CYSTISOMATIDAE 269 



The present collection contains 22 specimens, of which six are $$, one is a <$, and 

 the rest are more or less immature. Valuable as this collection is, nevertheless it un- 

 fortunately helps us very little forward. Two forms are easily recognized and referred 

 to already known forms. But with the hope of bringing a little more definiteness into 

 the definitions of the species of this genus, I have subjected the descriptions of all known 

 specimens to tabular analysis. The result scarcely justifies the time expended, but for 

 what they may be worth I add a few notes and comments. 



Firstly, as regards C. longipes (Bov.), which Bovallius instituted (1886) on one speci- 

 men, later (1889) called Specimen A (Indian Ocean) (Stephensen, 1918, p. 60). The 

 fact that Bovallius later (1889) referred another specimen (B) to this same species, in- 

 correctly as later authors have recognised, is no reason why he should be deprived of 

 the authorship of the species. Woltereck (1903) clearly did not mean his bovallii to 

 supersede longipes, to which he does not refer except under a different heading, and to 

 state (p. 459) that he suspects parkinsoni to be a synonym of it. Later Stephensen dis- 

 covered the type and paratype (specimens A and B) of Bovallius' longipes and found 

 them to be different species ; but he incorrectly used (1918, pp. 59, 60) Woltereck's name 

 for the type (specimen A) of longipes "to avoid further complication ".He has, however, 

 accomplished exactly what he wished to avoid; clearly longipes, Bov., must stand, what- 

 ever happens to bovallii, Wolt. 



C. bovallii is in fact a species insufficiently described by its original author, for the 

 number of ventral spines is not stated. Unless Stephensen has examined the type 

 specimen of bovallii (and he does not state that he has done so), the identification of 

 the Thor specimens from Stations 61, 73, 76 with Woltereck's species is risky. On 

 the other hand, if there are the same number of ventral spines as in longipes, then it 

 would seem that bovallii is really only the female of longipes. The 'Valdivia' took 

 2 $? bovallii in the Atlantic, and the ' Thor ' took 2 S3 longipes in the Atlantic ; with the 

 exception of Bovallius' type specimen A neither longipes nor bovallii is recorded from 

 the Indian Ocean. 



I am unable to pronounce definitely on the status of bovallii, but wish to vindicate 

 Bovallius' name longipes. 



Secondly, C. coalitum (Wolt.) is another species which, as regards the number of 

 mandibular spines, is inadequately described. If there is only one spine the resemblance 

 to fabricii, Stebb., becomes remarkably close. The coalescence of the peraeon segments 

 seems to be a character on which too much reliance should not be placed. I entirely 

 agree with Stebbing's remarks (1888, p. 13 19). It is certainly not always easy to de- 

 termine beyond cavil the presence or absence of a suture. Bovallius found a suture 

 between peraeon segments 1 and 2 in both the specimens he assigned to longipes, 

 whereas Stephensen (19 18, p. 60) says there are no sutures. Spandl (1927) has recorded 

 a specimen of magnum in which he finds a faint but distinct line of articulation between 

 the first two segments. In the 52 mm. long specimen from Discovery Station 81 the 

 first two segments seem to be quite distinct. Probably it was some such (aberrant) 

 specimen on which Bovallius founded loveni, but as the type specimen is apparently 



