Madreporarian Genua Turbinaria. 503 



might be obliterated. It seems to me easier to believe tiiat 

 the costffi have secondarily lost their connexions with the septa 

 than that in those Turbinarians in which they are regularly 

 continuous the arrangement has been secondarily accjuired. 

 It is dilHeult to believe that it is not the primitive arrange- 

 ment. 



77<e Buddinf) of the Axial Polyp. — The axial polyp in the 

 stalk of a minute Turbinarian colony buds lalevalhj^ the buds 

 forming a simple ring round the axial polyp. The new- 

 polyps radiate upwards and outwards around the axial 

 polyj), which either does not grow any more or else grows 

 very slowly. It is obvious that this single ring of daughter- 

 polyps, cemented together by coenenchyma, which appears to 

 stream down round the axial polyp, forms, togetlier with the 

 axial l)olyp, a stalked cup. Tiiis cup, according to the angle 

 the ring of buds makes with the axial polyp, and also with 

 the regularity of the ring, may vary considerably in shape. 



It is to be noted that the budding from the axial polyp is 

 lateral, as in Madrepora, and not basal, as Dana, following 

 Ehrenberg's " Stolonformation," described. The error of 

 these distinguished naturalists in this respect was most natural, 

 and arose from the fact that they examined only sections of 

 fronds, not of young cups. A section of a frond alone cer- 

 tainly seems to show at lirst sight that the budding is basal. 

 This mistake led Dana to place the genus Turbinaria [Oem- 

 mijwra) in a different tribe from that of the genus Madrepora. 

 In the former the budding was thought to be basal, while in 



the latter it was lateral, whereas in both cases the buddin<»- 

 11 ° 



18 lateral. 



Comparison between Turbinaria and ^ladrepora. — This 

 central parent polyp of the Turbinarians appears to me, then, 

 in every way comparable with an axial j)olyp of a typical 

 Madrepore, and the fundamental difference between the Turbi- 

 narians and the Madrepores is due to their different methods of 

 buddinc/ [cf. PI. XIX. figs. 3, 3 a). 



In tlie iMadrepores the buds appear laterally on the wall of 

 the axial polyj), and the higher tiiis grows the more buds are 

 produced, till each axial polyp is thickly crowded with 

 daughter-polyps, radiating out from it in all directions. 

 Nutrient fluids stream down the channels between the costaj 

 forming new layers of coenenchyma round the lower portions 

 of the stock, which may increase in thickness so greatly as to 

 submerge the lower and first-formed buds. h the stem 

 branches, one of the buds becomes in its turn an axial polyp 

 and gives off buds ; otherwise the buds do not, as a rule 

 themselves again bud. In the Turbinarians, on the other 



