306 DROMALIA ALEXANDRI. 



The portion of the pneumatosaccus which is invaginated to form the inner 

 wall of the aurophore, is connected with the outer wall, the pneumatocodon, 

 by a series of radial septa, just as it is in Rhodalia and Archangelopsis. Trans- 

 verse sections (Plate 24, fig. 7) in the mid-region of the aurophore show from 

 12 to 16 septa.. Both entoderm and supporting layer are concerned in the forma- 

 tion of these septa, just as they are in the septa which subdivide the peri- 

 cystic space surrounding the distal portion of the pneumatosaccus in other 

 Physophores. Of course the cavity of the aurophore is nothing more than 

 an evaginated portion of the pericystic space. 



Appendages of the aurophore. These are simple papilliform structures; 

 externally smooth walled. They are hollow, and their cavities communicate 

 freely with that portion of the general pericystic space lying within the auro- 

 phore. Their cavities open to the exterior by terminal pores (Plate 24, fig. 9). 

 The external layer of ectoderm is composed of tile-shaped cells except at the 

 distal extremity where the cells are higher; the lining entoderm layer, which is 

 of course continuous with the entoderm of the pneumatocodon, is much thicker 

 and composed of columnar cells with very conspicuous nuclei. The two cell- 

 layers are separated from each other by a well-developed supporting layer. 



Zone of proliferation of nedophores and cormidia. Two diametrically opposed 

 accounts of the zone of proliferation have been published. According to Haeckel 

 this region is opposite the aurophore ; on the other hand Lens and Van Riems- 

 dijk maintain that the aurophore itself is the zone of proliferation, its papilliform 

 appendages being the young cormidia and nectophores. The conditions in our 

 specimens of Dromalia give a very decided answer to this question, entirely 

 bearing out Haeckel's statement. This is of course a question of great theoretic 

 interest, since on it depends the interpretation of the dorsoventral symmetry 

 of the Rhodaliidae. Therefore it is very fortunate that the specimens ai-e so 

 large, and the actual appearances so easily shown by photographs, that there is 

 no difficulty in tracing the location of the various structures outlined below. 



A view of the conn, facing the side on which the aurophore is located 

 (Plate 23, fig. 7) shows that directly below that structure there is a vacant 

 zone, extending the whole length of the nectosome, on which neither cormidia, 

 nectophores, nor buds of any kind are to be seen. Neither are there any indica- 

 tions that any such have been detached. On the contrary, in all the specimens 

 the surface in this region is perfectly smooth. Flanking this naked zone on either 

 side, are the longitudinal muscular ridges to which the nectophores were attached 

 in life. The presence of the naked zone, and its relation to these muscular 



