406 MR. A. W. WATERS ON THE RELATIONSHIPS 



zoarmm floats with the apex at the bottom, and we are awaiting proofs as to 

 how this takes place, as it is difficult to understand. Whitelegge *, in a 

 postscript, says that he has had C. pTiilippinensis under examination, and that 

 " nearly every specimen possesses a pair of tubular filaments inserted on each 

 side of the zoarium " ; also he thinks " the tube appears to grow out of an 

 aviculurium." Will it not be found that they grow from the semilunar slit? 

 Harmer t, speaking of the Selenariadpe, says that he has some evidence 

 that they may be attached to the ooze by means of very delicate flexible 

 rooting processes, but he does not indicate the species or genus to which he 

 is referring. From fossils and dried specimens of Conescharellina the con- 

 clusion come to independently, is that there have been radicle processes, but 

 in Cupularia, Selenaria, and Lunulites there is no indication of anything 

 similar. 



Classification. 



We may now turn to he classification, as these investigations were made 

 to find a natural one, and certainly we are brought up against a most 

 difficult problem. Generally in Cupularia, Selenaria, and Lunulites there is 

 a thick under surface, through which long tubes may pass, or there may be 

 a series of chambers ; further, the arrangement of the back as well as the 

 front is radial, all of which seems to be quite different from anything known 

 in other .genera. In many cases in the Cheilostomata, such differences as 

 whether the zooecia are uni- or bilateral, or whether they are adnate or 

 erect, are purely zoarial characters, of no or but slight value in classification, 

 but the characters on the under surface, now dealt with, are not zoarial in 

 the same sense, but are in connection with the zooecia. 



Looking at Cupularia and Selenaria with their similar opercula, similar 

 lower surface, in most cases with central small zooecia, often closed, with 

 vibracula in both genera, there can be no hesitation in placing them in the 

 same family; in Lunulites*, the lower surface is radial and thick, long pores 

 pass through it as in the last two genera, the opercula are similar, but 

 there is a more solid calcareous frontal wall without perforations, though it 

 clearly belongs to the same family. Lunulites, as generally understood, 

 requires separating into several genera, as already indicated. 



ConescJiarellina differs from Cupularia and Selenaria in many particulars, 

 such as the shape of the separable operculum, the reversal of the position of 

 the operculum, the semilunar slit, absence of radial under surface, although 

 the zooecia are placed radially ; on the other hand, in the lower part there are 

 a number of vertical chambers, which seem comparable with the horizontal 

 ones of Cupularia canariensis, and it would be strange that species not 



* " Australian Polyzoa," p. 317. 1887. 



t Presidential Address, Brit. Assoc. Zool. Section, p. (9). 1908. 



