A. JEJ. Verrill — Mevision Genera and Species of Starfishes. 147 



This brief review is, however, by no means intended as a complete 

 history of the subject. Generally only the works that seem essential 

 to the object in view will be referred to here, A fuller discussion 

 must be left to a much more extensive work on American starfishes, 

 which is now well advanced towards completion and in which most 

 of the genera and species will be well illustrated. 



I. The first generic name applied by binomial writers to any sub- 

 division of the old genus Asterias (Linne), and pertaining to the 

 present family, was Goniaster. This name was proposed by L. 

 Agassiz, in 1836, for the pentagonal starfishes, collectively, includ- 

 ing representatives of more than one family. 



This name was adopted by Forbes in 1841 (Brit. Starfishes), in 

 the same sense, for he included in it such diverse forms as Hippas- 

 teria and Astero^ysis, without assigning to it any definite type. 

 Miiller and Troschel used it in the same way, in 1840. Dr. J. E. 

 Gray, in 1 840, adopted the name for a very restricted group, with a 

 definite diagnosis, and named as a type, G. cuspidatus, a well- 

 known species and one of those given by Agassiz as examples of his 

 genus. This species should, therefore, remain as the type of the 

 restricted genus. 



In 1842, Miiller and Troschel reunited Goniaster, Pentagonaster, 

 Tosia, and Hippasteria of Gray into a single genus, to which they 

 applied the new name, Astrogonitmi. If these four groups really 

 constitute only a single genus, it is evident that Goniaster (emended) 

 should have been adopted as its name. 



II. The name Pentagonaster was first used, under the binomial 

 system, by Gray, in 1840. He applied it to a particular type (P. 



applied. In other words, a generic name correctly applied to a restricted group 

 has just as much claim to priority, in the new sense, as a neiu name would have. 



E. — When a generic name is a real synonym of another earlier one it should be 

 dropped from the system, unless it had a different type-species when first pro- 

 posed. In case the two types belong to different subdivisions of a composite 

 genus both names may be retained in a modified sense. In cases where two 

 names are only jjartially synonymous, both may be used if they can be properly 

 restricted to distinct subdivisions of the groups to which they may have been 

 originally applied (in accordance with rule C). 



F. — The application of an old or discarded name to a species or group not 

 included in the group to which it was originally applied is to be avoided as lead- 

 ing to confusion and instability. A name once dropped from the system, for 

 good cause, should fall into disuse in every other sense. To use a discarded 

 generic name for a new genus in the same class or order (as if it were a new 

 name), should never be thought of, for it is sure to cause confusion. [Goniaster 

 and Astrogonium among starfishes afford examples of incorrect transposition of 

 names.) 



