146 A. E. Verrill — Revision Genera and Species of Starfishes. 



uncertainty. This applies strikingly to Goniaster, Astrogonium, 

 Dorigona, etc. 



A few recent and prominent writers, especially Perrier and Sladen, 

 have restored the ancient names given by Linck (1733) to certain 

 genera and species of starfishes, thus displacing names well estab- 

 lished under the binomial system. Linck was a very able naturalist, 

 for his period, but he was not a binomial writer, and his names cannot 

 properly be allowed priority over those established under the binomial 

 system. 



The name Pentagonaster is the only generic name in this family to 

 which this remark applies. 



Perrier himself, although he restores several of Linck's names of 

 species, does not go so far as to try to restore others that have equal 

 claims to priority, for to do this would overthrow the well known 

 names of several common European species.* Nor has he proposed 

 to restore the names of Seba, which have equal claims to recognition. 



In the following pages I propose to briefly review the history of 

 some of the earlier names and of the more important groups to which 

 they have been applied by various writers, in order to show, if possi- 

 ble, to what particular groups certain of these names ought rightly 

 to be applied, in accordance with the generally accepted rules of 

 biological nomenclature.f 



* Among the names adopted by Perrier, and also by Sladen, from Linck, are 

 oculata, under Cribrella; planus, under Hippasterias ; corniculatus, under 

 Ctenodiscus. Neither of these can be justified. 



f Among the recognized rules that I follow, and which need to be applied to 

 this group, are the following : 



A. — Strict priority to be apjjlied to all names properly published in actual 

 binomial works, in general dating only from Ed. X. of the Syst. Nat. of Linne. 



B. — Exclusion from the rule of priority of names taken from earlier polynomial 

 writers, unless adopted by later binomial writers. In that case they should date 

 only from their introduction into binomial literature. 



C. — When an old composite genus has been divided by a later writer, the 

 original name must be kept for one of the component groups, and for one or 

 more of the species originally included by name. If a definite generic tyjje was 

 given by the original author, the name must remain with that type. If no type 

 was mentioned, the mere position on the page cannot fix the type. Nor does it 

 follow that the first s|pecies named was the type, unless so stated originally, for 

 many early writers arranged their species alphabetically, or in some other arbi- 

 trary way. 



D. — A composite genus having been subdivided and the original name definitely 

 applied to one of its parts (in accordance with rule C), it must ever after be kept 

 for that group (or some part of it) just the same as if it had been originally so 



