286 My. 8. O. Ridley on Growth and Budding 
evidence he adduces, viz. the growth of these species in 
erect or incrusting plates, and the absence of an apical 
polype, for the simple fact is, that, though in the species re- 
ferred to there is no one apical polype, there are instead 
several. I have examined two of these species, labrosa and 
cuneata, and find that the ends of the branches are well 
covered by large calicles, at the sides (and in cuneata, at any 
rate, from the sides) of which originate young calicles. The 
fact that there is no one apical calicle appears to be due to the 
fact that on the broad ends of the lobes all calicles are equally 
circumstanced, whereas in the pointed-ended Madrepore the 
terminal calicles stand alone in position and circumstances. 
In point of fact the most essential distinction between 
Madrepora and. Montipora is thus overlooked by Dana. In 
Madrepora (as may be seen at once by examining the ends of 
branches of any species except labrosa, securis, cuneata) one 
or more calicles take the lead in the growth, and others origi- 
nate below them, constituting a centrifugal method of budding ; 
in Montipora an undifferentiated apex of coenenchyma takes 
the lead in the growth (as may be well seen in both the widely 
different species J/. foliosa, Pallas, and digitata, Dana), and 
new calices originate in this coonenchyma above the already 
formed calicles; in other words, the budding is centripetal. 
This distinction lies so deeply rooted in the structure and 
physiology of these corals that it is difficult to see how a 
directly ‘ connecting-link”’ between the two types can be 
found. Ishould rather expect to find the connecting-point far 
back in some common form in remote geological time. 
The distinetion is the same as that denoted in flowering 
plants by the terms “determinate” and ‘indeterminate in- 
florescence.”’ In a determinate inflorescence growth is centri- 
fugal, the first flower being formed at the apex; in an inde- 
terminate inflorescence the first flowers are formed at the sides 
and they successively approach the centre or apex of the spike, 
The distinction appears to me to furnish a good character 
by which to divide the Madreporme from the Montiporine 
(sufficiently closely allied to each other and removed from the 
Poritide, as it seems to me, by the possession of a spongy 
coenenchyma, of a well-developed and deep calicle, devoid of 
columella and pali) ; hence I would classify Madreporide as 
follows :— 
Subfam. 1. Madreporine.—Gemmation centrifugal, from 
the sides of terminal calicles. 
Subfam. 2. Montiporine.—Gemmation centripetal, from 
a terminal coenenchymal mass. 
A new genus, which I describe below under the name Ana- 
