Manchester Memoirs, Vol. /xii, (191 7) -No. 4 3 



Where the above methods failed to reveal filaments by reason of 

 the contracted condition of the tissues, or of the presence of 

 foreign matter in the coelenteric cavities, it was found that their 

 presence could be demonstrated in longitudinal sections of the 

 siphonozooids about 0.5 m.m. in thickness, cut by band from a 

 piece of tissue previously hardened for two or three hours in 

 absolute alcohol. 



Structure of Siphonozooids. 



The siphonozooids of Pennatulacea are distinguishable from the 

 autozooids by reason of their lack of tentacles (with the exception 

 of certain zooids in Umbellula and Chunella which possess one 

 tentacle) : the presence of a wide ciliated groove, the siphonogylph, 

 down the ventral length of the stomodaeum : the comparatively 

 slight development of the eight mesenteries : and the absence of 

 longitudinal retractor muscles and gonads on the mesenteries. 

 Mesenteric filaments are never found on the six ventral mesen- 

 teries ; they may or may not occur on the dorsal mesenteries. 



The structure of the dorsal filaments and their ectodermjc origin 

 is well described by Wilson (1884, p. 12) and little need be added to 

 his account. 



It may be noted throughout the order that there is extraordinarily 

 little variation in the structure and shape of the filament : the 

 Y-shape in transverse section is constant, the two lateral lobes 

 containing nuclei which stain deeply, the medium groove appear- 

 ing clear. The diagrams given are typical for many species of the 

 order. (Figs. I. and II.) 



There is a slight variation in the width of the filament in different 

 species, and to a greater extent in the length ; also in certain 

 species the filaments appear straight or slightly curved, and in 

 others they are considerably convoluted. How far this latter con- 

 dition is natural, or to what extent produced by killing and fixing 

 I am unable to state. It is recorded of fully expanded living polyps 

 of Alcyonium or Paralcyonium (Wilson 1884, p. 13) that the ventral 

 endodermic filaments constantly change their form, being thrown 

 into convolutions by the contractions of the mesenteries, but "the 

 straight ectodermic filaments present a very different appearance." 

 If this is correct for the siphonozooids it might be expected that the 

 part of the mesentery below the stomodaeum would shew signs of 

 contraction in the tentacular zooids of Umbellula Carpenteri for 

 example, but there is no appearance of such contraction. Marshall 

 figures a siphonozooid of U. gracilis (1883 Plate XXV., fig. 33 ^ 

 with the filaments considerably coiled. 



