12 8IK C. ELIOT. 



distortion. The following is an analysis of a radula of 32 rows, row No. i. being 

 the hindmost : — 



Eows i-ii . 8 18, the median tooth being smaller in i than in ii. 



Kows iii-ix . 8 8. 



Eow X . . 7r r7, where r indicates a rudimentary lateral.^ 



Rows xi-xiv . 7 7. 



Row XX • . 6r r6 „ „ ,, „ 



Rows xxi-xxvi. 6 6. 



Rows xxvii-xxx 5 5. 



Row xxxi . 4 4. 



Row xxxii . 3 3. 



The shape of the teeth is much as in Clione limacina, except that the median teeth 

 (llgh) are markedly smaller and less distinctly cuspid, and that the laterals in the 

 hindmost rows (lief) have forked bases. It is noticeable that several specimens have 

 been preserved in the act of holding and apparently eating small elongate fishes, 

 which are still attached to them (fig. 12). 



In CI. limacina the penis is large, frequently found exserted, and provided with 

 a large accessory organ {vide Boas's figure reproduced by Meisenheimer in Die 

 Arktischen Pteropoden, p. 415) ; but these characters are not found in any of the 

 present specimens of CI. antarctica. The penis is small, invariably retracted, and no 

 accessory organs were discovered. Too much stress should not be laid on this 

 difi'erence, for no doubt the accessory organs may have a very different appearance 

 in the same species when exserted and when retracted. 



The presence of the larval rings and the small size of the penis naturally suggest 

 that this form is immature, and it is also noticeable that the specimens were all captured 

 between the months of November and March (though in different years), most of them 

 in January and February. This certainly makes it probable that they are all in the 

 same and possibly not adult stage of development. On the other hand, it is clear from 

 a comparison of Arctic and Antarctic specimens of the same size that the young Clione 

 limacina is not more like CI. antarctica than are the full-grown individuals. Also 

 the type specimens of Clione antarctica, captured by the ' Southern Cross ' ofi" 

 Cape Adare and lent to me for examination by the kindness of Mr. E. A. Smith, I.S.O., 

 resemble those found by the ' Discovery ' in most external points except the 

 colour, which is dirty green with purplish spots, a difi'erence no doubt due to 

 the method of preservation. The length is 20 mm. or a little less, the larval rings 

 are more or less distinctly visible, the fins, foot, and buccal cones are as in my 

 specimens. 



' The conformation of the radula may be due to one or both of two causes : — {a) In most opisthobranchs 

 the front, that is the earliest rows of the radula, contain fewer teeth than the posterior or later rows. (6) As, in 

 this case, the radula is split into two portions, it is possible that not only the central tooth, but the inner laterals 

 on either side are worn away, in which case r would indicate a vestigial tooth. 



