ANTHOPHYSA ROSEA. 297 



muscular bands arching over the pneumatophore which Fewkes ('88b) described 

 as the "hood, elsewhere unknown among the physophores." And it is the 

 region near the apex where they arise and where new bracts are budded, which 

 was the nipple-like structure noted by Lesson ('43, p. 496) in his account of 

 Angela cytherea. Bedot has already corrected Haeckel's statement that there 

 are invariably four muscle-bands in each group. In his specimens he noted four 

 or five, but in the large individuals collected by the "Albatross" there are 

 usually six, sometimes seven. In large specimens there are eight or nine groups 

 of muscle-bands separated by as many naked zones and, in addition, there are 

 many young lamellae along the ventral face of the pneumatophore, i. e. at the 

 zone of proliferation. When expanded the lamellae take the form of broad, 

 thin sheets (Plate 23, fig. 3). An accurate account of their muscular structure 

 has been given by Bedot (: 04, p. 6, pi. 1, fig. 14). 



The bracts themselves, in their spatulate form and in the presence of a 

 prominent tooth on either margin (Plate 23, fig. 3), agree with Fewkes's 

 figures. They have five longitudinal nematocyst ridges. They change in 

 outline with growth, j'ounger ones being short and broad, older ones longer, 

 narrower and proportionately shallower at the region of attachment. 



Cormidia. The zone of proliferation of the cormidia lies on the ventral 

 face of the corm immediately below the region where the bracts are formed. 

 Haeckel believed that he could discern evidences of bilateral symmetry in this 

 region, but although I have examined the "AJbatross" specimens with care, 

 I am unable to corroborate him. 



The arrangement of the cormidia on the siphosome in the specimens studied 

 does not agree with Haeckel's ('88b) account which has so far formed the chief 

 basis for our knowledge. According to him the conxiidia in the Anthophysidae 

 in general ('88b, p. 271), are arranged symmetrically in a flat spiral which is 

 twisted around the broad base of the shortened vascular stem. The condition 

 in the "Albatross" specimens, easily seen in two examples with the corm denuded 

 of its appendages (Plate 23, fig. 5), is quite different. Instead of two rows of 

 siphons there is a single row, not in a spiral, but in a straight line extending 

 from the ventral side over the lower surface of the corm to the dorsal side. The 

 youngest siphons, of course, lie on the ventral face, the oldest one on the dorsal 

 face just below the level of the bracts. Moreover, the eight or nine well-devel- 

 oped siphons alternate in position with the groups of bracts. The gonodendra, 

 situated as observed by Chun ('97b) in pairs, 9 and cT, immediately below the 

 bracts, lie opposite the latter, and thus alternate with the siphons. The very 



