15 



ridges is total and tlie one ridge composed of these two has a convex course. They form with 

 the opposite ones, the outer lines of the hydroecial canal. 



Facet A' is (as may be concluded from the description of its two neighbours D and D') 

 a very doubtful facet, as we may never speak about a facet when we mean a cavity. 



The posterior points, the bases of the longitudinal eight ridges are all very slightly 

 serrated; this is only to be seen by magnifying about 30 times. 



This schematical description should be considered as a reconstruction of a nectophore, 

 which through bad preservation (it is very much altered especially a little waj- under the apex) 

 does not show all the particularities in a normal way. 



We spoke of the hydroecial canal which is open from the base to 5 mm. under the 

 apex (length of ridge a^^') and we will now try to make clear what the interior of the nectophore 

 is like (PL I, fig. 9). First of all the contours, the exact position and shape of the nectosac 

 are absolutely invisible. The aperture has probably been between ridge e, f, f as p- and <r' 

 belong more to the hydroecium, that is to say the antero-ventral part of the nectophore. At 

 the basal and ventral part of the nectophore (immediately below the point where the lateral 

 walls of the hydroecium grow divergent) there is a remarkable structure, which magnified shows 

 the following shape (PI. II, fig. 12). 



There are three canals diverging from a central membranous part. The anterior one 

 (PI. II, fig. 12 a.c>) goes up to the apex of the nectophore, grows thinner and thinner and 

 disappears, then the contours are still to be seen, though very indistinctly and it seems as if 

 the canal widens first and then narrows again and the canal gets as thread-like as in the 

 beginning and disappears gradually near the top, the exact position not to be found. 



The right lateral one (PI. II, fig. 12 r.c.) loses itself almost immediately, the left lateral 

 one (PL II, fig. 12 /.<;.) passes very soon under another appendage and ends abruptly only just 

 appearing on the other side. 



This appendage, so strangely divided into three, may we call it a somatocyst? 

 The stem is not so enigmatical. Its appendages consist of a multitude of buds at the 

 basal undivided part of the so-called somatocyst and the true stem begins at the back side of 

 this same structure. But at this point all the appendages except a few buds, placed on regular 

 intervals one from the other, disappear. In fact every part of stem and appendages is 

 remarkably incomplete. 



Inferior nectophore. (PL II, figg. 13, 14, 15). 

 The loose inferior nectophore has the same outward appearance as the superior one, 

 as concern its elasticity and the brownish colour of the prominent ridges. 

 Its length is 22 mm., its breadth 7 mm. 



At its top we see three ridges, a difterence with the upper nectophore where there 

 were four. 



These three ridges bifurcate on difterent heights. The dorsal ridge (textfig. 1 9) of facet 

 A {a') divides itself about 5 mm. from the top and the two side ridges {a and a') orow divergent 

 gradually. At the base is ridge d which is curved, so that the two side ridges end pointed. 

 The shape of this facet is much the same as the one in the upper nectocalyx. Through the 



