. 
J. D. Dana on: Zoophytes. 19 
short distance from the edge, and there the polyp is developed ; 
while in the Gemmipora the polyp cell opens at the very edge it- 
self, and is continuous through the lower extremity of its visceral 
cavity with that of preceding polyps. 
71. The Echinopora referred to is an example of prolate 
growth without acrogenous. In the Astreeas, both modes are 
combined. The margin of an Astrea argus grows essentially 
like the Echinopora. But buds also mul- Fig. 37. ‘ 
tiply over the growing surface, as illus- : 
trated in the annexed figure. The growth, 
enlarging the surface, tends to enlarge the 
polyps and widen (in this species) the in- ¢ 
tervals between the disks. But this 
widening has its limits, determined by the 
normal size of the polyps, and when these 
limits begin to be exceeded, a new polyp 
buds out in the interval. An example of = 
this is shown near the middle of the an- a 
nexed figure. This process is constantly ASR ATEM 
going on, and by means of it the symmetry of form which be- 
longs to these hemispherical corals, is retained. 
72. 'This prolate growth of the summits, instead of being con- 
fined to the parts exterior to the disks, belongs in many species 
to the disks themselves; and the consequence is that there is a 
tendency in these disks (and not in the intervals between 
them) to exceed their normal size. The result, namely site 
opening of a new polyp, has already been described in $ 47; a 
“6 takes place with the greatest regularity over the hemispherical 
treeas, 
73. In figures 28 to 31 we have illustrated the formation o 
a new polyp, and its subsequent separation from the parent. 
Suppose that instead of separating as soon as formed, the disk in 
figure 29 should continue widening, till another and another 
mouth opened before the subdivision (illustrated in the figures,) 
should commence. This is no hypothetical case, or the M 
drinas are all examples of it and in this simple particular alone, 
they differ from the common Astreas. There are some Astraas 
in which two or three mouths, or even four, occasionally open in 
a single widening disk, and it is dificult to say whether they 
should be arranged with the Meandrinas or not. There are Me- 
andrinas in which the disk is prolonged indefinitely or subdivides 
only at long intervals. However great, therefore, the dissimilari- 
ty in the coralla of a Meandrina and an Astrea, their actual re- 
lations are extremely close. In some instances: there are simple 
and meandrine species in the same genus (Mussa), for the reason 
that no line of division can be laid down. 
