6 2 CERIANTHARIA. 



as shewn also in its shape into two spirocyst-glandular portions, separated from each other by a 

 deep groove. This groove here consists exclusively of covering cells which are considerably lower 

 than the other ectoderm cells in the median streak, but bear longer cilia than they, in short this 

 depression is a ciliated groove. In Arachnanthus oligopodus and sarsi as also in Arachnactis albida 

 the groove between the two portions of the median streak is also strongly marked, but its ectoderm 

 is higher than in A. lobiancoi (Textfigure 8). In C. mcmbranaccus, lloydii, valdiviac, P. solitarius, mul- 

 tiplicatus, maua, C. americanus we may also trace this subdivision of the median streak, in the fact 

 that in these cases also a depression, though not so deep as in the cases previously mentioned, runs 

 down the middle of the median streak (the depression is plainer in C. lloydii, valdiviae, P. multipli- 

 catus, maua and C. americanus than in C. membranaccus and P. solitarius]. In this depression, and 

 consequently in the mid-plane of the mesentery, the ectoderm cells consist chiefly of covering cells 

 with strong cilia; it is noticeable however that between the supporting cells especially in C. nicmbra- 

 naceus 2 ) and C. lloydii scattered gland cells and nematocysts are interposed (Textfigure 9). In Botru- 

 cnidifer norvegicus on the other hand (Textfigure 10, PI. 5, fig. 9) no trace is found of such a differen- 

 tiation of the median streak. It has in transverse section a tongue-like appearance and in outline is 

 suggestive of a cnido-glandular tract. This applies also to the mesogloea, which is not in this species 

 divided into two lamellae, as is the case with the species mentioned above. The median streaks thus 

 shew in Ceriantharia grades of transition from a simple undivided structure (in Botrucnidifer norvegicus) 

 to a separation into two well marked portions running parallel and divided by a differentiated ciliated 

 groove (in Arachnanthus oligopodus and Arachnactis albida and lobiancoi]. 



In the aboral part of the ciliated tract region there is a gradual disappearance of this sub- 

 division of the median streak in the series C. mcmbranacc^^s C. americanus, so that the median streak 

 here assumes the same appearance as the median streak in Botriicnidifcr. In Arachnactis lobiancoi 

 and albida and also in Arachnanthus the median streak shews on the contrary a bi-partition through- 

 out the whole course of the ciliated tract region, so that a well marked ciliated groove is always 

 found between the two portions of the median streak. 



We now pass on to consider the ciliated tracts which consist only of ciliated supporting cells. 



Just before the stomatodaeum terminates, we have seen that its longitudinal furrows acquire 

 longer cilia than higher up. When the ridges of the stomatodaeum at the transition into the 

 filaments divide in the middle of the longitudinal furrows, the ectoderm in the bottom and a portion 

 of the sides of the furrow on either side of the ridges passes over into the ciliated tracts. The 

 ciliated cells in the sides of the furrows, through the flattening out of the ridges, come to lie on the 

 underside of the median streak, whilst the ciliated cells at the bottom form a process, supported by a 



J ) Where the stomatodaeum passes over into the filament the median streak has often not retained this flattening 

 and differentiation. Thus in the case of C. lloydii, in the median streak on the mesenteries of the third and fourth cycles 

 immediately below the stomatodaeum, not a trace is to be seen of a partition into streaks, though further down these median 

 streaks also assume the same appearance as on the other mesenteries. 



2 ) Me. Murrich (1910) mentions the same arrangement in Pachyccrianthus fimbriatus and Ceriantheopsis americaints. 

 Me. Murrich distinguishes five streaks, one median (my median ciliated groove), two lateral distal streaks (my two portions 

 of the median streak) and two lateral proximal streaks (the ciliated streaks of other writers). It seems to me hardly fitting 

 to designate Me. Mur rich's first three streaks as separate streaks, as the whole transition from a single median streak to 

 a separation of the median streak into two distinct portions divided by a well marked ciliated groove takes place in the 

 Ceriantharia. Topographically of course Me. Murrich's terms may be made use of. 



