138 MR. G. H. FOWLER ON A NEW PENNATULA. [Feb. 21, 



The appearances seen in a transverse section of the immature 

 autozooids are of some interest, and bear out the conchisions of 

 Hickson {loc. cit.) and of Wilson (" Development oi' Renilla," Phil. 

 Trans, vol. 174, p. 723). These immature polyps are not as yet 

 provided with tentacles, but possess stomatodsea and the usual eight 

 mesenteries. Of these latter, the two axial (dorsal) mesenteries, in 

 the youngest autozooids, as in the mature siphonozooids, alone 

 exhibit mesenterial filaments of the characteristic bilobate shape ; 

 the cells of which the lobes are constituted contain deeply staining 

 nuclei, and histologically agree with the ectoderm ot the stomato- 

 dseum from which they are derived. As was shown by Hicksou, no 

 siphonoglyphe is recognizable in the mature autozooids o\' Pemiatula ; 

 but in the youngest autozooids, which are much compressed at right 

 angles to the usual direction {i.e., in an axial-abaxial plane), I find 

 that the whole of the abaxial side of the stomatodseum is clothed by 

 very long columnar cells, with numerous deeply staining nuclei, and 

 bearing long stout cilia, these ajipearances being entirely character- 

 istic of a siphoiioglyphe ; this region is marked off from the remain- 

 ing three filths of the stomatodseum by a ridge on either side, and 

 constitutes indisputably a true siphonoglyphe. The somewhat older 

 autozooids, which are nevertheless hardly mature, exhibit filaments 

 on the lower six, as well as on the axial two, mesenteries ; the^e 

 filaments are ol endodermic origin, and consist merely of an aggrega- 

 tion of pyriform cells, resembling those figured by Wilson (/. c. 

 pi. Ix. figs. 145, 146), but more swollen. Proportionately to their 

 age and development, the siphonoglyphe of the autozooids becomes 

 less and less marked, till in the fully mature polyps no trace of it is 

 to be recognized. 



This occurrence of a siphonoglyphe in the ontogeny of the auto- 

 zooids is a point of some considerable interest. Knowing so little as 

 we do of the stin.ulus that cautes a developing ovum to recapitulate 

 its ancestial history, we should hardly be justified in asserting that 

 such recapitulation might not also ociur in asexual reproduction. 

 The colonial ancettor ol the OLtactinise (x'\lcyonaria) resembled the 

 existing Alcyoniimi'm the absence of dimorphism and the possession 

 of a siphonoglyj he ; morphological differentiation, correlated with 

 division of labour, resulted, among such descendants as the Penna- 

 tulida, in the production of autozooids (nutritive and sexual polyps, 

 de\oid of a siphonoglyphe) and of siphonozooids (circulatory or 

 resj)iratory polyjs, incapable of generation and of nutrition, but 

 provided with a strong siphonoglyphe in order to effect the circula- 

 tion of "chylaijueous fluid" tlirough the colony). If then, as appeals 

 to n;e to be the case, the very considerable size and number of the 

 siphonozooids contrasted with the paucity of the immature autozooids, 

 together with the loss of the siphonoglyphe as maturity is attained, 

 jjonit to the functional uselessness of this organ in the ytung buds, 

 we are driven to the conclusion that we have here to deal with a case 

 ill which asexi al ontogeny is repeating phylogeny. 



As was desciibed by Wilson (Mitth. Zool. Sta. Ncapel, v.) in the 

 buds of Reiulhi, the two tixial nu senterial filan.ents appear before 



