85 



PdtdUi also feeds in another way by grippiiifv a piece 

 of seaweed with its outer lips (aided hy the palate) and 

 then scraping off fragments from it by the rasping action 

 of the radnla. The mechanism for retracting the 

 odontophore is not as manifest as that for its extrusion, 

 and this process is probably in part a leturn to normal 

 conditions from a strained state of expansion. The 

 slightly developed retractor fibres, the cessation of con- 

 traction of the protractors, and the antero-lateral muscles 

 probably all help, and assistance must also be given by the 

 contraction of the muscular snout. 



While the tip of the radula is scraping food from the 

 rock, another part of it is, by the same motion, working 

 against the jaw, or palatal plate, and thus helping to grind 

 the food already obtained. This working of the radula 

 against the jaws continues after actual feeding has ceased, 

 and the oesophageal pouches seem to retain the food frag- 

 ments temporarily till this function has been completed, 

 and till they can be passed on into the crop. This passing 

 on is necessarily slow on account of the many folds project- 

 ing into the cavity of the crop. Meanwhile, also, the food 

 becomes mixed with the secretion of the buccal glands. 



It is noteworthy that the oesophageal pouches of Patella 

 do not show the development of folds and papillae which 

 characterises presumably homologous structures in the 

 primitive Ehipidoglossa, and another feature is that the 

 dorsal and ventral valves, which in the latter, and even in 

 Acmiva (dorsal valve), prevent passage of the food from 

 the pouches back into the buccal cavity, are very much 

 leduced or absent. The pouches are therefore proba'bly, 

 in part, spare tissue and, in part, temporary stores, and 

 not highly specialised secretory regions, as in TIaVwth, and 

 other Rhipidoglossa. The corresponding secretory func- 

 tion is performed by the crop, which is extremely 



