lO 



3. Sagifta fcrox Doncaster. 



L. DoN'CASTER. Op. cit., p. 212. 



No species in the collection gave greater trouble than this and its ally robusta of the 

 same author. Neither of the two was adequately described in the first instance; and the figure 

 of robusta contradicted the text in two important particiilars, — namely, the proportionate length 

 of the tail segment, and the posterior extension of the posterior fins ; even the number of jaws 

 was not correctly drawn. Like most freehand drawings of Chaetognatha, the figure cannot be 

 trusted to grive the real outline of the animal. Doncaster's beautiful figures are imich more 

 likij Chaetognatha than my camera outlines, but no artist can draw with the precision of the 

 camera. Of ferox only the front end was figured. A very large number of likely specimens 

 have therefore been carefully examined with a view of placing these alleged species on a more 

 .satisfactory footing. As the resuh I beHeve that I have found, among the formalin material, 

 representatives of both fcrox and robttsta ; but whether they are separate species, or the one 

 a precocious variety of the other, or both together forming a highly variable species in the act 

 of splitting up, I have no clear idea. Doncaster himself suggested that fcrox "should possibly 

 be classed with robusta'. 



There can be no cjuestion that the specimens assigned below to ferox are readily 

 distinguishable from every other species hitherto described. Those listed as robusta generalh" 

 exhibit one or two more jaws, and one or two more posterior teeth, at a similar length ; they 

 are apparently nearly sexually ripe at about 12 mm., hut fcrox at about 15 mm. (to this there 

 seem to be exceptions on both sides). Robusta retains its .sense papillae much better than fcrox., 

 and oftcn has a "hispid" appearance. I do not find the constant difference in the tips of the 

 jaws mentioned by Doncaster, although it happens that in some specimens they have been 

 more worn, and are shorter and blunter, than in others. On the whole it seems best to retain 

 both the species, in the hope that further collections may .settle their exact relationship. 



The "characters" assigned below to robusta and ferox have naturally been drawn from 

 extreme examples, in order to empha.sise the differences as much as possible. 



Characters. Head large, in expansion broader than the body at its broadest. Body 

 firm, broad, opaque owing to the marked development of the longitudinal muscles; trunk of 

 nearly uniform width from in front of the ganglion to the tail septum, tapering very gradually 

 forwards to the neck, more suddenly backwards from the tail septum. Laleral hclds narrow. 

 There is a well developed collarette, which extends to, or nearly to, tlie antcrior fins, and 

 is powerfully developed in adult spe'cimens. Tail segment about 26 — 36 per cent. of the 

 total length. 



Anterior fins long, rounded, widest po.steriorly and narrowing slightly forwards, commencing 

 about on a level with the abdominal ganglion. Posterior fins roundctl, sliglu]\- shorter than the 

 anterior, more on the tail than on the trunk, extending posteriorly to the vesiculae seminales 

 when these are tumid, widest behind ihe se|Hum. Tail (\n generally roundeil, meeting the 

 vesiculae seminales when tumid. 



Jaws short, strong, broad at the base, .strongly curved ; the tip small. Anterior teeth 



