148 



calicoblastic ectoderm merges gradually into the external 

 ectoderm, which only differs from that of the adult in being 

 less vacuolated and in the comparative absence of mucous 

 and granular gland cells, 



A later stage, which I have also examined by serial 

 sections, is that of a still attached polyp, long axis of calicle 

 5 mm., and short axis 2*5 mm. It has 48 septa, of which 12 

 alone fuse in the axial fossa, the stage being hence a little 

 earlier than fig. 15. The stomodoem is an extremely 

 shallow invagination. There is as yet no trace of generative 

 organs, the polyp being otherwise, even in its minute 

 anatomy, similar to the adult. The tissues surrounding the 

 septa fuse across below the mass of trabeculae in the axial 

 fossa, thus extending to the very bottom of the corallite. 

 The minute examination has shown no cause which can in 

 any way produce the breaking-off of the calicle from its 

 stalk. Some of the tissues would undoubtedly seem to remain 

 in the pedicle, but, as already mentioned, I have found no 

 evidence of its possible future growth to form a fresb 

 corallite. 



9. CONCLUSIONS RELATING TO THE GENUS 

 FLABELLUM. 



In the preceding pages an attempt has been made to in- 

 vestigate and determine as far as possible the variation in' 

 the skeleton of two very dissimilar species of the genus 

 Flabellmn. The variation is of two kinds, specific and 

 normal.* The latter may be best seen by reference to the 

 synonomy. It is in these two species enormous, and the 

 study affords some data for considering the possible and 

 probable variations not only in the same but in other genera. 

 In F. rnbntin there apjDear to be three distinct specific or dis- 

 continuous variations, between which the intermediates do 

 not form more than \-i per cent, of the total number of 

 specimens. This, so far as I am aware, is the first suggestion 

 of the presence of distinct varieties in the ]\Iadreporaria. 



The question as to whether the skeleton in the axial 

 fossa represents a true columella, built by deposition on the 

 basal plate, or is merely formed by trabeculae from the 

 septal edges appears to me of morphological and generic 

 importance. The development shows clearly that there is 

 no such true columella in Flabc/In/u.'^ Trabeculae 



* See '* Her«dity, Diffttrentiation and other Conceptions of Biology," by W. 

 Bateeon, Proc. R. Soc , vol. 09, p. 193, 



tin this conneL'tion it is interesting to note that series of sections of even 

 moderate sized, attached corallites show at different heights the develoi nient of 

 the septa equally as well as separate corallites in different stages. There is in 

 F. rubiiim no obliteration of the axial foisa by coralluni betweeu the earliest tra- 

 beculae that arise and the basal plate. 



