290 MR. W. D. LANG ON GROWTH-STAGES [ Mar. 2 
presented in its baldest and most diagrammatic aspect, the extra- 
ordinary regularity in the genetic sequence of the costal character 
may be lost sight of in the ‘greater or less amount of development 
and condensation of the various characters, and in certain irregu- 
larities in their actual sequence due to environmental and other 
causes. One such case will be next considered. 
III. RecaprruLation oF GRowTH-STAGES AT REJUVENESCENCE. 
The life-history of the costa in an actual specimen of each of 
the described forms of English Chalk Purasmilia will be given 
later on, that of the type-specimen when available, and in the 
cases where this is not in the British Museum, one of Duncan’s 
figured specimens if possible. In two cases specimens are de- 
scribed which have no historical value in point of view of identi- 
fication. But before describing these actual cases, an important 
modification in the regular sequence of growth-stages must be 
noted, Periodically in the growth of the corallum constrictions 
arise, marking points in the life of the coral when fresh growth 
took ‘place after a period of rest. Such revivals of erowth are 
known as Rejuvenescence*, a term very happily introduced, 
because the coral literally renews its youth at these points, and 
expresses the renewal by recapitulating earlier growth-stages in a 
condensed form; at least this is so in the costal characters of 
Parasmiliay. A glance at the actual costal life-histories given 
hereafter (see este: figs. 39-46) will show the extent of this 
recapitulation. 
There is no reason for supposing that the recapitulation of 
younger stages at rejuvenescence is confined to the two characters 
observed, nor only to the genus in question; rather, it probably 
extends to the internal structure and is of general application. 
The anomaly then in Caninia cornucopie Michelin, described by 
Carruthers and mentioned early in this paper, may be explicable 
by this phenomenon. The coral had developed as far as the 
nystiana-stage, had next entered a period of rest, and on re- 
Juvenescence, with renewed growth, had begun in the dumonti- 
stage, recapitulating its grow uh from that stage to the nystiana- 
stage again, and then proceeding, perhaps through further 
interruptions of rejuvenescence and consequent recapitulations, 
to the edwardsiana-stage. 
TV. Bernarp’s THeory or THE NATURE OF REJUVENESCENCE 
It is difficult to dissociate rejuvenescence from branching when 
viewed in connection with recapitulation of earlier stages. Robert 
« Verjtingung-process, ” Milaschevitch, 1876, Palaontographica, vol. xxi. p. 194-5. 
3 ates Tomes, 1882, Quart. Journ. Geol. Soc. » p. 409; 1884, op. cit. 
p. 363; 1886, Geol. Mag., p. 394; 1888, op. cit. p. 208 ; 1899, op. cit. p- 305. 
iP It is the phenomenon here described that Bernard refers to iia a footnote, 
British Museum Catalogue of Madreporarian Corals, 1906, vol. vi. p. 
