448 NOTES ON PROSOBRANCHIATA, 



they make Septa the type genus of the family (Septid?e) [8,416, 

 et seq.]. 



Perry enumerated six species of the genus Septa; they are, in 

 the order he gave them : — 



Septa parkinsoniana = Triton fusi for mis, Kiener. 

 ,, scarlatina = ,, r^ibecida, Linn. 



,, sp>engleri — ,, spengleri, Chemn. 



,, rubicunda = ., nodiferus, Lamk. 

 ,, rubecula = ,, pilearis, Linn. 



,, triangularis = ,, costatus, Born. 



Lotoriimi parkinsoniarium is not related to the tritonis-group 

 (vide post, p. 475), but makes with some of the Australian Tertiary 

 species a rather distinct section, which is, however, connected 

 with L. waterhousei. Thus the first three species are all referable 

 to the so-called genus Laynpusia, which is, therefore, a synonym 

 of Septa. These again are generically inseparable from Lotoriam. 

 Therefore it is only by admitting Lotoriium as the type of the 

 family, and removing Perry's first three species thereto, that the 

 name Sep)ta can be applied to the tritonis-gvou^. It is obvious 

 that they cannot be referable to Lampusia or Ranularia, both of 

 these being later names. If, therefore. Septan are typical forms 

 (and parkinsonianum is the type thereof), Lotorium must be 

 regarded as a synonym, which has been shown to be impossible. 

 Should m}'- "lines of generic similarity" not be considered 

 sufficient!}^ complete, nor the other evidence conclusive, then the 

 name Septa must be applied to the group formed by pjarkinsoni- 

 anwn and the Australian fossils, this section being more distinct 

 than any of the others. ■ 



The tritonis-gvoviY) cannot be retained as the typical section of 

 the genus; it has not yet received a name which can be used. 

 Unless we are to regard L. parkiyisonianum as the type of the 

 genus, the name Septidm cannot be adopted; under these circum- 

 stances it seems far more reasonable to accept Harris's term 

 Lotoriidce. 



