Madreporarian Genus Tarbinaria. 507 



merging the polyps. In these cases it is not infrequent to 

 find such places very thickly studded with minute polyps. 

 So numerous are they that active budding can alone account 

 for them. Without having much actual information to give 

 as to the real origin of these adventitious buds, there seems 

 to be little doubt that while normally the polyps of the 

 Turbinarian colony merely produce their single ring (axial 

 polyp) or portions of a ring (radial polyps), when submerged 

 by coenenchyma they may continue to put out buds. It is 

 true that there is evidence to show that submerged polyps, if 

 not too deeply covered, may break through again. But when, 

 for instance, the axial and earliest-formed radial polyps are 

 completely submerged in the bottom of the cup by an enor- 

 mous thickness of coenenchyma, which is nevertheless thickly 

 studded with minute buds, we can hardly escape from the 

 conclusion that these are due to secondary buddings of the 

 submerged polyps. Further, some glomerate forms, which are 

 characterized by an enormous thickness of the coenenchyma, 

 show a tendency to secondary budding of the polyps, which 

 can be easily seen. In this connexion it is worth remarking 

 that the limitation of the buds to one ring or to a portion of a 

 ring is probably a derived condition, while the power of pro- 

 ducing an indefinite number of buds one above the other, 

 shown in Ifadrepora, is the more primitive. That this limita- 

 tion actually exists there is abundant evidence, as may be 

 gathered, for instance, from the very uniformity of the 

 earliest cup- or disk-shape of the corallum, and again from 

 sections of glomerate Turbinarians (fig. 7), which show enor- 

 mous thickening of the coenenchyma, with corresponding 

 lengthening of the polyp-cavities, often without any traces 

 of secondary budding. But that this adventitious budding 

 undoubtedly exists and plays a part in the ultimate forms of 

 the coralla, perhaps as a return to more primitive conditions, 

 there can be no doubt. It seems, however, to play but a 

 subordinate part, and in discussing the morphological basis of 

 the classification of the genus it may be temporarily ignored. 

 The two prime factors above mentioned are sufficient. 



Variations in the Form of the Cup due to subsequent Growth. 

 — Of the three initial forms, dependent in the first place 

 on the angle at which the first buds leave the parent polyp 

 (or, perhaps, on the curving of the daughter-polyp outwards), 

 the cup-shape, the disk-shape, and the hemispherical, the first 

 is that most liable to great modification during the subse- 

 quent growth, while the last is naturally that in which form- 

 changes are least to be expected. 



The second or disk-shape, inasmuch as it hovers on the 



35* 



