558 U. Green 4" C. D. Sherboni— 



of recent corals already appear in cycles, and other complications mark 

 their periods. 



It is easy to see how, by tachygenesis, a period of, for instance, 

 six separately appearing mesenteries (or septa) may be condensed 

 into one of three stages where the septa appear in pairs ; and this, in 

 turn, to a single cycle. But tachygenesis generally is shown by the 

 greater condensation of the earlier stages. In the case considered, 

 liowever, it is the earlier periods that are the least condensed and the 

 later progressively more so. Again, one period may tend to overlap 

 a previous period, as, for instance, the cyclic anticipating the end of 

 the fossular period in Engosa. This is quite in keeping with the 

 trend of tachygenesis. 



Bernard's evolutionary units are themselves the totals formed of 

 units of a former period. So it is in ontogeny. The post-embryonic 

 life-history of a coral involves a radially symmetrical tree-swimming 

 stage (very much reduced in recent corals'), a stage of bilateral 

 symmetry, shown by tlie elongate shape of the mouth and by the 

 position and arrangement of the muscle bands of the mesenteries, and 

 a secondarily radial ('biradial') stage in which a radial symmetry 

 tends to be resumed. Similarly, Zaphrentis (to take the example of 

 septal development we have already cited) presumably during the 

 free period of its life passed through a stage with no septa, then, on 

 becoming fixed, one in which two, four, and finally six septa are 

 bilaterally arranged, to a stage which has become again radially 

 symmetrical by tlie equalization in length and spacing of the septa. 

 Tlie fossular period is the anabasis of a second wave of symmetry, 

 bringing the radial up to tlie acme of bilateral symmetry, and the 

 cyclic period its katabasis down to a secondarily radial symmetry. 

 That is, the first-described periods of septal arrangement themselves 

 become stages in periods of alternating bilateral and radial symmetry.^ 

 A similar rhythmic process of periods has been demonstrated in the 

 costal development of Parasmilia.^ Thus Bernard's periodic rhythm 

 of units in animal evolution as a whole finds corroboration in the 

 facts of coral development, and, as a point of view, is useful for 

 estimating the relations of various phenomena in phylogenetics. 



IV. — Note on the Poli.ueian-Tkkwavas Coast Section, Cornwall. 



By IJPFiELD Green and C. Davies Sherboen.'* 



1 SSUMING that the Lizard Peninsula is a pre-Cambrian complex 



j\_ (Flett), and that the Hornblende Schists of Pollurian have 



nothing to do with the northern sediments, it will be well to call 



^ Lacaze-Duthiers, op. cit., pi. xiii, fig. 9. 



- These alternating periods were fully appreciated by Duerden, " The 

 Moi-phology of the Madreporaria. — V. Septal Sequence " : Biol. Bull., vol. vii, 

 pp. 99-101, 1904. 



■* Lang, "Growth-stages in the British species of the Coral genws Parasmilia'''': 

 Proc. Zool. Soc. Lend., 1909, pp. 287-90. 



■* This paper was written in May, 1910. In the previous month the Director 

 of the Geological Survey kindly allowed us to look at the unpublished map, for 

 which favour we undertook to give priority to the Survey. The Survey Memoir 

 having now appeared, we print our views as then written. 



