93 



pneuniatophore {pn.) having attained a length of lo mm., a breadth of 6 mm. It is very muscular, 

 shows lines and furrows on its top which are simply thickenings of its outer wall. As it is 

 flattened apically, it seems very much larger than in Cat. loA. We will speak later of the 

 bunch of appendages situated to the right side of the flattened pneumatophore. 



The lower part of the nectosome as in Cat. lo A. shows itself as a well-marked muscular 

 wall underneath the pneumatophore; it shows the same linear radial elevations as in Cat. lo A. 

 and seems larger in comparison to the breadth of the pneumatophore. But the appendages of 

 the siphosome being less voluminous and abundant, may perhaps give the impression of its 

 unusual dimensions. No nectophores are left. 



The siphosome shows more or less the same appendages as Haeckel 88 b gives on 

 PI. \\\ fig. 15 for Rhodalia miranda. Not one appendage can be distinguished from the other; 

 they constitute probably much contracted siphons and gonostyles. This specimen, preserved in 

 formaldehyd 47^, proved on section to be histologically even worse than the two others 

 preserved in ale. 90°j^. 



As one can judge already by comparing the three sketches of the three specimens, there 

 is no aurophore and we came to the conclusion that the bunch of appendages, visible in two 

 sketches on the right side of the pneumatophore (in one on the left side Cat. 10 B.), could only 

 be the zone of proHferation, the spot where the appendages, either nectophores or appendao-es 

 of the siphosome, or both, begin their very first stages of development. 



Not finding any appendage having the faintest resemblance to an aurophore on the other 



side of this bunch of elongate appendages, we decided to make sections of the three specimens, 



so as to get a clear idea of the real position of both pneumatophore and young appendao-es. 



The three specimens were therefore stained with alum carmine and cut into sections of 



1 5 p. thickness. 



The respective numbers of sections made, were 



for Cat. 10 A 1143 sections, 



for Cat. 10 B 394 sections, 



for Cat. 22 1597 sections. 



The sections were all made in a longitudinal direction; so, that in the sketch made ot 

 10 A. (PI. XVni, fig. 137) the whole specimen ought to be turned a little more to the riaht, 

 the bunch of appendages thus coming more out of sight, the nectophores coming more in 

 advance. Placed thus the section was made in the direction of the arrow. 



In main points the sections of all three specimens agree; in minor details there are of course 

 difterences, but not important enough, to need different specific denominations. Histoloo-ically the 

 material is very unsatisfactory, whole layers being destroyed. Especially Cat. 22, the formaldehyd 

 specimen hardly shows any well-defined cells ; the different cavities are filled with their remnants. 

 After careful research and examination of the 1143 sections of Cat. 10 A. (which proved to be 

 the best preserved and the best specimen fit for section) we were able to reconstruct the whole 

 and we have come to the result that not only in Archangelopsis, but in all Angelidae 

 (Haeckel's Anronectae) there is no particular medusoid structure, no aurophore 

 to be found in them. The so-called aurophore constitutes externally the 



