AMPHITRETIDAE 



38s 



1886, p. 67) from a small specimen obtained off the Kermadec Islands. In 1902 Ijima 

 and Ikeda (1902, p. 85) described a female obtained in 1897 in the Sagami Sea, Japan. 

 Chun (1900, fig. on p. 535) and Thiele {loc. cit.) figured and described a third specimen 

 from the Agulhas Stream (34° 31' S, 26° 2' E). Sasaki (1917, p. 361) described a male, 

 also from the Sagami Sea. This and the other Japanese specimen were again described 

 by Sasaki (1929, p. 16). 



Now it is, to my mind, very uncertain whether all these descriptions relate to one and 

 the same species. Sasaki had the advantage of seeing Ijima and Ikeda's specimen, and 

 regarded it as conspecific with his own. But the relation of 

 the Japanese, the South African and the Kermadec specimens 

 is very uncertain. Hoyle's specimen is very small and in a poor 

 condition, and Thiele only described the radula, mandibles 

 (not known in the Japanese forms) and eye of the 'Valdivia' 

 example, so that we are plainly not dealing with comparable 

 data. 



Whatever we may think of these forms I am quite convinced 

 that the specimen obtained by the ' Discovery ' is not con- 

 specific with any of the previously described forms. This is 

 all the more striking, as it was obtained at no great distance 

 from the spot at which the 'Valdivia' specimen was taken. 

 It is very singular that two different species of this very rare 



genus should be taken more or less in the same area. 



TT -1111 -1 I i<T^- , Fip.13. Amphitretus thielei. 



However, it should be pomted out that the Discovery Female gonaduct, etc. X4. 

 specimen was taken to the west of Cape Town, i.e. in the 



Benguela Stream, and the 'Valdivia' one on the other side of the Agulhas divide. 

 The following table will make clear the differences : 



In addition, the radula and mandibles of thielei and the ' Valdivia ' specimen are singularly 

 unlike, and the funnel-organ of thielei agrees with that of neither the type of pelagiais 

 nor of the Japanese forms. 



Too much store need not be set on these characters, especially as only single specimens 

 are involved. In the ' Yaldivia' pelagicus the only useful systematic data we know are 

 of the radula, jaws, general shape and web (as seen in the figure of Thiele, loc. cit. 

 pi. xci, fig. 6). The web is more like that of thielei, about half as long as the arms, not 



