302 Mr. F. A. Bather on Shell-growth 



septa sometimes occurs as a purely pathological episode in 

 middle age. Further, these normal signs of a period when 

 powers are weakened outwardly repeat those of a period when 

 powers have not yet acquired strength. Old age " is second 

 childishness." But there is an old age of the Phylum as well 

 as of the Individual; it is brought about by an acceleration 

 of development and by an absorption into the mature form of 

 senile characters inherited from its ancestors. Hence the 

 laws that govern the life of the Individual govern that of the 

 Phylum. Phylogenetic old age repeats the characters of 

 phylogenetic youth. The gradual coiling of the Amnio- 

 noidea, followed by a converse uncoiling, is a well-known 

 instance. Another instance is afforded by the character now 

 under consideration. 



Hyatt (5, p. 328) has pointed out that the most antique 

 Ce])lialopoda known to us are certain North-American species 

 of Emioceras, Piloceras, and allied genera ; in these forms the 

 septa are very close to one another throughout. From this 

 fact, as well as from a consideration of septa in certain other 

 molluscan shells and the tabula in Ccelenterate skeletons, we 

 may infer that when first the Cephalopod shell became cham- 

 bered the septa were close together. In the three main lines 

 of descent from such an ancestor the septa came to lie further 

 apart. In the Kautiloidea, which seem to have been the 

 last to specialize, the septa are still far apart, but approach in 

 old age. The Ammonoidea differed at an earlier stage from the 

 parent stem ; so early as the Goniatites the septa are far apart 

 in proportion to the diameter of the whorl ; senile characters 

 gradually appear, and among them this one^ the approxima- 

 tion of the septa ; it is gradually absorbed into the mature 

 forms, and in the retrograde Cretaceous species all of the septa 

 are again closer together. The straight forms in which the 

 protoconch is protected by a sheath {Aulacoceras, Belemnites^ 

 &c.) form another genetic series parallel with the Ammo- 

 noidea, for which I propose the name Coleoidea *. The 

 same process takes place here, along two lines, as described 

 in IV. ; at the end of one of these comes Sejna, and it is clear 

 that the closely pressed lamellae of the " pad " t ai"e nothing 

 else than septa in which this retrogression, started by phylo- 

 genesis, has been enhanced by natural selection on account 

 of its adding strength to the now internal shell. It is therefore 

 the general approximation ot all septa in Sepia that is homo- 



* KoXfdy, sheath, elbos, form. See further (14). 



t " Pad" = German Wulst; suggested as shorter and less misleading 

 than " spongioid tissue."' 



