42 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 73 



ions 35 and 46. For determination of this point it is not necessary 

 to follow the literatnre further and the fact that U. irrostraius has been 

 used as type by some authors is irrelevant as the case now stands. 



2. Family name. A complication has arisen because of the fact that 

 U. irrostratns has been used as type * of Urotlwe. 



Stebbing (1906, Das Tierreich, vol. 21, p. 131) retains U. irrostra- 

 tns in Urothoe, family Haustoriidae. and classifies (idem., p. 146) 

 U. rostratus in Pontharpinia Stebbing, 1897, mt. pinqxiis, family 

 Phoxocephalidae. Thus a typical " transfer case " is presented. 



Pirlot raises an important question in regard to Phoxocephalidae, 

 namely : 



I. Must the oldest included generic name be taken as type for the 

 family name? To this, the answer is in the negative. 



Article 4 of the Rules reads : " The name of a family is formed by 

 adding the ending idae, the name of a subfamily by adding inae, to 

 the stem of the name of its type genus." 



This rule does not prescribe how the type genus of a family is to be 

 selected ; and in the absence of restrictions covering this point it is to 

 be assumed that, in accordance with custom, the original author is 

 free to select as type genus any generic unit which he prefers. This 

 is in harmony with the spirit of Article 30 which obviously leaves an 

 original author of a genus entirely free to select as type species any 

 species he wishes thus to designate. If the original author of a family 

 (or of a genus) were compelled to select as type the oldest genus (or 

 the oldest species) in the proposed family (or genus), this might 

 confine his choice to a little known and very rare taxonomic unit — 

 a restriction which would obviously be contrary to the interest both 

 of taxonomy and of nomenclature. In this connection it is to be 

 recalled that the " tyi>e " selected is the nomenclatorial type as dis- 

 tinguished from the assumed anatomical norm. 



Since (with the exception of isolated instances by early authors) 

 family names are based upon the name of the respective type genus, 

 such family name constitutes, ipso facto, a definite designation of the 

 type genus. For instance, Musca is definitely and unambiguously des- 

 ignated generic type by the use of the family Muscidae, Hotno of 

 Hominidae, Ascaris of Ascaridae, etc. It would be a nomenclatorial 

 reductio ad absurdum to consider any other genus as type of any of 

 these families. The concepts of a given family are not identical as 

 adopted by different authors and if the rule obtained that the oldest 



'Stebbing, 1891, on the genus Urothoe [etc.]. Trans. Zool. Soc. London, 

 vol. 13, no. I, p. 10: "This, which has become the type species of this genus." 



