20 BULLETIN 75, UNITED STATES NATIONAL MUSEUM. 



The genus Ophiomusium throws considerable Ught on the subject. 

 The species cancellatum and jolliensis are very closely allied, and 

 there is little doubt that the latter is more nearly related to the former 

 than to any other species. Yet they were taken together at stations 

 3738, 4965, 5079, and 5091. At station 4933, 0. liltkeni was taken, 

 while at the neighboring station, 4934, under the same conditions, 

 0. lunare, a very closely related species, was taken. The three species, 

 lymani, simplex, and trychnum, are very nearly allied and the two 

 latter probably find their nearest relative in the former. At station 

 4933, simplex and lymani were taken together, while trychnum and 

 lymani were found at stations 3704 and 5092. Furthermore, lymani 

 was found at station 4956 and at station 4959, only 5 miles away, 

 under nearly identical conditions, trychnum occurred. 



In the genus Amphiura the species hellis and diomedese are very 

 nearly related; they were taken together at station 3697, and while 

 diomedese occurred at station 5082, hellis was found under identical 

 conditions close by at station 5083. Two species of Amphiodia 

 {macraspis and eunjaspis) were taken together at stations 4862 and 

 4997, but the value of this occurrence is vitiated by the fact that these 

 two so-called species may be identical. At station 3713 Amphiodia 

 ancistrota and A. digitula were found together. 



In the genus Ophiacantha many of the species are none too well 

 defined, and this is particularly true of the pentagona group, including 

 among others rhachophora and levispina. Nevertheless it is of some 

 interest to record that pentagona and levispina occurred together at 

 station 4916, while rhachophora and pentagona were found associated 

 at stations 3698, 3713, 4893, 4902, 4903, 4965, and 4967. Although 

 Ophiacantha cataleimmoida may be most nearly allied to 0. relicta, it is 

 very near normani, and its occurrence with the latter at stations 2853, 

 3332, 3338, 3340, and 4980 is worthy of note. 



Of the genus Ophiolehes, four species (tylota, paucispina, diaphora, 

 and hrevispina) are so nearly alhed that it is hard to determine just 

 what their interrelationships are. At station 3480, hrevispina and 

 paucispina were found together; at station 3599, tylota and diaphora 

 occurred; at station 4781, tylota, paucispina, and diaphora were all 

 three taken; and at station 4784 tylota and diaphora were again found. 



The two species of Ophiomyxa, australis, and anisacantha are so 

 closely allied there is no doubt that the former is the nearest relative 

 of the latter; australis was found at stations 4894 and 4895, wliile 

 anisacantha was taken at stations 4892 and 4893; yet external con- 

 ditions at the four stations were not essentially different. 



The Japanese basket-fish, Gorgonocephalus sagaminus, certainly 

 finds its nearest ally in G. caryi, yet the two are perfectly distinct. 

 They were taken together, however, at three stations (3707, 4986, 



