502 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 2 



to justify his new species on differences that he observed between his 

 own material and the description of Linton. The present writer has 

 access to some of the cotypes of G. medius. These include the two speci- 

 mens that were restudied in 1918 when medius was reassigned to the 

 genus Rhadinorhynchus and three additional individuals that had been 

 preserved in alcohol but are now stained and mounted for microscopical 

 study. Measurements and other observations taken on this cotype mate- 

 rial fail to agree with the data cited by Linton in the original charac- 

 terization of the species. 



The original description of G. medius gives 45 /a as length of the 

 hooks on the base of the proboscis. On the cotypes these measure 59 to 70 

 /A, thus including the 70 fx cited by Chandler for G. gibber. Likewise, 

 other hooks on the proboscis of G. medius were recorded as 60 ju, while 

 reaching 70 to 80 p. in G. gibber. Most of the hooks measured on the 

 cotypes of G. medius were at least 70 /x long. Similarly, the spines on the 

 body were recorded as distinctly larger in G. gibber j but spines 60 jx were 

 vtry common on the cotypes of G. medius, thus eliminating spine size as a 

 possible specific difference. In like manner, the cited range of hook 

 formulas for the two presumed species does not fall beyond the range of 

 individual variability commonly encountered in Gorgorhynchidae. Linton 

 cited "about 22 ventral rows of hooks," while Chandler listed an un- 

 qualified 22 for medius and cited 24 for G. gibber. Bodily measurements 

 and proportions cited in Chandler's comparison of the two presumed 

 species are not sufficient ground on which to base specific distinction 

 when other more stable characters, not affected by growth stages, are 

 removed from the list of differences. 



On the basis of the foregoing analysis the present writer is convinced 

 that G. gibber must be considered as a direct synonym of G. medius 

 (Linton, 1907). Since G. gibber had been cited as genotype by original 

 designation, its valid synonym G. medius (Linton, 1907) becomes the 

 type of the genus Gorgorhynchus. In a paper now in press (Van Cleave 

 and Lincicome), the genus Gorgorhynchus is separated from the family 

 Rhadinorhynchidae and is designated as type of a new family Gorgo- 

 rhynchidae, separable from the Rhadinorhynchidae chiefly on the basis 

 that the Gorgorhynchidae have males with four cement glands, while 

 the Rhadinorhynchidae have males with eight cement glands. 



Two distinct species of Gorgorhynchus were taken from marine 

 fishes of Albemarle Island by the Hancock Expedition. These are de- 

 scribed in the following section. 



i 



