuliporoid bryozoan. The corallites of the different 
colonies vary greatly in size, but the members of any colony 
are practically equal, showing them to be derived from polyps 
of a similar age; there is no suggestion of a larger parent 
corallite and smaller buds. They seem best regarded as aggre- 
gated colonies. 
The common West Indian coral, Siderastrea radians (Pallas), 
in the course of its development, has recently provided me 
with actual instances of colony formation by the fusion of 
originally free, distinct larve. The process of growth has 
been followed from the time of fixation of the free-swimming 
individuals as far as the production of the early skeleton. 
Small, unfixed, adult colonies of S. radiaus occur in plenty 
in the shallow waters of Kingston harbor, Jamaica, and are 
easily kept alive in aquaria in the laboratory. The physiologi- 
cal activity of numbers of unicellular algz, symbiotic within 
the endodermal tissues, renders unnecessary any artificial means 
of aération of the water. Colonies are occasionally met with in 
which the internal cavity of many of the polyps is crowded with 
larva. These are shot out from time to time and are thus 
free to establish new colonies in turn. As a rule, only one or 
two larvz are extruded at a time, but upon disturbance of a 
fully charged colony a score or so would appear together. In 
one colony the intermittent liberation of fresh individuals was 
continued for about a month. 
Immediately on being set free within the water the larvae 
were able to swim about, and revolved around their principal 
axis, being provided with a uniform layer of cilia. They 
measured two to three millimeters in length and were mostly 
pear-shaped, the narrow aboral end being anterior, or forward, 
in swimming. The broad oral end was posterior and dark 
colored, the latter condition owing to the presence of zoóxan- 
thellze (yellow cells) within the outer ectoderm. In a few larve 
the forward aboral end was the broader, as appears to be more 
usually the case in actinian and madreporarian larvae when 
first extruded. The majority were opaque, without any out- 
ward indication of internal mesenteries, but occasionally a 
distended specimen was met with in which the walls were 
