The diaphid fishes, a rather difficult group taxonomically, are still not well understood 

 throughout the world oceans. Although some species are quite distinctive, most are difficult to 

 distinguish, a difficulty due in part to a very extensive overlap in many characters. The 

 number of fin rays and anal photophores (AOa and AOp) are, with few exceptions, rather 

 constant among species. The number of gill rakers and the positions of certain photophores 

 (PLO, VLO, SAOi and SAO3, Pol and Prc4, and the AOa and AOp series) are of greatest use in 

 separating species. However, there is variation in numbers of gill rakers within species, 

 particularly among those of worldwide distribution, and there is some inconstancy in the 

 positioning of the photophores named above. Seldom are the relative positions of the fins of 

 taxonomic value. 



If only adults or young adults in good condition are considered, the luminous organs of the 

 head (variations of the Dn and Vn organs common to other genera of the family) will in most 

 cases serve to delimit species groups, and some species. But, in some groups there are often 

 distinct differences in these organs between adults and juveniles and between males and 

 females. Unfortunately, these organs are tender and easily damaged or lost. However, despite 

 these obstacles, it is possible to arrange the species into groups (albeit rather broad groups) on 

 the basis of the kind and arrangement of the luminous organs of the head; the principal 

 categories of the key to species used here are based on these arrangements (for adults and 

 young adults only). This was attempted by Fraser-Brunner (1949), who employed four sub- 

 genera, Hyperphotops Fraser-Brunner, 1949; Partthophos Jordan and Hubbs, 1925; Lam- 

 prossa Jordan and Hubbs, 1925; and Diaphus Eigenmann and Eigenmann, 1890. 



Bolin (1959) segregated into the genus Lobianchia some species formerly placed in sub- 

 genus Hyperphotops, primarily on the basis of the presence of supracaudal and infracaudal 

 luminous glands — glands not present in other diaphid species. Also, Bolin, on the basis of 

 dentition, placed the remaining species in two genera, Aethoprora Goode and Bean, 1896, and 

 Diaphus Eigenmann and Eigenmann, 1890. The genus Diaphus was characterised as having 

 an inner series of broad-based, sharply recurved teeth on the posterior part of the upper jaw. 

 Teeth of the outer series of the upper jaw and dentary of genus Aethoprora were described as 

 small and closely set, with those of the inner series somewhat enlarged and tending to be more 

 widely spaced, particularly on the dentary. 



The types of luminous organs of the head appear to coincide with the respective types of 

 dentition in that the genus Diaphus (as restricted by Bolin) has the Dn and Vn widely sepa- 

 rated and, in addition, a more or less prominent suborbital organ (So); the genus Aethoprora 

 has a well-defined Dn and often has the Vn elongated and extending upward and contiguous or 

 confluent with Dn, but has no So. 



Moser and Ahlstrom (1974) supported the concept of three genera by demonstrating that 

 the larvae of Lobianchia, Aethoprora, and Diaphus differ recognizably. However, they re- 

 ported on the larvae of only one species each of the latter plus two genera — Diaphus theta and 

 D. paciftcus (Aethoprora group). 



Nafpaktitis (1968) suggested that only the two genera Lobianchia and Diaphus be consid- 

 ered valid. His examination of large numbers of specimens of diaphid series revealed that the 

 premaxillary teeth displayed a series of gradually changing shapes and that species with 

 luminous organs of the head that differed considerably from those species assigned to genus 

 Diaphus (as restricted by Bolin) had dentition closely resembling that of that genus. Nafpak- 

 titis concluded, in addition, that although these luminous organs were generally useful in 

 separating species groups, their structure, development, and considerable variation in appear- 

 ance, plus certain instances of striking sexual dimorphism, were at present too poorly under- 

 stood to warrant separation into genera. 



1 concur in this opinion. The finding of overlap in types of dentition, the incomplete 

 knowledge of development of luminous organs of the head and transitional states of develop- 

 ment between the two proposed genera, plus the as yet meager (although cogent) evidence of 

 differences in larval types, indicate that more study is needed before adding more confusion to 



92 



