a sufficiently difficult group. Therefore, only the genera Lobianchia and Diaphus are recog- 

 nized herein. 



The species oi Diaphus treated herein are segregated more or less by the type of luminous 

 organs of the head (see Fig. 83) only as a convenience to the student, and there is no intention 

 to represent a truly phylogenetic relationship. I have seen only adult, or young adult, speci- 

 mens, and information on the development with age of the head organs of the very young of 

 most species must await further study. 



It is important to remember that the species treated in the following key to identification 

 may not represent the total actually occupying the eastern Pacific Ocean, since the south- 

 central and western portions are still poorly collected. Many of the species were originally 

 described and known only from the Atlantic Ocean, or from the Indo-Pacific region and from 

 near New Zealand. Often these were represented by only a few specimens, frequently in poor 

 condition, and it was not possible to compare specimens from widely separate localities. Thus, 

 there was seldom an indication as to whether those found in the eastern Pacific were different 

 or truly identical with those of distant waters, and, perhaps, were expatriates. An attempt has 

 been made to include species which may conceivably occur in the eastern Pacific, largely 

 because they are known to occur in water masses that impinge on those of the area of study. 



Predictably, several taxonomic problems were encountered. One that I am presently un- 

 able to solve involves at least seven apparently distinct forms, or species. Four of these conform 

 reasonably well to Brauer's (1906) description and figure of Diaphus fulgens (from the tropical 

 Indian Ocean), and one may be that species; three other forms are very similar to D. rafines- 

 quii (Cocco, 1838), presently known only from the Atlantic Ocean. Two species described by 

 Gilbert (1908) and based on juveniles from near Nukuhiva, Marquesas, Diaphus nanus and D. 

 agassizi, the types respectively 17 and 21 mm long, also present problems. One of the fulgens- 

 like forms may represent D. nanus, but I cannot at all place D. agassizi. 



Fraser-Brunner (1949) placed/), agassizi in the subgenus Hyperphotops (= Lobianchia), 

 apparently because of the presence of only a Dn, as described and figured by Gilbert. But, as 

 shown by Nafpaktitis (1968), very often diaphid fishes of about 20 mm or less have only the Dn 

 visible, regardless of the ultimate development of other head organs. It is thus quite probable 

 that D. agassizi should not be placed in that group having only a Dn, a character now consid- 

 ered as confined to the genus Lobianchia. Indeed, the patterns of body photophores are quite 

 unlike that of L. gemellarii, L. dofleini, or L. urolampa, although they are basically more like 

 those of the latter species. I have seen no large specimens bearing only a Dn that conform to 

 the description and figure presented by Gilbert ( 1908) for the diminutive holotype ofD. agas- 

 sizi . 



It is hoped that the obvious shortcomings and inadequacies of this effort on diaphid fishes 

 of the eastern Pacific Ocean will induce others to undertake studies on this very interesting 

 but quite exasperating group of fishes. 



Lobianchia 



Gatti, 1903 



Five VO, the first 3 usually in an ascending oblique line ( VO2 raised to level of VO:j in one 

 species ). Luminous organs of head limited to a small, inconspicuous Dn. No luminous scales at 

 PLO. A long supracaudal luminous gland present in adult males; a shorter, rather inconspicu- 

 ous infracaudal gland present in adult females. One Pol; 4 Pre. 



Key to species of Lobianchia 



la. All upper photophores at or very near lateral line. First 3 VO not in a straight ascending 

 series, VO;. raised to level of VO.,. Only 2 SAO photophores. First AOa highly 

 elevated L. urolampa 



lb. All photophores well below lateral line. First 3 VO in a straight, ascending series. Three 

 SAO. First AOa not, or but little, elevated 2 



93 



