DISCUSSION AND COI^CLUSIONS 365 



have all been shown to be in stfict agreement with undoubted Bryozoa. 

 The presence of the protoecium alone is practically conclusive on this 

 point, since it represents a definite stage in the metamorphosis of the 

 larva, not represented in the corals. In the latter the development from 

 the instant when the planula becomes sedentary is direct. Again, the 

 restriction of budding to the neck region oT the ancestrnhi finds no 

 counterpart in the corals, where buds come oif symmetrically all around 

 the primary individual of the colony (cf. figures 9 and 12). This feature 

 of coral budding is beautifully shown in the figure of Pleurodictyum, 

 reproduced herewith, and still more clearly in the diagrams of Michelinia 

 given by Beecher (3). It also characterizes the recent red coral (Coral- 

 liuni), according to Lacaze-Duthiers (8), and Renilla, as described by 

 Wilson (15). Bernard mentions symmetrical budding in Turbinaria 

 (4) and in Montipora (5). No doubt it occurs generally in corals, as 

 the construction of the coral polyp would lead us to expect. '^J'he litera- 

 ture of corals is, however, for the most part singularly silent on the 

 subject of early colonial development. 



In view of this conclusive evidence from development in regard to the 

 systematic position of the Trepostomata, it is scarcely necessary to re- 

 vamp the evidence variously presented by Ulrich (12), Lindstrom (9), 

 and others on this point, or to review the adverse opinions of Nicholson 

 (10), Sardeson (11), and others. Lindstrom's views do receive a new 

 interest from the present studies. While I do not believe with him that 

 there is any direct relation between Ceramopora and Monticulipora, he 

 should receive the credit for having attacked the question of systematic 

 position from the right direction and for having made suggestions of 

 great value. Had his suggestions been carefully followed up, the mystery 

 of the Trepostomata would ha\(' been cleared up long ago. 



Ulrich has given a powerful presentation ol* the evidence of mor- 

 phology bearing on the systematic position of the group, and recent 

 studies of Bassler (2) and myself have materially strengthened this class 

 of evidence. 1 have, for example, found that communication pores are 

 present in a considerable Jiuniber of genera, as Dckaijia, Ihiiosloma, 

 Hythopora, Cailopora, Eridoirjjpa, Montimlipora, Nicliolsonella, and 

 Peronopora in addition to the long known cases of Homolrypa. The 

 evidence on this point is now in press. 



A word may be said with reference to the views of Sardeson. First of 

 all, he stakes, his case on the proposition that the l^repostoniata and 

 Cryptostomata are very intimately related. This is ]x^rhaps more than 

 the most devoted students of the Bryozoa would be willing to gi*ant. 

 But, accepting it at its face value, his case for the coral affinities of the 



