The mesenterial Filaments of the Alcyonaria. 5 



smoothly rounded, in other cases are attenuated and have short proces- 

 ses which are perha])s coutinuous with nervous fibrils. 



Although typical examples of these two forms of cells are very 

 distinct , there are nevertheless intermediate forms , and 1 am inclined 

 to regard them as only dififerent couditions in the activity of the same 

 kind of cell. In many cases the cell is sharply contracted in the middle, 

 so as to assume an hour-glass form , the lower division containing the 

 nucleus. In such cells both divisious are sometimes filled with the re- 

 fringent spheroids, but in other cases the latter are coufined to the 

 outer di vision, and the basal part is clear or slightly granular. Other 

 cells have the usuai form but their conteuts are confusedly granular, 

 showing neither the sharply defined spheroids of the one form nor the 

 uniform structure of the other. In many cases a dose examination of 

 the clear cells shows that they also are filled vs^ith the spheroidal bo- 

 dies, but these are so slightly stained as to be barely visible. 



These cells are evidently of the same nature as the two forms of 

 »gland-cells« described by the Hertwig brothers in the filaments of the 

 Actinians, with which they agree in all respects except that I have 

 been unable to sec the protoplasmic network of the clear cells as plainly 

 as it is figured by these autliors. 



Scattered irregularly through the filament are minute nettle-cap- 

 sules [n] . They are remarkable for their very small size, being smaller 

 than the nuclei of the cntoderm cells. They have an oval form and 

 each contains a spirally coiled filament. In the minuteness and rarity 

 of the nettle-capsules the mesenterial filaments of the Alcyonaria diifer 

 decidedly from those of the Actinians, and it seems possible that in the 

 former group they are to be regarded as rudimentary organs. 



Sensory cells, like those of the Actinians , I have been unable to 

 find. The centre of the filament is occupied by a clear, apparently 

 fibrous mass, in which are considerable numbers of characteristic ento- 

 derm nuclei. 



b) Development. 



I have already described the development of these filaments in Me- 

 nilla (1. c.) — though I failed to perceive their fundamental difi'erence from 

 the dorsal filaments — and obtained stroug evidence of their entodermic 

 origin as locai thickenings upon the edges of the septa. I found that the 

 rudiments of the filaments, in some cases, made their appearance before 

 the stomodaeum broke through, and while the invaginated ectoderm 



