W. 2). Lang — BenKtrcVs Biological Theories of Fossils. 557 



developmental aim is to produce mesenteries in pairs, shown in the 

 second phase (acme), and so perfected in the third phase (katabasis) 

 that this phase reproduces the idea of the first, since the two 

 mesenteries appear practically as a single unit. Our new structural 

 unit is now a pair of mesenteries, and the second six mesenteries 

 appear in pairs and constitute a second period. We are, however, 

 not directly concerned with the further development of mesenteries, 

 for it is in fossils that we are looking for periodicity in development. 

 But, since the septa follow and to some extent are determined by the 

 mesenteries in number, position, and time of appearance, it is well to 

 realize that mesenteries have a period in development whose structural 

 unit is a single mesentery, prior to one with a pair of mesenteries for 

 its structural unit. The most primitive septal development yet 

 described is that o^ Zaphrejifds} Carruthei's figured three stages in 

 a ' paired ' period in this form, namely, stages with, two, four, and 

 six septa respectively. The next period in Zaplirentis and in Rugose 

 corals generally has still a pair of septa for its structural unit, but 

 instead of being added to the previous septa as if these constituted 

 but one system, the new septa appear in pairs independently at two 

 levels, one in the ' dorsal ' and the other in the ' ventral ' parts of the 

 corallum^ as if the previous septa fornied two systems. Since the 

 resulting appearance is three fossulse, this period may be called 

 the ' fossular ' period. After a time, new septa no longer are added 

 in pairs about the two levels, but a ' cyclic ' period is introduced, the 

 structural unit being a complete cycle of septa interposed between 

 those already existing. Three points are remax'kable in the cyclic 

 period ; in, at any rate, most Rugose corals but one cycle is thus 

 produced; often the cycle is not perfect in its time of appearance, but 

 the septa in the neighbourhood of the counter-septum progressively 

 anticipate those in the other half, showing liow the cyclic method 

 originated ; there is often, too, an acceleration of the cyclic period 

 upon the fossular period, since the first cyclic septa may appear before 

 the last fossular.^ 



Recent corals start with six or twelve septa appearing simultaneously, 

 doubtless owing to the fact that, on becoming fixed, the larva already 

 has twelve or more mesenteries.^ It would be impracticable for the 

 complete appearance of the septa, like the mesenteries, to be pushed 

 back by tachy genesis into the larval period, for the little coral, 

 embarrassed with a cycle of septa, Avould be too straight laced to 

 swim in comfort. The septa are therefore suppressed until they may 

 conveniently appear, and then they arrive in numbers appropriate to 

 the stage of development attained by the soft parts. The septa, then, 



^ Carruthers, " The Primary Septal Plan of the Eugosa " : Ann. Mag. Nat. 

 Hist., ser. vil, vol. xviii, pp. 356-63, 1906. 



-For example of these points see Duerden, "The Morphology of the 

 Madreporaria. — VI. The Fossula in Eugose Corals": Biol. Bull., vol. ix, 

 pp. 34-7, figs. 7-10, 1905. 



^ e.g. in Astroides. Lacaze-Duthiers states : " C'est lorsque le nombre 12 

 des divisions du polype est atteint . . . que se produisent les depots calcaires 

 a peu pres en meme temps et de meme dans toute la zone " (" Developpement 

 des Coraillaires, Deuxieme Memoire " : Archives de Zoologie experimentale et 

 generale, vol. ii, p. 328, 1873). 



