NO. 3574 COPEPOD CRUSTACEANS — LEWIS 65 



the process has a single, posteriorly du-ected spine, an inwardly di- 

 rected bifid spine, and a single spine arising between the preceding 

 two. In a second female, from the same host specimen, the left 

 postantennal process is similar to that figm-ed by Shiino (1954b, fig. 

 le) and by Hewitt (1964a, fig. 13), having only the posteriorly di- 

 rected spine and the bifid inwardly directed spine. In the same female 

 specimen, the postantennal process on the right side has a posteriorly 

 du-ected spine and a bifid inwardly directed spine, but the tip of the 

 anteriormost tine of the latter is bifid, giving an appearance similar 

 to that shown in figure 2Sd although the bifurcation is not as well 

 developed. Many of the characteristics used to separate the species 

 within the genus Gloiopotes have been shown to be variable (Hemtt, 

 1964a). Hewitt, however (p. 93), states that "the shape of the 

 plates [alae] on the fom-th thoracic [fourth pedigerous] segment does 

 not appear to vary but they may be inclined at various angles . . .". 

 The figured female specimen (figs. 22a, c) exhibits alae vnih a posterior 

 projection while the alae of some of the other Hawaiian specimens do 

 not have this projection and are rounded posteriorly or are almost 

 flat. It is felt that the use of alae shape as the primary means of 

 distinguishing between species is open to question, as is the use of 

 the inclination of these plates and the amount and position of plu- 

 mosities on the dorsal surface of the cephalothorax (HcAvitt, 1964a). 

 Hewitt also discusses the variation in the number of spines present 

 in the material that he had available for study. This variation is 

 exhibited by the Hawaiian specunens and, from an examination of 

 material in the U.S. National Musemn, appears to be characteristic 

 of members of the genus. This author feels that spines may be of 

 value, but only if a complete series is either absent or present (see 

 discussion of G. ornatus below). One of the problems in using this 

 characteristic, however, is that many of the spinules or even small 

 spines are not figiu-ed or described in the literature (see discussion of 

 G. auriculatus below.) 



An examination of the type material of G. costatus Wilson, 1919 

 (USNM 49772, 49773, not 51040 and 51041 as Wilson indicates 

 [p. 313]), and other material identified as this species, indicated that 

 the variation that prompted Wilson to erect this species is also found 

 in the Hawaiian specunens of G. huttoni (the variation in the number 

 and position of spinules and the shape and mclination of the alae). 

 One of the differences used by Shiino (1954b, p. 278) to separate 

 G. longicaudatus (Marukawa) from G. costatus Wilson is the tripartite 

 spine figured by Wilson on the first segment of the exopodite of the 

 thii'd leg of G. costatus. The female and male cotype specimens 

 used by Wilson for his original description comprised a copulating 

 pair, and the original specimens were in copulo until the present 



226-321—67 5 



