28 PERCY SLADEN TRUST EXPEDITION 



Nematocysts are restricted to the median region of the filament, usually somewhat 

 scarce in its straight region, but when present may belong to any of the three types 

 previously described ; in the coiled region, they occur in large numbers, usually arranged 

 close together in the form of batteries (figs. 17 and 59). Type I is less frequently present 

 than II and III. From histological appearances it seems probable that the straight 

 region takes part mainly in digestion, while the convoluted region is for ofiFensive or 

 defensive purposes. This suggestion may account for the fact that the convolutions are 

 frequently found protruded through the oral-disc and through the stomodseum. On some of 

 the mesenteries that do not reach the stomodseum filaments are rudimentary or even absent. 



H. V. Wilson regarded the mesenterial filament of Manicina as equivalent to the 

 trilobed filament of an Actinian as described by the Hertwigs, the ventro-lateral tracts 

 corresponding histologically with the lateral ciliated bands (" Flimmerstreifen") and the 

 median portion with the central glandular lobe (" Nesseldriisenstreif "), whereas according 

 to Duerden an entire coral filament represents only the middle lobe of the trifid Actinian 

 filament and that in the former " there is nothing which corresponds morphologically with 

 the lateral lobes " (32, p. 472). 



Various views have been held on the origin of the filaments in the Anthozoa. 

 H. V. Wilson in 1889 concluded that the filaments were ectodermal structures, those of 

 the primary mesenteries being extensions of the stomodaeal ectoderm along the free 

 mesenterial margins, while those of the incomplete mesenteries were formed by the 

 reflection of the stomodseal ectoderm towards the oral-disc and thence across its inner 

 surface to the free edges of the mesenteries. He further held that the three lobes of an 

 Actinian filament were also ectodermal in origin and suggested that they could be derived 

 from the simpler larval filament of Manicina ; in the latter the two lateral ciliated tracts 

 and the median secretory tract are already differentiated ; "to produce the trifid filament 

 it is only necessary for these tracts to become separated by the division of the mesoderm 

 ( = mesogl8ea) into three lobes" (122, p. 228). E. B. Wilson in 1884 had regarded the 

 "Flimmerstreifen" of the Hertwigs to be ectodermal and homologous with the two longer 

 dorsal filaments of an Alcyonarian, which were downgrowths of the stomodseal ectoderm, 

 provided with long cilia but with no gland cells and functioning as organs of circulation ; 

 the same author believed the " Nesseldrtisenstreifen " to be endodermal and homologous 

 with the remaining six shorter Alcyonarian filaments which were thickenings of the 

 mesenterial edges containing gland cells and subserving a digestive function. McMurrich 

 in 1891 disagreed completely from H. V. Wilson's and favoured E. B. Wilson's view. 

 According to him the median tract of a filament of Aulactinia was formed first by a 

 differentiation of the endoderm at the free margin of the mesentery while the two lateral 

 lobes were downgrowths of the stomodseal ectoderm* ; he further suggested that the same 

 order was followed in the phylogenetic history of the filament. In a later paper, while 

 still maintaining that the median region and the lateral tracts arose independently in 

 ontogeny, he did not regard the ectoderm and endoderm of Coelenterates as having 

 reached in phylogeny a degree of differentiation equal to that of the epiblast and hypo- 



* E. B. Wilson's figs. 20 and 23, transverse sections of the embryonic filaments of Aulactinia, are similar to 



those of adult coral filaments. 



