32 UmTACRmUS: ITS STRUCTUEE AND RELATIONS. 



form; (2) their intercalation in the monocydic form; or (3) the coalescence 

 of infrabasals with the centrale. 



To either of the first two, the orientation of the centrale presents a mor- 

 phological difficulty so great that I do not see how they can be entertained. 

 A resorption of the infrabasals in the dicyclic form would leave the centrale 

 interradial in the resulting monocyclic base, which is contrary to the fact; 

 the intercalation of infrabasals in the monocyclic form would leave a radial 

 centrale in the resulting dicyclic base, which is equally contrary to the 

 fact, — unless, during this process in either case, the centrale should at the 

 same time undergo a revolution from interradial to radial, or vice versa^ or 

 should be reduced by the truncation of its angles until only an interradial 

 pentagon should be left. That such a movement or reduction may have 

 taken place in the larval stage — governed by causes we know nothing 

 about — need not be doubted. But that they could have occurred among 

 the hard parts of the Crinoids, after they had reached any stage of growth 

 such as we find in the fossil condition, is contrary to everything hitherto 

 known touching the mode of growth of these organisms. There are a few 

 cases, as before stated, in which the infrabasals are very minute, not large 

 enough to affect the position of the centrale by crowding it, — but there is 

 no sign of any revolution or shifting ; and the effect of the truncation of 

 the angles in these cases is to make the centrale a decagon, — practically 

 almost round. 



A coalescence or fusion of the infrabasals with the centrale would not be 

 subject to the same objection on the ground of orientation. The resulting 

 plate would be radial, as it should be in a monocyclic form. But there are 

 other difficulties in the way of this method equally insurmountable, in my 

 opinion. Among the 268 dicyclic specimens examined by me I can scarcely 

 point out one in which the coalescence of the infrabasals and centrale would 

 produce a plate having the same outline as the centrale has, or which would 

 not be entirely distinct in shape from the centrale in most of the 275 mono- 

 cyclic specimens in the collection. The centrale is never stellate, while the 

 infrabasal ring is almost always more or less so. 



Hence in order to successfully transform the infrabasals plus the centrale 

 into a centrale alone, the process must change not only the outline of the 

 infrabasals, but that of the surrounding plates as well, — which would mean 

 not only the fusion of infrabasals with centrale, but a modification of the 

 shape of the basals also. The relative size of these plates is also against it. 



