288 MR, .W. D. LANG ON GROWTH-STAGES | Mar. 2, 
division. In each species the coste, in passing from the proximal 
to the distal end of the corallum, exhibit very definite growth- 
stages; and these costal stages are dependent on changes in two 
characters—the general shape of the costa and its ornamentation. 
The changes in the general shape of the costa are from a broad, 
low, indistinct, to a comparatively narrow, high, and clearly 
marked costa, and a corresponding katagenetic phase. The kata- 
genesis * is nearly always rapid compared with the anagenesis *, 
and each is represented on the diagram, text-fig. 38, by a steep limb 
of the wave-like progression. It appears that thrice over in the 
series under consideration has this period of costal rise and fall been 
repeated, and each time with a different type of ornamentation. 
When present, the ornamentation of the costa is of two kinds. 
In the most primitive stages the costs are plain, or very slightly 
rough, and only change in height. ‘This period is therefore here 
called the Plain Period (text-fig. 38). The first marked orna- 
mentation is a longitudinal etching as if the costal surface had 
been corroded. As this becomes more intense, the ridges tend to 
mass together and fuse, until the highest development is a wavy 
ridge down the middle of the costa. Correlated with this etched 
ornamentation is a second period of costal rise, the broader lower 
costa having the irregular etching, and the high narrow costa the 
ridged or nainent : this is the Etched Period. The second and 
more advanced type of ornament is the granule, and the Granular 
Period corresponds with a third costal rise. 
Nine main stages may thus be formulated in this series of 
Corals, namely,a low, medium, and high stage in the Plain, Etched, 
and Granular Periods. Katagenetic stages have not been con- 
sidered, because as a rule they are passed through too quickly to 
be appreciated. Of course, no one species of Parasmilia shows 
all nine stages; in fact the first three are so condensed in P. cen- 
tralis (Mantell), one of the more primitive species, that a Plain 
Period would not have been suspected to have had a separate 
existence from the Etched Period, had not the scheme suggested 
it when the other stages were marshalled in order. A more 
careful examination ot P. centralis (Mantell) then showed a 
heightening of the costa (costal-stage III.) before the normal low 
etched stage (costal-stage TV.), thins revealing a Plain Period, 
characteristic of P. serpentina Kdwards & Haime, preceding the 
characteristic Htched Period of P. centralis (see text-fig. 40, F, G, 
p- 295). Such Tachygenesis fT is, however, to be looked for when 
cases of individual development in other phyla are remembered. 
Again, the greatest gap in the series is between P. centralis 
~ (Mantell), the most advanced costal-stage of which is the high stage 
of the Etched Period (costal-stage VI.), and P. fittont Edwards & 
Haime, the greater extent of whose costz is in the low normal stage 
* A, Hyatt, 1889, “Genesis of the Arietide,’ Smithsonian Contributions to 
Knowledge, Washington, xxvi. Art. 2, pp. 71-74. 
+ A. Hyatt, 1893, “ Bioplastology and the related branches of Biologic research,” 
Proc. Boston Soc. Nat. Hist., vol. xxvi. p. 77. 
