334 THYTIVOROITS MOLLUSCA 



but you must remember that the dendritic character of 

 the system is not always so much exaggerated, if I may 

 use the expression. It varies, in fact, considerably in the 

 different genera, passing through some intermediate steps 

 to its full developement, and again receding to normal sim- 

 plicity in other genera, as in Limapontia and Chalides, in 

 whom there are only two large pouches for the retention 

 and elaboration of the alimentary matters.* 



Now, you must beware of believing in the existence of 

 any unbroken barrier between the mollusca that feed on 

 herbage and the mollusca that feed on flesh. That certain 

 families are herbaceous and others carnivorous, under or- 

 dinary circumstances, is very true ; but in every order or 

 class you will find some genera or species that break the 

 rule, and indulge in a diet repulsive to the class as a body. 

 Perhaps no mollusks are so pre-eminently carnivorous as 

 the Cephalopods, but I know that some of them do feed 

 occasionally, — not on olives, as Pliny tells us,f but on 

 sea-weed, for I have found the stomach of Loligo sagit- 

 tata crammed full of fragments of the midrib of Alaria 

 esculenta, and pieces of the same were sticking between 

 the mandibles, as if the creature had been killed in the act 

 of eating. The Scalaria, Turritella, Velutina, Ianthina, and 

 Sty lifer, genera which the characters of their shells would 

 decide to be herbivorous, are really and exclusively carni- 

 vorous. The favourite food of the Ianthina appears to be 

 the gelatinous Velella, whence, it is said, that it derives 

 the blue tincture of its shell ; and the Stilifer lives a para- 

 site amid the forest of spines that clothes the Echinus, or 

 burrows under the skin of the star-fish, upon whose juices 

 there is reason to believe that it feeds. "With that in- 

 stinct of self-preservation imparted to all parasites whose 

 existence depends upon that of their nidus, the Stilifer, 

 like the Ichneumon among insects, appears to avoid the 

 vital parts ; for in no instance did Mr. Cuming find it 

 imbedded anywhere save in the rays, though some had 

 penetrated at their base and very near the pelvis." J 



Of the Nudibranches, which are usually, and, with regard 

 to many of them, properly classified with phytivorous mol- 

 lusks, the greater number appear to be animal feeders. 



* Ann. des Sc. Nat. (1844) i. 16. Also the vol. for 1845, p. 275. 



t Of the Octopodus, Rondeletius says, — " lis mangent les coquilles de mer, 

 lis aiment fort les branches de l'olivier, et par cette friandise on les prend, 

 ils aiment aussi le figuicr." — Hist, des Pois. i. 373. 



+ Bioderip in Proc. Zool. Soc. ii. 60. 



