200 GRIFFIN. 



cies which he described as Valencinia aimandi n. sp. The 

 careful description of this form by its discoverer (Macintosh '75) 

 and the able anatomical investigations of Oudemans ('85) made 

 it clear that Valencinia arniandi is not only the representative of 

 a distinct type (allied to CepJialotJndx)^ but a form in many re- 

 spects intermediate between Carinclla and other Nemerteans. 

 The special interest centers in the fact that anteriorly the nerve- 

 cords lie in a similar position to those of Carinella (although 

 surrounded by a thin layer of longitudinal muscles), while more 

 posteriorly they break through the outer circular layer and lie 

 for the rest of their course within the longitudinal layer. Oude- 

 mans was thus thoroughly justified in creating the new genus 

 Carinonia for its reception. For twenty years the form remained 

 the sole representative of its genus. In 1895 Burger described 

 the C. patagonica from some very scanty material collected at 

 Punta Arenas, Patagonia. Of this material he observes : ** Uber 

 ihr Aussehen im Leben fehlen leider Angaben." In C. pata- 

 gonica the nerves lie wholly within the longitudinal muscle layer^ 

 so that within the limits of the genus Cariiionia we have accom- 

 plished the theoretically required migration of the nerves through 

 the circular milscle layer. It now became easy^ to derive the 

 Enopla directly from Carinclla through Carijwnia and CepJialo- 

 thrix^ while the ScJiizonemertean type (including the Eupolidae) 

 comes off as an independent side branch from an ances- 

 tor of Carinoma, which retained the nerve-cords outside of the 

 circular muscles, but lost the inner circular layer and developed 

 a new longitudinal layer beneath the basal membrane of which 

 Carinoma arnianda shows rudiments.^ 



These points are all clearly recognized in Burger's taxonomic 

 system. Carinclla with Carinonia and HidirccJitia constitute the 

 first and most primitive order Protonemertini ; Carinoma and 

 Cephalothrix are ranked as an independent order Mesonemertini; 



1 Cf. Biirger '95 (2). 



2 Carinoma, while more primitive as regards the nerve-cords and presence of 

 nephridia, seems to have lost the cephalic organs still retained in CephalotJwix (com- 

 pare Joubin '90). 



3 Such an ancestor Burger believes to be realized in Hubrechtia desiderata (v» 

 Kennel). 



