42 



teeth resting upon d, 2 upon the sickles a and a rectangular one b. This row of teeth does 

 not correspond neither with the first, nor with the second row of "gutturosum" (Kowalevsky 

 13, fig. 25, pag. 276—277). 



Though there are some points of resemblance a comparison is extremely difficult on 

 account of my sections being seriously injured. From the structure of the large tooth dt, which 

 in both species is somewhat fork-shaped, it is evident, that there may be some relation. 



The question also arises whether Kowalevsky is justified in calling the pieces d and 

 d teeth ; is it not possible that they may be compared just as well with the mandibulae of 

 "radulifera" as with the cuticular "Leisten" of "nitidulum" (mandibulae, Kowalevsky). The same 

 is the case with the large tooth dt-, it appears to me that we have good reason not to consider 

 it as belonging to the radula and to compare it with the side-pieces (mandibulae?) d and d '. 



In account of the piece 6 (fig. 192) the radula of "Lovéni" approaches more closely that 

 of "radulifera" where a similar piece is found in the middle of the row of teeth (Kowalevsky 13, 

 fig. 9 pc). Fig. 191 shows that there may have been still another row of teeth; an oblong piece ó, 

 which may better be compared with the piece pc of "radulifera", is likewise present. It is evident 

 therefore that the radula of "Lovéni" may be compared with that of "radulifera" and "gutturosum". 



The radula of "Wiréni" presents other peculiarities; here the large tooth dt is of the ordinary 

 structure, but the two pieces d are, no doubt, chitinous, that is, surrounded by a chitinous mantle. 

 The latter may be compared with the large tooth dt of "nitidulum" and "canadense" which also 

 consists of cuticular matter, surrounded by a thin chitinous layer v rendering the tooth very brittle 

 after isolation (cf. fig. 21 6). Of this radula the sickles a are moreover quite chitinous and differ 

 therefore from those of "Lovéni" and "radulifera" (?), which are cuticular, whilst the teeth they 

 carry are. chitinous. For the rest there is nothing more to be said of the radula of "Wiréni". 



We may compare the radula of the species of Chaetoderma (present of Rev. Norman) 

 with that of "radulifera", but a closer resemblance exists between it and "Lovéni"; the middle 

 piece è is heart-shaped but the 2 sickles are wanting; the teeth e have been situated perhaps 

 upon the side-pieces d and the teeth a may be compared with those on the sickles of "Lovéni"; 

 these are however mere suppositions. 



The radula of "nitidulum", "procluctum" and "canadense" may have developed from a 

 radula like that of "Wiréni"; the little teeth have disappeared ; only the 2 chitinous sickles 

 remain, whilst the large tooth and the side-pieces are still present. 



At any rate it is evident, that the radula of Chaetoderma is of great importance : any 

 contribution to its knowledge is heartily welcome. A division into a number of smaller genera 

 may be expected in the future; the liver also presents important differences. Chaetoderma 

 radulifera with its well-developed radula and extremely small liver is perhaps the most primitive form ; 

 the other forms have developed with a rudimentary radula and strongly developed liver ] ). The 

 question to which species of the Neomeniidae Chaetoderma is related, must still remain unanswered. 



1) When this article was passing through the press, I examined a new species of Chaetoderma, brought home by the Challenger- 

 Expedition, which I intend to call Chaetoderma Challengeri. This species a description of which I hope soon to publish, has a purely 

 distichous radula, consistirjg of many rows of teeth, one behind the other, and a well-developed radula-sac. The liver on the contrary 

 is hardly developed at all. 



