Q4 CORALS AND CORAL ISLANDS. 
here by Moseley of the Challenger Expedition. The eight 
tentacles are pinnately fringed as in other Alcyonoids, and 
are wholly retractile by mtroversion. /eliolites, a Paleozoic 
genus, is supposed to be related to Heliopora. The most 
common species of Heliopora is the //. cwrulea of the East 
Indies. 
IV. LIFE AND DEATH IN CONCURRENT PROGRESS IN CORAL 
ZOOPHYTES. 
The large, massive forms of stony corals would not exist, 
and the tree-shaped and other kinds would be of diminutive size, 
were it not for the fact that, in the living zodphyte, death and 
life are going on together, pari passu. This condition of 
growth is favored by the coral secretions; for these give a 
chance for the polyp to mount upward on the coral, as it 
lengthens it by secretions at the top. But, to be successful in 
this ascending process, either the polyp must have the power 
of indefinite elongation, or it must desert the lower part of 
the corallum as growth goes forward; and this last is what 
happens. In some instances, a polyp, but a fourth of an inch 
long, or even shorter, is finally found at the top of a stem 
many inches in height. ‘The following figure represents a case 
of this kind; for all is dead coral, excepting less than an 
inch at the extremity of each branch. The tissues that once 
filled the cells of the rest of the corallum have dried away, 
as increase went on above. Another example is shown 
on page 54, in which the living part had a length of one 
eighth of an inch. The Goniopora, on page 52, is still an- 
other example of the process; but here the living part com- 
bines a great number of polyps: these are growing and _ bud- 
ding with all the exuberance of life, while below, the old _pol- 
